diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,27356 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 3902,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005125576627370579,
+      "grad_norm": 6.828924103980576,
+      "learning_rate": 1.6949152542372883e-07,
+      "loss": 1.1498,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010251153254741158,
+      "grad_norm": 7.819575513427743,
+      "learning_rate": 3.3898305084745766e-07,
+      "loss": 1.2452,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015376729882111738,
+      "grad_norm": 7.516775560171066,
+      "learning_rate": 5.084745762711865e-07,
+      "loss": 1.2206,
+      "step": 3
+    },
+    {
+      "epoch": 0.0020502306509482316,
+      "grad_norm": 7.133213585773929,
+      "learning_rate": 6.779661016949153e-07,
+      "loss": 1.1874,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025627883136852894,
+      "grad_norm": 6.753132765039162,
+      "learning_rate": 8.474576271186441e-07,
+      "loss": 1.1121,
+      "step": 5
+    },
+    {
+      "epoch": 0.0030753459764223477,
+      "grad_norm": 7.064184575724668,
+      "learning_rate": 1.016949152542373e-06,
+      "loss": 1.2437,
+      "step": 6
+    },
+    {
+      "epoch": 0.0035879036391594054,
+      "grad_norm": 6.047750996847248,
+      "learning_rate": 1.186440677966102e-06,
+      "loss": 1.1526,
+      "step": 7
+    },
+    {
+      "epoch": 0.004100461301896463,
+      "grad_norm": 6.0089801992484615,
+      "learning_rate": 1.3559322033898307e-06,
+      "loss": 1.1468,
+      "step": 8
+    },
+    {
+      "epoch": 0.0046130189646335215,
+      "grad_norm": 4.447082698725203,
+      "learning_rate": 1.5254237288135596e-06,
+      "loss": 1.0696,
+      "step": 9
+    },
+    {
+      "epoch": 0.005125576627370579,
+      "grad_norm": 3.901247717019762,
+      "learning_rate": 1.6949152542372882e-06,
+      "loss": 1.0203,
+      "step": 10
+    },
+    {
+      "epoch": 0.005638134290107637,
+      "grad_norm": 3.785155388206917,
+      "learning_rate": 1.8644067796610171e-06,
+      "loss": 1.1358,
+      "step": 11
+    },
+    {
+      "epoch": 0.006150691952844695,
+      "grad_norm": 2.60587246095241,
+      "learning_rate": 2.033898305084746e-06,
+      "loss": 0.9632,
+      "step": 12
+    },
+    {
+      "epoch": 0.006663249615581753,
+      "grad_norm": 2.7080976605100373,
+      "learning_rate": 2.203389830508475e-06,
+      "loss": 0.9762,
+      "step": 13
+    },
+    {
+      "epoch": 0.007175807278318811,
+      "grad_norm": 2.66345949646028,
+      "learning_rate": 2.372881355932204e-06,
+      "loss": 0.9515,
+      "step": 14
+    },
+    {
+      "epoch": 0.007688364941055869,
+      "grad_norm": 2.860062645984448,
+      "learning_rate": 2.5423728813559323e-06,
+      "loss": 1.0176,
+      "step": 15
+    },
+    {
+      "epoch": 0.008200922603792926,
+      "grad_norm": 1.8895151804905954,
+      "learning_rate": 2.7118644067796613e-06,
+      "loss": 0.8846,
+      "step": 16
+    },
+    {
+      "epoch": 0.008713480266529985,
+      "grad_norm": 1.9829614193667728,
+      "learning_rate": 2.8813559322033903e-06,
+      "loss": 0.9046,
+      "step": 17
+    },
+    {
+      "epoch": 0.009226037929267043,
+      "grad_norm": 2.0582741856513014,
+      "learning_rate": 3.0508474576271192e-06,
+      "loss": 0.9076,
+      "step": 18
+    },
+    {
+      "epoch": 0.009738595592004101,
+      "grad_norm": 1.7985178320233453,
+      "learning_rate": 3.2203389830508473e-06,
+      "loss": 0.8289,
+      "step": 19
+    },
+    {
+      "epoch": 0.010251153254741158,
+      "grad_norm": 1.5024494437749527,
+      "learning_rate": 3.3898305084745763e-06,
+      "loss": 0.799,
+      "step": 20
+    },
+    {
+      "epoch": 0.010763710917478216,
+      "grad_norm": 1.3917398592506678,
+      "learning_rate": 3.5593220338983053e-06,
+      "loss": 0.8384,
+      "step": 21
+    },
+    {
+      "epoch": 0.011276268580215274,
+      "grad_norm": 1.5876865207771096,
+      "learning_rate": 3.7288135593220342e-06,
+      "loss": 0.7568,
+      "step": 22
+    },
+    {
+      "epoch": 0.011788826242952332,
+      "grad_norm": 2.6002458370470745,
+      "learning_rate": 3.898305084745763e-06,
+      "loss": 0.8207,
+      "step": 23
+    },
+    {
+      "epoch": 0.01230138390568939,
+      "grad_norm": 1.7838132382849243,
+      "learning_rate": 4.067796610169492e-06,
+      "loss": 0.7958,
+      "step": 24
+    },
+    {
+      "epoch": 0.012813941568426449,
+      "grad_norm": 1.2910689421769297,
+      "learning_rate": 4.23728813559322e-06,
+      "loss": 0.7978,
+      "step": 25
+    },
+    {
+      "epoch": 0.013326499231163505,
+      "grad_norm": 1.2626599535372367,
+      "learning_rate": 4.40677966101695e-06,
+      "loss": 0.8031,
+      "step": 26
+    },
+    {
+      "epoch": 0.013839056893900564,
+      "grad_norm": 0.9264599967438035,
+      "learning_rate": 4.576271186440678e-06,
+      "loss": 0.7325,
+      "step": 27
+    },
+    {
+      "epoch": 0.014351614556637622,
+      "grad_norm": 1.0430084018663441,
+      "learning_rate": 4.745762711864408e-06,
+      "loss": 0.7796,
+      "step": 28
+    },
+    {
+      "epoch": 0.01486417221937468,
+      "grad_norm": 0.8983419063746677,
+      "learning_rate": 4.915254237288136e-06,
+      "loss": 0.7086,
+      "step": 29
+    },
+    {
+      "epoch": 0.015376729882111738,
+      "grad_norm": 0.9324663597506756,
+      "learning_rate": 5.084745762711865e-06,
+      "loss": 0.6899,
+      "step": 30
+    },
+    {
+      "epoch": 0.015889287544848796,
+      "grad_norm": 0.9171825226261315,
+      "learning_rate": 5.254237288135594e-06,
+      "loss": 0.7368,
+      "step": 31
+    },
+    {
+      "epoch": 0.016401845207585853,
+      "grad_norm": 0.8798330884499767,
+      "learning_rate": 5.423728813559323e-06,
+      "loss": 0.7099,
+      "step": 32
+    },
+    {
+      "epoch": 0.016914402870322913,
+      "grad_norm": 0.8877917139995327,
+      "learning_rate": 5.593220338983051e-06,
+      "loss": 0.7468,
+      "step": 33
+    },
+    {
+      "epoch": 0.01742696053305997,
+      "grad_norm": 0.9502455022163924,
+      "learning_rate": 5.7627118644067805e-06,
+      "loss": 0.7451,
+      "step": 34
+    },
+    {
+      "epoch": 0.017939518195797026,
+      "grad_norm": 0.9013785549236059,
+      "learning_rate": 5.932203389830509e-06,
+      "loss": 0.6931,
+      "step": 35
+    },
+    {
+      "epoch": 0.018452075858534086,
+      "grad_norm": 0.7749891470744573,
+      "learning_rate": 6.1016949152542385e-06,
+      "loss": 0.668,
+      "step": 36
+    },
+    {
+      "epoch": 0.018964633521271142,
+      "grad_norm": 0.8922549326376857,
+      "learning_rate": 6.271186440677966e-06,
+      "loss": 0.7472,
+      "step": 37
+    },
+    {
+      "epoch": 0.019477191184008202,
+      "grad_norm": 0.8074634290979792,
+      "learning_rate": 6.440677966101695e-06,
+      "loss": 0.7201,
+      "step": 38
+    },
+    {
+      "epoch": 0.01998974884674526,
+      "grad_norm": 0.8410207684408009,
+      "learning_rate": 6.610169491525424e-06,
+      "loss": 0.7085,
+      "step": 39
+    },
+    {
+      "epoch": 0.020502306509482315,
+      "grad_norm": 0.7741971999945924,
+      "learning_rate": 6.779661016949153e-06,
+      "loss": 0.6274,
+      "step": 40
+    },
+    {
+      "epoch": 0.021014864172219375,
+      "grad_norm": 0.7910617727808956,
+      "learning_rate": 6.949152542372882e-06,
+      "loss": 0.7166,
+      "step": 41
+    },
+    {
+      "epoch": 0.021527421834956432,
+      "grad_norm": 0.7691918476967737,
+      "learning_rate": 7.1186440677966106e-06,
+      "loss": 0.6762,
+      "step": 42
+    },
+    {
+      "epoch": 0.022039979497693492,
+      "grad_norm": 0.7667705263191332,
+      "learning_rate": 7.288135593220339e-06,
+      "loss": 0.6943,
+      "step": 43
+    },
+    {
+      "epoch": 0.02255253716043055,
+      "grad_norm": 0.7161850680216578,
+      "learning_rate": 7.4576271186440685e-06,
+      "loss": 0.6559,
+      "step": 44
+    },
+    {
+      "epoch": 0.023065094823167605,
+      "grad_norm": 0.8056990033404475,
+      "learning_rate": 7.627118644067797e-06,
+      "loss": 0.6868,
+      "step": 45
+    },
+    {
+      "epoch": 0.023577652485904665,
+      "grad_norm": 0.7022126380647868,
+      "learning_rate": 7.796610169491526e-06,
+      "loss": 0.6263,
+      "step": 46
+    },
+    {
+      "epoch": 0.02409021014864172,
+      "grad_norm": 0.7867660389218702,
+      "learning_rate": 7.966101694915255e-06,
+      "loss": 0.7121,
+      "step": 47
+    },
+    {
+      "epoch": 0.02460276781137878,
+      "grad_norm": 0.6982411068325368,
+      "learning_rate": 8.135593220338983e-06,
+      "loss": 0.618,
+      "step": 48
+    },
+    {
+      "epoch": 0.025115325474115838,
+      "grad_norm": 0.762602814769391,
+      "learning_rate": 8.305084745762712e-06,
+      "loss": 0.6738,
+      "step": 49
+    },
+    {
+      "epoch": 0.025627883136852898,
+      "grad_norm": 0.7420405387657487,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 0.6552,
+      "step": 50
+    },
+    {
+      "epoch": 0.026140440799589954,
+      "grad_norm": 0.8156620888823138,
+      "learning_rate": 8.64406779661017e-06,
+      "loss": 0.6542,
+      "step": 51
+    },
+    {
+      "epoch": 0.02665299846232701,
+      "grad_norm": 0.7628615209816971,
+      "learning_rate": 8.8135593220339e-06,
+      "loss": 0.7107,
+      "step": 52
+    },
+    {
+      "epoch": 0.02716555612506407,
+      "grad_norm": 0.6857111123018776,
+      "learning_rate": 8.983050847457628e-06,
+      "loss": 0.6138,
+      "step": 53
+    },
+    {
+      "epoch": 0.027678113787801127,
+      "grad_norm": 0.7410815455138027,
+      "learning_rate": 9.152542372881356e-06,
+      "loss": 0.6243,
+      "step": 54
+    },
+    {
+      "epoch": 0.028190671450538187,
+      "grad_norm": 0.7370248668464935,
+      "learning_rate": 9.322033898305085e-06,
+      "loss": 0.6596,
+      "step": 55
+    },
+    {
+      "epoch": 0.028703229113275244,
+      "grad_norm": 0.7628016934923556,
+      "learning_rate": 9.491525423728815e-06,
+      "loss": 0.6469,
+      "step": 56
+    },
+    {
+      "epoch": 0.0292157867760123,
+      "grad_norm": 0.7623191764504682,
+      "learning_rate": 9.661016949152544e-06,
+      "loss": 0.695,
+      "step": 57
+    },
+    {
+      "epoch": 0.02972834443874936,
+      "grad_norm": 0.7381305900352751,
+      "learning_rate": 9.830508474576272e-06,
+      "loss": 0.6246,
+      "step": 58
+    },
+    {
+      "epoch": 0.030240902101486417,
+      "grad_norm": 0.7265935136412146,
+      "learning_rate": 1e-05,
+      "loss": 0.6165,
+      "step": 59
+    },
+    {
+      "epoch": 0.030753459764223477,
+      "grad_norm": 0.7451766944147012,
+      "learning_rate": 1.016949152542373e-05,
+      "loss": 0.6254,
+      "step": 60
+    },
+    {
+      "epoch": 0.031266017426960536,
+      "grad_norm": 0.7096224688731553,
+      "learning_rate": 1.0338983050847458e-05,
+      "loss": 0.6343,
+      "step": 61
+    },
+    {
+      "epoch": 0.03177857508969759,
+      "grad_norm": 0.7810001573031462,
+      "learning_rate": 1.0508474576271188e-05,
+      "loss": 0.6268,
+      "step": 62
+    },
+    {
+      "epoch": 0.03229113275243465,
+      "grad_norm": 0.793070763665909,
+      "learning_rate": 1.0677966101694917e-05,
+      "loss": 0.655,
+      "step": 63
+    },
+    {
+      "epoch": 0.032803690415171706,
+      "grad_norm": 0.7250330420578672,
+      "learning_rate": 1.0847457627118645e-05,
+      "loss": 0.6374,
+      "step": 64
+    },
+    {
+      "epoch": 0.03331624807790876,
+      "grad_norm": 0.7220582373247162,
+      "learning_rate": 1.1016949152542374e-05,
+      "loss": 0.6126,
+      "step": 65
+    },
+    {
+      "epoch": 0.033828805740645826,
+      "grad_norm": 0.7336393723109893,
+      "learning_rate": 1.1186440677966102e-05,
+      "loss": 0.5833,
+      "step": 66
+    },
+    {
+      "epoch": 0.03434136340338288,
+      "grad_norm": 0.7087490867628216,
+      "learning_rate": 1.1355932203389833e-05,
+      "loss": 0.5926,
+      "step": 67
+    },
+    {
+      "epoch": 0.03485392106611994,
+      "grad_norm": 0.7329593111711369,
+      "learning_rate": 1.1525423728813561e-05,
+      "loss": 0.5965,
+      "step": 68
+    },
+    {
+      "epoch": 0.035366478728856995,
+      "grad_norm": 0.7037691207381355,
+      "learning_rate": 1.169491525423729e-05,
+      "loss": 0.6161,
+      "step": 69
+    },
+    {
+      "epoch": 0.03587903639159405,
+      "grad_norm": 0.8455423523934471,
+      "learning_rate": 1.1864406779661018e-05,
+      "loss": 0.7017,
+      "step": 70
+    },
+    {
+      "epoch": 0.036391594054331115,
+      "grad_norm": 0.7587405811685459,
+      "learning_rate": 1.2033898305084745e-05,
+      "loss": 0.604,
+      "step": 71
+    },
+    {
+      "epoch": 0.03690415171706817,
+      "grad_norm": 0.732002502525456,
+      "learning_rate": 1.2203389830508477e-05,
+      "loss": 0.5648,
+      "step": 72
+    },
+    {
+      "epoch": 0.03741670937980523,
+      "grad_norm": 0.75148930281832,
+      "learning_rate": 1.2372881355932205e-05,
+      "loss": 0.5636,
+      "step": 73
+    },
+    {
+      "epoch": 0.037929267042542285,
+      "grad_norm": 0.7563177066630524,
+      "learning_rate": 1.2542372881355932e-05,
+      "loss": 0.6097,
+      "step": 74
+    },
+    {
+      "epoch": 0.03844182470527934,
+      "grad_norm": 0.7062856749021322,
+      "learning_rate": 1.2711864406779661e-05,
+      "loss": 0.5962,
+      "step": 75
+    },
+    {
+      "epoch": 0.038954382368016405,
+      "grad_norm": 0.786643973966164,
+      "learning_rate": 1.288135593220339e-05,
+      "loss": 0.6115,
+      "step": 76
+    },
+    {
+      "epoch": 0.03946694003075346,
+      "grad_norm": 0.7621449769925668,
+      "learning_rate": 1.305084745762712e-05,
+      "loss": 0.6267,
+      "step": 77
+    },
+    {
+      "epoch": 0.03997949769349052,
+      "grad_norm": 0.7369511292408728,
+      "learning_rate": 1.3220338983050848e-05,
+      "loss": 0.6101,
+      "step": 78
+    },
+    {
+      "epoch": 0.040492055356227574,
+      "grad_norm": 0.7582620942012265,
+      "learning_rate": 1.3389830508474577e-05,
+      "loss": 0.6336,
+      "step": 79
+    },
+    {
+      "epoch": 0.04100461301896463,
+      "grad_norm": 1.040289084728475,
+      "learning_rate": 1.3559322033898305e-05,
+      "loss": 0.6404,
+      "step": 80
+    },
+    {
+      "epoch": 0.041517170681701694,
+      "grad_norm": 0.7672603827593716,
+      "learning_rate": 1.3728813559322034e-05,
+      "loss": 0.5961,
+      "step": 81
+    },
+    {
+      "epoch": 0.04202972834443875,
+      "grad_norm": 0.7374245673111222,
+      "learning_rate": 1.3898305084745764e-05,
+      "loss": 0.6151,
+      "step": 82
+    },
+    {
+      "epoch": 0.04254228600717581,
+      "grad_norm": 0.7115166771273684,
+      "learning_rate": 1.4067796610169493e-05,
+      "loss": 0.6066,
+      "step": 83
+    },
+    {
+      "epoch": 0.043054843669912864,
+      "grad_norm": 0.7712106873566231,
+      "learning_rate": 1.4237288135593221e-05,
+      "loss": 0.5708,
+      "step": 84
+    },
+    {
+      "epoch": 0.04356740133264992,
+      "grad_norm": 0.696613902157862,
+      "learning_rate": 1.440677966101695e-05,
+      "loss": 0.5688,
+      "step": 85
+    },
+    {
+      "epoch": 0.044079958995386984,
+      "grad_norm": 0.6890719514913243,
+      "learning_rate": 1.4576271186440678e-05,
+      "loss": 0.5541,
+      "step": 86
+    },
+    {
+      "epoch": 0.04459251665812404,
+      "grad_norm": 0.7429210938066708,
+      "learning_rate": 1.4745762711864408e-05,
+      "loss": 0.6287,
+      "step": 87
+    },
+    {
+      "epoch": 0.0451050743208611,
+      "grad_norm": 0.7222843734382394,
+      "learning_rate": 1.4915254237288137e-05,
+      "loss": 0.5639,
+      "step": 88
+    },
+    {
+      "epoch": 0.04561763198359815,
+      "grad_norm": 0.7479848515191257,
+      "learning_rate": 1.5084745762711865e-05,
+      "loss": 0.5613,
+      "step": 89
+    },
+    {
+      "epoch": 0.04613018964633521,
+      "grad_norm": 0.872510095222935,
+      "learning_rate": 1.5254237288135594e-05,
+      "loss": 0.5573,
+      "step": 90
+    },
+    {
+      "epoch": 0.04664274730907227,
+      "grad_norm": 0.7653826265479223,
+      "learning_rate": 1.5423728813559326e-05,
+      "loss": 0.624,
+      "step": 91
+    },
+    {
+      "epoch": 0.04715530497180933,
+      "grad_norm": 0.7635281439229672,
+      "learning_rate": 1.5593220338983053e-05,
+      "loss": 0.5674,
+      "step": 92
+    },
+    {
+      "epoch": 0.047667862634546386,
+      "grad_norm": 0.7522452247482868,
+      "learning_rate": 1.576271186440678e-05,
+      "loss": 0.5884,
+      "step": 93
+    },
+    {
+      "epoch": 0.04818042029728344,
+      "grad_norm": 0.770014182225227,
+      "learning_rate": 1.593220338983051e-05,
+      "loss": 0.5894,
+      "step": 94
+    },
+    {
+      "epoch": 0.0486929779600205,
+      "grad_norm": 0.7135537572232957,
+      "learning_rate": 1.6101694915254237e-05,
+      "loss": 0.5799,
+      "step": 95
+    },
+    {
+      "epoch": 0.04920553562275756,
+      "grad_norm": 0.7003786369390538,
+      "learning_rate": 1.6271186440677967e-05,
+      "loss": 0.5593,
+      "step": 96
+    },
+    {
+      "epoch": 0.04971809328549462,
+      "grad_norm": 0.7539610695623407,
+      "learning_rate": 1.6440677966101697e-05,
+      "loss": 0.5643,
+      "step": 97
+    },
+    {
+      "epoch": 0.050230650948231675,
+      "grad_norm": 0.7483251207792487,
+      "learning_rate": 1.6610169491525424e-05,
+      "loss": 0.5503,
+      "step": 98
+    },
+    {
+      "epoch": 0.05074320861096873,
+      "grad_norm": 1.5526109460138462,
+      "learning_rate": 1.6779661016949154e-05,
+      "loss": 0.6229,
+      "step": 99
+    },
+    {
+      "epoch": 0.051255766273705795,
+      "grad_norm": 0.7798011641844642,
+      "learning_rate": 1.694915254237288e-05,
+      "loss": 0.5607,
+      "step": 100
+    },
+    {
+      "epoch": 0.05176832393644285,
+      "grad_norm": 0.7482900057069612,
+      "learning_rate": 1.711864406779661e-05,
+      "loss": 0.5606,
+      "step": 101
+    },
+    {
+      "epoch": 0.05228088159917991,
+      "grad_norm": 0.7392563438263927,
+      "learning_rate": 1.728813559322034e-05,
+      "loss": 0.5861,
+      "step": 102
+    },
+    {
+      "epoch": 0.052793439261916965,
+      "grad_norm": 0.8155713047619292,
+      "learning_rate": 1.745762711864407e-05,
+      "loss": 0.5659,
+      "step": 103
+    },
+    {
+      "epoch": 0.05330599692465402,
+      "grad_norm": 0.8419926077497434,
+      "learning_rate": 1.76271186440678e-05,
+      "loss": 0.5283,
+      "step": 104
+    },
+    {
+      "epoch": 0.053818554587391085,
+      "grad_norm": 0.7231639928690438,
+      "learning_rate": 1.7796610169491526e-05,
+      "loss": 0.5868,
+      "step": 105
+    },
+    {
+      "epoch": 0.05433111225012814,
+      "grad_norm": 0.7365010474323096,
+      "learning_rate": 1.7966101694915256e-05,
+      "loss": 0.5833,
+      "step": 106
+    },
+    {
+      "epoch": 0.0548436699128652,
+      "grad_norm": 0.8054916659519644,
+      "learning_rate": 1.8135593220338986e-05,
+      "loss": 0.5696,
+      "step": 107
+    },
+    {
+      "epoch": 0.055356227575602254,
+      "grad_norm": 0.7285094461842039,
+      "learning_rate": 1.8305084745762713e-05,
+      "loss": 0.5431,
+      "step": 108
+    },
+    {
+      "epoch": 0.05586878523833931,
+      "grad_norm": 0.7267403035728092,
+      "learning_rate": 1.8474576271186443e-05,
+      "loss": 0.5618,
+      "step": 109
+    },
+    {
+      "epoch": 0.056381342901076374,
+      "grad_norm": 0.8484036844060503,
+      "learning_rate": 1.864406779661017e-05,
+      "loss": 0.5686,
+      "step": 110
+    },
+    {
+      "epoch": 0.05689390056381343,
+      "grad_norm": 0.7005124510226866,
+      "learning_rate": 1.88135593220339e-05,
+      "loss": 0.574,
+      "step": 111
+    },
+    {
+      "epoch": 0.05740645822655049,
+      "grad_norm": 0.7812320139757176,
+      "learning_rate": 1.898305084745763e-05,
+      "loss": 0.6093,
+      "step": 112
+    },
+    {
+      "epoch": 0.057919015889287544,
+      "grad_norm": 0.7766953885601624,
+      "learning_rate": 1.9152542372881357e-05,
+      "loss": 0.5664,
+      "step": 113
+    },
+    {
+      "epoch": 0.0584315735520246,
+      "grad_norm": 0.7136070455557347,
+      "learning_rate": 1.9322033898305087e-05,
+      "loss": 0.5265,
+      "step": 114
+    },
+    {
+      "epoch": 0.058944131214761664,
+      "grad_norm": 0.7631408681814531,
+      "learning_rate": 1.9491525423728814e-05,
+      "loss": 0.6164,
+      "step": 115
+    },
+    {
+      "epoch": 0.05945668887749872,
+      "grad_norm": 0.7711046604890244,
+      "learning_rate": 1.9661016949152545e-05,
+      "loss": 0.5836,
+      "step": 116
+    },
+    {
+      "epoch": 0.05996924654023578,
+      "grad_norm": 0.7457924612011373,
+      "learning_rate": 1.9830508474576275e-05,
+      "loss": 0.5616,
+      "step": 117
+    },
+    {
+      "epoch": 0.06048180420297283,
+      "grad_norm": 0.7036275755014029,
+      "learning_rate": 2e-05,
+      "loss": 0.5975,
+      "step": 118
+    },
+    {
+      "epoch": 0.06099436186570989,
+      "grad_norm": 0.7209795882650869,
+      "learning_rate": 1.9999996553585817e-05,
+      "loss": 0.5475,
+      "step": 119
+    },
+    {
+      "epoch": 0.06150691952844695,
+      "grad_norm": 0.7074784324506611,
+      "learning_rate": 1.999998621434563e-05,
+      "loss": 0.5582,
+      "step": 120
+    },
+    {
+      "epoch": 0.06201947719118401,
+      "grad_norm": 0.7327102051701987,
+      "learning_rate": 1.9999968982286576e-05,
+      "loss": 0.5532,
+      "step": 121
+    },
+    {
+      "epoch": 0.06253203485392107,
+      "grad_norm": 0.7218402431366631,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 0.582,
+      "step": 122
+    },
+    {
+      "epoch": 0.06304459251665813,
+      "grad_norm": 0.7924663020341974,
+      "learning_rate": 1.9999913839764117e-05,
+      "loss": 0.6016,
+      "step": 123
+    },
+    {
+      "epoch": 0.06355715017939519,
+      "grad_norm": 0.7556628505812407,
+      "learning_rate": 1.9999875929338718e-05,
+      "loss": 0.5492,
+      "step": 124
+    },
+    {
+      "epoch": 0.06406970784213224,
+      "grad_norm": 0.7023854898264759,
+      "learning_rate": 1.999983112617047e-05,
+      "loss": 0.5446,
+      "step": 125
+    },
+    {
+      "epoch": 0.0645822655048693,
+      "grad_norm": 0.723214644183677,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 0.5511,
+      "step": 126
+    },
+    {
+      "epoch": 0.06509482316760636,
+      "grad_norm": 0.7082656460674889,
+      "learning_rate": 1.9999720841733688e-05,
+      "loss": 0.5611,
+      "step": 127
+    },
+    {
+      "epoch": 0.06560738083034341,
+      "grad_norm": 0.6992645557001506,
+      "learning_rate": 1.999965536054117e-05,
+      "loss": 0.5532,
+      "step": 128
+    },
+    {
+      "epoch": 0.06611993849308047,
+      "grad_norm": 0.7247470923924793,
+      "learning_rate": 1.999958298675784e-05,
+      "loss": 0.5178,
+      "step": 129
+    },
+    {
+      "epoch": 0.06663249615581752,
+      "grad_norm": 0.6886069726166952,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.5082,
+      "step": 130
+    },
+    {
+      "epoch": 0.06714505381855458,
+      "grad_norm": 0.7100425557900145,
+      "learning_rate": 1.999941756162301e-05,
+      "loss": 0.5636,
+      "step": 131
+    },
+    {
+      "epoch": 0.06765761148129165,
+      "grad_norm": 0.7513963784115057,
+      "learning_rate": 1.9999324510385543e-05,
+      "loss": 0.5616,
+      "step": 132
+    },
+    {
+      "epoch": 0.06817016914402871,
+      "grad_norm": 0.6831788883562648,
+      "learning_rate": 1.9999224566785302e-05,
+      "loss": 0.5451,
+      "step": 133
+    },
+    {
+      "epoch": 0.06868272680676576,
+      "grad_norm": 0.6499219027870621,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 0.5612,
+      "step": 134
+    },
+    {
+      "epoch": 0.06919528446950282,
+      "grad_norm": 0.6740350074359596,
+      "learning_rate": 1.9999004002776822e-05,
+      "loss": 0.5349,
+      "step": 135
+    },
+    {
+      "epoch": 0.06970784213223988,
+      "grad_norm": 0.6912654417569759,
+      "learning_rate": 1.999888338252061e-05,
+      "loss": 0.5813,
+      "step": 136
+    },
+    {
+      "epoch": 0.07022039979497693,
+      "grad_norm": 0.6673567319885729,
+      "learning_rate": 1.9998755870205694e-05,
+      "loss": 0.5761,
+      "step": 137
+    },
+    {
+      "epoch": 0.07073295745771399,
+      "grad_norm": 0.6931871009176531,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 0.5472,
+      "step": 138
+    },
+    {
+      "epoch": 0.07124551512045105,
+      "grad_norm": 0.6939008424252235,
+      "learning_rate": 1.9998480169756052e-05,
+      "loss": 0.5421,
+      "step": 139
+    },
+    {
+      "epoch": 0.0717580727831881,
+      "grad_norm": 0.7267338303000498,
+      "learning_rate": 1.999833198181137e-05,
+      "loss": 0.6071,
+      "step": 140
+    },
+    {
+      "epoch": 0.07227063044592516,
+      "grad_norm": 0.730514756584142,
+      "learning_rate": 1.9998176902188042e-05,
+      "loss": 0.5428,
+      "step": 141
+    },
+    {
+      "epoch": 0.07278318810866223,
+      "grad_norm": 0.6580913025879824,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 0.5211,
+      "step": 142
+    },
+    {
+      "epoch": 0.07329574577139929,
+      "grad_norm": 0.6151036821302706,
+      "learning_rate": 1.999784606833781e-05,
+      "loss": 0.533,
+      "step": 143
+    },
+    {
+      "epoch": 0.07380830343413634,
+      "grad_norm": 0.6911226232647965,
+      "learning_rate": 1.9997670314338936e-05,
+      "loss": 0.5551,
+      "step": 144
+    },
+    {
+      "epoch": 0.0743208610968734,
+      "grad_norm": 0.6919112455938906,
+      "learning_rate": 1.9997487669117506e-05,
+      "loss": 0.5728,
+      "step": 145
+    },
+    {
+      "epoch": 0.07483341875961046,
+      "grad_norm": 0.7079136585889061,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 0.5852,
+      "step": 146
+    },
+    {
+      "epoch": 0.07534597642234751,
+      "grad_norm": 0.6845247929382975,
+      "learning_rate": 1.999710170551529e-05,
+      "loss": 0.5822,
+      "step": 147
+    },
+    {
+      "epoch": 0.07585853408508457,
+      "grad_norm": 0.6635595088188522,
+      "learning_rate": 1.999689838740054e-05,
+      "loss": 0.537,
+      "step": 148
+    },
+    {
+      "epoch": 0.07637109174782163,
+      "grad_norm": 0.8045407290957151,
+      "learning_rate": 1.9996688178595303e-05,
+      "loss": 0.5664,
+      "step": 149
+    },
+    {
+      "epoch": 0.07688364941055868,
+      "grad_norm": 0.7781789356613706,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 0.5463,
+      "step": 150
+    },
+    {
+      "epoch": 0.07739620707329574,
+      "grad_norm": 0.6682878546095715,
+      "learning_rate": 1.9996247089497703e-05,
+      "loss": 0.5407,
+      "step": 151
+    },
+    {
+      "epoch": 0.07790876473603281,
+      "grad_norm": 0.6648947611393514,
+      "learning_rate": 1.999601620950937e-05,
+      "loss": 0.5339,
+      "step": 152
+    },
+    {
+      "epoch": 0.07842132239876987,
+      "grad_norm": 0.8051214468472842,
+      "learning_rate": 1.999577843943863e-05,
+      "loss": 0.5753,
+      "step": 153
+    },
+    {
+      "epoch": 0.07893388006150692,
+      "grad_norm": 0.6302912999279268,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 0.5062,
+      "step": 154
+    },
+    {
+      "epoch": 0.07944643772424398,
+      "grad_norm": 0.7642829965098711,
+      "learning_rate": 1.9995282229710208e-05,
+      "loss": 0.6174,
+      "step": 155
+    },
+    {
+      "epoch": 0.07995899538698104,
+      "grad_norm": 0.7206363707809622,
+      "learning_rate": 1.9995023790394558e-05,
+      "loss": 0.5637,
+      "step": 156
+    },
+    {
+      "epoch": 0.08047155304971809,
+      "grad_norm": 0.6693816108955785,
+      "learning_rate": 1.999475846168055e-05,
+      "loss": 0.5129,
+      "step": 157
+    },
+    {
+      "epoch": 0.08098411071245515,
+      "grad_norm": 0.7538668286416843,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 0.5588,
+      "step": 158
+    },
+    {
+      "epoch": 0.0814966683751922,
+      "grad_norm": 0.7006079780738972,
+      "learning_rate": 1.9994207136793763e-05,
+      "loss": 0.5828,
+      "step": 159
+    },
+    {
+      "epoch": 0.08200922603792926,
+      "grad_norm": 0.6739314364214601,
+      "learning_rate": 1.9993921141001003e-05,
+      "loss": 0.5425,
+      "step": 160
+    },
+    {
+      "epoch": 0.08252178370066633,
+      "grad_norm": 0.7024072648197256,
+      "learning_rate": 1.999362825656992e-05,
+      "loss": 0.5427,
+      "step": 161
+    },
+    {
+      "epoch": 0.08303434136340339,
+      "grad_norm": 0.6542017259403593,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 0.5011,
+      "step": 162
+    },
+    {
+      "epoch": 0.08354689902614044,
+      "grad_norm": 0.6649256890311436,
+      "learning_rate": 1.999302182260506e-05,
+      "loss": 0.5666,
+      "step": 163
+    },
+    {
+      "epoch": 0.0840594566888775,
+      "grad_norm": 0.6860291304313701,
+      "learning_rate": 1.999270827348929e-05,
+      "loss": 0.5125,
+      "step": 164
+    },
+    {
+      "epoch": 0.08457201435161456,
+      "grad_norm": 0.6863439776413617,
+      "learning_rate": 1.999238783657121e-05,
+      "loss": 0.523,
+      "step": 165
+    },
+    {
+      "epoch": 0.08508457201435161,
+      "grad_norm": 0.6564423475660008,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 0.5282,
+      "step": 166
+    },
+    {
+      "epoch": 0.08559712967708867,
+      "grad_norm": 0.6644587141806786,
+      "learning_rate": 1.999172630021635e-05,
+      "loss": 0.5343,
+      "step": 167
+    },
+    {
+      "epoch": 0.08610968733982573,
+      "grad_norm": 0.7065954318987528,
+      "learning_rate": 1.9991385201235552e-05,
+      "loss": 0.5575,
+      "step": 168
+    },
+    {
+      "epoch": 0.08662224500256278,
+      "grad_norm": 0.6585648826605895,
+      "learning_rate": 1.999103721536442e-05,
+      "loss": 0.5435,
+      "step": 169
+    },
+    {
+      "epoch": 0.08713480266529984,
+      "grad_norm": 0.680449951703073,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 0.5642,
+      "step": 170
+    },
+    {
+      "epoch": 0.08764736032803691,
+      "grad_norm": 0.67469015211052,
+      "learning_rate": 1.999032058391532e-05,
+      "loss": 0.55,
+      "step": 171
+    },
+    {
+      "epoch": 0.08815991799077397,
+      "grad_norm": 0.7063725403188603,
+      "learning_rate": 1.9989951938831315e-05,
+      "loss": 0.6137,
+      "step": 172
+    },
+    {
+      "epoch": 0.08867247565351102,
+      "grad_norm": 0.6669432859699835,
+      "learning_rate": 1.9989576407844894e-05,
+      "loss": 0.5529,
+      "step": 173
+    },
+    {
+      "epoch": 0.08918503331624808,
+      "grad_norm": 0.6660604453171053,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 0.547,
+      "step": 174
+    },
+    {
+      "epoch": 0.08969759097898514,
+      "grad_norm": 0.6646005522732114,
+      "learning_rate": 1.9988804689204934e-05,
+      "loss": 0.5591,
+      "step": 175
+    },
+    {
+      "epoch": 0.0902101486417222,
+      "grad_norm": 0.6872639685657017,
+      "learning_rate": 1.9988408502083328e-05,
+      "loss": 0.5537,
+      "step": 176
+    },
+    {
+      "epoch": 0.09072270630445925,
+      "grad_norm": 0.7186103976751318,
+      "learning_rate": 1.998800543012317e-05,
+      "loss": 0.5457,
+      "step": 177
+    },
+    {
+      "epoch": 0.0912352639671963,
+      "grad_norm": 0.735453128764158,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 0.5306,
+      "step": 178
+    },
+    {
+      "epoch": 0.09174782162993336,
+      "grad_norm": 0.729514659848969,
+      "learning_rate": 1.9987178632803265e-05,
+      "loss": 0.5211,
+      "step": 179
+    },
+    {
+      "epoch": 0.09226037929267042,
+      "grad_norm": 0.7584869087874441,
+      "learning_rate": 1.9986754908013415e-05,
+      "loss": 0.592,
+      "step": 180
+    },
+    {
+      "epoch": 0.09277293695540749,
+      "grad_norm": 0.6909185237722874,
+      "learning_rate": 1.9986324299524807e-05,
+      "loss": 0.5367,
+      "step": 181
+    },
+    {
+      "epoch": 0.09328549461814455,
+      "grad_norm": 0.6846532867784723,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 0.517,
+      "step": 182
+    },
+    {
+      "epoch": 0.0937980522808816,
+      "grad_norm": 0.6636201979800417,
+      "learning_rate": 1.99854424326433e-05,
+      "loss": 0.5507,
+      "step": 183
+    },
+    {
+      "epoch": 0.09431060994361866,
+      "grad_norm": 0.7177468928341916,
+      "learning_rate": 1.998499117485826e-05,
+      "loss": 0.5278,
+      "step": 184
+    },
+    {
+      "epoch": 0.09482316760635572,
+      "grad_norm": 0.6917182244561318,
+      "learning_rate": 1.998453303459017e-05,
+      "loss": 0.5078,
+      "step": 185
+    },
+    {
+      "epoch": 0.09533572526909277,
+      "grad_norm": 0.643347076699553,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 0.5302,
+      "step": 186
+    },
+    {
+      "epoch": 0.09584828293182983,
+      "grad_norm": 0.6744486049815316,
+      "learning_rate": 1.9983596107872748e-05,
+      "loss": 0.5113,
+      "step": 187
+    },
+    {
+      "epoch": 0.09636084059456688,
+      "grad_norm": 0.6400172644212885,
+      "learning_rate": 1.9983117322069226e-05,
+      "loss": 0.5527,
+      "step": 188
+    },
+    {
+      "epoch": 0.09687339825730394,
+      "grad_norm": 0.6757124370912334,
+      "learning_rate": 1.9982631655074264e-05,
+      "loss": 0.4712,
+      "step": 189
+    },
+    {
+      "epoch": 0.097385955920041,
+      "grad_norm": 0.6987970258104467,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 0.5061,
+      "step": 190
+    },
+    {
+      "epoch": 0.09789851358277807,
+      "grad_norm": 0.7201890118912728,
+      "learning_rate": 1.998163967885384e-05,
+      "loss": 0.5535,
+      "step": 191
+    },
+    {
+      "epoch": 0.09841107124551512,
+      "grad_norm": 0.708720879254577,
+      "learning_rate": 1.9981133370312123e-05,
+      "loss": 0.539,
+      "step": 192
+    },
+    {
+      "epoch": 0.09892362890825218,
+      "grad_norm": 0.7206502407042601,
+      "learning_rate": 1.9980620181946476e-05,
+      "loss": 0.5527,
+      "step": 193
+    },
+    {
+      "epoch": 0.09943618657098924,
+      "grad_norm": 0.612440621653419,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 0.4841,
+      "step": 194
+    },
+    {
+      "epoch": 0.0999487442337263,
+      "grad_norm": 0.715744977664602,
+      "learning_rate": 1.997957316716307e-05,
+      "loss": 0.582,
+      "step": 195
+    },
+    {
+      "epoch": 0.10046130189646335,
+      "grad_norm": 0.6710781887626952,
+      "learning_rate": 1.9979039341466997e-05,
+      "loss": 0.508,
+      "step": 196
+    },
+    {
+      "epoch": 0.10097385955920041,
+      "grad_norm": 0.6513860306538726,
+      "learning_rate": 1.9978498637390375e-05,
+      "loss": 0.5443,
+      "step": 197
+    },
+    {
+      "epoch": 0.10148641722193746,
+      "grad_norm": 0.6130586691384516,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 0.4765,
+      "step": 198
+    },
+    {
+      "epoch": 0.10199897488467452,
+      "grad_norm": 0.676169327961081,
+      "learning_rate": 1.997739659559101e-05,
+      "loss": 0.5275,
+      "step": 199
+    },
+    {
+      "epoch": 0.10251153254741159,
+      "grad_norm": 0.6966156340106725,
+      "learning_rate": 1.9976835258627884e-05,
+      "loss": 0.5331,
+      "step": 200
+    },
+    {
+      "epoch": 0.10302409021014865,
+      "grad_norm": 0.629015203742023,
+      "learning_rate": 1.9976267044803442e-05,
+      "loss": 0.5058,
+      "step": 201
+    },
+    {
+      "epoch": 0.1035366478728857,
+      "grad_norm": 0.6552827219254702,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 0.5215,
+      "step": 202
+    },
+    {
+      "epoch": 0.10404920553562276,
+      "grad_norm": 0.698955262967572,
+      "learning_rate": 1.9975109988142e-05,
+      "loss": 0.5009,
+      "step": 203
+    },
+    {
+      "epoch": 0.10456176319835982,
+      "grad_norm": 0.6498219812295211,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 0.4722,
+      "step": 204
+    },
+    {
+      "epoch": 0.10507432086109687,
+      "grad_norm": 0.6076056274373424,
+      "learning_rate": 1.9973925428796837e-05,
+      "loss": 0.5222,
+      "step": 205
+    },
+    {
+      "epoch": 0.10558687852383393,
+      "grad_norm": 0.7212238574021528,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 0.5595,
+      "step": 206
+    },
+    {
+      "epoch": 0.10609943618657099,
+      "grad_norm": 0.6977384930068656,
+      "learning_rate": 1.9972713370033937e-05,
+      "loss": 0.5257,
+      "step": 207
+    },
+    {
+      "epoch": 0.10661199384930804,
+      "grad_norm": 0.6289257483067097,
+      "learning_rate": 1.997209702941219e-05,
+      "loss": 0.4849,
+      "step": 208
+    },
+    {
+      "epoch": 0.1071245515120451,
+      "grad_norm": 0.6173675269284898,
+      "learning_rate": 1.9971473815195106e-05,
+      "loss": 0.5078,
+      "step": 209
+    },
+    {
+      "epoch": 0.10763710917478217,
+      "grad_norm": 0.7070346286820632,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 0.5537,
+      "step": 210
+    },
+    {
+      "epoch": 0.10814966683751923,
+      "grad_norm": 0.7535798456935602,
+      "learning_rate": 1.9970206767697958e-05,
+      "loss": 0.5352,
+      "step": 211
+    },
+    {
+      "epoch": 0.10866222450025628,
+      "grad_norm": 0.6487988566136014,
+      "learning_rate": 1.9969562935291248e-05,
+      "loss": 0.5128,
+      "step": 212
+    },
+    {
+      "epoch": 0.10917478216299334,
+      "grad_norm": 0.6804722412677658,
+      "learning_rate": 1.996891223103591e-05,
+      "loss": 0.5492,
+      "step": 213
+    },
+    {
+      "epoch": 0.1096873398257304,
+      "grad_norm": 0.7671766457441788,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 0.5607,
+      "step": 214
+    },
+    {
+      "epoch": 0.11019989748846745,
+      "grad_norm": 0.660991395641086,
+      "learning_rate": 1.996759020877817e-05,
+      "loss": 0.5009,
+      "step": 215
+    },
+    {
+      "epoch": 0.11071245515120451,
+      "grad_norm": 0.6482136792868417,
+      "learning_rate": 1.996691889168701e-05,
+      "loss": 0.5021,
+      "step": 216
+    },
+    {
+      "epoch": 0.11122501281394157,
+      "grad_norm": 0.629465375637083,
+      "learning_rate": 1.9966240704569722e-05,
+      "loss": 0.5056,
+      "step": 217
+    },
+    {
+      "epoch": 0.11173757047667862,
+      "grad_norm": 0.6634827126342779,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 0.5134,
+      "step": 218
+    },
+    {
+      "epoch": 0.11225012813941568,
+      "grad_norm": 0.8127651370532675,
+      "learning_rate": 1.996486372213133e-05,
+      "loss": 0.505,
+      "step": 219
+    },
+    {
+      "epoch": 0.11276268580215275,
+      "grad_norm": 0.6786610117758186,
+      "learning_rate": 1.9964164927759354e-05,
+      "loss": 0.5261,
+      "step": 220
+    },
+    {
+      "epoch": 0.1132752434648898,
+      "grad_norm": 0.6745463977467945,
+      "learning_rate": 1.9963459265259512e-05,
+      "loss": 0.5038,
+      "step": 221
+    },
+    {
+      "epoch": 0.11378780112762686,
+      "grad_norm": 0.6523477429199502,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 0.5004,
+      "step": 222
+    },
+    {
+      "epoch": 0.11430035879036392,
+      "grad_norm": 0.6098862420301526,
+      "learning_rate": 1.9962027337826538e-05,
+      "loss": 0.4761,
+      "step": 223
+    },
+    {
+      "epoch": 0.11481291645310097,
+      "grad_norm": 0.6488002034265402,
+      "learning_rate": 1.9961301073880413e-05,
+      "loss": 0.5234,
+      "step": 224
+    },
+    {
+      "epoch": 0.11532547411583803,
+      "grad_norm": 0.7344102525393421,
+      "learning_rate": 1.9960567943780423e-05,
+      "loss": 0.5524,
+      "step": 225
+    },
+    {
+      "epoch": 0.11583803177857509,
+      "grad_norm": 0.6491015924662693,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.5022,
+      "step": 226
+    },
+    {
+      "epoch": 0.11635058944131214,
+      "grad_norm": 0.6263604527869331,
+      "learning_rate": 1.995908108714491e-05,
+      "loss": 0.505,
+      "step": 227
+    },
+    {
+      "epoch": 0.1168631471040492,
+      "grad_norm": 0.6495312412001931,
+      "learning_rate": 1.9958327361634248e-05,
+      "loss": 0.5174,
+      "step": 228
+    },
+    {
+      "epoch": 0.11737570476678626,
+      "grad_norm": 0.6307154699675626,
+      "learning_rate": 1.995756677201945e-05,
+      "loss": 0.4989,
+      "step": 229
+    },
+    {
+      "epoch": 0.11788826242952333,
+      "grad_norm": 0.6725773965783972,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 0.5446,
+      "step": 230
+    },
+    {
+      "epoch": 0.11840082009226038,
+      "grad_norm": 0.680312475332318,
+      "learning_rate": 1.9956025002579214e-05,
+      "loss": 0.5184,
+      "step": 231
+    },
+    {
+      "epoch": 0.11891337775499744,
+      "grad_norm": 0.6271095811416424,
+      "learning_rate": 1.995524382381649e-05,
+      "loss": 0.5035,
+      "step": 232
+    },
+    {
+      "epoch": 0.1194259354177345,
+      "grad_norm": 0.6372204881318652,
+      "learning_rate": 1.995445578307506e-05,
+      "loss": 0.5063,
+      "step": 233
+    },
+    {
+      "epoch": 0.11993849308047155,
+      "grad_norm": 0.6960348354656597,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 0.5222,
+      "step": 234
+    },
+    {
+      "epoch": 0.12045105074320861,
+      "grad_norm": 0.679134064099199,
+      "learning_rate": 1.9952859117833526e-05,
+      "loss": 0.5014,
+      "step": 235
+    },
+    {
+      "epoch": 0.12096360840594567,
+      "grad_norm": 0.6327161248901884,
+      "learning_rate": 1.9952050494433983e-05,
+      "loss": 0.4691,
+      "step": 236
+    },
+    {
+      "epoch": 0.12147616606868272,
+      "grad_norm": 0.6813855330325674,
+      "learning_rate": 1.9951235011256836e-05,
+      "loss": 0.5383,
+      "step": 237
+    },
+    {
+      "epoch": 0.12198872373141978,
+      "grad_norm": 0.6597145908391994,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.5464,
+      "step": 238
+    },
+    {
+      "epoch": 0.12250128139415685,
+      "grad_norm": 0.7066103631239254,
+      "learning_rate": 1.9949583467822863e-05,
+      "loss": 0.5497,
+      "step": 239
+    },
+    {
+      "epoch": 0.1230138390568939,
+      "grad_norm": 0.6701145448173735,
+      "learning_rate": 1.9948747408704415e-05,
+      "loss": 0.5037,
+      "step": 240
+    },
+    {
+      "epoch": 0.12352639671963096,
+      "grad_norm": 0.6250706226615986,
+      "learning_rate": 1.9947904492085122e-05,
+      "loss": 0.5187,
+      "step": 241
+    },
+    {
+      "epoch": 0.12403895438236802,
+      "grad_norm": 0.6867781020389341,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.5363,
+      "step": 242
+    },
+    {
+      "epoch": 0.12455151204510508,
+      "grad_norm": 0.6306425786466211,
+      "learning_rate": 1.9946198088672776e-05,
+      "loss": 0.5093,
+      "step": 243
+    },
+    {
+      "epoch": 0.12506406970784215,
+      "grad_norm": 0.6346298914327985,
+      "learning_rate": 1.994533460305591e-05,
+      "loss": 0.5008,
+      "step": 244
+    },
+    {
+      "epoch": 0.1255766273705792,
+      "grad_norm": 0.6469682114122414,
+      "learning_rate": 1.9944464262290596e-05,
+      "loss": 0.502,
+      "step": 245
+    },
+    {
+      "epoch": 0.12608918503331626,
+      "grad_norm": 0.6602600643969688,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 0.5037,
+      "step": 246
+    },
+    {
+      "epoch": 0.12660174269605332,
+      "grad_norm": 0.7013954571944024,
+      "learning_rate": 1.9942703017718977e-05,
+      "loss": 0.546,
+      "step": 247
+    },
+    {
+      "epoch": 0.12711430035879037,
+      "grad_norm": 0.6330691055310902,
+      "learning_rate": 1.9941812115126664e-05,
+      "loss": 0.4822,
+      "step": 248
+    },
+    {
+      "epoch": 0.12762685802152743,
+      "grad_norm": 0.7465507079573863,
+      "learning_rate": 1.994091435981389e-05,
+      "loss": 0.5068,
+      "step": 249
+    },
+    {
+      "epoch": 0.12813941568426448,
+      "grad_norm": 0.6588584086156924,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 0.505,
+      "step": 250
+    },
+    {
+      "epoch": 0.12865197334700154,
+      "grad_norm": 0.6238320566016561,
+      "learning_rate": 1.9939098293506906e-05,
+      "loss": 0.4721,
+      "step": 251
+    },
+    {
+      "epoch": 0.1291645310097386,
+      "grad_norm": 0.6141978790791331,
+      "learning_rate": 1.993817998376448e-05,
+      "loss": 0.4966,
+      "step": 252
+    },
+    {
+      "epoch": 0.12967708867247565,
+      "grad_norm": 0.6630522703493429,
+      "learning_rate": 1.9937254823805156e-05,
+      "loss": 0.5074,
+      "step": 253
+    },
+    {
+      "epoch": 0.1301896463352127,
+      "grad_norm": 0.6583018700340271,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 0.4684,
+      "step": 254
+    },
+    {
+      "epoch": 0.13070220399794977,
+      "grad_norm": 0.616541829968061,
+      "learning_rate": 1.9935383955791326e-05,
+      "loss": 0.5004,
+      "step": 255
+    },
+    {
+      "epoch": 0.13121476166068682,
+      "grad_norm": 0.6240159279994848,
+      "learning_rate": 1.993443824902638e-05,
+      "loss": 0.5059,
+      "step": 256
+    },
+    {
+      "epoch": 0.13172731932342388,
+      "grad_norm": 0.7389765767116779,
+      "learning_rate": 1.993348569462365e-05,
+      "loss": 0.5301,
+      "step": 257
+    },
+    {
+      "epoch": 0.13223987698616094,
+      "grad_norm": 0.6312388765525858,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 0.4571,
+      "step": 258
+    },
+    {
+      "epoch": 0.132752434648898,
+      "grad_norm": 0.7205280765556014,
+      "learning_rate": 1.9931560045535873e-05,
+      "loss": 0.5226,
+      "step": 259
+    },
+    {
+      "epoch": 0.13326499231163505,
+      "grad_norm": 0.6516614143258374,
+      "learning_rate": 1.9930586952178146e-05,
+      "loss": 0.5024,
+      "step": 260
+    },
+    {
+      "epoch": 0.1337775499743721,
+      "grad_norm": 0.6377697330787861,
+      "learning_rate": 1.9929607013837268e-05,
+      "loss": 0.5101,
+      "step": 261
+    },
+    {
+      "epoch": 0.13429010763710916,
+      "grad_norm": 0.6646512898208231,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 0.4676,
+      "step": 262
+    },
+    {
+      "epoch": 0.13480266529984622,
+      "grad_norm": 0.6993937394984001,
+      "learning_rate": 1.9927626604912594e-05,
+      "loss": 0.5191,
+      "step": 263
+    },
+    {
+      "epoch": 0.1353152229625833,
+      "grad_norm": 1.1843461216931792,
+      "learning_rate": 1.9926626135693866e-05,
+      "loss": 0.5108,
+      "step": 264
+    },
+    {
+      "epoch": 0.13582778062532036,
+      "grad_norm": 0.5971471863080159,
+      "learning_rate": 1.9925618824222103e-05,
+      "loss": 0.479,
+      "step": 265
+    },
+    {
+      "epoch": 0.13634033828805742,
+      "grad_norm": 0.6592370619698613,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 0.4879,
+      "step": 266
+    },
+    {
+      "epoch": 0.13685289595079447,
+      "grad_norm": 0.6885722570360087,
+      "learning_rate": 1.9923583677301507e-05,
+      "loss": 0.5072,
+      "step": 267
+    },
+    {
+      "epoch": 0.13736545361353153,
+      "grad_norm": 0.6685845211235351,
+      "learning_rate": 1.9922555843255463e-05,
+      "loss": 0.5398,
+      "step": 268
+    },
+    {
+      "epoch": 0.13787801127626859,
+      "grad_norm": 0.6939145733335776,
+      "learning_rate": 1.9921521169761974e-05,
+      "loss": 0.5145,
+      "step": 269
+    },
+    {
+      "epoch": 0.13839056893900564,
+      "grad_norm": 0.6237277152885383,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 0.5174,
+      "step": 270
+    },
+    {
+      "epoch": 0.1389031266017427,
+      "grad_norm": 0.6383264317519571,
+      "learning_rate": 1.9919431307290105e-05,
+      "loss": 0.4815,
+      "step": 271
+    },
+    {
+      "epoch": 0.13941568426447976,
+      "grad_norm": 0.6210999210346629,
+      "learning_rate": 1.991837611975223e-05,
+      "loss": 0.4845,
+      "step": 272
+    },
+    {
+      "epoch": 0.1399282419272168,
+      "grad_norm": 0.6077454588818982,
+      "learning_rate": 1.991731409564792e-05,
+      "loss": 0.4665,
+      "step": 273
+    },
+    {
+      "epoch": 0.14044079958995387,
+      "grad_norm": 0.6198399236028331,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.5196,
+      "step": 274
+    },
+    {
+      "epoch": 0.14095335725269093,
+      "grad_norm": 0.6159802987653653,
+      "learning_rate": 1.9915169540672857e-05,
+      "loss": 0.4962,
+      "step": 275
+    },
+    {
+      "epoch": 0.14146591491542798,
+      "grad_norm": 0.6495827440294281,
+      "learning_rate": 1.9914087011280308e-05,
+      "loss": 0.5211,
+      "step": 276
+    },
+    {
+      "epoch": 0.14197847257816504,
+      "grad_norm": 0.6588618150127611,
+      "learning_rate": 1.9912997648277736e-05,
+      "loss": 0.5162,
+      "step": 277
+    },
+    {
+      "epoch": 0.1424910302409021,
+      "grad_norm": 0.6675920774530805,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.5025,
+      "step": 278
+    },
+    {
+      "epoch": 0.14300358790363915,
+      "grad_norm": 0.7346623848197696,
+      "learning_rate": 1.9910798424450734e-05,
+      "loss": 0.5172,
+      "step": 279
+    },
+    {
+      "epoch": 0.1435161455663762,
+      "grad_norm": 0.6269270776086969,
+      "learning_rate": 1.99096885651422e-05,
+      "loss": 0.4756,
+      "step": 280
+    },
+    {
+      "epoch": 0.14402870322911326,
+      "grad_norm": 0.5936653898303434,
+      "learning_rate": 1.9908571875255416e-05,
+      "loss": 0.4551,
+      "step": 281
+    },
+    {
+      "epoch": 0.14454126089185032,
+      "grad_norm": 0.6200701056247242,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.4681,
+      "step": 282
+    },
+    {
+      "epoch": 0.1450538185545874,
+      "grad_norm": 0.6681084059608245,
+      "learning_rate": 1.990631800683066e-05,
+      "loss": 0.4627,
+      "step": 283
+    },
+    {
+      "epoch": 0.14556637621732446,
+      "grad_norm": 0.6432743061184111,
+      "learning_rate": 1.990518082984624e-05,
+      "loss": 0.4927,
+      "step": 284
+    },
+    {
+      "epoch": 0.14607893388006152,
+      "grad_norm": 0.6156175489099207,
+      "learning_rate": 1.9904036825390672e-05,
+      "loss": 0.5336,
+      "step": 285
+    },
+    {
+      "epoch": 0.14659149154279857,
+      "grad_norm": 0.6903312126087798,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 0.4969,
+      "step": 286
+    },
+    {
+      "epoch": 0.14710404920553563,
+      "grad_norm": 0.6419713846931644,
+      "learning_rate": 1.990172833722498e-05,
+      "loss": 0.5052,
+      "step": 287
+    },
+    {
+      "epoch": 0.1476166068682727,
+      "grad_norm": 0.6116829607362324,
+      "learning_rate": 1.9900563855106055e-05,
+      "loss": 0.4883,
+      "step": 288
+    },
+    {
+      "epoch": 0.14812916453100974,
+      "grad_norm": 0.6277739420919254,
+      "learning_rate": 1.9899392548698385e-05,
+      "loss": 0.4716,
+      "step": 289
+    },
+    {
+      "epoch": 0.1486417221937468,
+      "grad_norm": 0.6449379304077641,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4742,
+      "step": 290
+    },
+    {
+      "epoch": 0.14915427985648386,
+      "grad_norm": 0.6642940416469129,
+      "learning_rate": 1.9897029466250955e-05,
+      "loss": 0.4721,
+      "step": 291
+    },
+    {
+      "epoch": 0.1496668375192209,
+      "grad_norm": 0.6514359627116323,
+      "learning_rate": 1.989583769184003e-05,
+      "loss": 0.4921,
+      "step": 292
+    },
+    {
+      "epoch": 0.15017939518195797,
+      "grad_norm": 0.7076913089622164,
+      "learning_rate": 1.9894639096398022e-05,
+      "loss": 0.5051,
+      "step": 293
+    },
+    {
+      "epoch": 0.15069195284469503,
+      "grad_norm": 0.6114615595051633,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.437,
+      "step": 294
+    },
+    {
+      "epoch": 0.15120451050743208,
+      "grad_norm": 0.650873650947476,
+      "learning_rate": 1.989222144573015e-05,
+      "loss": 0.5056,
+      "step": 295
+    },
+    {
+      "epoch": 0.15171706817016914,
+      "grad_norm": 0.5703657220976504,
+      "learning_rate": 1.9891002392170727e-05,
+      "loss": 0.4351,
+      "step": 296
+    },
+    {
+      "epoch": 0.1522296258329062,
+      "grad_norm": 0.6917487380587324,
+      "learning_rate": 1.988977652091311e-05,
+      "loss": 0.4892,
+      "step": 297
+    },
+    {
+      "epoch": 0.15274218349564325,
+      "grad_norm": 0.6705027025919212,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 0.5207,
+      "step": 298
+    },
+    {
+      "epoch": 0.1532547411583803,
+      "grad_norm": 0.604687850643197,
+      "learning_rate": 1.9887304328687892e-05,
+      "loss": 0.49,
+      "step": 299
+    },
+    {
+      "epoch": 0.15376729882111737,
+      "grad_norm": 0.6248051110328243,
+      "learning_rate": 1.9886058009424323e-05,
+      "loss": 0.5208,
+      "step": 300
+    },
+    {
+      "epoch": 0.15427985648385442,
+      "grad_norm": 0.6290556632217106,
+      "learning_rate": 1.9884804875870645e-05,
+      "loss": 0.4558,
+      "step": 301
+    },
+    {
+      "epoch": 0.15479241414659148,
+      "grad_norm": 0.6106409542400119,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 0.4691,
+      "step": 302
+    },
+    {
+      "epoch": 0.15530497180932856,
+      "grad_norm": 0.6385125473408743,
+      "learning_rate": 1.988227816935269e-05,
+      "loss": 0.501,
+      "step": 303
+    },
+    {
+      "epoch": 0.15581752947206562,
+      "grad_norm": 0.616704669183617,
+      "learning_rate": 1.988100459813003e-05,
+      "loss": 0.4492,
+      "step": 304
+    },
+    {
+      "epoch": 0.15633008713480268,
+      "grad_norm": 0.6376072991379901,
+      "learning_rate": 1.9879724216100488e-05,
+      "loss": 0.4959,
+      "step": 305
+    },
+    {
+      "epoch": 0.15684264479753973,
+      "grad_norm": 0.6004130762695509,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 0.4407,
+      "step": 306
+    },
+    {
+      "epoch": 0.1573552024602768,
+      "grad_norm": 0.6657147777612079,
+      "learning_rate": 1.987714302315562e-05,
+      "loss": 0.5007,
+      "step": 307
+    },
+    {
+      "epoch": 0.15786776012301384,
+      "grad_norm": 0.6876255381162313,
+      "learning_rate": 1.987584221401947e-05,
+      "loss": 0.4594,
+      "step": 308
+    },
+    {
+      "epoch": 0.1583803177857509,
+      "grad_norm": 0.6517658513384645,
+      "learning_rate": 1.9874534597634776e-05,
+      "loss": 0.4717,
+      "step": 309
+    },
+    {
+      "epoch": 0.15889287544848796,
+      "grad_norm": 0.6521463712386892,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 0.5476,
+      "step": 310
+    },
+    {
+      "epoch": 0.15940543311122501,
+      "grad_norm": 0.6797507301820702,
+      "learning_rate": 1.9871898946729728e-05,
+      "loss": 0.4748,
+      "step": 311
+    },
+    {
+      "epoch": 0.15991799077396207,
+      "grad_norm": 0.6021339426111404,
+      "learning_rate": 1.9870570914026074e-05,
+      "loss": 0.4753,
+      "step": 312
+    },
+    {
+      "epoch": 0.16043054843669913,
+      "grad_norm": 0.6754150579528226,
+      "learning_rate": 1.9869236077707302e-05,
+      "loss": 0.5252,
+      "step": 313
+    },
+    {
+      "epoch": 0.16094310609943618,
+      "grad_norm": 0.5627040109659946,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 0.4543,
+      "step": 314
+    },
+    {
+      "epoch": 0.16145566376217324,
+      "grad_norm": 0.6314191614667306,
+      "learning_rate": 1.9866545997909387e-05,
+      "loss": 0.4708,
+      "step": 315
+    },
+    {
+      "epoch": 0.1619682214249103,
+      "grad_norm": 0.5816206496061312,
+      "learning_rate": 1.9865190756284467e-05,
+      "loss": 0.4659,
+      "step": 316
+    },
+    {
+      "epoch": 0.16248077908764735,
+      "grad_norm": 0.638327085484802,
+      "learning_rate": 1.9863828714752877e-05,
+      "loss": 0.5074,
+      "step": 317
+    },
+    {
+      "epoch": 0.1629933367503844,
+      "grad_norm": 0.6613323722211508,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 0.4961,
+      "step": 318
+    },
+    {
+      "epoch": 0.16350589441312147,
+      "grad_norm": 0.5673579399611183,
+      "learning_rate": 1.986108423572968e-05,
+      "loss": 0.4545,
+      "step": 319
+    },
+    {
+      "epoch": 0.16401845207585852,
+      "grad_norm": 0.6205816945594128,
+      "learning_rate": 1.9859701800129797e-05,
+      "loss": 0.491,
+      "step": 320
+    },
+    {
+      "epoch": 0.16453100973859558,
+      "grad_norm": 0.5997925524988109,
+      "learning_rate": 1.9858312568406686e-05,
+      "loss": 0.4694,
+      "step": 321
+    },
+    {
+      "epoch": 0.16504356740133266,
+      "grad_norm": 0.6063288202415337,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 0.4368,
+      "step": 322
+    },
+    {
+      "epoch": 0.16555612506406972,
+      "grad_norm": 0.6135835207451752,
+      "learning_rate": 1.9855513720425742e-05,
+      "loss": 0.4954,
+      "step": 323
+    },
+    {
+      "epoch": 0.16606868272680678,
+      "grad_norm": 0.6173614044473401,
+      "learning_rate": 1.985410410609711e-05,
+      "loss": 0.5019,
+      "step": 324
+    },
+    {
+      "epoch": 0.16658124038954383,
+      "grad_norm": 0.619538753062264,
+      "learning_rate": 1.9852687699503638e-05,
+      "loss": 0.4555,
+      "step": 325
+    },
+    {
+      "epoch": 0.1670937980522809,
+      "grad_norm": 0.68074284722669,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 0.5044,
+      "step": 326
+    },
+    {
+      "epoch": 0.16760635571501795,
+      "grad_norm": 0.6520918111081759,
+      "learning_rate": 1.9849834513432084e-05,
+      "loss": 0.502,
+      "step": 327
+    },
+    {
+      "epoch": 0.168118913377755,
+      "grad_norm": 0.6590103682355488,
+      "learning_rate": 1.9848397735920656e-05,
+      "loss": 0.4841,
+      "step": 328
+    },
+    {
+      "epoch": 0.16863147104049206,
+      "grad_norm": 0.5934596207654561,
+      "learning_rate": 1.984695417007769e-05,
+      "loss": 0.4498,
+      "step": 329
+    },
+    {
+      "epoch": 0.16914402870322912,
+      "grad_norm": 0.6762688194653954,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 0.4832,
+      "step": 330
+    },
+    {
+      "epoch": 0.16965658636596617,
+      "grad_norm": 0.594236743160809,
+      "learning_rate": 1.9844046677381936e-05,
+      "loss": 0.4698,
+      "step": 331
+    },
+    {
+      "epoch": 0.17016914402870323,
+      "grad_norm": 0.6646941168517261,
+      "learning_rate": 1.984258275253323e-05,
+      "loss": 0.5177,
+      "step": 332
+    },
+    {
+      "epoch": 0.17068170169144029,
+      "grad_norm": 0.6360962994508872,
+      "learning_rate": 1.984111204336116e-05,
+      "loss": 0.5142,
+      "step": 333
+    },
+    {
+      "epoch": 0.17119425935417734,
+      "grad_norm": 0.6300660481979834,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.4698,
+      "step": 334
+    },
+    {
+      "epoch": 0.1717068170169144,
+      "grad_norm": 0.5770762563190722,
+      "learning_rate": 1.983815027610654e-05,
+      "loss": 0.4797,
+      "step": 335
+    },
+    {
+      "epoch": 0.17221937467965145,
+      "grad_norm": 0.6595136054732988,
+      "learning_rate": 1.983665922006548e-05,
+      "loss": 0.4832,
+      "step": 336
+    },
+    {
+      "epoch": 0.1727319323423885,
+      "grad_norm": 0.6216416755679994,
+      "learning_rate": 1.983516138378404e-05,
+      "loss": 0.4944,
+      "step": 337
+    },
+    {
+      "epoch": 0.17324449000512557,
+      "grad_norm": 0.6386997911623253,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 0.4842,
+      "step": 338
+    },
+    {
+      "epoch": 0.17375704766786262,
+      "grad_norm": 0.6979708540975553,
+      "learning_rate": 1.9832145374634445e-05,
+      "loss": 0.5141,
+      "step": 339
+    },
+    {
+      "epoch": 0.17426960533059968,
+      "grad_norm": 0.6606037280736322,
+      "learning_rate": 1.9830627203845164e-05,
+      "loss": 0.4954,
+      "step": 340
+    },
+    {
+      "epoch": 0.17478216299333674,
+      "grad_norm": 0.6692402903530307,
+      "learning_rate": 1.9829102256973274e-05,
+      "loss": 0.4582,
+      "step": 341
+    },
+    {
+      "epoch": 0.17529472065607382,
+      "grad_norm": 0.6099113227401556,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 0.473,
+      "step": 342
+    },
+    {
+      "epoch": 0.17580727831881088,
+      "grad_norm": 0.6394252573766227,
+      "learning_rate": 1.9826032039190805e-05,
+      "loss": 0.4886,
+      "step": 343
+    },
+    {
+      "epoch": 0.17631983598154793,
+      "grad_norm": 0.6160079194195135,
+      "learning_rate": 1.9824486770396477e-05,
+      "loss": 0.4996,
+      "step": 344
+    },
+    {
+      "epoch": 0.176832393644285,
+      "grad_norm": 0.6920029905813989,
+      "learning_rate": 1.9822934729752036e-05,
+      "loss": 0.4834,
+      "step": 345
+    },
+    {
+      "epoch": 0.17734495130702205,
+      "grad_norm": 0.6534013277728197,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 0.4797,
+      "step": 346
+    },
+    {
+      "epoch": 0.1778575089697591,
+      "grad_norm": 0.5984153075309845,
+      "learning_rate": 1.9819810337196644e-05,
+      "loss": 0.4975,
+      "step": 347
+    },
+    {
+      "epoch": 0.17837006663249616,
+      "grad_norm": 0.6441658598317781,
+      "learning_rate": 1.9818237987439287e-05,
+      "loss": 0.4903,
+      "step": 348
+    },
+    {
+      "epoch": 0.17888262429523322,
+      "grad_norm": 0.6337149631987146,
+      "learning_rate": 1.981665887013899e-05,
+      "loss": 0.438,
+      "step": 349
+    },
+    {
+      "epoch": 0.17939518195797027,
+      "grad_norm": 0.6628922229916709,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 0.5167,
+      "step": 350
+    },
+    {
+      "epoch": 0.17990773962070733,
+      "grad_norm": 0.6151554170908305,
+      "learning_rate": 1.981348033726809e-05,
+      "loss": 0.4531,
+      "step": 351
+    },
+    {
+      "epoch": 0.1804202972834444,
+      "grad_norm": 0.62528112900968,
+      "learning_rate": 1.9811880923888388e-05,
+      "loss": 0.4812,
+      "step": 352
+    },
+    {
+      "epoch": 0.18093285494618144,
+      "grad_norm": 0.6321409797169452,
+      "learning_rate": 1.9810274747347565e-05,
+      "loss": 0.5085,
+      "step": 353
+    },
+    {
+      "epoch": 0.1814454126089185,
+      "grad_norm": 0.6005054631588977,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 0.4688,
+      "step": 354
+    },
+    {
+      "epoch": 0.18195797027165556,
+      "grad_norm": 0.6283363410749165,
+      "learning_rate": 1.9807042109215656e-05,
+      "loss": 0.4708,
+      "step": 355
+    },
+    {
+      "epoch": 0.1824705279343926,
+      "grad_norm": 0.572601869954044,
+      "learning_rate": 1.9805415649852767e-05,
+      "loss": 0.4348,
+      "step": 356
+    },
+    {
+      "epoch": 0.18298308559712967,
+      "grad_norm": 0.5630858605938541,
+      "learning_rate": 1.980378243178516e-05,
+      "loss": 0.421,
+      "step": 357
+    },
+    {
+      "epoch": 0.18349564325986673,
+      "grad_norm": 0.6100273836490916,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 0.4437,
+      "step": 358
+    },
+    {
+      "epoch": 0.18400820092260378,
+      "grad_norm": 0.7000740886691411,
+      "learning_rate": 1.980049572404344e-05,
+      "loss": 0.5082,
+      "step": 359
+    },
+    {
+      "epoch": 0.18452075858534084,
+      "grad_norm": 0.5927330616788398,
+      "learning_rate": 1.9798842236634797e-05,
+      "loss": 0.423,
+      "step": 360
+    },
+    {
+      "epoch": 0.18503331624807792,
+      "grad_norm": 0.6323939188280061,
+      "learning_rate": 1.9797181995052374e-05,
+      "loss": 0.462,
+      "step": 361
+    },
+    {
+      "epoch": 0.18554587391081498,
+      "grad_norm": 0.6757164905934084,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 0.4894,
+      "step": 362
+    },
+    {
+      "epoch": 0.18605843157355204,
+      "grad_norm": 0.6782002882356951,
+      "learning_rate": 1.979384125394835e-05,
+      "loss": 0.4741,
+      "step": 363
+    },
+    {
+      "epoch": 0.1865709892362891,
+      "grad_norm": 0.6328866984448983,
+      "learning_rate": 1.9792160756729468e-05,
+      "loss": 0.4913,
+      "step": 364
+    },
+    {
+      "epoch": 0.18708354689902615,
+      "grad_norm": 0.60307457026725,
+      "learning_rate": 1.9790473509942233e-05,
+      "loss": 0.4337,
+      "step": 365
+    },
+    {
+      "epoch": 0.1875961045617632,
+      "grad_norm": 0.6787116137544283,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 0.4508,
+      "step": 366
+    },
+    {
+      "epoch": 0.18810866222450026,
+      "grad_norm": 0.6023679405000496,
+      "learning_rate": 1.9787078772319328e-05,
+      "loss": 0.4662,
+      "step": 367
+    },
+    {
+      "epoch": 0.18862121988723732,
+      "grad_norm": 0.610470544055691,
+      "learning_rate": 1.978537128382359e-05,
+      "loss": 0.4559,
+      "step": 368
+    },
+    {
+      "epoch": 0.18913377754997437,
+      "grad_norm": 0.659239601809496,
+      "learning_rate": 1.978365705043937e-05,
+      "loss": 0.4549,
+      "step": 369
+    },
+    {
+      "epoch": 0.18964633521271143,
+      "grad_norm": 0.7177316010966338,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 0.4855,
+      "step": 370
+    },
+    {
+      "epoch": 0.1901588928754485,
+      "grad_norm": 0.6348151143499278,
+      "learning_rate": 1.9780208353736493e-05,
+      "loss": 0.483,
+      "step": 371
+    },
+    {
+      "epoch": 0.19067145053818554,
+      "grad_norm": 0.583862284865993,
+      "learning_rate": 1.977847389279497e-05,
+      "loss": 0.4766,
+      "step": 372
+    },
+    {
+      "epoch": 0.1911840082009226,
+      "grad_norm": 0.6491110007878494,
+      "learning_rate": 1.9776732691719215e-05,
+      "loss": 0.4555,
+      "step": 373
+    },
+    {
+      "epoch": 0.19169656586365966,
+      "grad_norm": 0.6075796737073911,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 0.4525,
+      "step": 374
+    },
+    {
+      "epoch": 0.1922091235263967,
+      "grad_norm": 0.6297178409888091,
+      "learning_rate": 1.977323007397038e-05,
+      "loss": 0.4649,
+      "step": 375
+    },
+    {
+      "epoch": 0.19272168118913377,
+      "grad_norm": 0.6216896420885621,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.4289,
+      "step": 376
+    },
+    {
+      "epoch": 0.19323423885187083,
+      "grad_norm": 0.585240895254178,
+      "learning_rate": 1.9769700510147173e-05,
+      "loss": 0.4621,
+      "step": 377
+    },
+    {
+      "epoch": 0.19374679651460788,
+      "grad_norm": 0.6137001756831676,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 0.4741,
+      "step": 378
+    },
+    {
+      "epoch": 0.19425935417734494,
+      "grad_norm": 0.5980522770629302,
+      "learning_rate": 1.976614400998105e-05,
+      "loss": 0.4807,
+      "step": 379
+    },
+    {
+      "epoch": 0.194771911840082,
+      "grad_norm": 0.6401846123325262,
+      "learning_rate": 1.9764355661830796e-05,
+      "loss": 0.4443,
+      "step": 380
+    },
+    {
+      "epoch": 0.19528446950281908,
+      "grad_norm": 0.6902112030324087,
+      "learning_rate": 1.9762560583277763e-05,
+      "loss": 0.4893,
+      "step": 381
+    },
+    {
+      "epoch": 0.19579702716555614,
+      "grad_norm": 0.6651211724555203,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.4935,
+      "step": 382
+    },
+    {
+      "epoch": 0.1963095848282932,
+      "grad_norm": 0.5795902806473132,
+      "learning_rate": 1.975895023991728e-05,
+      "loss": 0.4535,
+      "step": 383
+    },
+    {
+      "epoch": 0.19682214249103025,
+      "grad_norm": 0.6388766712581653,
+      "learning_rate": 1.9757134977598374e-05,
+      "loss": 0.4709,
+      "step": 384
+    },
+    {
+      "epoch": 0.1973347001537673,
+      "grad_norm": 0.6318231915819886,
+      "learning_rate": 1.975531298985379e-05,
+      "loss": 0.4739,
+      "step": 385
+    },
+    {
+      "epoch": 0.19784725781650436,
+      "grad_norm": 0.6471959352883845,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.4988,
+      "step": 386
+    },
+    {
+      "epoch": 0.19835981547924142,
+      "grad_norm": 0.6652940485551666,
+      "learning_rate": 1.975164884311567e-05,
+      "loss": 0.4548,
+      "step": 387
+    },
+    {
+      "epoch": 0.19887237314197848,
+      "grad_norm": 0.6928660681758323,
+      "learning_rate": 1.974980668664777e-05,
+      "loss": 0.488,
+      "step": 388
+    },
+    {
+      "epoch": 0.19938493080471553,
+      "grad_norm": 0.644086195630129,
+      "learning_rate": 1.974795780980545e-05,
+      "loss": 0.473,
+      "step": 389
+    },
+    {
+      "epoch": 0.1998974884674526,
+      "grad_norm": 0.6521309942939905,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 0.471,
+      "step": 390
+    },
+    {
+      "epoch": 0.20041004613018965,
+      "grad_norm": 0.5952672211396001,
+      "learning_rate": 1.974423990009979e-05,
+      "loss": 0.448,
+      "step": 391
+    },
+    {
+      "epoch": 0.2009226037929267,
+      "grad_norm": 0.6493957269440347,
+      "learning_rate": 1.9742370869799147e-05,
+      "loss": 0.4576,
+      "step": 392
+    },
+    {
+      "epoch": 0.20143516145566376,
+      "grad_norm": 0.5876646519544126,
+      "learning_rate": 1.9740495124249462e-05,
+      "loss": 0.4429,
+      "step": 393
+    },
+    {
+      "epoch": 0.20194771911840081,
+      "grad_norm": 0.6269547524071647,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 0.4808,
+      "step": 394
+    },
+    {
+      "epoch": 0.20246027678113787,
+      "grad_norm": 0.6355777129066641,
+      "learning_rate": 1.9736723492579295e-05,
+      "loss": 0.5028,
+      "step": 395
+    },
+    {
+      "epoch": 0.20297283444387493,
+      "grad_norm": 0.6568949945234874,
+      "learning_rate": 1.9734827609058533e-05,
+      "loss": 0.4648,
+      "step": 396
+    },
+    {
+      "epoch": 0.20348539210661198,
+      "grad_norm": 0.621824221497376,
+      "learning_rate": 1.9732925015488174e-05,
+      "loss": 0.4163,
+      "step": 397
+    },
+    {
+      "epoch": 0.20399794976934904,
+      "grad_norm": 0.6196400169541889,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.4613,
+      "step": 398
+    },
+    {
+      "epoch": 0.2045105074320861,
+      "grad_norm": 0.7589239049987565,
+      "learning_rate": 1.9729099703449e-05,
+      "loss": 0.4622,
+      "step": 399
+    },
+    {
+      "epoch": 0.20502306509482318,
+      "grad_norm": 0.613422834370645,
+      "learning_rate": 1.9727176987616897e-05,
+      "loss": 0.4404,
+      "step": 400
+    },
+    {
+      "epoch": 0.20553562275756024,
+      "grad_norm": 0.6584844933973998,
+      "learning_rate": 1.972524756700865e-05,
+      "loss": 0.4558,
+      "step": 401
+    },
+    {
+      "epoch": 0.2060481804202973,
+      "grad_norm": 0.6260484483459837,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 0.4848,
+      "step": 402
+    },
+    {
+      "epoch": 0.20656073808303435,
+      "grad_norm": 0.6433572683712755,
+      "learning_rate": 1.9721368616787976e-05,
+      "loss": 0.4758,
+      "step": 403
+    },
+    {
+      "epoch": 0.2070732957457714,
+      "grad_norm": 0.6631166765181412,
+      "learning_rate": 1.971941908984925e-05,
+      "loss": 0.462,
+      "step": 404
+    },
+    {
+      "epoch": 0.20758585340850846,
+      "grad_norm": 0.6693319444778841,
+      "learning_rate": 1.9717462863481753e-05,
+      "loss": 0.4518,
+      "step": 405
+    },
+    {
+      "epoch": 0.20809841107124552,
+      "grad_norm": 0.6044628117278833,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 0.4478,
+      "step": 406
+    },
+    {
+      "epoch": 0.20861096873398258,
+      "grad_norm": 0.6026188665858787,
+      "learning_rate": 1.9713530317858647e-05,
+      "loss": 0.452,
+      "step": 407
+    },
+    {
+      "epoch": 0.20912352639671963,
+      "grad_norm": 0.5881271457407786,
+      "learning_rate": 1.9711554001313674e-05,
+      "loss": 0.4425,
+      "step": 408
+    },
+    {
+      "epoch": 0.2096360840594567,
+      "grad_norm": 0.6109666412341543,
+      "learning_rate": 1.9709570990761207e-05,
+      "loss": 0.4248,
+      "step": 409
+    },
+    {
+      "epoch": 0.21014864172219375,
+      "grad_norm": 0.6509284072938516,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 0.4723,
+      "step": 410
+    },
+    {
+      "epoch": 0.2106611993849308,
+      "grad_norm": 0.6459053744798665,
+      "learning_rate": 1.970558489310581e-05,
+      "loss": 0.5412,
+      "step": 411
+    },
+    {
+      "epoch": 0.21117375704766786,
+      "grad_norm": 0.7369672572600167,
+      "learning_rate": 1.9703581808750436e-05,
+      "loss": 0.5338,
+      "step": 412
+    },
+    {
+      "epoch": 0.21168631471040492,
+      "grad_norm": 0.6176454355201523,
+      "learning_rate": 1.9701572035882658e-05,
+      "loss": 0.4478,
+      "step": 413
+    },
+    {
+      "epoch": 0.21219887237314197,
+      "grad_norm": 0.6525491064500083,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 0.4634,
+      "step": 414
+    },
+    {
+      "epoch": 0.21271143003587903,
+      "grad_norm": 0.5831422894545593,
+      "learning_rate": 1.9697532430155718e-05,
+      "loss": 0.434,
+      "step": 415
+    },
+    {
+      "epoch": 0.21322398769861609,
+      "grad_norm": 0.6265787824703533,
+      "learning_rate": 1.9695502600080983e-05,
+      "loss": 0.4731,
+      "step": 416
+    },
+    {
+      "epoch": 0.21373654536135314,
+      "grad_norm": 0.5955421973236761,
+      "learning_rate": 1.969346608706271e-05,
+      "loss": 0.406,
+      "step": 417
+    },
+    {
+      "epoch": 0.2142491030240902,
+      "grad_norm": 0.6061327598596417,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.4395,
+      "step": 418
+    },
+    {
+      "epoch": 0.21476166068682725,
+      "grad_norm": 0.6956343169709611,
+      "learning_rate": 1.9689373017815076e-05,
+      "loss": 0.4621,
+      "step": 419
+    },
+    {
+      "epoch": 0.21527421834956434,
+      "grad_norm": 0.6757933934629281,
+      "learning_rate": 1.9687316464406997e-05,
+      "loss": 0.49,
+      "step": 420
+    },
+    {
+      "epoch": 0.2157867760123014,
+      "grad_norm": 0.6491171208955874,
+      "learning_rate": 1.9685253233697943e-05,
+      "loss": 0.4334,
+      "step": 421
+    },
+    {
+      "epoch": 0.21629933367503845,
+      "grad_norm": 0.6208804480015244,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 0.4186,
+      "step": 422
+    },
+    {
+      "epoch": 0.2168118913377755,
+      "grad_norm": 0.6724451704426618,
+      "learning_rate": 1.9681106746070096e-05,
+      "loss": 0.4961,
+      "step": 423
+    },
+    {
+      "epoch": 0.21732444900051257,
+      "grad_norm": 0.6682974431641722,
+      "learning_rate": 1.967902349200941e-05,
+      "loss": 0.4598,
+      "step": 424
+    },
+    {
+      "epoch": 0.21783700666324962,
+      "grad_norm": 0.6270765318778506,
+      "learning_rate": 1.9676933566363943e-05,
+      "loss": 0.4811,
+      "step": 425
+    },
+    {
+      "epoch": 0.21834956432598668,
+      "grad_norm": 0.7010331590031008,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 0.4216,
+      "step": 426
+    },
+    {
+      "epoch": 0.21886212198872373,
+      "grad_norm": 0.6283443612318651,
+      "learning_rate": 1.9672733706085488e-05,
+      "loss": 0.4508,
+      "step": 427
+    },
+    {
+      "epoch": 0.2193746796514608,
+      "grad_norm": 0.6885445673206841,
+      "learning_rate": 1.9670623774347387e-05,
+      "loss": 0.4713,
+      "step": 428
+    },
+    {
+      "epoch": 0.21988723731419785,
+      "grad_norm": 0.6530128107819684,
+      "learning_rate": 1.9668507176814295e-05,
+      "loss": 0.4531,
+      "step": 429
+    },
+    {
+      "epoch": 0.2203997949769349,
+      "grad_norm": 0.6256988964032321,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 0.4427,
+      "step": 430
+    },
+    {
+      "epoch": 0.22091235263967196,
+      "grad_norm": 0.5853150136127208,
+      "learning_rate": 1.9664253990203456e-05,
+      "loss": 0.4345,
+      "step": 431
+    },
+    {
+      "epoch": 0.22142491030240902,
+      "grad_norm": 0.6302565356245501,
+      "learning_rate": 1.966211740405736e-05,
+      "loss": 0.4325,
+      "step": 432
+    },
+    {
+      "epoch": 0.22193746796514607,
+      "grad_norm": 0.6478629858337103,
+      "learning_rate": 1.9659974157979566e-05,
+      "loss": 0.4707,
+      "step": 433
+    },
+    {
+      "epoch": 0.22245002562788313,
+      "grad_norm": 0.6418563530889652,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.4519,
+      "step": 434
+    },
+    {
+      "epoch": 0.2229625832906202,
+      "grad_norm": 0.6908388316328086,
+      "learning_rate": 1.9655667691942684e-05,
+      "loss": 0.5089,
+      "step": 435
+    },
+    {
+      "epoch": 0.22347514095335724,
+      "grad_norm": 0.6737356495103414,
+      "learning_rate": 1.9653504474951965e-05,
+      "loss": 0.4617,
+      "step": 436
+    },
+    {
+      "epoch": 0.2239876986160943,
+      "grad_norm": 0.6364533533894546,
+      "learning_rate": 1.9651334603966298e-05,
+      "loss": 0.4704,
+      "step": 437
+    },
+    {
+      "epoch": 0.22450025627883136,
+      "grad_norm": 0.5692354648619666,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.4558,
+      "step": 438
+    },
+    {
+      "epoch": 0.22501281394156844,
+      "grad_norm": 0.6486712426209987,
+      "learning_rate": 1.9646974905997295e-05,
+      "loss": 0.4865,
+      "step": 439
+    },
+    {
+      "epoch": 0.2255253716043055,
+      "grad_norm": 0.6202997282852948,
+      "learning_rate": 1.964478508201903e-05,
+      "loss": 0.4972,
+      "step": 440
+    },
+    {
+      "epoch": 0.22603792926704255,
+      "grad_norm": 0.6093182724947406,
+      "learning_rate": 1.964258861005594e-05,
+      "loss": 0.4152,
+      "step": 441
+    },
+    {
+      "epoch": 0.2265504869297796,
+      "grad_norm": 0.6048676189486943,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 0.4405,
+      "step": 442
+    },
+    {
+      "epoch": 0.22706304459251667,
+      "grad_norm": 0.6164807699320765,
+      "learning_rate": 1.9638175728235817e-05,
+      "loss": 0.4608,
+      "step": 443
+    },
+    {
+      "epoch": 0.22757560225525372,
+      "grad_norm": 0.59766898337836,
+      "learning_rate": 1.9635959321420514e-05,
+      "loss": 0.4577,
+      "step": 444
+    },
+    {
+      "epoch": 0.22808815991799078,
+      "grad_norm": 0.6626611938365956,
+      "learning_rate": 1.9633736272703826e-05,
+      "loss": 0.4648,
+      "step": 445
+    },
+    {
+      "epoch": 0.22860071758072784,
+      "grad_norm": 0.6358146972842108,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.4846,
+      "step": 446
+    },
+    {
+      "epoch": 0.2291132752434649,
+      "grad_norm": 0.6721239580896927,
+      "learning_rate": 1.9629270255700127e-05,
+      "loss": 0.4891,
+      "step": 447
+    },
+    {
+      "epoch": 0.22962583290620195,
+      "grad_norm": 0.5849079705907492,
+      "learning_rate": 1.962702729049146e-05,
+      "loss": 0.4739,
+      "step": 448
+    },
+    {
+      "epoch": 0.230138390568939,
+      "grad_norm": 0.6092593254754007,
+      "learning_rate": 1.9624777689538107e-05,
+      "loss": 0.4524,
+      "step": 449
+    },
+    {
+      "epoch": 0.23065094823167606,
+      "grad_norm": 0.5924642120215301,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 0.3968,
+      "step": 450
+    },
+    {
+      "epoch": 0.23116350589441312,
+      "grad_norm": 0.5699330884950117,
+      "learning_rate": 1.9620258586604364e-05,
+      "loss": 0.4094,
+      "step": 451
+    },
+    {
+      "epoch": 0.23167606355715017,
+      "grad_norm": 0.5905000266168502,
+      "learning_rate": 1.9617989087738913e-05,
+      "loss": 0.4609,
+      "step": 452
+    },
+    {
+      "epoch": 0.23218862121988723,
+      "grad_norm": 0.5673706974693065,
+      "learning_rate": 1.9615712959358656e-05,
+      "loss": 0.4207,
+      "step": 453
+    },
+    {
+      "epoch": 0.2327011788826243,
+      "grad_norm": 0.6766218195684616,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 0.4813,
+      "step": 454
+    },
+    {
+      "epoch": 0.23321373654536134,
+      "grad_norm": 0.6051195350091134,
+      "learning_rate": 1.9611140820333868e-05,
+      "loss": 0.4437,
+      "step": 455
+    },
+    {
+      "epoch": 0.2337262942080984,
+      "grad_norm": 0.5741096479477734,
+      "learning_rate": 1.960884481284084e-05,
+      "loss": 0.4741,
+      "step": 456
+    },
+    {
+      "epoch": 0.23423885187083546,
+      "grad_norm": 0.6125195512454823,
+      "learning_rate": 1.960654218213599e-05,
+      "loss": 0.4552,
+      "step": 457
+    },
+    {
+      "epoch": 0.2347514095335725,
+      "grad_norm": 0.6107999250650776,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.4535,
+      "step": 458
+    },
+    {
+      "epoch": 0.2352639671963096,
+      "grad_norm": 0.6370720671415663,
+      "learning_rate": 1.960191705744407e-05,
+      "loss": 0.4646,
+      "step": 459
+    },
+    {
+      "epoch": 0.23577652485904665,
+      "grad_norm": 0.604945046143909,
+      "learning_rate": 1.9599594566645016e-05,
+      "loss": 0.4707,
+      "step": 460
+    },
+    {
+      "epoch": 0.2362890825217837,
+      "grad_norm": 0.637961674521814,
+      "learning_rate": 1.9597265459010177e-05,
+      "loss": 0.4395,
+      "step": 461
+    },
+    {
+      "epoch": 0.23680164018452077,
+      "grad_norm": 0.5885230384704008,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4616,
+      "step": 462
+    },
+    {
+      "epoch": 0.23731419784725782,
+      "grad_norm": 0.6205131611510152,
+      "learning_rate": 1.959258739965938e-05,
+      "loss": 0.4858,
+      "step": 463
+    },
+    {
+      "epoch": 0.23782675550999488,
+      "grad_norm": 0.6224211011046287,
+      "learning_rate": 1.9590238451167927e-05,
+      "loss": 0.4764,
+      "step": 464
+    },
+    {
+      "epoch": 0.23833931317273194,
+      "grad_norm": 0.582407693338383,
+      "learning_rate": 1.95878828922897e-05,
+      "loss": 0.4449,
+      "step": 465
+    },
+    {
+      "epoch": 0.238851870835469,
+      "grad_norm": 0.6742276236968419,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 0.4202,
+      "step": 466
+    },
+    {
+      "epoch": 0.23936442849820605,
+      "grad_norm": 0.6700245015307164,
+      "learning_rate": 1.9583151949872083e-05,
+      "loss": 0.4873,
+      "step": 467
+    },
+    {
+      "epoch": 0.2398769861609431,
+      "grad_norm": 0.5485873113979571,
+      "learning_rate": 1.9580776569593646e-05,
+      "loss": 0.4133,
+      "step": 468
+    },
+    {
+      "epoch": 0.24038954382368016,
+      "grad_norm": 0.6201629632763813,
+      "learning_rate": 1.9578394585450354e-05,
+      "loss": 0.4996,
+      "step": 469
+    },
+    {
+      "epoch": 0.24090210148641722,
+      "grad_norm": 0.6107893057531669,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 0.4469,
+      "step": 470
+    },
+    {
+      "epoch": 0.24141465914915428,
+      "grad_norm": 0.5726429871122204,
+      "learning_rate": 1.9573610812141183e-05,
+      "loss": 0.4203,
+      "step": 471
+    },
+    {
+      "epoch": 0.24192721681189133,
+      "grad_norm": 0.6129382690277111,
+      "learning_rate": 1.9571209026272683e-05,
+      "loss": 0.4338,
+      "step": 472
+    },
+    {
+      "epoch": 0.2424397744746284,
+      "grad_norm": 0.6029244858612489,
+      "learning_rate": 1.9568800643134073e-05,
+      "loss": 0.4532,
+      "step": 473
+    },
+    {
+      "epoch": 0.24295233213736545,
+      "grad_norm": 0.6218124045562262,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.4566,
+      "step": 474
+    },
+    {
+      "epoch": 0.2434648898001025,
+      "grad_norm": 0.6162459615056012,
+      "learning_rate": 1.9563964091691275e-05,
+      "loss": 0.5049,
+      "step": 475
+    },
+    {
+      "epoch": 0.24397744746283956,
+      "grad_norm": 0.6000826803513242,
+      "learning_rate": 1.9561535926720846e-05,
+      "loss": 0.4458,
+      "step": 476
+    },
+    {
+      "epoch": 0.24449000512557661,
+      "grad_norm": 0.5501876773826728,
+      "learning_rate": 1.95591011711478e-05,
+      "loss": 0.4575,
+      "step": 477
+    },
+    {
+      "epoch": 0.2450025627883137,
+      "grad_norm": 0.6453642773226369,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 0.4458,
+      "step": 478
+    },
+    {
+      "epoch": 0.24551512045105076,
+      "grad_norm": 0.5264095076022388,
+      "learning_rate": 1.9554211894911363e-05,
+      "loss": 0.4118,
+      "step": 479
+    },
+    {
+      "epoch": 0.2460276781137878,
+      "grad_norm": 0.5887926300665813,
+      "learning_rate": 1.9551757377618056e-05,
+      "loss": 0.4596,
+      "step": 480
+    },
+    {
+      "epoch": 0.24654023577652487,
+      "grad_norm": 0.5878975740898159,
+      "learning_rate": 1.9549296276462326e-05,
+      "loss": 0.4267,
+      "step": 481
+    },
+    {
+      "epoch": 0.24705279343926193,
+      "grad_norm": 0.643861220423485,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.4248,
+      "step": 482
+    },
+    {
+      "epoch": 0.24756535110199898,
+      "grad_norm": 0.5826232259336361,
+      "learning_rate": 1.95443543293537e-05,
+      "loss": 0.4688,
+      "step": 483
+    },
+    {
+      "epoch": 0.24807790876473604,
+      "grad_norm": 0.6134986746537208,
+      "learning_rate": 1.954187348680721e-05,
+      "loss": 0.4738,
+      "step": 484
+    },
+    {
+      "epoch": 0.2485904664274731,
+      "grad_norm": 0.6063272631875714,
+      "learning_rate": 1.953938606721108e-05,
+      "loss": 0.4535,
+      "step": 485
+    },
+    {
+      "epoch": 0.24910302409021015,
+      "grad_norm": 0.5962880742796507,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.4585,
+      "step": 486
+    },
+    {
+      "epoch": 0.2496155817529472,
+      "grad_norm": 0.6228082960810553,
+      "learning_rate": 1.9534391503732617e-05,
+      "loss": 0.4615,
+      "step": 487
+    },
+    {
+      "epoch": 0.2501281394156843,
+      "grad_norm": 0.6529481476983112,
+      "learning_rate": 1.9531884363292944e-05,
+      "loss": 0.4285,
+      "step": 488
+    },
+    {
+      "epoch": 0.25064069707842135,
+      "grad_norm": 0.5915296949219395,
+      "learning_rate": 1.952937065268897e-05,
+      "loss": 0.4626,
+      "step": 489
+    },
+    {
+      "epoch": 0.2511532547411584,
+      "grad_norm": 0.5895861237283015,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.4362,
+      "step": 490
+    },
+    {
+      "epoch": 0.25166581240389546,
+      "grad_norm": 0.6249915098353631,
+      "learning_rate": 1.9524323527923284e-05,
+      "loss": 0.4775,
+      "step": 491
+    },
+    {
+      "epoch": 0.2521783700666325,
+      "grad_norm": 0.6119729376497774,
+      "learning_rate": 1.9521790117240472e-05,
+      "loss": 0.4458,
+      "step": 492
+    },
+    {
+      "epoch": 0.2526909277293696,
+      "grad_norm": 0.624555256034914,
+      "learning_rate": 1.9519250143351146e-05,
+      "loss": 0.4252,
+      "step": 493
+    },
+    {
+      "epoch": 0.25320348539210663,
+      "grad_norm": 0.6659347059999439,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 0.4285,
+      "step": 494
+    },
+    {
+      "epoch": 0.2537160430548437,
+      "grad_norm": 0.5915654353496794,
+      "learning_rate": 1.951415051296054e-05,
+      "loss": 0.4562,
+      "step": 495
+    },
+    {
+      "epoch": 0.25422860071758074,
+      "grad_norm": 0.587920365732569,
+      "learning_rate": 1.9511590859974344e-05,
+      "loss": 0.4939,
+      "step": 496
+    },
+    {
+      "epoch": 0.2547411583803178,
+      "grad_norm": 0.6435427251762267,
+      "learning_rate": 1.9509024650811813e-05,
+      "loss": 0.442,
+      "step": 497
+    },
+    {
+      "epoch": 0.25525371604305486,
+      "grad_norm": 0.6507212922963379,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.5095,
+      "step": 498
+    },
+    {
+      "epoch": 0.2557662737057919,
+      "grad_norm": 0.6502918197172725,
+      "learning_rate": 1.9503872571037637e-05,
+      "loss": 0.4695,
+      "step": 499
+    },
+    {
+      "epoch": 0.25627883136852897,
+      "grad_norm": 0.6256347910525726,
+      "learning_rate": 1.9501286703977232e-05,
+      "loss": 0.4255,
+      "step": 500
+    },
+    {
+      "epoch": 0.256791389031266,
+      "grad_norm": 0.6127276791972335,
+      "learning_rate": 1.9498694287842972e-05,
+      "loss": 0.4857,
+      "step": 501
+    },
+    {
+      "epoch": 0.2573039466940031,
+      "grad_norm": 0.5968094447628093,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.4686,
+      "step": 502
+    },
+    {
+      "epoch": 0.25781650435674014,
+      "grad_norm": 0.5837601607888753,
+      "learning_rate": 1.949348981550502e-05,
+      "loss": 0.4683,
+      "step": 503
+    },
+    {
+      "epoch": 0.2583290620194772,
+      "grad_norm": 0.6230462602794585,
+      "learning_rate": 1.9490877762888685e-05,
+      "loss": 0.4203,
+      "step": 504
+    },
+    {
+      "epoch": 0.25884161968221425,
+      "grad_norm": 0.615621216267251,
+      "learning_rate": 1.9488259168373198e-05,
+      "loss": 0.4692,
+      "step": 505
+    },
+    {
+      "epoch": 0.2593541773449513,
+      "grad_norm": 0.6283108244427518,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.4592,
+      "step": 506
+    },
+    {
+      "epoch": 0.25986673500768837,
+      "grad_norm": 0.6158722837043947,
+      "learning_rate": 1.9483002360869082e-05,
+      "loss": 0.4324,
+      "step": 507
+    },
+    {
+      "epoch": 0.2603792926704254,
+      "grad_norm": 0.6969841615709742,
+      "learning_rate": 1.9480364151503876e-05,
+      "loss": 0.4728,
+      "step": 508
+    },
+    {
+      "epoch": 0.2608918503331625,
+      "grad_norm": 0.5890364309083278,
+      "learning_rate": 1.9477719407486372e-05,
+      "loss": 0.4307,
+      "step": 509
+    },
+    {
+      "epoch": 0.26140440799589953,
+      "grad_norm": 0.6071151627134794,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.4342,
+      "step": 510
+    },
+    {
+      "epoch": 0.2619169656586366,
+      "grad_norm": 0.6197548578293316,
+      "learning_rate": 1.947241032279087e-05,
+      "loss": 0.3978,
+      "step": 511
+    },
+    {
+      "epoch": 0.26242952332137365,
+      "grad_norm": 0.6368730416037391,
+      "learning_rate": 1.9469745985772333e-05,
+      "loss": 0.4556,
+      "step": 512
+    },
+    {
+      "epoch": 0.2629420809841107,
+      "grad_norm": 0.6248785771884786,
+      "learning_rate": 1.9467075121420415e-05,
+      "loss": 0.4169,
+      "step": 513
+    },
+    {
+      "epoch": 0.26345463864684776,
+      "grad_norm": 0.6150958126123865,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.4549,
+      "step": 514
+    },
+    {
+      "epoch": 0.2639671963095848,
+      "grad_norm": 0.6368667681310836,
+      "learning_rate": 1.9461713818084857e-05,
+      "loss": 0.4752,
+      "step": 515
+    },
+    {
+      "epoch": 0.2644797539723219,
+      "grad_norm": 0.6034828205714066,
+      "learning_rate": 1.9459023382796667e-05,
+      "loss": 0.4447,
+      "step": 516
+    },
+    {
+      "epoch": 0.26499231163505893,
+      "grad_norm": 0.5638833308623431,
+      "learning_rate": 1.9456326427566002e-05,
+      "loss": 0.4264,
+      "step": 517
+    },
+    {
+      "epoch": 0.265504869297796,
+      "grad_norm": 0.6258474187529659,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.4278,
+      "step": 518
+    },
+    {
+      "epoch": 0.26601742696053304,
+      "grad_norm": 0.617285642403996,
+      "learning_rate": 1.9450912964717604e-05,
+      "loss": 0.4425,
+      "step": 519
+    },
+    {
+      "epoch": 0.2665299846232701,
+      "grad_norm": 0.6681377361854317,
+      "learning_rate": 1.9448196460831274e-05,
+      "loss": 0.4746,
+      "step": 520
+    },
+    {
+      "epoch": 0.26704254228600716,
+      "grad_norm": 0.6179386710802767,
+      "learning_rate": 1.944547344446528e-05,
+      "loss": 0.4044,
+      "step": 521
+    },
+    {
+      "epoch": 0.2675550999487442,
+      "grad_norm": 0.598004741655023,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.4426,
+      "step": 522
+    },
+    {
+      "epoch": 0.26806765761148127,
+      "grad_norm": 0.6344917084877345,
+      "learning_rate": 1.9440007881806502e-05,
+      "loss": 0.4386,
+      "step": 523
+    },
+    {
+      "epoch": 0.2685802152742183,
+      "grad_norm": 0.5950613210628302,
+      "learning_rate": 1.9437265339281034e-05,
+      "loss": 0.4304,
+      "step": 524
+    },
+    {
+      "epoch": 0.2690927729369554,
+      "grad_norm": 0.6302842422315329,
+      "learning_rate": 1.943451629181054e-05,
+      "loss": 0.4761,
+      "step": 525
+    },
+    {
+      "epoch": 0.26960533059969244,
+      "grad_norm": 0.6420481269048874,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.4439,
+      "step": 526
+    },
+    {
+      "epoch": 0.27011788826242955,
+      "grad_norm": 0.631310547353491,
+      "learning_rate": 1.9428998689618428e-05,
+      "loss": 0.4127,
+      "step": 527
+    },
+    {
+      "epoch": 0.2706304459251666,
+      "grad_norm": 0.5492598863534703,
+      "learning_rate": 1.9426230138699998e-05,
+      "loss": 0.417,
+      "step": 528
+    },
+    {
+      "epoch": 0.27114300358790366,
+      "grad_norm": 0.5612843643374066,
+      "learning_rate": 1.9423455090442917e-05,
+      "loss": 0.4129,
+      "step": 529
+    },
+    {
+      "epoch": 0.2716555612506407,
+      "grad_norm": 0.5140746603197311,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 0.4186,
+      "step": 530
+    },
+    {
+      "epoch": 0.2721681189133778,
+      "grad_norm": 0.6084287597909264,
+      "learning_rate": 1.9417885509568433e-05,
+      "loss": 0.4347,
+      "step": 531
+    },
+    {
+      "epoch": 0.27268067657611483,
+      "grad_norm": 0.632991225722545,
+      "learning_rate": 1.941509098079005e-05,
+      "loss": 0.455,
+      "step": 532
+    },
+    {
+      "epoch": 0.2731932342388519,
+      "grad_norm": 0.577648962203757,
+      "learning_rate": 1.9412289962351042e-05,
+      "loss": 0.4518,
+      "step": 533
+    },
+    {
+      "epoch": 0.27370579190158895,
+      "grad_norm": 0.6543761263375933,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.4322,
+      "step": 534
+    },
+    {
+      "epoch": 0.274218349564326,
+      "grad_norm": 0.6369697245546415,
+      "learning_rate": 1.94066684642184e-05,
+      "loss": 0.4949,
+      "step": 535
+    },
+    {
+      "epoch": 0.27473090722706306,
+      "grad_norm": 0.6784857258677417,
+      "learning_rate": 1.940384798839957e-05,
+      "loss": 0.4305,
+      "step": 536
+    },
+    {
+      "epoch": 0.2752434648898001,
+      "grad_norm": 0.5880142418496587,
+      "learning_rate": 1.9401021030669715e-05,
+      "loss": 0.4335,
+      "step": 537
+    },
+    {
+      "epoch": 0.27575602255253717,
+      "grad_norm": 0.5968609196445768,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 0.481,
+      "step": 538
+    },
+    {
+      "epoch": 0.27626858021527423,
+      "grad_norm": 0.5669131808617776,
+      "learning_rate": 1.9395347677275695e-05,
+      "loss": 0.4256,
+      "step": 539
+    },
+    {
+      "epoch": 0.2767811378780113,
+      "grad_norm": 0.6423439744984043,
+      "learning_rate": 1.9392501285522078e-05,
+      "loss": 0.45,
+      "step": 540
+    },
+    {
+      "epoch": 0.27729369554074834,
+      "grad_norm": 0.541765694872795,
+      "learning_rate": 1.9389648419678523e-05,
+      "loss": 0.4258,
+      "step": 541
+    },
+    {
+      "epoch": 0.2778062532034854,
+      "grad_norm": 0.5710935780029611,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.4484,
+      "step": 542
+    },
+    {
+      "epoch": 0.27831881086622245,
+      "grad_norm": 0.612895438580781,
+      "learning_rate": 1.9383923273591793e-05,
+      "loss": 0.43,
+      "step": 543
+    },
+    {
+      "epoch": 0.2788313685289595,
+      "grad_norm": 0.5677104455977324,
+      "learning_rate": 1.9381050997294864e-05,
+      "loss": 0.4318,
+      "step": 544
+    },
+    {
+      "epoch": 0.27934392619169657,
+      "grad_norm": 0.6063879197983586,
+      "learning_rate": 1.9378172254800487e-05,
+      "loss": 0.4525,
+      "step": 545
+    },
+    {
+      "epoch": 0.2798564838544336,
+      "grad_norm": 0.6615740693516219,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 0.4634,
+      "step": 546
+    },
+    {
+      "epoch": 0.2803690415171707,
+      "grad_norm": 0.6452884500438248,
+      "learning_rate": 1.9372395379160913e-05,
+      "loss": 0.4229,
+      "step": 547
+    },
+    {
+      "epoch": 0.28088159917990774,
+      "grad_norm": 0.6182213742360952,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 0.4654,
+      "step": 548
+    },
+    {
+      "epoch": 0.2813941568426448,
+      "grad_norm": 0.6378501956960184,
+      "learning_rate": 1.9366592662600678e-05,
+      "loss": 0.4757,
+      "step": 549
+    },
+    {
+      "epoch": 0.28190671450538185,
+      "grad_norm": 0.6396660491069084,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 0.453,
+      "step": 550
+    },
+    {
+      "epoch": 0.2824192721681189,
+      "grad_norm": 0.6137200363876782,
+      "learning_rate": 1.936076412111862e-05,
+      "loss": 0.4683,
+      "step": 551
+    },
+    {
+      "epoch": 0.28293182983085596,
+      "grad_norm": 0.6057969691318004,
+      "learning_rate": 1.9357840171051025e-05,
+      "loss": 0.4479,
+      "step": 552
+    },
+    {
+      "epoch": 0.283444387493593,
+      "grad_norm": 0.5908008891493759,
+      "learning_rate": 1.9354909770784808e-05,
+      "loss": 0.3982,
+      "step": 553
+    },
+    {
+      "epoch": 0.2839569451563301,
+      "grad_norm": 0.584125698056456,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 0.4509,
+      "step": 554
+    },
+    {
+      "epoch": 0.28446950281906713,
+      "grad_norm": 0.627288188127924,
+      "learning_rate": 1.9349029627740434e-05,
+      "loss": 0.4088,
+      "step": 555
+    },
+    {
+      "epoch": 0.2849820604818042,
+      "grad_norm": 0.6025723941857102,
+      "learning_rate": 1.9346079889015366e-05,
+      "loss": 0.42,
+      "step": 556
+    },
+    {
+      "epoch": 0.28549461814454125,
+      "grad_norm": 0.5975382452270548,
+      "learning_rate": 1.9343123708197834e-05,
+      "loss": 0.4465,
+      "step": 557
+    },
+    {
+      "epoch": 0.2860071758072783,
+      "grad_norm": 0.6380520083061455,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.4738,
+      "step": 558
+    },
+    {
+      "epoch": 0.28651973347001536,
+      "grad_norm": 0.6048391717599558,
+      "learning_rate": 1.9337192028440395e-05,
+      "loss": 0.4457,
+      "step": 559
+    },
+    {
+      "epoch": 0.2870322911327524,
+      "grad_norm": 0.6201425378022258,
+      "learning_rate": 1.9334216533589095e-05,
+      "loss": 0.3925,
+      "step": 560
+    },
+    {
+      "epoch": 0.28754484879548947,
+      "grad_norm": 0.558057808803036,
+      "learning_rate": 1.933123460482254e-05,
+      "loss": 0.4691,
+      "step": 561
+    },
+    {
+      "epoch": 0.28805740645822653,
+      "grad_norm": 0.5925177437934743,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.4647,
+      "step": 562
+    },
+    {
+      "epoch": 0.2885699641209636,
+      "grad_norm": 0.6530918685374633,
+      "learning_rate": 1.9325251453769657e-05,
+      "loss": 0.4132,
+      "step": 563
+    },
+    {
+      "epoch": 0.28908252178370064,
+      "grad_norm": 0.6572098857193267,
+      "learning_rate": 1.9322250235607418e-05,
+      "loss": 0.4308,
+      "step": 564
+    },
+    {
+      "epoch": 0.2895950794464377,
+      "grad_norm": 0.645794470212144,
+      "learning_rate": 1.9319242591778086e-05,
+      "loss": 0.4355,
+      "step": 565
+    },
+    {
+      "epoch": 0.2901076371091748,
+      "grad_norm": 0.5633105204185109,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.4215,
+      "step": 566
+    },
+    {
+      "epoch": 0.29062019477191187,
+      "grad_norm": 0.5578176196297239,
+      "learning_rate": 1.9313208035415042e-05,
+      "loss": 0.3915,
+      "step": 567
+    },
+    {
+      "epoch": 0.2911327524346489,
+      "grad_norm": 0.6259086766923461,
+      "learning_rate": 1.9310181127040842e-05,
+      "loss": 0.4954,
+      "step": 568
+    },
+    {
+      "epoch": 0.291645310097386,
+      "grad_norm": 0.5836414024637069,
+      "learning_rate": 1.9307147801318585e-05,
+      "loss": 0.4467,
+      "step": 569
+    },
+    {
+      "epoch": 0.29215786776012304,
+      "grad_norm": 0.611226088578877,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 0.4297,
+      "step": 570
+    },
+    {
+      "epoch": 0.2926704254228601,
+      "grad_norm": 0.5942928619678638,
+      "learning_rate": 1.9301061906197572e-05,
+      "loss": 0.4414,
+      "step": 571
+    },
+    {
+      "epoch": 0.29318298308559715,
+      "grad_norm": 0.5820823736484783,
+      "learning_rate": 1.9298009340993725e-05,
+      "loss": 0.3906,
+      "step": 572
+    },
+    {
+      "epoch": 0.2936955407483342,
+      "grad_norm": 0.6132051407034689,
+      "learning_rate": 1.9294950366831617e-05,
+      "loss": 0.4347,
+      "step": 573
+    },
+    {
+      "epoch": 0.29420809841107126,
+      "grad_norm": 0.6283560881886736,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 0.4525,
+      "step": 574
+    },
+    {
+      "epoch": 0.2947206560738083,
+      "grad_norm": 0.6447821516036115,
+      "learning_rate": 1.9288813200071035e-05,
+      "loss": 0.4618,
+      "step": 575
+    },
+    {
+      "epoch": 0.2952332137365454,
+      "grad_norm": 0.6264204098211594,
+      "learning_rate": 1.9285735011702803e-05,
+      "loss": 0.434,
+      "step": 576
+    },
+    {
+      "epoch": 0.29574577139928243,
+      "grad_norm": 0.7748735290581021,
+      "learning_rate": 1.928265042283679e-05,
+      "loss": 0.4204,
+      "step": 577
+    },
+    {
+      "epoch": 0.2962583290620195,
+      "grad_norm": 0.5410170902781122,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 0.4459,
+      "step": 578
+    },
+    {
+      "epoch": 0.29677088672475654,
+      "grad_norm": 0.6077314637720823,
+      "learning_rate": 1.9276462052120475e-05,
+      "loss": 0.4535,
+      "step": 579
+    },
+    {
+      "epoch": 0.2972834443874936,
+      "grad_norm": 0.5745924041130557,
+      "learning_rate": 1.9273358274535703e-05,
+      "loss": 0.4291,
+      "step": 580
+    },
+    {
+      "epoch": 0.29779600205023066,
+      "grad_norm": 0.5868069454397189,
+      "learning_rate": 1.927024810498423e-05,
+      "loss": 0.4575,
+      "step": 581
+    },
+    {
+      "epoch": 0.2983085597129677,
+      "grad_norm": 0.5601690317696364,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.4691,
+      "step": 582
+    },
+    {
+      "epoch": 0.29882111737570477,
+      "grad_norm": 0.5665603886797981,
+      "learning_rate": 1.9264008598560727e-05,
+      "loss": 0.4667,
+      "step": 583
+    },
+    {
+      "epoch": 0.2993336750384418,
+      "grad_norm": 0.5422278508187873,
+      "learning_rate": 1.9260879265989477e-05,
+      "loss": 0.4423,
+      "step": 584
+    },
+    {
+      "epoch": 0.2998462327011789,
+      "grad_norm": 0.5705977719160407,
+      "learning_rate": 1.9257743550053097e-05,
+      "loss": 0.3852,
+      "step": 585
+    },
+    {
+      "epoch": 0.30035879036391594,
+      "grad_norm": 0.6435895750969524,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.438,
+      "step": 586
+    },
+    {
+      "epoch": 0.300871348026653,
+      "grad_norm": 0.6474791178058024,
+      "learning_rate": 1.92514529767349e-05,
+      "loss": 0.3939,
+      "step": 587
+    },
+    {
+      "epoch": 0.30138390568939005,
+      "grad_norm": 0.6060603233207734,
+      "learning_rate": 1.924829812368907e-05,
+      "loss": 0.4124,
+      "step": 588
+    },
+    {
+      "epoch": 0.3018964633521271,
+      "grad_norm": 0.5863986654878778,
+      "learning_rate": 1.924513689595007e-05,
+      "loss": 0.4544,
+      "step": 589
+    },
+    {
+      "epoch": 0.30240902101486417,
+      "grad_norm": 0.6839014416620915,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.4219,
+      "step": 590
+    },
+    {
+      "epoch": 0.3029215786776012,
+      "grad_norm": 0.6312012370374425,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.4733,
+      "step": 591
+    },
+    {
+      "epoch": 0.3034341363403383,
+      "grad_norm": 0.5868053634268785,
+      "learning_rate": 1.9235614986385806e-05,
+      "loss": 0.445,
+      "step": 592
+    },
+    {
+      "epoch": 0.30394669400307534,
+      "grad_norm": 0.5932118924604575,
+      "learning_rate": 1.923242828170784e-05,
+      "loss": 0.448,
+      "step": 593
+    },
+    {
+      "epoch": 0.3044592516658124,
+      "grad_norm": 0.6284282603128786,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.4676,
+      "step": 594
+    },
+    {
+      "epoch": 0.30497180932854945,
+      "grad_norm": 0.5690425692889751,
+      "learning_rate": 1.9226035783289748e-05,
+      "loss": 0.4449,
+      "step": 595
+    },
+    {
+      "epoch": 0.3054843669912865,
+      "grad_norm": 0.5268041530286178,
+      "learning_rate": 1.922282999395586e-05,
+      "loss": 0.4457,
+      "step": 596
+    },
+    {
+      "epoch": 0.30599692465402356,
+      "grad_norm": 0.6073222383089534,
+      "learning_rate": 1.9219617847483553e-05,
+      "loss": 0.4273,
+      "step": 597
+    },
+    {
+      "epoch": 0.3065094823167606,
+      "grad_norm": 0.5392832971846893,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.4307,
+      "step": 598
+    },
+    {
+      "epoch": 0.3070220399794977,
+      "grad_norm": 0.5474068712055952,
+      "learning_rate": 1.9213174491984342e-05,
+      "loss": 0.4005,
+      "step": 599
+    },
+    {
+      "epoch": 0.30753459764223473,
+      "grad_norm": 0.6463555538796705,
+      "learning_rate": 1.9209943287398738e-05,
+      "loss": 0.4195,
+      "step": 600
+    },
+    {
+      "epoch": 0.3080471553049718,
+      "grad_norm": 0.6116730833944737,
+      "learning_rate": 1.9206705734557293e-05,
+      "loss": 0.371,
+      "step": 601
+    },
+    {
+      "epoch": 0.30855971296770884,
+      "grad_norm": 0.5509072111530304,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.3702,
+      "step": 602
+    },
+    {
+      "epoch": 0.3090722706304459,
+      "grad_norm": 0.6107722409459658,
+      "learning_rate": 1.9200211593037612e-05,
+      "loss": 0.4782,
+      "step": 603
+    },
+    {
+      "epoch": 0.30958482829318296,
+      "grad_norm": 0.6315211247177352,
+      "learning_rate": 1.9196955008835678e-05,
+      "loss": 0.4135,
+      "step": 604
+    },
+    {
+      "epoch": 0.31009738595592007,
+      "grad_norm": 0.6038653158650275,
+      "learning_rate": 1.9193692085330503e-05,
+      "loss": 0.4288,
+      "step": 605
+    },
+    {
+      "epoch": 0.3106099436186571,
+      "grad_norm": 0.6200185721624881,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 0.4562,
+      "step": 606
+    },
+    {
+      "epoch": 0.3111225012813942,
+      "grad_norm": 0.5638958714712866,
+      "learning_rate": 1.9187147229411094e-05,
+      "loss": 0.4042,
+      "step": 607
+    },
+    {
+      "epoch": 0.31163505894413124,
+      "grad_norm": 0.6112725266369267,
+      "learning_rate": 1.918386530150812e-05,
+      "loss": 0.4484,
+      "step": 608
+    },
+    {
+      "epoch": 0.3121476166068683,
+      "grad_norm": 0.5467835050873157,
+      "learning_rate": 1.9180577043324415e-05,
+      "loss": 0.4318,
+      "step": 609
+    },
+    {
+      "epoch": 0.31266017426960535,
+      "grad_norm": 0.5403377003778622,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.3892,
+      "step": 610
+    },
+    {
+      "epoch": 0.3131727319323424,
+      "grad_norm": 0.5619343707963038,
+      "learning_rate": 1.9173981545185327e-05,
+      "loss": 0.4201,
+      "step": 611
+    },
+    {
+      "epoch": 0.31368528959507946,
+      "grad_norm": 0.5458763699345418,
+      "learning_rate": 1.917067430977611e-05,
+      "loss": 0.4155,
+      "step": 612
+    },
+    {
+      "epoch": 0.3141978472578165,
+      "grad_norm": 0.5814352621413827,
+      "learning_rate": 1.916736075317848e-05,
+      "loss": 0.4072,
+      "step": 613
+    },
+    {
+      "epoch": 0.3147104049205536,
+      "grad_norm": 0.591992082066158,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 0.4279,
+      "step": 614
+    },
+    {
+      "epoch": 0.31522296258329063,
+      "grad_norm": 0.6074856377216148,
+      "learning_rate": 1.9160714685558272e-05,
+      "loss": 0.4166,
+      "step": 615
+    },
+    {
+      "epoch": 0.3157355202460277,
+      "grad_norm": 0.5561512629608296,
+      "learning_rate": 1.9157382179116705e-05,
+      "loss": 0.4657,
+      "step": 616
+    },
+    {
+      "epoch": 0.31624807790876475,
+      "grad_norm": 0.6351905450255506,
+      "learning_rate": 1.915404336064877e-05,
+      "loss": 0.4542,
+      "step": 617
+    },
+    {
+      "epoch": 0.3167606355715018,
+      "grad_norm": 0.6110245905021758,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.4372,
+      "step": 618
+    },
+    {
+      "epoch": 0.31727319323423886,
+      "grad_norm": 0.5730028299876052,
+      "learning_rate": 1.9147346796843695e-05,
+      "loss": 0.4271,
+      "step": 619
+    },
+    {
+      "epoch": 0.3177857508969759,
+      "grad_norm": 0.5513761193489177,
+      "learning_rate": 1.9143989056122385e-05,
+      "loss": 0.4012,
+      "step": 620
+    },
+    {
+      "epoch": 0.318298308559713,
+      "grad_norm": 0.6028441900375847,
+      "learning_rate": 1.9140625012606353e-05,
+      "loss": 0.4094,
+      "step": 621
+    },
+    {
+      "epoch": 0.31881086622245003,
+      "grad_norm": 0.5899309030988729,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 0.4347,
+      "step": 622
+    },
+    {
+      "epoch": 0.3193234238851871,
+      "grad_norm": 0.5958027528079011,
+      "learning_rate": 1.913387802646958e-05,
+      "loss": 0.4332,
+      "step": 623
+    },
+    {
+      "epoch": 0.31983598154792414,
+      "grad_norm": 0.6010255400684187,
+      "learning_rate": 1.9130495088499417e-05,
+      "loss": 0.4514,
+      "step": 624
+    },
+    {
+      "epoch": 0.3203485392106612,
+      "grad_norm": 0.6198298526544422,
+      "learning_rate": 1.9127105857035697e-05,
+      "loss": 0.4158,
+      "step": 625
+    },
+    {
+      "epoch": 0.32086109687339825,
+      "grad_norm": 0.6264228444155231,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.4478,
+      "step": 626
+    },
+    {
+      "epoch": 0.3213736545361353,
+      "grad_norm": 0.5793420848614765,
+      "learning_rate": 1.9120308522976464e-05,
+      "loss": 0.3946,
+      "step": 627
+    },
+    {
+      "epoch": 0.32188621219887237,
+      "grad_norm": 0.621533057366719,
+      "learning_rate": 1.911690042506624e-05,
+      "loss": 0.4447,
+      "step": 628
+    },
+    {
+      "epoch": 0.3223987698616094,
+      "grad_norm": 0.5999262869873329,
+      "learning_rate": 1.9113486043033025e-05,
+      "loss": 0.4416,
+      "step": 629
+    },
+    {
+      "epoch": 0.3229113275243465,
+      "grad_norm": 0.5633558354528765,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.4523,
+      "step": 630
+    },
+    {
+      "epoch": 0.32342388518708354,
+      "grad_norm": 0.6232615762736277,
+      "learning_rate": 1.9106638436015842e-05,
+      "loss": 0.438,
+      "step": 631
+    },
+    {
+      "epoch": 0.3239364428498206,
+      "grad_norm": 0.564846105627677,
+      "learning_rate": 1.9103205215751815e-05,
+      "loss": 0.408,
+      "step": 632
+    },
+    {
+      "epoch": 0.32444900051255765,
+      "grad_norm": 0.5366213282534108,
+      "learning_rate": 1.9099765720804666e-05,
+      "loss": 0.4114,
+      "step": 633
+    },
+    {
+      "epoch": 0.3249615581752947,
+      "grad_norm": 0.6158895461047174,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4279,
+      "step": 634
+    },
+    {
+      "epoch": 0.32547411583803176,
+      "grad_norm": 0.6618097403768928,
+      "learning_rate": 1.909286791634848e-05,
+      "loss": 0.4084,
+      "step": 635
+    },
+    {
+      "epoch": 0.3259866735007688,
+      "grad_norm": 0.548396551906715,
+      "learning_rate": 1.908940961159397e-05,
+      "loss": 0.4159,
+      "step": 636
+    },
+    {
+      "epoch": 0.3264992311635059,
+      "grad_norm": 0.55211787697171,
+      "learning_rate": 1.9085945041665418e-05,
+      "loss": 0.454,
+      "step": 637
+    },
+    {
+      "epoch": 0.32701178882624293,
+      "grad_norm": 0.5868583308659561,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.4428,
+      "step": 638
+    },
+    {
+      "epoch": 0.32752434648898,
+      "grad_norm": 0.5679509607482779,
+      "learning_rate": 1.9078997115842767e-05,
+      "loss": 0.447,
+      "step": 639
+    },
+    {
+      "epoch": 0.32803690415171705,
+      "grad_norm": 0.5463426454337539,
+      "learning_rate": 1.9075513764737753e-05,
+      "loss": 0.4025,
+      "step": 640
+    },
+    {
+      "epoch": 0.3285494618144541,
+      "grad_norm": 0.6043201697346615,
+      "learning_rate": 1.907202415803686e-05,
+      "loss": 0.4216,
+      "step": 641
+    },
+    {
+      "epoch": 0.32906201947719116,
+      "grad_norm": 0.5771549273668641,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.4093,
+      "step": 642
+    },
+    {
+      "epoch": 0.3295745771399282,
+      "grad_norm": 0.6224383393282293,
+      "learning_rate": 1.906502618747306e-05,
+      "loss": 0.4254,
+      "step": 643
+    },
+    {
+      "epoch": 0.3300871348026653,
+      "grad_norm": 0.5627356971085579,
+      "learning_rate": 1.906151782843373e-05,
+      "loss": 0.4296,
+      "step": 644
+    },
+    {
+      "epoch": 0.3305996924654024,
+      "grad_norm": 0.5371401076861595,
+      "learning_rate": 1.9058003223445676e-05,
+      "loss": 0.4051,
+      "step": 645
+    },
+    {
+      "epoch": 0.33111225012813944,
+      "grad_norm": 0.5601201431016486,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.452,
+      "step": 646
+    },
+    {
+      "epoch": 0.3316248077908765,
+      "grad_norm": 0.5297619217891291,
+      "learning_rate": 1.9050955285317957e-05,
+      "loss": 0.4336,
+      "step": 647
+    },
+    {
+      "epoch": 0.33213736545361355,
+      "grad_norm": 0.5490606874746901,
+      "learning_rate": 1.90474219570363e-05,
+      "loss": 0.4172,
+      "step": 648
+    },
+    {
+      "epoch": 0.3326499231163506,
+      "grad_norm": 0.6299132424326019,
+      "learning_rate": 1.9043882392521973e-05,
+      "loss": 0.4841,
+      "step": 649
+    },
+    {
+      "epoch": 0.33316248077908767,
+      "grad_norm": 0.6120066547934971,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 0.4449,
+      "step": 650
+    },
+    {
+      "epoch": 0.3336750384418247,
+      "grad_norm": 0.5379854962410942,
+      "learning_rate": 1.9036784564558623e-05,
+      "loss": 0.4099,
+      "step": 651
+    },
+    {
+      "epoch": 0.3341875961045618,
+      "grad_norm": 0.6201678320374656,
+      "learning_rate": 1.9033226306002016e-05,
+      "loss": 0.4262,
+      "step": 652
+    },
+    {
+      "epoch": 0.33470015376729884,
+      "grad_norm": 0.5962893339574927,
+      "learning_rate": 1.9029661820997548e-05,
+      "loss": 0.3936,
+      "step": 653
+    },
+    {
+      "epoch": 0.3352127114300359,
+      "grad_norm": 0.5855371513118299,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.4684,
+      "step": 654
+    },
+    {
+      "epoch": 0.33572526909277295,
+      "grad_norm": 0.6045276973630284,
+      "learning_rate": 1.9022514181477085e-05,
+      "loss": 0.3962,
+      "step": 655
+    },
+    {
+      "epoch": 0.33623782675551,
+      "grad_norm": 0.6429421284330048,
+      "learning_rate": 1.901893103188783e-05,
+      "loss": 0.4562,
+      "step": 656
+    },
+    {
+      "epoch": 0.33675038441824706,
+      "grad_norm": 0.594498116766538,
+      "learning_rate": 1.9015341665704206e-05,
+      "loss": 0.4695,
+      "step": 657
+    },
+    {
+      "epoch": 0.3372629420809841,
+      "grad_norm": 0.587716326536086,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.4329,
+      "step": 658
+    },
+    {
+      "epoch": 0.3377754997437212,
+      "grad_norm": 0.5884276782613048,
+      "learning_rate": 1.9008144293454482e-05,
+      "loss": 0.4003,
+      "step": 659
+    },
+    {
+      "epoch": 0.33828805740645823,
+      "grad_norm": 0.6131282404550585,
+      "learning_rate": 1.9004536292349406e-05,
+      "loss": 0.4552,
+      "step": 660
+    },
+    {
+      "epoch": 0.3388006150691953,
+      "grad_norm": 0.5849695091451211,
+      "learning_rate": 1.9000922084572008e-05,
+      "loss": 0.4449,
+      "step": 661
+    },
+    {
+      "epoch": 0.33931317273193234,
+      "grad_norm": 0.5949443520063821,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 0.4526,
+      "step": 662
+    },
+    {
+      "epoch": 0.3398257303946694,
+      "grad_norm": 0.5244905721390859,
+      "learning_rate": 1.899367505896936e-05,
+      "loss": 0.4406,
+      "step": 663
+    },
+    {
+      "epoch": 0.34033828805740646,
+      "grad_norm": 0.59114772192471,
+      "learning_rate": 1.899004224613936e-05,
+      "loss": 0.4224,
+      "step": 664
+    },
+    {
+      "epoch": 0.3408508457201435,
+      "grad_norm": 0.6028734569553604,
+      "learning_rate": 1.898640323662753e-05,
+      "loss": 0.3917,
+      "step": 665
+    },
+    {
+      "epoch": 0.34136340338288057,
+      "grad_norm": 0.5544857812450461,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 0.4018,
+      "step": 666
+    },
+    {
+      "epoch": 0.3418759610456176,
+      "grad_norm": 0.5786947719361171,
+      "learning_rate": 1.8979106637595895e-05,
+      "loss": 0.4172,
+      "step": 667
+    },
+    {
+      "epoch": 0.3423885187083547,
+      "grad_norm": 0.6467565027385662,
+      "learning_rate": 1.8975449053105505e-05,
+      "loss": 0.4521,
+      "step": 668
+    },
+    {
+      "epoch": 0.34290107637109174,
+      "grad_norm": 0.5917380401204393,
+      "learning_rate": 1.8971785281992124e-05,
+      "loss": 0.4136,
+      "step": 669
+    },
+    {
+      "epoch": 0.3434136340338288,
+      "grad_norm": 0.6073539623670516,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.3893,
+      "step": 670
+    },
+    {
+      "epoch": 0.34392619169656585,
+      "grad_norm": 0.5994076980088598,
+      "learning_rate": 1.8964439190002156e-05,
+      "loss": 0.4366,
+      "step": 671
+    },
+    {
+      "epoch": 0.3444387493593029,
+      "grad_norm": 0.6057580904564751,
+      "learning_rate": 1.8960756874189104e-05,
+      "loss": 0.4027,
+      "step": 672
+    },
+    {
+      "epoch": 0.34495130702203997,
+      "grad_norm": 0.5891968273478849,
+      "learning_rate": 1.895706838188013e-05,
+      "loss": 0.4204,
+      "step": 673
+    },
+    {
+      "epoch": 0.345463864684777,
+      "grad_norm": 0.5460509133780702,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.4114,
+      "step": 674
+    },
+    {
+      "epoch": 0.3459764223475141,
+      "grad_norm": 0.5494046728693905,
+      "learning_rate": 1.8949672877948324e-05,
+      "loss": 0.4576,
+      "step": 675
+    },
+    {
+      "epoch": 0.34648898001025114,
+      "grad_norm": 0.5641972727538962,
+      "learning_rate": 1.894596587142309e-05,
+      "loss": 0.4435,
+      "step": 676
+    },
+    {
+      "epoch": 0.3470015376729882,
+      "grad_norm": 0.5251236050352944,
+      "learning_rate": 1.8942252698597113e-05,
+      "loss": 0.3919,
+      "step": 677
+    },
+    {
+      "epoch": 0.34751409533572525,
+      "grad_norm": 0.5463096597342325,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 0.3953,
+      "step": 678
+    },
+    {
+      "epoch": 0.3480266529984623,
+      "grad_norm": 0.551642975227663,
+      "learning_rate": 1.8934807864284904e-05,
+      "loss": 0.4104,
+      "step": 679
+    },
+    {
+      "epoch": 0.34853921066119936,
+      "grad_norm": 0.5921893038445039,
+      "learning_rate": 1.8931076207930265e-05,
+      "loss": 0.4141,
+      "step": 680
+    },
+    {
+      "epoch": 0.3490517683239364,
+      "grad_norm": 0.6171364527081707,
+      "learning_rate": 1.892733839553808e-05,
+      "loss": 0.4621,
+      "step": 681
+    },
+    {
+      "epoch": 0.3495643259866735,
+      "grad_norm": 0.5544680485387425,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.382,
+      "step": 682
+    },
+    {
+      "epoch": 0.3500768836494106,
+      "grad_norm": 0.6375192051284304,
+      "learning_rate": 1.8919844312950937e-05,
+      "loss": 0.4252,
+      "step": 683
+    },
+    {
+      "epoch": 0.35058944131214764,
+      "grad_norm": 0.6110068450765196,
+      "learning_rate": 1.8916088047921527e-05,
+      "loss": 0.4199,
+      "step": 684
+    },
+    {
+      "epoch": 0.3511019989748847,
+      "grad_norm": 0.5547590632152561,
+      "learning_rate": 1.8912325637185647e-05,
+      "loss": 0.4088,
+      "step": 685
+    },
+    {
+      "epoch": 0.35161455663762176,
+      "grad_norm": 0.5459227047551427,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.3989,
+      "step": 686
+    },
+    {
+      "epoch": 0.3521271143003588,
+      "grad_norm": 0.6316850248774554,
+      "learning_rate": 1.8904782388972186e-05,
+      "loss": 0.4378,
+      "step": 687
+    },
+    {
+      "epoch": 0.35263967196309587,
+      "grad_norm": 0.6183752012045562,
+      "learning_rate": 1.890100155669403e-05,
+      "loss": 0.4127,
+      "step": 688
+    },
+    {
+      "epoch": 0.3531522296258329,
+      "grad_norm": 0.6131551668773976,
+      "learning_rate": 1.889721458910827e-05,
+      "loss": 0.4539,
+      "step": 689
+    },
+    {
+      "epoch": 0.35366478728857,
+      "grad_norm": 0.6124161384824086,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 0.4511,
+      "step": 690
+    },
+    {
+      "epoch": 0.35417734495130704,
+      "grad_norm": 0.5345476343867226,
+      "learning_rate": 1.8889622258459317e-05,
+      "loss": 0.418,
+      "step": 691
+    },
+    {
+      "epoch": 0.3546899026140441,
+      "grad_norm": 0.5743131926100472,
+      "learning_rate": 1.888581690062939e-05,
+      "loss": 0.4152,
+      "step": 692
+    },
+    {
+      "epoch": 0.35520246027678115,
+      "grad_norm": 0.6114331097114022,
+      "learning_rate": 1.8882005417958376e-05,
+      "loss": 0.4033,
+      "step": 693
+    },
+    {
+      "epoch": 0.3557150179395182,
+      "grad_norm": 0.5560554726307178,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 0.4337,
+      "step": 694
+    },
+    {
+      "epoch": 0.35622757560225526,
+      "grad_norm": 0.531546245897098,
+      "learning_rate": 1.887436408860607e-05,
+      "loss": 0.409,
+      "step": 695
+    },
+    {
+      "epoch": 0.3567401332649923,
+      "grad_norm": 0.5605142867171921,
+      "learning_rate": 1.8870534247191815e-05,
+      "loss": 0.3703,
+      "step": 696
+    },
+    {
+      "epoch": 0.3572526909277294,
+      "grad_norm": 0.6179873535419057,
+      "learning_rate": 1.886669829147055e-05,
+      "loss": 0.3983,
+      "step": 697
+    },
+    {
+      "epoch": 0.35776524859046643,
+      "grad_norm": 0.5788851461967123,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 0.4519,
+      "step": 698
+    },
+    {
+      "epoch": 0.3582778062532035,
+      "grad_norm": 0.5838609397284729,
+      "learning_rate": 1.885900804768742e-05,
+      "loss": 0.405,
+      "step": 699
+    },
+    {
+      "epoch": 0.35879036391594055,
+      "grad_norm": 0.5911129434746643,
+      "learning_rate": 1.8855153764926307e-05,
+      "loss": 0.415,
+      "step": 700
+    },
+    {
+      "epoch": 0.3593029215786776,
+      "grad_norm": 0.5379608897228639,
+      "learning_rate": 1.8851293378459685e-05,
+      "loss": 0.394,
+      "step": 701
+    },
+    {
+      "epoch": 0.35981547924141466,
+      "grad_norm": 0.6067752243562266,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 0.4268,
+      "step": 702
+    },
+    {
+      "epoch": 0.3603280369041517,
+      "grad_norm": 0.5915114590429891,
+      "learning_rate": 1.88435543050577e-05,
+      "loss": 0.4032,
+      "step": 703
+    },
+    {
+      "epoch": 0.3608405945668888,
+      "grad_norm": 0.6805795617268559,
+      "learning_rate": 1.883967562345675e-05,
+      "loss": 0.4236,
+      "step": 704
+    },
+    {
+      "epoch": 0.36135315222962583,
+      "grad_norm": 0.5499984257142561,
+      "learning_rate": 1.8835790848819105e-05,
+      "loss": 0.4139,
+      "step": 705
+    },
+    {
+      "epoch": 0.3618657098923629,
+      "grad_norm": 0.6285778260843062,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 0.4599,
+      "step": 706
+    },
+    {
+      "epoch": 0.36237826755509994,
+      "grad_norm": 0.6259975237757512,
+      "learning_rate": 1.8828003031148764e-05,
+      "loss": 0.4076,
+      "step": 707
+    },
+    {
+      "epoch": 0.362890825217837,
+      "grad_norm": 0.5357632600393464,
+      "learning_rate": 1.8824099993484075e-05,
+      "loss": 0.4232,
+      "step": 708
+    },
+    {
+      "epoch": 0.36340338288057406,
+      "grad_norm": 0.550434776226621,
+      "learning_rate": 1.8820190873518708e-05,
+      "loss": 0.4432,
+      "step": 709
+    },
+    {
+      "epoch": 0.3639159405433111,
+      "grad_norm": 0.5448543780939162,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 0.4367,
+      "step": 710
+    },
+    {
+      "epoch": 0.36442849820604817,
+      "grad_norm": 0.5190246299243453,
+      "learning_rate": 1.8812354397468076e-05,
+      "loss": 0.3985,
+      "step": 711
+    },
+    {
+      "epoch": 0.3649410558687852,
+      "grad_norm": 0.6395703948095185,
+      "learning_rate": 1.8808427046784365e-05,
+      "loss": 0.4366,
+      "step": 712
+    },
+    {
+      "epoch": 0.3654536135315223,
+      "grad_norm": 0.5801656720492908,
+      "learning_rate": 1.8804493624603064e-05,
+      "loss": 0.3985,
+      "step": 713
+    },
+    {
+      "epoch": 0.36596617119425934,
+      "grad_norm": 0.5703313056638967,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 0.4123,
+      "step": 714
+    },
+    {
+      "epoch": 0.3664787288569964,
+      "grad_norm": 0.5972274404025266,
+      "learning_rate": 1.879660857659685e-05,
+      "loss": 0.4494,
+      "step": 715
+    },
+    {
+      "epoch": 0.36699128651973345,
+      "grad_norm": 0.59192510573632,
+      "learning_rate": 1.879265695620696e-05,
+      "loss": 0.4288,
+      "step": 716
+    },
+    {
+      "epoch": 0.3675038441824705,
+      "grad_norm": 0.5515798101330228,
+      "learning_rate": 1.8788699275189538e-05,
+      "loss": 0.4536,
+      "step": 717
+    },
+    {
+      "epoch": 0.36801640184520756,
+      "grad_norm": 0.5695673945758788,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.4347,
+      "step": 718
+    },
+    {
+      "epoch": 0.3685289595079446,
+      "grad_norm": 0.5781498934841822,
+      "learning_rate": 1.8780765742188113e-05,
+      "loss": 0.4336,
+      "step": 719
+    },
+    {
+      "epoch": 0.3690415171706817,
+      "grad_norm": 0.5010931922208206,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.4346,
+      "step": 720
+    },
+    {
+      "epoch": 0.36955407483341873,
+      "grad_norm": 0.5435576750179325,
+      "learning_rate": 1.8772807999466362e-05,
+      "loss": 0.4162,
+      "step": 721
+    },
+    {
+      "epoch": 0.37006663249615585,
+      "grad_norm": 0.6009294076668399,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.3953,
+      "step": 722
+    },
+    {
+      "epoch": 0.3705791901588929,
+      "grad_norm": 0.5935276062190282,
+      "learning_rate": 1.876482606896482e-05,
+      "loss": 0.4144,
+      "step": 723
+    },
+    {
+      "epoch": 0.37109174782162996,
+      "grad_norm": 0.5553384828946372,
+      "learning_rate": 1.8760826040171286e-05,
+      "loss": 0.4024,
+      "step": 724
+    },
+    {
+      "epoch": 0.371604305484367,
+      "grad_norm": 0.570133910472963,
+      "learning_rate": 1.875681997269072e-05,
+      "loss": 0.4238,
+      "step": 725
+    },
+    {
+      "epoch": 0.37211686314710407,
+      "grad_norm": 0.6065166690729229,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.4116,
+      "step": 726
+    },
+    {
+      "epoch": 0.3726294208098411,
+      "grad_norm": 0.5561949844982605,
+      "learning_rate": 1.874878973271791e-05,
+      "loss": 0.4134,
+      "step": 727
+    },
+    {
+      "epoch": 0.3731419784725782,
+      "grad_norm": 0.5735530058545172,
+      "learning_rate": 1.8744765565760778e-05,
+      "loss": 0.4258,
+      "step": 728
+    },
+    {
+      "epoch": 0.37365453613531524,
+      "grad_norm": 0.512990960862343,
+      "learning_rate": 1.8740735371186823e-05,
+      "loss": 0.3549,
+      "step": 729
+    },
+    {
+      "epoch": 0.3741670937980523,
+      "grad_norm": 0.5537090470754283,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 0.4366,
+      "step": 730
+    },
+    {
+      "epoch": 0.37467965146078935,
+      "grad_norm": 0.5687808431411854,
+      "learning_rate": 1.8732656910304378e-05,
+      "loss": 0.4221,
+      "step": 731
+    },
+    {
+      "epoch": 0.3751922091235264,
+      "grad_norm": 0.6131881958545728,
+      "learning_rate": 1.872860864956423e-05,
+      "loss": 0.4557,
+      "step": 732
+    },
+    {
+      "epoch": 0.37570476678626347,
+      "grad_norm": 0.6059001950802133,
+      "learning_rate": 1.872455437234395e-05,
+      "loss": 0.457,
+      "step": 733
+    },
+    {
+      "epoch": 0.3762173244490005,
+      "grad_norm": 0.5583534562666203,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.3903,
+      "step": 734
+    },
+    {
+      "epoch": 0.3767298821117376,
+      "grad_norm": 0.6223786591919854,
+      "learning_rate": 1.8716427779645303e-05,
+      "loss": 0.3841,
+      "step": 735
+    },
+    {
+      "epoch": 0.37724243977447464,
+      "grad_norm": 0.5844228197074263,
+      "learning_rate": 1.8712355469768454e-05,
+      "loss": 0.4635,
+      "step": 736
+    },
+    {
+      "epoch": 0.3777549974372117,
+      "grad_norm": 0.5947456290480222,
+      "learning_rate": 1.870827715461451e-05,
+      "loss": 0.4402,
+      "step": 737
+    },
+    {
+      "epoch": 0.37826755509994875,
+      "grad_norm": 0.5639620591588684,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 0.3993,
+      "step": 738
+    },
+    {
+      "epoch": 0.3787801127626858,
+      "grad_norm": 0.561627827407368,
+      "learning_rate": 1.8700102519723912e-05,
+      "loss": 0.4016,
+      "step": 739
+    },
+    {
+      "epoch": 0.37929267042542286,
+      "grad_norm": 0.5460620139805518,
+      "learning_rate": 1.8696006205621898e-05,
+      "loss": 0.4105,
+      "step": 740
+    },
+    {
+      "epoch": 0.3798052280881599,
+      "grad_norm": 0.5446070787275512,
+      "learning_rate": 1.869190389751205e-05,
+      "loss": 0.3993,
+      "step": 741
+    },
+    {
+      "epoch": 0.380317785750897,
+      "grad_norm": 0.5826790756088154,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.4279,
+      "step": 742
+    },
+    {
+      "epoch": 0.38083034341363403,
+      "grad_norm": 0.6126902647426233,
+      "learning_rate": 1.86836813105836e-05,
+      "loss": 0.4313,
+      "step": 743
+    },
+    {
+      "epoch": 0.3813429010763711,
+      "grad_norm": 0.5181698232142076,
+      "learning_rate": 1.867956103743268e-05,
+      "loss": 0.4226,
+      "step": 744
+    },
+    {
+      "epoch": 0.38185545873910814,
+      "grad_norm": 0.5524453524574379,
+      "learning_rate": 1.8675434781609305e-05,
+      "loss": 0.3774,
+      "step": 745
+    },
+    {
+      "epoch": 0.3823680164018452,
+      "grad_norm": 0.5012869618045329,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.3662,
+      "step": 746
+    },
+    {
+      "epoch": 0.38288057406458226,
+      "grad_norm": 0.5196022577041353,
+      "learning_rate": 1.8667164333325928e-05,
+      "loss": 0.4128,
+      "step": 747
+    },
+    {
+      "epoch": 0.3833931317273193,
+      "grad_norm": 0.5145324534748185,
+      "learning_rate": 1.8663020146566606e-05,
+      "loss": 0.3976,
+      "step": 748
+    },
+    {
+      "epoch": 0.38390568939005637,
+      "grad_norm": 0.5283943550544717,
+      "learning_rate": 1.8658869988536175e-05,
+      "loss": 0.4109,
+      "step": 749
+    },
+    {
+      "epoch": 0.3844182470527934,
+      "grad_norm": 0.5773631296122648,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.4126,
+      "step": 750
+    },
+    {
+      "epoch": 0.3849308047155305,
+      "grad_norm": 0.6248646685722752,
+      "learning_rate": 1.8650551770108644e-05,
+      "loss": 0.3887,
+      "step": 751
+    },
+    {
+      "epoch": 0.38544336237826754,
+      "grad_norm": 0.5235748666439245,
+      "learning_rate": 1.8646383715445144e-05,
+      "loss": 0.4139,
+      "step": 752
+    },
+    {
+      "epoch": 0.3859559200410046,
+      "grad_norm": 0.5512798178867581,
+      "learning_rate": 1.8642209700977745e-05,
+      "loss": 0.3633,
+      "step": 753
+    },
+    {
+      "epoch": 0.38646847770374165,
+      "grad_norm": 0.5877811982011446,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 0.4262,
+      "step": 754
+    },
+    {
+      "epoch": 0.3869810353664787,
+      "grad_norm": 0.5587780524191297,
+      "learning_rate": 1.8633843804143663e-05,
+      "loss": 0.4187,
+      "step": 755
+    },
+    {
+      "epoch": 0.38749359302921577,
+      "grad_norm": 0.5744290103407644,
+      "learning_rate": 1.8629651927543447e-05,
+      "loss": 0.4447,
+      "step": 756
+    },
+    {
+      "epoch": 0.3880061506919528,
+      "grad_norm": 0.5817177185216379,
+      "learning_rate": 1.8625454102672266e-05,
+      "loss": 0.4508,
+      "step": 757
+    },
+    {
+      "epoch": 0.3885187083546899,
+      "grad_norm": 0.5938629617857264,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 0.4586,
+      "step": 758
+    },
+    {
+      "epoch": 0.38903126601742694,
+      "grad_norm": 0.5095491986181975,
+      "learning_rate": 1.8617040619695055e-05,
+      "loss": 0.3953,
+      "step": 759
+    },
+    {
+      "epoch": 0.389543823680164,
+      "grad_norm": 0.5722821066616003,
+      "learning_rate": 1.8612824967388297e-05,
+      "loss": 0.3944,
+      "step": 760
+    },
+    {
+      "epoch": 0.3900563813429011,
+      "grad_norm": 0.5772913913060717,
+      "learning_rate": 1.8608603378409108e-05,
+      "loss": 0.4175,
+      "step": 761
+    },
+    {
+      "epoch": 0.39056893900563816,
+      "grad_norm": 0.5741845546040107,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.3701,
+      "step": 762
+    },
+    {
+      "epoch": 0.3910814966683752,
+      "grad_norm": 0.5016023760629328,
+      "learning_rate": 1.8600142402077006e-05,
+      "loss": 0.3663,
+      "step": 763
+    },
+    {
+      "epoch": 0.3915940543311123,
+      "grad_norm": 0.5699229918774636,
+      "learning_rate": 1.8595903020556098e-05,
+      "loss": 0.4481,
+      "step": 764
+    },
+    {
+      "epoch": 0.39210661199384933,
+      "grad_norm": 0.5574809558406526,
+      "learning_rate": 1.8591657714026764e-05,
+      "loss": 0.4032,
+      "step": 765
+    },
+    {
+      "epoch": 0.3926191696565864,
+      "grad_norm": 0.6019188453256485,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 0.4439,
+      "step": 766
+    },
+    {
+      "epoch": 0.39313172731932344,
+      "grad_norm": 0.5883513231095905,
+      "learning_rate": 1.858314933765178e-05,
+      "loss": 0.4268,
+      "step": 767
+    },
+    {
+      "epoch": 0.3936442849820605,
+      "grad_norm": 0.552437502538594,
+      "learning_rate": 1.8578886273670807e-05,
+      "loss": 0.4104,
+      "step": 768
+    },
+    {
+      "epoch": 0.39415684264479756,
+      "grad_norm": 0.5782116296513438,
+      "learning_rate": 1.857461729641076e-05,
+      "loss": 0.4452,
+      "step": 769
+    },
+    {
+      "epoch": 0.3946694003075346,
+      "grad_norm": 0.5469500367348614,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.4048,
+      "step": 770
+    },
+    {
+      "epoch": 0.39518195797027167,
+      "grad_norm": 0.5304622145661465,
+      "learning_rate": 1.856606161382766e-05,
+      "loss": 0.3829,
+      "step": 771
+    },
+    {
+      "epoch": 0.3956945156330087,
+      "grad_norm": 0.6173543930461453,
+      "learning_rate": 1.8561774914401883e-05,
+      "loss": 0.4604,
+      "step": 772
+    },
+    {
+      "epoch": 0.3962070732957458,
+      "grad_norm": 0.5761643461549815,
+      "learning_rate": 1.8557482313491607e-05,
+      "loss": 0.4239,
+      "step": 773
+    },
+    {
+      "epoch": 0.39671963095848284,
+      "grad_norm": 0.6128102918517123,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.3944,
+      "step": 774
+    },
+    {
+      "epoch": 0.3972321886212199,
+      "grad_norm": 0.57393444618412,
+      "learning_rate": 1.854887941905687e-05,
+      "loss": 0.4566,
+      "step": 775
+    },
+    {
+      "epoch": 0.39774474628395695,
+      "grad_norm": 0.5452497982563483,
+      "learning_rate": 1.8544569131462235e-05,
+      "loss": 0.392,
+      "step": 776
+    },
+    {
+      "epoch": 0.398257303946694,
+      "grad_norm": 0.5075161540447584,
+      "learning_rate": 1.8540252954242742e-05,
+      "loss": 0.3809,
+      "step": 777
+    },
+    {
+      "epoch": 0.39876986160943106,
+      "grad_norm": 0.5818668690288266,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.3843,
+      "step": 778
+    },
+    {
+      "epoch": 0.3992824192721681,
+      "grad_norm": 0.5357152390212772,
+      "learning_rate": 1.8531602942833525e-05,
+      "loss": 0.4365,
+      "step": 779
+    },
+    {
+      "epoch": 0.3997949769349052,
+      "grad_norm": 0.6024287247716755,
+      "learning_rate": 1.8527269114606102e-05,
+      "loss": 0.4069,
+      "step": 780
+    },
+    {
+      "epoch": 0.40030753459764223,
+      "grad_norm": 0.5719802209308483,
+      "learning_rate": 1.8522929408678426e-05,
+      "loss": 0.396,
+      "step": 781
+    },
+    {
+      "epoch": 0.4008200922603793,
+      "grad_norm": 0.5835004661701976,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.4191,
+      "step": 782
+    },
+    {
+      "epoch": 0.40133264992311635,
+      "grad_norm": 0.5848806985885683,
+      "learning_rate": 1.851423237569152e-05,
+      "loss": 0.4263,
+      "step": 783
+    },
+    {
+      "epoch": 0.4018452075858534,
+      "grad_norm": 0.5547860144281307,
+      "learning_rate": 1.8509875054627e-05,
+      "loss": 0.4307,
+      "step": 784
+    },
+    {
+      "epoch": 0.40235776524859046,
+      "grad_norm": 0.547806238403973,
+      "learning_rate": 1.8505511867851656e-05,
+      "loss": 0.3628,
+      "step": 785
+    },
+    {
+      "epoch": 0.4028703229113275,
+      "grad_norm": 0.5983826823530151,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 0.4394,
+      "step": 786
+    },
+    {
+      "epoch": 0.4033828805740646,
+      "grad_norm": 0.544639235929749,
+      "learning_rate": 1.8496767909202423e-05,
+      "loss": 0.4458,
+      "step": 787
+    },
+    {
+      "epoch": 0.40389543823680163,
+      "grad_norm": 0.5414267442546881,
+      "learning_rate": 1.8492387143355594e-05,
+      "loss": 0.4061,
+      "step": 788
+    },
+    {
+      "epoch": 0.4044079958995387,
+      "grad_norm": 0.5184015584896399,
+      "learning_rate": 1.848800052385206e-05,
+      "loss": 0.3614,
+      "step": 789
+    },
+    {
+      "epoch": 0.40492055356227574,
+      "grad_norm": 0.5139466026430121,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.4343,
+      "step": 790
+    },
+    {
+      "epoch": 0.4054331112250128,
+      "grad_norm": 0.5427123153261502,
+      "learning_rate": 1.847920973597339e-05,
+      "loss": 0.4484,
+      "step": 791
+    },
+    {
+      "epoch": 0.40594566888774986,
+      "grad_norm": 0.5945948791067623,
+      "learning_rate": 1.84748055736576e-05,
+      "loss": 0.3911,
+      "step": 792
+    },
+    {
+      "epoch": 0.4064582265504869,
+      "grad_norm": 0.5398923972234162,
+      "learning_rate": 1.847039556980377e-05,
+      "loss": 0.4493,
+      "step": 793
+    },
+    {
+      "epoch": 0.40697078421322397,
+      "grad_norm": 0.5479412250309055,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.404,
+      "step": 794
+    },
+    {
+      "epoch": 0.407483341875961,
+      "grad_norm": 0.5864622205399839,
+      "learning_rate": 1.846155804964501e-05,
+      "loss": 0.4171,
+      "step": 795
+    },
+    {
+      "epoch": 0.4079958995386981,
+      "grad_norm": 0.5429255302512453,
+      "learning_rate": 1.8457130539431623e-05,
+      "loss": 0.4297,
+      "step": 796
+    },
+    {
+      "epoch": 0.40850845720143514,
+      "grad_norm": 0.5742919569301188,
+      "learning_rate": 1.8452697199863305e-05,
+      "loss": 0.3758,
+      "step": 797
+    },
+    {
+      "epoch": 0.4090210148641722,
+      "grad_norm": 0.5068727734440271,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 0.3761,
+      "step": 798
+    },
+    {
+      "epoch": 0.40953357252690925,
+      "grad_norm": 0.5675979071672879,
+      "learning_rate": 1.8443813044889182e-05,
+      "loss": 0.4473,
+      "step": 799
+    },
+    {
+      "epoch": 0.41004613018964636,
+      "grad_norm": 0.5504670514756405,
+      "learning_rate": 1.8439362235607074e-05,
+      "loss": 0.4024,
+      "step": 800
+    },
+    {
+      "epoch": 0.4105586878523834,
+      "grad_norm": 0.5191565579196622,
+      "learning_rate": 1.8434905609217415e-05,
+      "loss": 0.4065,
+      "step": 801
+    },
+    {
+      "epoch": 0.4110712455151205,
+      "grad_norm": 0.566533978042447,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.4061,
+      "step": 802
+    },
+    {
+      "epoch": 0.41158380317785753,
+      "grad_norm": 0.567847079766571,
+      "learning_rate": 1.8425974917406974e-05,
+      "loss": 0.4392,
+      "step": 803
+    },
+    {
+      "epoch": 0.4120963608405946,
+      "grad_norm": 0.5619814864200016,
+      "learning_rate": 1.842150085814196e-05,
+      "loss": 0.3812,
+      "step": 804
+    },
+    {
+      "epoch": 0.41260891850333165,
+      "grad_norm": 0.5912774820973237,
+      "learning_rate": 1.8417020994080945e-05,
+      "loss": 0.412,
+      "step": 805
+    },
+    {
+      "epoch": 0.4131214761660687,
+      "grad_norm": 0.5333811929951495,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4647,
+      "step": 806
+    },
+    {
+      "epoch": 0.41363403382880576,
+      "grad_norm": 0.5506531724798861,
+      "learning_rate": 1.8408043863926462e-05,
+      "loss": 0.3362,
+      "step": 807
+    },
+    {
+      "epoch": 0.4141465914915428,
+      "grad_norm": 0.5754788507146248,
+      "learning_rate": 1.840354660402078e-05,
+      "loss": 0.3846,
+      "step": 808
+    },
+    {
+      "epoch": 0.41465914915427987,
+      "grad_norm": 0.5223028041054586,
+      "learning_rate": 1.8399043551694655e-05,
+      "loss": 0.3796,
+      "step": 809
+    },
+    {
+      "epoch": 0.41517170681701693,
+      "grad_norm": 0.5755787632697099,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.3895,
+      "step": 810
+    },
+    {
+      "epoch": 0.415684264479754,
+      "grad_norm": 0.5571177364909065,
+      "learning_rate": 1.8390020082200553e-05,
+      "loss": 0.4176,
+      "step": 811
+    },
+    {
+      "epoch": 0.41619682214249104,
+      "grad_norm": 0.5391652261539842,
+      "learning_rate": 1.8385499671252305e-05,
+      "loss": 0.4203,
+      "step": 812
+    },
+    {
+      "epoch": 0.4167093798052281,
+      "grad_norm": 0.5423651769080541,
+      "learning_rate": 1.8380973480323052e-05,
+      "loss": 0.4153,
+      "step": 813
+    },
+    {
+      "epoch": 0.41722193746796515,
+      "grad_norm": 0.5921023630847237,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 0.4106,
+      "step": 814
+    },
+    {
+      "epoch": 0.4177344951307022,
+      "grad_norm": 0.5052383567635528,
+      "learning_rate": 1.8371903771004812e-05,
+      "loss": 0.3849,
+      "step": 815
+    },
+    {
+      "epoch": 0.41824705279343927,
+      "grad_norm": 0.5599085188025777,
+      "learning_rate": 1.8367360258867418e-05,
+      "loss": 0.4236,
+      "step": 816
+    },
+    {
+      "epoch": 0.4187596104561763,
+      "grad_norm": 0.5915843872611799,
+      "learning_rate": 1.8362810979252206e-05,
+      "loss": 0.4248,
+      "step": 817
+    },
+    {
+      "epoch": 0.4192721681189134,
+      "grad_norm": 0.6029662182905523,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.4306,
+      "step": 818
+    },
+    {
+      "epoch": 0.41978472578165044,
+      "grad_norm": 0.5428849482593042,
+      "learning_rate": 1.8353695130135262e-05,
+      "loss": 0.379,
+      "step": 819
+    },
+    {
+      "epoch": 0.4202972834443875,
+      "grad_norm": 0.5725630246023327,
+      "learning_rate": 1.834912856691692e-05,
+      "loss": 0.4086,
+      "step": 820
+    },
+    {
+      "epoch": 0.42080984110712455,
+      "grad_norm": 0.5572637159342845,
+      "learning_rate": 1.8344556248787556e-05,
+      "loss": 0.3924,
+      "step": 821
+    },
+    {
+      "epoch": 0.4213223987698616,
+      "grad_norm": 0.5453531878328126,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.4269,
+      "step": 822
+    },
+    {
+      "epoch": 0.42183495643259866,
+      "grad_norm": 0.5162141232250845,
+      "learning_rate": 1.8335394360406184e-05,
+      "loss": 0.4006,
+      "step": 823
+    },
+    {
+      "epoch": 0.4223475140953357,
+      "grad_norm": 0.5415131263522793,
+      "learning_rate": 1.8330804796469314e-05,
+      "loss": 0.3956,
+      "step": 824
+    },
+    {
+      "epoch": 0.4228600717580728,
+      "grad_norm": 0.5412065460240969,
+      "learning_rate": 1.8326209490251678e-05,
+      "loss": 0.4536,
+      "step": 825
+    },
+    {
+      "epoch": 0.42337262942080983,
+      "grad_norm": 0.581595492234363,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.4322,
+      "step": 826
+    },
+    {
+      "epoch": 0.4238851870835469,
+      "grad_norm": 0.5220890042910279,
+      "learning_rate": 1.831700166364792e-05,
+      "loss": 0.3662,
+      "step": 827
+    },
+    {
+      "epoch": 0.42439774474628394,
+      "grad_norm": 0.5648836909253335,
+      "learning_rate": 1.8312389149608595e-05,
+      "loss": 0.3974,
+      "step": 828
+    },
+    {
+      "epoch": 0.424910302409021,
+      "grad_norm": 0.592854764179354,
+      "learning_rate": 1.8307770905982092e-05,
+      "loss": 0.437,
+      "step": 829
+    },
+    {
+      "epoch": 0.42542286007175806,
+      "grad_norm": 0.5489537837692718,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.3884,
+      "step": 830
+    },
+    {
+      "epoch": 0.4259354177344951,
+      "grad_norm": 0.5596209554593033,
+      "learning_rate": 1.8298517242704607e-05,
+      "loss": 0.412,
+      "step": 831
+    },
+    {
+      "epoch": 0.42644797539723217,
+      "grad_norm": 0.6086311679551524,
+      "learning_rate": 1.8293881829432015e-05,
+      "loss": 0.4283,
+      "step": 832
+    },
+    {
+      "epoch": 0.4269605330599692,
+      "grad_norm": 0.5584641876093719,
+      "learning_rate": 1.828924069932902e-05,
+      "loss": 0.4137,
+      "step": 833
+    },
+    {
+      "epoch": 0.4274730907227063,
+      "grad_norm": 0.5587825611963028,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 0.4032,
+      "step": 834
+    },
+    {
+      "epoch": 0.42798564838544334,
+      "grad_norm": 0.5722478550641511,
+      "learning_rate": 1.827994130143198e-05,
+      "loss": 0.3978,
+      "step": 835
+    },
+    {
+      "epoch": 0.4284982060481804,
+      "grad_norm": 0.5952211492380136,
+      "learning_rate": 1.8275283040047848e-05,
+      "loss": 0.4147,
+      "step": 836
+    },
+    {
+      "epoch": 0.42901076371091745,
+      "grad_norm": 0.5381454578732445,
+      "learning_rate": 1.8270619074653143e-05,
+      "loss": 0.4111,
+      "step": 837
+    },
+    {
+      "epoch": 0.4295233213736545,
+      "grad_norm": 0.4779361697510905,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 0.3957,
+      "step": 838
+    },
+    {
+      "epoch": 0.4300358790363916,
+      "grad_norm": 0.5545979293162127,
+      "learning_rate": 1.8261274044695106e-05,
+      "loss": 0.3982,
+      "step": 839
+    },
+    {
+      "epoch": 0.4305484366991287,
+      "grad_norm": 0.5885974075242077,
+      "learning_rate": 1.825659298657314e-05,
+      "loss": 0.4309,
+      "step": 840
+    },
+    {
+      "epoch": 0.43106099436186573,
+      "grad_norm": 0.53164461508697,
+      "learning_rate": 1.8251906237323338e-05,
+      "loss": 0.4022,
+      "step": 841
+    },
+    {
+      "epoch": 0.4315735520246028,
+      "grad_norm": 0.5327739501895454,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.3923,
+      "step": 842
+    },
+    {
+      "epoch": 0.43208610968733985,
+      "grad_norm": 0.5448087961332365,
+      "learning_rate": 1.8242515678366114e-05,
+      "loss": 0.3947,
+      "step": 843
+    },
+    {
+      "epoch": 0.4325986673500769,
+      "grad_norm": 0.5689288250137065,
+      "learning_rate": 1.8237811875131446e-05,
+      "loss": 0.4164,
+      "step": 844
+    },
+    {
+      "epoch": 0.43311122501281396,
+      "grad_norm": 0.5236068358051442,
+      "learning_rate": 1.8233102393714435e-05,
+      "loss": 0.4382,
+      "step": 845
+    },
+    {
+      "epoch": 0.433623782675551,
+      "grad_norm": 0.5633261078871528,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.4356,
+      "step": 846
+    },
+    {
+      "epoch": 0.4341363403382881,
+      "grad_norm": 0.5509339876020969,
+      "learning_rate": 1.8223666409321953e-05,
+      "loss": 0.4259,
+      "step": 847
+    },
+    {
+      "epoch": 0.43464889800102513,
+      "grad_norm": 0.5793512720207437,
+      "learning_rate": 1.8218939912850547e-05,
+      "loss": 0.3949,
+      "step": 848
+    },
+    {
+      "epoch": 0.4351614556637622,
+      "grad_norm": 0.611416072860162,
+      "learning_rate": 1.8214207751204917e-05,
+      "loss": 0.4455,
+      "step": 849
+    },
+    {
+      "epoch": 0.43567401332649924,
+      "grad_norm": 0.5507891047460124,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.4282,
+      "step": 850
+    },
+    {
+      "epoch": 0.4361865709892363,
+      "grad_norm": 0.5016988087415162,
+      "learning_rate": 1.8204726445442084e-05,
+      "loss": 0.3944,
+      "step": 851
+    },
+    {
+      "epoch": 0.43669912865197336,
+      "grad_norm": 0.6094335090073131,
+      "learning_rate": 1.8199977307860178e-05,
+      "loss": 0.3703,
+      "step": 852
+    },
+    {
+      "epoch": 0.4372116863147104,
+      "grad_norm": 0.5568068533408395,
+      "learning_rate": 1.8195222518174654e-05,
+      "loss": 0.39,
+      "step": 853
+    },
+    {
+      "epoch": 0.43772424397744747,
+      "grad_norm": 0.5188088259300201,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.4073,
+      "step": 854
+    },
+    {
+      "epoch": 0.4382368016401845,
+      "grad_norm": 0.5664827932580417,
+      "learning_rate": 1.8185695995606196e-05,
+      "loss": 0.3957,
+      "step": 855
+    },
+    {
+      "epoch": 0.4387493593029216,
+      "grad_norm": 0.5439411686837821,
+      "learning_rate": 1.8180924269289737e-05,
+      "loss": 0.4113,
+      "step": 856
+    },
+    {
+      "epoch": 0.43926191696565864,
+      "grad_norm": 0.5094644471260413,
+      "learning_rate": 1.8176146904002587e-05,
+      "loss": 0.3851,
+      "step": 857
+    },
+    {
+      "epoch": 0.4397744746283957,
+      "grad_norm": 0.529703636667814,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.4607,
+      "step": 858
+    },
+    {
+      "epoch": 0.44028703229113275,
+      "grad_norm": 0.5055358093347538,
+      "learning_rate": 1.8166575269691914e-05,
+      "loss": 0.4064,
+      "step": 859
+    },
+    {
+      "epoch": 0.4407995899538698,
+      "grad_norm": 0.5703455801694665,
+      "learning_rate": 1.8161781007265958e-05,
+      "loss": 0.415,
+      "step": 860
+    },
+    {
+      "epoch": 0.44131214761660686,
+      "grad_norm": 0.4921904501327304,
+      "learning_rate": 1.815698111906443e-05,
+      "loss": 0.3788,
+      "step": 861
+    },
+    {
+      "epoch": 0.4418247052793439,
+      "grad_norm": 0.5911517274669054,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 0.4244,
+      "step": 862
+    },
+    {
+      "epoch": 0.442337262942081,
+      "grad_norm": 0.5111406802237394,
+      "learning_rate": 1.8147364478572463e-05,
+      "loss": 0.4332,
+      "step": 863
+    },
+    {
+      "epoch": 0.44284982060481803,
+      "grad_norm": 0.5414954608871917,
+      "learning_rate": 1.8142547732910606e-05,
+      "loss": 0.3824,
+      "step": 864
+    },
+    {
+      "epoch": 0.4433623782675551,
+      "grad_norm": 0.6708447534588579,
+      "learning_rate": 1.813772537473035e-05,
+      "loss": 0.3938,
+      "step": 865
+    },
+    {
+      "epoch": 0.44387493593029215,
+      "grad_norm": 0.5165784639169279,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.4164,
+      "step": 866
+    },
+    {
+      "epoch": 0.4443874935930292,
+      "grad_norm": 0.5428609530942549,
+      "learning_rate": 1.812806383411436e-05,
+      "loss": 0.4258,
+      "step": 867
+    },
+    {
+      "epoch": 0.44490005125576626,
+      "grad_norm": 0.544886078797771,
+      "learning_rate": 1.8123224658338166e-05,
+      "loss": 0.421,
+      "step": 868
+    },
+    {
+      "epoch": 0.4454126089185033,
+      "grad_norm": 0.589538637149286,
+      "learning_rate": 1.8118379883362626e-05,
+      "loss": 0.4271,
+      "step": 869
+    },
+    {
+      "epoch": 0.4459251665812404,
+      "grad_norm": 0.5099692704104994,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.3802,
+      "step": 870
+    },
+    {
+      "epoch": 0.44643772424397743,
+      "grad_norm": 0.5309698857943974,
+      "learning_rate": 1.810867354917507e-05,
+      "loss": 0.3796,
+      "step": 871
+    },
+    {
+      "epoch": 0.4469502819067145,
+      "grad_norm": 0.5599668622186548,
+      "learning_rate": 1.8103811996653458e-05,
+      "loss": 0.3893,
+      "step": 872
+    },
+    {
+      "epoch": 0.44746283956945154,
+      "grad_norm": 0.5661966378840165,
+      "learning_rate": 1.809894485831332e-05,
+      "loss": 0.3898,
+      "step": 873
+    },
+    {
+      "epoch": 0.4479753972321886,
+      "grad_norm": 0.5811711580201032,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.3971,
+      "step": 874
+    },
+    {
+      "epoch": 0.44848795489492566,
+      "grad_norm": 0.5686045986879311,
+      "learning_rate": 1.8089193837600653e-05,
+      "loss": 0.3914,
+      "step": 875
+    },
+    {
+      "epoch": 0.4490005125576627,
+      "grad_norm": 0.5183914825772079,
+      "learning_rate": 1.8084309961949338e-05,
+      "loss": 0.3704,
+      "step": 876
+    },
+    {
+      "epoch": 0.44951307022039977,
+      "grad_norm": 0.5605266585770253,
+      "learning_rate": 1.8079420513921913e-05,
+      "loss": 0.4069,
+      "step": 877
+    },
+    {
+      "epoch": 0.4500256278831369,
+      "grad_norm": 0.5395915635630752,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.396,
+      "step": 878
+    },
+    {
+      "epoch": 0.45053818554587394,
+      "grad_norm": 0.5885298913013474,
+      "learning_rate": 1.8069624914223425e-05,
+      "loss": 0.3658,
+      "step": 879
+    },
+    {
+      "epoch": 0.451050743208611,
+      "grad_norm": 0.5289629203771873,
+      "learning_rate": 1.8064718769304304e-05,
+      "loss": 0.3974,
+      "step": 880
+    },
+    {
+      "epoch": 0.45156330087134805,
+      "grad_norm": 0.5181668579785994,
+      "learning_rate": 1.8059807065512945e-05,
+      "loss": 0.3993,
+      "step": 881
+    },
+    {
+      "epoch": 0.4520758585340851,
+      "grad_norm": 0.5088859935129094,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.3526,
+      "step": 882
+    },
+    {
+      "epoch": 0.45258841619682216,
+      "grad_norm": 0.5110127506352814,
+      "learning_rate": 1.8049966994859565e-05,
+      "loss": 0.4224,
+      "step": 883
+    },
+    {
+      "epoch": 0.4531009738595592,
+      "grad_norm": 0.5081586072228838,
+      "learning_rate": 1.8045038634780136e-05,
+      "loss": 0.4024,
+      "step": 884
+    },
+    {
+      "epoch": 0.4536135315222963,
+      "grad_norm": 0.54376535814758,
+      "learning_rate": 1.8040104729393652e-05,
+      "loss": 0.3815,
+      "step": 885
+    },
+    {
+      "epoch": 0.45412608918503333,
+      "grad_norm": 0.5526348643617117,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.4362,
+      "step": 886
+    },
+    {
+      "epoch": 0.4546386468477704,
+      "grad_norm": 0.5333193846459844,
+      "learning_rate": 1.8030220296306752e-05,
+      "loss": 0.3864,
+      "step": 887
+    },
+    {
+      "epoch": 0.45515120451050745,
+      "grad_norm": 0.585576956135994,
+      "learning_rate": 1.802526977541951e-05,
+      "loss": 0.4216,
+      "step": 888
+    },
+    {
+      "epoch": 0.4556637621732445,
+      "grad_norm": 0.5613111073554425,
+      "learning_rate": 1.8020313722851547e-05,
+      "loss": 0.3942,
+      "step": 889
+    },
+    {
+      "epoch": 0.45617631983598156,
+      "grad_norm": 0.5828497064103694,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.4003,
+      "step": 890
+    },
+    {
+      "epoch": 0.4566888774987186,
+      "grad_norm": 0.5759713628274081,
+      "learning_rate": 1.801038503634176e-05,
+      "loss": 0.3892,
+      "step": 891
+    },
+    {
+      "epoch": 0.45720143516145567,
+      "grad_norm": 0.5562786656120058,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.4381,
+      "step": 892
+    },
+    {
+      "epoch": 0.45771399282419273,
+      "grad_norm": 0.5079014912042032,
+      "learning_rate": 1.8000434264152074e-05,
+      "loss": 0.3664,
+      "step": 893
+    },
+    {
+      "epoch": 0.4582265504869298,
+      "grad_norm": 0.5484511544904725,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.4182,
+      "step": 894
+    },
+    {
+      "epoch": 0.45873910814966684,
+      "grad_norm": 0.5145051112101483,
+      "learning_rate": 1.799046143371807e-05,
+      "loss": 0.4061,
+      "step": 895
+    },
+    {
+      "epoch": 0.4592516658124039,
+      "grad_norm": 0.5348969719431181,
+      "learning_rate": 1.7985466755249703e-05,
+      "loss": 0.3822,
+      "step": 896
+    },
+    {
+      "epoch": 0.45976422347514095,
+      "grad_norm": 0.5113561192826979,
+      "learning_rate": 1.798046657253615e-05,
+      "loss": 0.3868,
+      "step": 897
+    },
+    {
+      "epoch": 0.460276781137878,
+      "grad_norm": 0.565048256573343,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.3754,
+      "step": 898
+    },
+    {
+      "epoch": 0.46078933880061507,
+      "grad_norm": 0.552662950715967,
+      "learning_rate": 1.7970449708163455e-05,
+      "loss": 0.413,
+      "step": 899
+    },
+    {
+      "epoch": 0.4613018964633521,
+      "grad_norm": 0.5241796484090322,
+      "learning_rate": 1.796543303340876e-05,
+      "loss": 0.4112,
+      "step": 900
+    },
+    {
+      "epoch": 0.4618144541260892,
+      "grad_norm": 0.602093833380555,
+      "learning_rate": 1.7960410868217786e-05,
+      "loss": 0.3906,
+      "step": 901
+    },
+    {
+      "epoch": 0.46232701178882624,
+      "grad_norm": 0.5806177743165531,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.4089,
+      "step": 902
+    },
+    {
+      "epoch": 0.4628395694515633,
+      "grad_norm": 0.5590288247560901,
+      "learning_rate": 1.7950350080377545e-05,
+      "loss": 0.3564,
+      "step": 903
+    },
+    {
+      "epoch": 0.46335212711430035,
+      "grad_norm": 0.5814015521586882,
+      "learning_rate": 1.7945311464663002e-05,
+      "loss": 0.396,
+      "step": 904
+    },
+    {
+      "epoch": 0.4638646847770374,
+      "grad_norm": 0.6318960275714032,
+      "learning_rate": 1.794026737238163e-05,
+      "loss": 0.3967,
+      "step": 905
+    },
+    {
+      "epoch": 0.46437724243977446,
+      "grad_norm": 0.5540391732609651,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.3882,
+      "step": 906
+    },
+    {
+      "epoch": 0.4648898001025115,
+      "grad_norm": 0.5534008353697359,
+      "learning_rate": 1.79301627720294e-05,
+      "loss": 0.417,
+      "step": 907
+    },
+    {
+      "epoch": 0.4654023577652486,
+      "grad_norm": 0.5522674280230327,
+      "learning_rate": 1.7925102270923465e-05,
+      "loss": 0.3833,
+      "step": 908
+    },
+    {
+      "epoch": 0.46591491542798563,
+      "grad_norm": 0.5036492808662042,
+      "learning_rate": 1.7920036307180554e-05,
+      "loss": 0.3951,
+      "step": 909
+    },
+    {
+      "epoch": 0.4664274730907227,
+      "grad_norm": 0.5621934736151116,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.3662,
+      "step": 910
+    },
+    {
+      "epoch": 0.46694003075345974,
+      "grad_norm": 0.5588700040114184,
+      "learning_rate": 1.7909888005755077e-05,
+      "loss": 0.4107,
+      "step": 911
+    },
+    {
+      "epoch": 0.4674525884161968,
+      "grad_norm": 0.5555050727031499,
+      "learning_rate": 1.7904805675067568e-05,
+      "loss": 0.433,
+      "step": 912
+    },
+    {
+      "epoch": 0.46796514607893386,
+      "grad_norm": 0.5791315359682462,
+      "learning_rate": 1.7899717895733175e-05,
+      "loss": 0.4234,
+      "step": 913
+    },
+    {
+      "epoch": 0.4684777037416709,
+      "grad_norm": 0.5456562146845993,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.3982,
+      "step": 914
+    },
+    {
+      "epoch": 0.46899026140440797,
+      "grad_norm": 0.5545462212402139,
+      "learning_rate": 1.788952600515516e-05,
+      "loss": 0.3897,
+      "step": 915
+    },
+    {
+      "epoch": 0.469502819067145,
+      "grad_norm": 0.5631489665463837,
+      "learning_rate": 1.788442190093664e-05,
+      "loss": 0.3906,
+      "step": 916
+    },
+    {
+      "epoch": 0.47001537672988214,
+      "grad_norm": 0.5515924742672705,
+      "learning_rate": 1.7879312362121414e-05,
+      "loss": 0.4113,
+      "step": 917
+    },
+    {
+      "epoch": 0.4705279343926192,
+      "grad_norm": 0.567316433857189,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.3949,
+      "step": 918
+    },
+    {
+      "epoch": 0.47104049205535625,
+      "grad_norm": 0.5510569794699807,
+      "learning_rate": 1.786907699479229e-05,
+      "loss": 0.4049,
+      "step": 919
+    },
+    {
+      "epoch": 0.4715530497180933,
+      "grad_norm": 0.5865313179638121,
+      "learning_rate": 1.786395117333345e-05,
+      "loss": 0.4156,
+      "step": 920
+    },
+    {
+      "epoch": 0.47206560738083037,
+      "grad_norm": 0.5307641052914804,
+      "learning_rate": 1.785881993138803e-05,
+      "loss": 0.4073,
+      "step": 921
+    },
+    {
+      "epoch": 0.4725781650435674,
+      "grad_norm": 0.5159262172484435,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4303,
+      "step": 922
+    },
+    {
+      "epoch": 0.4730907227063045,
+      "grad_norm": 0.5795957508700574,
+      "learning_rate": 1.7848541200188708e-05,
+      "loss": 0.3903,
+      "step": 923
+    },
+    {
+      "epoch": 0.47360328036904153,
+      "grad_norm": 0.578080160666759,
+      "learning_rate": 1.7843393718019753e-05,
+      "loss": 0.3822,
+      "step": 924
+    },
+    {
+      "epoch": 0.4741158380317786,
+      "grad_norm": 0.5450341484514143,
+      "learning_rate": 1.7838240829534125e-05,
+      "loss": 0.3527,
+      "step": 925
+    },
+    {
+      "epoch": 0.47462839569451565,
+      "grad_norm": 0.48089602092078687,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.3708,
+      "step": 926
+    },
+    {
+      "epoch": 0.4751409533572527,
+      "grad_norm": 0.5405123044261939,
+      "learning_rate": 1.7827918847823746e-05,
+      "loss": 0.3958,
+      "step": 927
+    },
+    {
+      "epoch": 0.47565351101998976,
+      "grad_norm": 0.5413455978106134,
+      "learning_rate": 1.7822749761713767e-05,
+      "loss": 0.3719,
+      "step": 928
+    },
+    {
+      "epoch": 0.4761660686827268,
+      "grad_norm": 0.5302326112630525,
+      "learning_rate": 1.7817575283516637e-05,
+      "loss": 0.4032,
+      "step": 929
+    },
+    {
+      "epoch": 0.4766786263454639,
+      "grad_norm": 0.47861249508516585,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.4156,
+      "step": 930
+    },
+    {
+      "epoch": 0.47719118400820093,
+      "grad_norm": 0.5321847420933574,
+      "learning_rate": 1.7807210165131352e-05,
+      "loss": 0.4028,
+      "step": 931
+    },
+    {
+      "epoch": 0.477703741670938,
+      "grad_norm": 0.5766605559540371,
+      "learning_rate": 1.7802019532087692e-05,
+      "loss": 0.3976,
+      "step": 932
+    },
+    {
+      "epoch": 0.47821629933367504,
+      "grad_norm": 0.6382893532337649,
+      "learning_rate": 1.7796823521245876e-05,
+      "loss": 0.4556,
+      "step": 933
+    },
+    {
+      "epoch": 0.4787288569964121,
+      "grad_norm": 0.510132467559282,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.359,
+      "step": 934
+    },
+    {
+      "epoch": 0.47924141465914916,
+      "grad_norm": 0.5137288619054848,
+      "learning_rate": 1.778641538049755e-05,
+      "loss": 0.3827,
+      "step": 935
+    },
+    {
+      "epoch": 0.4797539723218862,
+      "grad_norm": 0.5631724989541013,
+      "learning_rate": 1.7781203257765197e-05,
+      "loss": 0.3984,
+      "step": 936
+    },
+    {
+      "epoch": 0.48026652998462327,
+      "grad_norm": 0.5023978174684799,
+      "learning_rate": 1.7775985771582986e-05,
+      "loss": 0.3651,
+      "step": 937
+    },
+    {
+      "epoch": 0.4807790876473603,
+      "grad_norm": 0.5034800755462775,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.3669,
+      "step": 938
+    },
+    {
+      "epoch": 0.4812916453100974,
+      "grad_norm": 0.5722652656502255,
+      "learning_rate": 1.7765534723257972e-05,
+      "loss": 0.3992,
+      "step": 939
+    },
+    {
+      "epoch": 0.48180420297283444,
+      "grad_norm": 0.5253986179030253,
+      "learning_rate": 1.7760301168318897e-05,
+      "loss": 0.415,
+      "step": 940
+    },
+    {
+      "epoch": 0.4823167606355715,
+      "grad_norm": 0.565869849609768,
+      "learning_rate": 1.7755062264337412e-05,
+      "loss": 0.4119,
+      "step": 941
+    },
+    {
+      "epoch": 0.48282931829830855,
+      "grad_norm": 0.5277468242360348,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.393,
+      "step": 942
+    },
+    {
+      "epoch": 0.4833418759610456,
+      "grad_norm": 0.5171619954275272,
+      "learning_rate": 1.7744568423695257e-05,
+      "loss": 0.3879,
+      "step": 943
+    },
+    {
+      "epoch": 0.48385443362378266,
+      "grad_norm": 0.5128587090145694,
+      "learning_rate": 1.7739313494267808e-05,
+      "loss": 0.3721,
+      "step": 944
+    },
+    {
+      "epoch": 0.4843669912865197,
+      "grad_norm": 0.5159173222836347,
+      "learning_rate": 1.7734053230264394e-05,
+      "loss": 0.3932,
+      "step": 945
+    },
+    {
+      "epoch": 0.4848795489492568,
+      "grad_norm": 0.5288365318528608,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.3882,
+      "step": 946
+    },
+    {
+      "epoch": 0.48539210661199383,
+      "grad_norm": 0.5321733718983899,
+      "learning_rate": 1.7723516713036588e-05,
+      "loss": 0.3924,
+      "step": 947
+    },
+    {
+      "epoch": 0.4859046642747309,
+      "grad_norm": 0.5156306189380441,
+      "learning_rate": 1.7718240467074833e-05,
+      "loss": 0.4263,
+      "step": 948
+    },
+    {
+      "epoch": 0.48641722193746795,
+      "grad_norm": 0.57173574290157,
+      "learning_rate": 1.7712958901062397e-05,
+      "loss": 0.3938,
+      "step": 949
+    },
+    {
+      "epoch": 0.486929779600205,
+      "grad_norm": 0.48660683480692396,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.4126,
+      "step": 950
+    },
+    {
+      "epoch": 0.48744233726294206,
+      "grad_norm": 0.5559786104360124,
+      "learning_rate": 1.7702379823451087e-05,
+      "loss": 0.424,
+      "step": 951
+    },
+    {
+      "epoch": 0.4879548949256791,
+      "grad_norm": 0.5392520810977104,
+      "learning_rate": 1.769708231914419e-05,
+      "loss": 0.3749,
+      "step": 952
+    },
+    {
+      "epoch": 0.4884674525884162,
+      "grad_norm": 0.5783241917556302,
+      "learning_rate": 1.769177950937056e-05,
+      "loss": 0.41,
+      "step": 953
+    },
+    {
+      "epoch": 0.48898001025115323,
+      "grad_norm": 0.6107674479551688,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.3796,
+      "step": 954
+    },
+    {
+      "epoch": 0.4894925679138903,
+      "grad_norm": 0.5668452211311245,
+      "learning_rate": 1.7681157988047272e-05,
+      "loss": 0.4291,
+      "step": 955
+    },
+    {
+      "epoch": 0.4900051255766274,
+      "grad_norm": 0.5495574913190155,
+      "learning_rate": 1.7675839283818848e-05,
+      "loss": 0.3692,
+      "step": 956
+    },
+    {
+      "epoch": 0.49051768323936445,
+      "grad_norm": 0.5455274276300597,
+      "learning_rate": 1.7670515288766146e-05,
+      "loss": 0.3649,
+      "step": 957
+    },
+    {
+      "epoch": 0.4910302409021015,
+      "grad_norm": 0.5116062843563923,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4018,
+      "step": 958
+    },
+    {
+      "epoch": 0.49154279856483857,
+      "grad_norm": 0.548707453759519,
+      "learning_rate": 1.7659851440870498e-05,
+      "loss": 0.4336,
+      "step": 959
+    },
+    {
+      "epoch": 0.4920553562275756,
+      "grad_norm": 0.5038749946475187,
+      "learning_rate": 1.7654511595377956e-05,
+      "loss": 0.4225,
+      "step": 960
+    },
+    {
+      "epoch": 0.4925679138903127,
+      "grad_norm": 0.5217673828026961,
+      "learning_rate": 1.7649166473761947e-05,
+      "loss": 0.3832,
+      "step": 961
+    },
+    {
+      "epoch": 0.49308047155304974,
+      "grad_norm": 0.5246802360034324,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.3976,
+      "step": 962
+    },
+    {
+      "epoch": 0.4935930292157868,
+      "grad_norm": 0.4783080409300549,
+      "learning_rate": 1.7638460416900355e-05,
+      "loss": 0.3859,
+      "step": 963
+    },
+    {
+      "epoch": 0.49410558687852385,
+      "grad_norm": 0.5356755247847834,
+      "learning_rate": 1.7633099489034272e-05,
+      "loss": 0.376,
+      "step": 964
+    },
+    {
+      "epoch": 0.4946181445412609,
+      "grad_norm": 0.5423069952409768,
+      "learning_rate": 1.7627733299803714e-05,
+      "loss": 0.4132,
+      "step": 965
+    },
+    {
+      "epoch": 0.49513070220399796,
+      "grad_norm": 0.571497789628928,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.4515,
+      "step": 966
+    },
+    {
+      "epoch": 0.495643259866735,
+      "grad_norm": 0.5739004661029871,
+      "learning_rate": 1.7616985152048092e-05,
+      "loss": 0.3839,
+      "step": 967
+    },
+    {
+      "epoch": 0.4961558175294721,
+      "grad_norm": 0.6272180380766607,
+      "learning_rate": 1.761160320093154e-05,
+      "loss": 0.3941,
+      "step": 968
+    },
+    {
+      "epoch": 0.49666837519220913,
+      "grad_norm": 0.5484069507773325,
+      "learning_rate": 1.760621600326754e-05,
+      "loss": 0.3946,
+      "step": 969
+    },
+    {
+      "epoch": 0.4971809328549462,
+      "grad_norm": 0.5437132656612906,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.4419,
+      "step": 970
+    },
+    {
+      "epoch": 0.49769349051768325,
+      "grad_norm": 0.5222388473892028,
+      "learning_rate": 1.759542588315401e-05,
+      "loss": 0.3738,
+      "step": 971
+    },
+    {
+      "epoch": 0.4982060481804203,
+      "grad_norm": 0.5141825986009737,
+      "learning_rate": 1.7590022968141928e-05,
+      "loss": 0.4025,
+      "step": 972
+    },
+    {
+      "epoch": 0.49871860584315736,
+      "grad_norm": 0.4805093833885646,
+      "learning_rate": 1.7584614821457275e-05,
+      "loss": 0.363,
+      "step": 973
+    },
+    {
+      "epoch": 0.4992311635058944,
+      "grad_norm": 0.4890742581087118,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.4022,
+      "step": 974
+    },
+    {
+      "epoch": 0.49974372116863147,
+      "grad_norm": 0.5112811140712663,
+      "learning_rate": 1.7573782847984848e-05,
+      "loss": 0.4036,
+      "step": 975
+    },
+    {
+      "epoch": 0.5002562788313686,
+      "grad_norm": 0.5053847693492911,
+      "learning_rate": 1.7568359028663365e-05,
+      "loss": 0.3816,
+      "step": 976
+    },
+    {
+      "epoch": 0.5007688364941056,
+      "grad_norm": 0.5570798698164371,
+      "learning_rate": 1.7562929992601895e-05,
+      "loss": 0.4337,
+      "step": 977
+    },
+    {
+      "epoch": 0.5012813941568427,
+      "grad_norm": 0.5501717531585665,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.3794,
+      "step": 978
+    },
+    {
+      "epoch": 0.5017939518195798,
+      "grad_norm": 0.5154269108789272,
+      "learning_rate": 1.7552056285231166e-05,
+      "loss": 0.4028,
+      "step": 979
+    },
+    {
+      "epoch": 0.5023065094823168,
+      "grad_norm": 0.536701110497156,
+      "learning_rate": 1.754661162141696e-05,
+      "loss": 0.4546,
+      "step": 980
+    },
+    {
+      "epoch": 0.5028190671450539,
+      "grad_norm": 0.5379302177897728,
+      "learning_rate": 1.7541161755852884e-05,
+      "loss": 0.4105,
+      "step": 981
+    },
+    {
+      "epoch": 0.5033316248077909,
+      "grad_norm": 0.5124172874104795,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.3924,
+      "step": 982
+    },
+    {
+      "epoch": 0.503844182470528,
+      "grad_norm": 0.7763047244903555,
+      "learning_rate": 1.75302464345047e-05,
+      "loss": 0.3763,
+      "step": 983
+    },
+    {
+      "epoch": 0.504356740133265,
+      "grad_norm": 0.5521282061173882,
+      "learning_rate": 1.7524780986244334e-05,
+      "loss": 0.4154,
+      "step": 984
+    },
+    {
+      "epoch": 0.5048692977960021,
+      "grad_norm": 0.6249124381069038,
+      "learning_rate": 1.751931035128158e-05,
+      "loss": 0.4329,
+      "step": 985
+    },
+    {
+      "epoch": 0.5053818554587391,
+      "grad_norm": 0.5515378956555349,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.3558,
+      "step": 986
+    },
+    {
+      "epoch": 0.5058944131214762,
+      "grad_norm": 0.5371032163209557,
+      "learning_rate": 1.750835353633574e-05,
+      "loss": 0.4313,
+      "step": 987
+    },
+    {
+      "epoch": 0.5064069707842133,
+      "grad_norm": 0.49715531694793696,
+      "learning_rate": 1.7502867363905e-05,
+      "loss": 0.3872,
+      "step": 988
+    },
+    {
+      "epoch": 0.5069195284469503,
+      "grad_norm": 0.5861100444366073,
+      "learning_rate": 1.749737601987655e-05,
+      "loss": 0.4469,
+      "step": 989
+    },
+    {
+      "epoch": 0.5074320861096874,
+      "grad_norm": 0.565221122679931,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.4203,
+      "step": 990
+    },
+    {
+      "epoch": 0.5079446437724244,
+      "grad_norm": 0.5502757906774514,
+      "learning_rate": 1.7486377832170463e-05,
+      "loss": 0.3965,
+      "step": 991
+    },
+    {
+      "epoch": 0.5084572014351615,
+      "grad_norm": 0.5749767361176171,
+      "learning_rate": 1.7480870996073682e-05,
+      "loss": 0.3903,
+      "step": 992
+    },
+    {
+      "epoch": 0.5089697590978985,
+      "grad_norm": 0.5670545096848201,
+      "learning_rate": 1.7475359003540916e-05,
+      "loss": 0.401,
+      "step": 993
+    },
+    {
+      "epoch": 0.5094823167606356,
+      "grad_norm": 0.5189511452329334,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.3804,
+      "step": 994
+    },
+    {
+      "epoch": 0.5099948744233727,
+      "grad_norm": 0.5162489732609535,
+      "learning_rate": 1.7464319564368268e-05,
+      "loss": 0.4108,
+      "step": 995
+    },
+    {
+      "epoch": 0.5105074320861097,
+      "grad_norm": 0.5226855220944094,
+      "learning_rate": 1.745879212533768e-05,
+      "loss": 0.4016,
+      "step": 996
+    },
+    {
+      "epoch": 0.5110199897488468,
+      "grad_norm": 0.5458723563873522,
+      "learning_rate": 1.7453259545089695e-05,
+      "loss": 0.3954,
+      "step": 997
+    },
+    {
+      "epoch": 0.5115325474115838,
+      "grad_norm": 0.49772786826425675,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.3985,
+      "step": 998
+    },
+    {
+      "epoch": 0.5120451050743209,
+      "grad_norm": 0.5175662846575434,
+      "learning_rate": 1.7442178976199112e-05,
+      "loss": 0.3713,
+      "step": 999
+    },
+    {
+      "epoch": 0.5125576627370579,
+      "grad_norm": 0.5454436983851327,
+      "learning_rate": 1.743663099519417e-05,
+      "loss": 0.429,
+      "step": 1000
+    },
+    {
+      "epoch": 0.513070220399795,
+      "grad_norm": 0.5662048962577753,
+      "learning_rate": 1.7431077888247108e-05,
+      "loss": 0.383,
+      "step": 1001
+    },
+    {
+      "epoch": 0.513582778062532,
+      "grad_norm": 0.5268501555002075,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.4309,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5140953357252691,
+      "grad_norm": 0.5420276664552458,
+      "learning_rate": 1.741995631184083e-05,
+      "loss": 0.3606,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5146078933880062,
+      "grad_norm": 0.552396557722035,
+      "learning_rate": 1.7414387850047522e-05,
+      "loss": 0.413,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5151204510507432,
+      "grad_norm": 0.5245224867288794,
+      "learning_rate": 1.740881427764392e-05,
+      "loss": 0.4347,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5156330087134803,
+      "grad_norm": 0.524689109833673,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.3743,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5161455663762173,
+      "grad_norm": 0.5550858480704632,
+      "learning_rate": 1.7397651816376423e-05,
+      "loss": 0.398,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5166581240389544,
+      "grad_norm": 0.595098522576283,
+      "learning_rate": 1.739206293520662e-05,
+      "loss": 0.359,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5171706817016914,
+      "grad_norm": 0.49559870681243984,
+      "learning_rate": 1.7386468958814706e-05,
+      "loss": 0.4148,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5176832393644285,
+      "grad_norm": 0.5676641449672227,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.4201,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5181957970271656,
+      "grad_norm": 0.5249022370208702,
+      "learning_rate": 1.7375265735791372e-05,
+      "loss": 0.3911,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5187083546899026,
+      "grad_norm": 0.5679932539570529,
+      "learning_rate": 1.736965649688214e-05,
+      "loss": 0.3939,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5192209123526397,
+      "grad_norm": 0.5114631232083119,
+      "learning_rate": 1.7364042178195176e-05,
+      "loss": 0.3934,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5197334700153767,
+      "grad_norm": 0.5782477191175208,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.4346,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5202460276781138,
+      "grad_norm": 0.5580082152170164,
+      "learning_rate": 1.735279831697093e-05,
+      "loss": 0.3997,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5207585853408508,
+      "grad_norm": 0.539794926736742,
+      "learning_rate": 1.7347168782183852e-05,
+      "loss": 0.4247,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5212711430035879,
+      "grad_norm": 0.564982913390649,
+      "learning_rate": 1.734153418311943e-05,
+      "loss": 0.4138,
+      "step": 1017
+    },
+    {
+      "epoch": 0.521783700666325,
+      "grad_norm": 0.5416814189266306,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.3644,
+      "step": 1018
+    },
+    {
+      "epoch": 0.522296258329062,
+      "grad_norm": 0.5633084971177932,
+      "learning_rate": 1.7330249807697374e-05,
+      "loss": 0.374,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5228088159917991,
+      "grad_norm": 0.5645876170393581,
+      "learning_rate": 1.7324600039117862e-05,
+      "loss": 0.4165,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5233213736545361,
+      "grad_norm": 0.6137078187336124,
+      "learning_rate": 1.7318945221817255e-05,
+      "loss": 0.4166,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5238339313172732,
+      "grad_norm": 0.5978134263880125,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.3914,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5243464889800102,
+      "grad_norm": 0.5464502279034424,
+      "learning_rate": 1.73076204566473e-05,
+      "loss": 0.3888,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5248590466427473,
+      "grad_norm": 0.5341309207344751,
+      "learning_rate": 1.7301950516583922e-05,
+      "loss": 0.3839,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5253716043054844,
+      "grad_norm": 0.5559685099700229,
+      "learning_rate": 1.729627554341137e-05,
+      "loss": 0.4021,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5258841619682214,
+      "grad_norm": 0.514849083753593,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.4131,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5263967196309585,
+      "grad_norm": 0.5325573417481815,
+      "learning_rate": 1.728491051338887e-05,
+      "loss": 0.4161,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5269092772936955,
+      "grad_norm": 0.5251036626125575,
+      "learning_rate": 1.7279220464372644e-05,
+      "loss": 0.4108,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5274218349564326,
+      "grad_norm": 0.5588747997857749,
+      "learning_rate": 1.7273525397914677e-05,
+      "loss": 0.4023,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5279343926191696,
+      "grad_norm": 0.529392665235331,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.4034,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5284469502819067,
+      "grad_norm": 0.5962675897898403,
+      "learning_rate": 1.7262120228379053e-05,
+      "loss": 0.4172,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5289595079446437,
+      "grad_norm": 0.5318666684802256,
+      "learning_rate": 1.725641013316277e-05,
+      "loss": 0.3825,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5294720656073808,
+      "grad_norm": 0.5712848918157315,
+      "learning_rate": 1.725069503622753e-05,
+      "loss": 0.4207,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5299846232701179,
+      "grad_norm": 0.5134761793702094,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.3804,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5304971809328549,
+      "grad_norm": 0.5737614167608907,
+      "learning_rate": 1.7239249852960863e-05,
+      "loss": 0.4207,
+      "step": 1035
+    },
+    {
+      "epoch": 0.531009738595592,
+      "grad_norm": 0.5392854813795418,
+      "learning_rate": 1.723351977451841e-05,
+      "loss": 0.3526,
+      "step": 1036
+    },
+    {
+      "epoch": 0.531522296258329,
+      "grad_norm": 0.5310210678805782,
+      "learning_rate": 1.7227784710134924e-05,
+      "loss": 0.3855,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5320348539210661,
+      "grad_norm": 0.5247829366570502,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.3813,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5325474115838031,
+      "grad_norm": 0.49805282588154315,
+      "learning_rate": 1.7216299639360605e-05,
+      "loss": 0.408,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5330599692465402,
+      "grad_norm": 0.5176943944331351,
+      "learning_rate": 1.7210549640886237e-05,
+      "loss": 0.3693,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5335725269092773,
+      "grad_norm": 0.5461801319736937,
+      "learning_rate": 1.7204794672303757e-05,
+      "loss": 0.4346,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5340850845720143,
+      "grad_norm": 0.5513009216483826,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 0.4047,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5345976422347514,
+      "grad_norm": 0.5371690745565119,
+      "learning_rate": 1.7193269840685074e-05,
+      "loss": 0.3996,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5351101998974884,
+      "grad_norm": 0.5051211583386515,
+      "learning_rate": 1.7187499985592743e-05,
+      "loss": 0.3637,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5356227575602255,
+      "grad_norm": 0.5380119807935765,
+      "learning_rate": 1.7181725176280028e-05,
+      "loss": 0.4472,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5361353152229625,
+      "grad_norm": 0.4868389115549939,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.3438,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5366478728856996,
+      "grad_norm": 0.5662004216193255,
+      "learning_rate": 1.717016071091877e-05,
+      "loss": 0.4595,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5371604305484367,
+      "grad_norm": 0.4911620639323809,
+      "learning_rate": 1.716437106284141e-05,
+      "loss": 0.4117,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5376729882111737,
+      "grad_norm": 0.5152686082246636,
+      "learning_rate": 1.715857647648604e-05,
+      "loss": 0.3987,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5381855458739108,
+      "grad_norm": 0.5305932570719928,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.3407,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5386981035366478,
+      "grad_norm": 0.5338134245112978,
+      "learning_rate": 1.7146972504921097e-05,
+      "loss": 0.4017,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5392106611993849,
+      "grad_norm": 0.5426552759662422,
+      "learning_rate": 1.714116312770994e-05,
+      "loss": 0.4337,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5397232188621219,
+      "grad_norm": 0.5497750181530805,
+      "learning_rate": 1.71353488282176e-05,
+      "loss": 0.4266,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5402357765248591,
+      "grad_norm": 0.5341392792654391,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.3538,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5407483341875962,
+      "grad_norm": 0.5731984587741572,
+      "learning_rate": 1.7123705478423552e-05,
+      "loss": 0.3916,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5412608918503332,
+      "grad_norm": 0.5305365493289652,
+      "learning_rate": 1.7117876436147404e-05,
+      "loss": 0.3743,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5417734495130703,
+      "grad_norm": 0.5340940417154323,
+      "learning_rate": 1.711204248764119e-05,
+      "loss": 0.4138,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5422860071758073,
+      "grad_norm": 0.5133163867986441,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.4135,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5427985648385444,
+      "grad_norm": 0.5402989775053251,
+      "learning_rate": 1.710035988802691e-05,
+      "loss": 0.3616,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5433111225012814,
+      "grad_norm": 0.549230590524337,
+      "learning_rate": 1.7094511244971456e-05,
+      "loss": 0.3914,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5438236801640185,
+      "grad_norm": 0.5575631893141015,
+      "learning_rate": 1.708865771179116e-05,
+      "loss": 0.3778,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5443362378267556,
+      "grad_norm": 0.584071750762465,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.4017,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5448487954894926,
+      "grad_norm": 0.5009961678303089,
+      "learning_rate": 1.7076935991198383e-05,
+      "loss": 0.4137,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5453613531522297,
+      "grad_norm": 0.5911758825913297,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.4324,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5458739108149667,
+      "grad_norm": 0.5299619491912527,
+      "learning_rate": 1.706519475856689e-05,
+      "loss": 0.3951,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5463864684777038,
+      "grad_norm": 0.49466513993622124,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.4243,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5468990261404408,
+      "grad_norm": 0.5051809356634179,
+      "learning_rate": 1.7053434046268793e-05,
+      "loss": 0.3888,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5474115838031779,
+      "grad_norm": 0.5154709468189551,
+      "learning_rate": 1.704754639537574e-05,
+      "loss": 0.3612,
+      "step": 1068
+    },
+    {
+      "epoch": 0.547924141465915,
+      "grad_norm": 0.5335905132951716,
+      "learning_rate": 1.704165388672992e-05,
+      "loss": 0.3577,
+      "step": 1069
+    },
+    {
+      "epoch": 0.548436699128652,
+      "grad_norm": 0.5197777302175073,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.4128,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5489492567913891,
+      "grad_norm": 0.5210413402642247,
+      "learning_rate": 1.7029854312429707e-05,
+      "loss": 0.3971,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5494618144541261,
+      "grad_norm": 0.5014321808379781,
+      "learning_rate": 1.7023947254908565e-05,
+      "loss": 0.3526,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5499743721168632,
+      "grad_norm": 0.5264831475375268,
+      "learning_rate": 1.701803535590113e-05,
+      "loss": 0.3881,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5504869297796002,
+      "grad_norm": 0.5258238525517556,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.3971,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5509994874423373,
+      "grad_norm": 0.5321459207119562,
+      "learning_rate": 1.70061970497306e-05,
+      "loss": 0.3863,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5515120451050743,
+      "grad_norm": 0.5429078655672077,
+      "learning_rate": 1.7000270650727446e-05,
+      "loss": 0.3539,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5520246027678114,
+      "grad_norm": 0.5336786275362252,
+      "learning_rate": 1.6994339426557872e-05,
+      "loss": 0.3678,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5525371604305485,
+      "grad_norm": 0.5141716081592486,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.4073,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5530497180932855,
+      "grad_norm": 0.5263653534688456,
+      "learning_rate": 1.6982462519075968e-05,
+      "loss": 0.3762,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5535622757560226,
+      "grad_norm": 0.515748519836586,
+      "learning_rate": 1.6976516843950187e-05,
+      "loss": 0.4171,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5540748334187596,
+      "grad_norm": 0.5345237111917727,
+      "learning_rate": 1.6970566360031077e-05,
+      "loss": 0.3982,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5545873910814967,
+      "grad_norm": 0.535305869138514,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.375,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5550999487442337,
+      "grad_norm": 0.5037417063954857,
+      "learning_rate": 1.6958650982222466e-05,
+      "loss": 0.3998,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5556125064069708,
+      "grad_norm": 0.5298492223703933,
+      "learning_rate": 1.695268609654602e-05,
+      "loss": 0.3952,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5561250640697079,
+      "grad_norm": 0.5363555447250289,
+      "learning_rate": 1.6946716418502386e-05,
+      "loss": 0.4287,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5566376217324449,
+      "grad_norm": 0.538682617509214,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.4201,
+      "step": 1086
+    },
+    {
+      "epoch": 0.557150179395182,
+      "grad_norm": 0.5492075785038313,
+      "learning_rate": 1.693476270177599e-05,
+      "loss": 0.4154,
+      "step": 1087
+    },
+    {
+      "epoch": 0.557662737057919,
+      "grad_norm": 0.5320603937598226,
+      "learning_rate": 1.692877867133273e-05,
+      "loss": 0.3711,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5581752947206561,
+      "grad_norm": 0.5335263226994902,
+      "learning_rate": 1.6922789865001254e-05,
+      "loss": 0.4281,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5586878523833931,
+      "grad_norm": 0.5344827745827052,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.3839,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5592004100461302,
+      "grad_norm": 0.5303324880865654,
+      "learning_rate": 1.6910797941188838e-05,
+      "loss": 0.3812,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5597129677088672,
+      "grad_norm": 0.49498782133072,
+      "learning_rate": 1.6904794831973733e-05,
+      "loss": 0.3467,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5602255253716043,
+      "grad_norm": 0.5146834563904041,
+      "learning_rate": 1.6898786963402057e-05,
+      "loss": 0.3762,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5607380830343414,
+      "grad_norm": 0.5476080815782917,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.36,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5612506406970784,
+      "grad_norm": 0.5608706466206013,
+      "learning_rate": 1.6886756964756743e-05,
+      "loss": 0.4214,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5617631983598155,
+      "grad_norm": 0.5557358156554991,
+      "learning_rate": 1.688073484297518e-05,
+      "loss": 0.3976,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5622757560225525,
+      "grad_norm": 0.5522313944560907,
+      "learning_rate": 1.687470797842118e-05,
+      "loss": 0.3738,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5627883136852896,
+      "grad_norm": 0.5673485888824737,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.4335,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5633008713480266,
+      "grad_norm": 0.4729209205361093,
+      "learning_rate": 1.6862640037616e-05,
+      "loss": 0.3551,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5638134290107637,
+      "grad_norm": 0.5504929907658597,
+      "learning_rate": 1.6856598969683045e-05,
+      "loss": 0.4207,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5643259866735008,
+      "grad_norm": 0.4942638827918055,
+      "learning_rate": 1.6850553175614093e-05,
+      "loss": 0.3903,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5648385443362378,
+      "grad_norm": 0.612521522471951,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.3788,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5653511019989749,
+      "grad_norm": 0.5495787335144179,
+      "learning_rate": 1.6838447425740524e-05,
+      "loss": 0.4408,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5658636596617119,
+      "grad_norm": 0.5081111364206597,
+      "learning_rate": 1.6832387478280185e-05,
+      "loss": 0.3894,
+      "step": 1104
+    },
+    {
+      "epoch": 0.566376217324449,
+      "grad_norm": 0.5226868931156972,
+      "learning_rate": 1.682632282137242e-05,
+      "loss": 0.3851,
+      "step": 1105
+    },
+    {
+      "epoch": 0.566888774987186,
+      "grad_norm": 0.5502399108984682,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.4281,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5674013326499231,
+      "grad_norm": 0.49962493977151723,
+      "learning_rate": 1.6814179395938915e-05,
+      "loss": 0.3884,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5679138903126602,
+      "grad_norm": 0.62144690846313,
+      "learning_rate": 1.680810063578342e-05,
+      "loss": 0.3607,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5684264479753972,
+      "grad_norm": 0.5067301962473104,
+      "learning_rate": 1.680201718292101e-05,
+      "loss": 0.3331,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5689390056381343,
+      "grad_norm": 0.5300382019026171,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.3651,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5694515633008713,
+      "grad_norm": 0.5077739839883454,
+      "learning_rate": 1.6789836215851525e-05,
+      "loss": 0.3947,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5699641209636084,
+      "grad_norm": 0.49916445779501656,
+      "learning_rate": 1.6783738710040588e-05,
+      "loss": 0.3956,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5704766786263454,
+      "grad_norm": 0.5102266048796649,
+      "learning_rate": 1.677763652831498e-05,
+      "loss": 0.3811,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5709892362890825,
+      "grad_norm": 0.47663258928212615,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.3791,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5715017939518195,
+      "grad_norm": 0.5039341333058842,
+      "learning_rate": 1.676541815394751e-05,
+      "loss": 0.3225,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5720143516145566,
+      "grad_norm": 0.4651210244429522,
+      "learning_rate": 1.675930196972756e-05,
+      "loss": 0.3487,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5725269092772937,
+      "grad_norm": 0.5059235314525273,
+      "learning_rate": 1.675318112643677e-05,
+      "loss": 0.3831,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5730394669400307,
+      "grad_norm": 0.5188994155831148,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.3374,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5735520246027678,
+      "grad_norm": 0.4897052607392741,
+      "learning_rate": 1.6740925479521844e-05,
+      "loss": 0.3732,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5740645822655048,
+      "grad_norm": 0.4612502027491828,
+      "learning_rate": 1.6734790684345322e-05,
+      "loss": 0.3815,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5745771399282419,
+      "grad_norm": 0.5013986597825788,
+      "learning_rate": 1.6728651246993165e-05,
+      "loss": 0.3482,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5750896975909789,
+      "grad_norm": 0.5218055061924448,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.3977,
+      "step": 1122
+    },
+    {
+      "epoch": 0.575602255253716,
+      "grad_norm": 0.4959030803859426,
+      "learning_rate": 1.671635846269239e-05,
+      "loss": 0.3745,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5761148129164531,
+      "grad_norm": 0.49884757756232834,
+      "learning_rate": 1.6710205124216973e-05,
+      "loss": 0.4064,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5766273705791901,
+      "grad_norm": 0.4943700093728223,
+      "learning_rate": 1.6704047160512333e-05,
+      "loss": 0.3828,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5771399282419272,
+      "grad_norm": 0.5162643194042498,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.4032,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5776524859046642,
+      "grad_norm": 0.5346174105734396,
+      "learning_rate": 1.669171737439687e-05,
+      "loss": 0.3928,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5781650435674013,
+      "grad_norm": 0.5059568853429284,
+      "learning_rate": 1.668554556048476e-05,
+      "loss": 0.4135,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5786776012301383,
+      "grad_norm": 0.5229161127766554,
+      "learning_rate": 1.6679369138340833e-05,
+      "loss": 0.3903,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5791901588928754,
+      "grad_norm": 0.5068576083937552,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.4076,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5797027165556125,
+      "grad_norm": 0.5181525152978642,
+      "learning_rate": 1.6667002486389924e-05,
+      "loss": 0.3615,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5802152742183496,
+      "grad_norm": 0.5454680562107808,
+      "learning_rate": 1.666081226510706e-05,
+      "loss": 0.3441,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5807278318810867,
+      "grad_norm": 0.5187398681980611,
+      "learning_rate": 1.6654617452640622e-05,
+      "loss": 0.3813,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5812403895438237,
+      "grad_norm": 0.5317672957413611,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.3799,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5817529472065608,
+      "grad_norm": 0.5452972814391692,
+      "learning_rate": 1.6642214071240085e-05,
+      "loss": 0.3709,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5822655048692978,
+      "grad_norm": 0.5250924255153342,
+      "learning_rate": 1.6636005510855425e-05,
+      "loss": 0.3534,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5827780625320349,
+      "grad_norm": 0.49044277119868973,
+      "learning_rate": 1.662979237638606e-05,
+      "loss": 0.3447,
+      "step": 1137
+    },
+    {
+      "epoch": 0.583290620194772,
+      "grad_norm": 0.542275764427163,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.3784,
+      "step": 1138
+    },
+    {
+      "epoch": 0.583803177857509,
+      "grad_norm": 0.5834443767937451,
+      "learning_rate": 1.661735240232679e-05,
+      "loss": 0.4215,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5843157355202461,
+      "grad_norm": 0.5279814421471729,
+      "learning_rate": 1.661112557131154e-05,
+      "loss": 0.3759,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5848282931829831,
+      "grad_norm": 0.502213023291551,
+      "learning_rate": 1.66048941833609e-05,
+      "loss": 0.3729,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5853408508457202,
+      "grad_norm": 0.5424908520102628,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.3646,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5858534085084572,
+      "grad_norm": 0.4520333891863473,
+      "learning_rate": 1.6592417753837338e-05,
+      "loss": 0.399,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5863659661711943,
+      "grad_norm": 0.452375279692537,
+      "learning_rate": 1.6586172720864206e-05,
+      "loss": 0.3582,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5868785238339314,
+      "grad_norm": 0.5397573117612195,
+      "learning_rate": 1.657992314815525e-05,
+      "loss": 0.3831,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5873910814966684,
+      "grad_norm": 0.5551574343024803,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.3833,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5879036391594055,
+      "grad_norm": 0.5133706468828918,
+      "learning_rate": 1.6567410400763902e-05,
+      "loss": 0.3582,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5884161968221425,
+      "grad_norm": 0.4821894092989885,
+      "learning_rate": 1.6561147234706333e-05,
+      "loss": 0.355,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5889287544848796,
+      "grad_norm": 0.5166775871562328,
+      "learning_rate": 1.6554879546162578e-05,
+      "loss": 0.3834,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5894413121476166,
+      "grad_norm": 0.5452101164729736,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.3616,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5899538698103537,
+      "grad_norm": 0.540195798651443,
+      "learning_rate": 1.654233061890048e-05,
+      "loss": 0.3874,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5904664274730907,
+      "grad_norm": 0.4995456612248197,
+      "learning_rate": 1.6536049388831897e-05,
+      "loss": 0.3604,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5909789851358278,
+      "grad_norm": 0.5424823783508574,
+      "learning_rate": 1.652976365357664e-05,
+      "loss": 0.3756,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5914915427985649,
+      "grad_norm": 0.5315310089100521,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.4055,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5920041004613019,
+      "grad_norm": 0.4891106513486986,
+      "learning_rate": 1.6517178684839834e-05,
+      "loss": 0.3551,
+      "step": 1155
+    },
+    {
+      "epoch": 0.592516658124039,
+      "grad_norm": 0.536958859715879,
+      "learning_rate": 1.651087946003288e-05,
+      "loss": 0.403,
+      "step": 1156
+    },
+    {
+      "epoch": 0.593029215786776,
+      "grad_norm": 0.5584729222354471,
+      "learning_rate": 1.6504575747388462e-05,
+      "loss": 0.3867,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5935417734495131,
+      "grad_norm": 0.5267732520111604,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.3892,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5940543311122501,
+      "grad_norm": 0.524380968665154,
+      "learning_rate": 1.6491954875970474e-05,
+      "loss": 0.3675,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5945668887749872,
+      "grad_norm": 0.5253225394407958,
+      "learning_rate": 1.6485637725896256e-05,
+      "loss": 0.3952,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5950794464377243,
+      "grad_norm": 0.4983643568246453,
+      "learning_rate": 1.6479316105383266e-05,
+      "loss": 0.3859,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5955920041004613,
+      "grad_norm": 0.4958896785376566,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.3853,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5961045617631984,
+      "grad_norm": 0.5092076340168136,
+      "learning_rate": 1.646665947047358e-05,
+      "loss": 0.4082,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5966171194259354,
+      "grad_norm": 0.5424245213683493,
+      "learning_rate": 1.6460324464800884e-05,
+      "loss": 0.3648,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5971296770886725,
+      "grad_norm": 0.5163830183308601,
+      "learning_rate": 1.6453985006137412e-05,
+      "loss": 0.3745,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5976422347514095,
+      "grad_norm": 0.4897241629005879,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.4025,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5981547924141466,
+      "grad_norm": 0.5372082666894953,
+      "learning_rate": 1.6441292747319923e-05,
+      "loss": 0.4207,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5986673500768837,
+      "grad_norm": 0.4779122639658493,
+      "learning_rate": 1.6434939955914466e-05,
+      "loss": 0.3916,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5991799077396207,
+      "grad_norm": 0.48045183273313896,
+      "learning_rate": 1.6428582729015333e-05,
+      "loss": 0.3963,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5996924654023578,
+      "grad_norm": 0.5317039238965953,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.38,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6002050230650948,
+      "grad_norm": 0.5179961637538172,
+      "learning_rate": 1.6415854986266824e-05,
+      "loss": 0.4212,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6007175807278319,
+      "grad_norm": 0.5416042081744197,
+      "learning_rate": 1.6409484479190457e-05,
+      "loss": 0.4151,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6012301383905689,
+      "grad_norm": 0.4905670085994225,
+      "learning_rate": 1.6403109554166447e-05,
+      "loss": 0.3655,
+      "step": 1173
+    },
+    {
+      "epoch": 0.601742696053306,
+      "grad_norm": 0.5107573069619062,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.4218,
+      "step": 1174
+    },
+    {
+      "epoch": 0.602255253716043,
+      "grad_norm": 0.512783253486487,
+      "learning_rate": 1.6390346467855027e-05,
+      "loss": 0.3694,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6027678113787801,
+      "grad_norm": 0.5440123972200835,
+      "learning_rate": 1.6383958315365e-05,
+      "loss": 0.3899,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6032803690415172,
+      "grad_norm": 0.5111047693855191,
+      "learning_rate": 1.6377565762522066e-05,
+      "loss": 0.3957,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6037929267042542,
+      "grad_norm": 0.47706169544518107,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.3933,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6043054843669913,
+      "grad_norm": 0.5150196956170625,
+      "learning_rate": 1.636476747340564e-05,
+      "loss": 0.3833,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6048180420297283,
+      "grad_norm": 0.5105602771513166,
+      "learning_rate": 1.6358361745953787e-05,
+      "loss": 0.3687,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6053305996924654,
+      "grad_norm": 0.4981761858249404,
+      "learning_rate": 1.6351951635792306e-05,
+      "loss": 0.3921,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6058431573552024,
+      "grad_norm": 0.5294546509639746,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.4212,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6063557150179395,
+      "grad_norm": 0.4996463203510194,
+      "learning_rate": 1.6339118285017005e-05,
+      "loss": 0.3473,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6068682726806766,
+      "grad_norm": 0.512453582924902,
+      "learning_rate": 1.633269505324899e-05,
+      "loss": 0.3998,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6073808303434136,
+      "grad_norm": 0.5110521396136901,
+      "learning_rate": 1.6326267456462965e-05,
+      "loss": 0.3561,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6078933880061507,
+      "grad_norm": 0.5750287013074954,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.4029,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6084059456688877,
+      "grad_norm": 0.518543008394888,
+      "learning_rate": 1.63133991855616e-05,
+      "loss": 0.3751,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6089185033316248,
+      "grad_norm": 0.5525667023888107,
+      "learning_rate": 1.6306958520316144e-05,
+      "loss": 0.4399,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6094310609943618,
+      "grad_norm": 0.5178422881325506,
+      "learning_rate": 1.6300513507792426e-05,
+      "loss": 0.3748,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6099436186570989,
+      "grad_norm": 0.5121882818129039,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.3715,
+      "step": 1190
+    },
+    {
+      "epoch": 0.610456176319836,
+      "grad_norm": 0.4888814924775512,
+      "learning_rate": 1.628761045868293e-05,
+      "loss": 0.3714,
+      "step": 1191
+    },
+    {
+      "epoch": 0.610968733982573,
+      "grad_norm": 0.5371936156473308,
+      "learning_rate": 1.6281152430991012e-05,
+      "loss": 0.4056,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6114812916453101,
+      "grad_norm": 0.530312855900763,
+      "learning_rate": 1.627469007380852e-05,
+      "loss": 0.3838,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6119938493080471,
+      "grad_norm": 0.52242391921988,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.4084,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6125064069707842,
+      "grad_norm": 0.553161074429206,
+      "learning_rate": 1.626175238879238e-05,
+      "loss": 0.3865,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6130189646335212,
+      "grad_norm": 0.5400993491378621,
+      "learning_rate": 1.6255277069876454e-05,
+      "loss": 0.4045,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6135315222962583,
+      "grad_norm": 0.5416506639918034,
+      "learning_rate": 1.6248797439305403e-05,
+      "loss": 0.4066,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6140440799589953,
+      "grad_norm": 0.5704599236645067,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.3554,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6145566376217324,
+      "grad_norm": 0.5081278984447437,
+      "learning_rate": 1.6235825261066078e-05,
+      "loss": 0.342,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6150691952844695,
+      "grad_norm": 0.5383571106281164,
+      "learning_rate": 1.6229332722339308e-05,
+      "loss": 0.3755,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6155817529472065,
+      "grad_norm": 0.5434318260071183,
+      "learning_rate": 1.6222835889840402e-05,
+      "loss": 0.3738,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6160943106099436,
+      "grad_norm": 0.5252727894180629,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.3596,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6166068682726806,
+      "grad_norm": 0.4858060799823988,
+      "learning_rate": 1.6209829361441767e-05,
+      "loss": 0.4078,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6171194259354177,
+      "grad_norm": 0.4667507886062558,
+      "learning_rate": 1.6203319674507216e-05,
+      "loss": 0.3825,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6176319835981547,
+      "grad_norm": 0.5247809391790367,
+      "learning_rate": 1.619680571173088e-05,
+      "loss": 0.3968,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6181445412608918,
+      "grad_norm": 0.501101643796346,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.3452,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6186570989236289,
+      "grad_norm": 0.5349870621485956,
+      "learning_rate": 1.618376497661564e-05,
+      "loss": 0.407,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6191696565863659,
+      "grad_norm": 0.5204807710635859,
+      "learning_rate": 1.6177238213265497e-05,
+      "loss": 0.391,
+      "step": 1208
+    },
+    {
+      "epoch": 0.619682214249103,
+      "grad_norm": 0.5462491096776055,
+      "learning_rate": 1.6170707192051067e-05,
+      "loss": 0.3917,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6201947719118401,
+      "grad_norm": 0.4902673270757968,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4198,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6207073295745772,
+      "grad_norm": 0.5302501426003751,
+      "learning_rate": 1.615763239403918e-05,
+      "loss": 0.3679,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6212198872373142,
+      "grad_norm": 0.4805279607937078,
+      "learning_rate": 1.6151088626253954e-05,
+      "loss": 0.3798,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6217324449000513,
+      "grad_norm": 0.48534440862134,
+      "learning_rate": 1.6144540618628907e-05,
+      "loss": 0.337,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6222450025627884,
+      "grad_norm": 0.5162626159422048,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.3659,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6227575602255254,
+      "grad_norm": 0.5184134181954713,
+      "learning_rate": 1.6131431901915987e-05,
+      "loss": 0.385,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6232701178882625,
+      "grad_norm": 0.5284547457941826,
+      "learning_rate": 1.612487120186373e-05,
+      "loss": 0.3714,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6237826755509995,
+      "grad_norm": 0.5544696045862556,
+      "learning_rate": 1.611830628004287e-05,
+      "loss": 0.4084,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6242952332137366,
+      "grad_norm": 0.5053428410992765,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.369,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6248077908764736,
+      "grad_norm": 0.5356908126676675,
+      "learning_rate": 1.6105163789198607e-05,
+      "loss": 0.4062,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6253203485392107,
+      "grad_norm": 0.5389466091071076,
+      "learning_rate": 1.6098586229234096e-05,
+      "loss": 0.4278,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6258329062019478,
+      "grad_norm": 0.5468696404571222,
+      "learning_rate": 1.609200446561877e-05,
+      "loss": 0.4016,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6263454638646848,
+      "grad_norm": 0.5486199572069123,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.3968,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6268580215274219,
+      "grad_norm": 0.49866229830737324,
+      "learning_rate": 1.6078828345585332e-05,
+      "loss": 0.3713,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6273705791901589,
+      "grad_norm": 0.5273356776895979,
+      "learning_rate": 1.6072233998249303e-05,
+      "loss": 0.4012,
+      "step": 1224
+    },
+    {
+      "epoch": 0.627883136852896,
+      "grad_norm": 0.5316618398492498,
+      "learning_rate": 1.606563546542659e-05,
+      "loss": 0.3579,
+      "step": 1225
+    },
+    {
+      "epoch": 0.628395694515633,
+      "grad_norm": 0.5292494817316022,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.4479,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6289082521783701,
+      "grad_norm": 0.5206144978857087,
+      "learning_rate": 1.6052425861517035e-05,
+      "loss": 0.4289,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6294208098411072,
+      "grad_norm": 0.5172716416029565,
+      "learning_rate": 1.604581479953534e-05,
+      "loss": 0.359,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6299333675038442,
+      "grad_norm": 0.5035541485913633,
+      "learning_rate": 1.603919957027727e-05,
+      "loss": 0.3475,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6304459251665813,
+      "grad_norm": 0.46498290949247323,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.3246,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6309584828293183,
+      "grad_norm": 0.4916388972729933,
+      "learning_rate": 1.6025956628173916e-05,
+      "loss": 0.4614,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6314710404920554,
+      "grad_norm": 0.48469735453605317,
+      "learning_rate": 1.6019328924456764e-05,
+      "loss": 0.3795,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6319835981547924,
+      "grad_norm": 0.49803965996167376,
+      "learning_rate": 1.6012697071719498e-05,
+      "loss": 0.357,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6324961558175295,
+      "grad_norm": 0.504083992738786,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.4134,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6330087134802665,
+      "grad_norm": 0.4903184725566439,
+      "learning_rate": 1.599942093747234e-05,
+      "loss": 0.3246,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6335212711430036,
+      "grad_norm": 0.5081392455793806,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.3852,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6340338288057407,
+      "grad_norm": 0.5143761046489013,
+      "learning_rate": 1.59861282620365e-05,
+      "loss": 0.4019,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6345463864684777,
+      "grad_norm": 0.4885182083731511,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.3813,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6350589441312148,
+      "grad_norm": 0.5045245884136816,
+      "learning_rate": 1.5972819082061605e-05,
+      "loss": 0.397,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6355715017939518,
+      "grad_norm": 0.527303161416555,
+      "learning_rate": 1.5966158314337472e-05,
+      "loss": 0.355,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6360840594566889,
+      "grad_norm": 0.5513818136436452,
+      "learning_rate": 1.595949343424281e-05,
+      "loss": 0.3612,
+      "step": 1241
+    },
+    {
+      "epoch": 0.636596617119426,
+      "grad_norm": 0.5018054801014589,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.3798,
+      "step": 1242
+    },
+    {
+      "epoch": 0.637109174782163,
+      "grad_norm": 0.5086674860920013,
+      "learning_rate": 1.5946151355320675e-05,
+      "loss": 0.3731,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6376217324449001,
+      "grad_norm": 0.5573839597281072,
+      "learning_rate": 1.593947416568967e-05,
+      "loss": 0.4114,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6381342901076371,
+      "grad_norm": 0.5358488262331278,
+      "learning_rate": 1.5932792882081062e-05,
+      "loss": 0.4204,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6386468477703742,
+      "grad_norm": 0.5105046290326648,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.3769,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6391594054331112,
+      "grad_norm": 0.48682215253572403,
+      "learning_rate": 1.591941805135502e-05,
+      "loss": 0.398,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6396719630958483,
+      "grad_norm": 0.488869587518554,
+      "learning_rate": 1.591272451345663e-05,
+      "loss": 0.3314,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6401845207585853,
+      "grad_norm": 0.4746500461265795,
+      "learning_rate": 1.5906026900018708e-05,
+      "loss": 0.4015,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6406970784213224,
+      "grad_norm": 0.5129205610124654,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.4006,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6412096360840595,
+      "grad_norm": 0.5395335654142048,
+      "learning_rate": 1.589261946499329e-05,
+      "loss": 0.3808,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6417221937467965,
+      "grad_norm": 0.5216621203475702,
+      "learning_rate": 1.588590965264731e-05,
+      "loss": 0.3779,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6422347514095336,
+      "grad_norm": 0.4954743740952468,
+      "learning_rate": 1.587919578324482e-05,
+      "loss": 0.4055,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6427473090722706,
+      "grad_norm": 0.5096102045393365,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.3803,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6432598667350077,
+      "grad_norm": 0.5106073058753074,
+      "learning_rate": 1.5865755891784137e-05,
+      "loss": 0.385,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6437724243977447,
+      "grad_norm": 0.533247222769959,
+      "learning_rate": 1.5859029878989828e-05,
+      "loss": 0.3292,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6442849820604818,
+      "grad_norm": 0.5122618427340082,
+      "learning_rate": 1.585229982766678e-05,
+      "loss": 0.3469,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6447975397232188,
+      "grad_norm": 0.5411857757893819,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.4014,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6453100973859559,
+      "grad_norm": 0.5136594246571953,
+      "learning_rate": 1.583882762799289e-05,
+      "loss": 0.3815,
+      "step": 1259
+    },
+    {
+      "epoch": 0.645822655048693,
+      "grad_norm": 0.5537698296543099,
+      "learning_rate": 1.58320854889282e-05,
+      "loss": 0.3988,
+      "step": 1260
+    },
+    {
+      "epoch": 0.64633521271143,
+      "grad_norm": 0.5413133506838096,
+      "learning_rate": 1.5825339329907074e-05,
+      "loss": 0.3994,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6468477703741671,
+      "grad_norm": 0.5306466173069707,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.3928,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6473603280369041,
+      "grad_norm": 0.4515671696465751,
+      "learning_rate": 1.581183497059834e-05,
+      "loss": 0.3639,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6478728856996412,
+      "grad_norm": 0.4763288063923676,
+      "learning_rate": 1.580507677961906e-05,
+      "loss": 0.3847,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6483854433623782,
+      "grad_norm": 0.5568631425117423,
+      "learning_rate": 1.5798314587299976e-05,
+      "loss": 0.3916,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6488980010251153,
+      "grad_norm": 0.46123688537260044,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.3591,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6494105586878524,
+      "grad_norm": 0.4921216170817472,
+      "learning_rate": 1.578477821728944e-05,
+      "loss": 0.3881,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6499231163505894,
+      "grad_norm": 0.46939168135101933,
+      "learning_rate": 1.577800404892838e-05,
+      "loss": 0.377,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6504356740133265,
+      "grad_norm": 0.5018350019892727,
+      "learning_rate": 1.5771225897888288e-05,
+      "loss": 0.3614,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6509482316760635,
+      "grad_norm": 0.47085527384637527,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.3992,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6514607893388006,
+      "grad_norm": 0.5388109993170335,
+      "learning_rate": 1.575765766646203e-05,
+      "loss": 0.3949,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6519733470015376,
+      "grad_norm": 0.49677295087357526,
+      "learning_rate": 1.5750867595428205e-05,
+      "loss": 0.3357,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6524859046642747,
+      "grad_norm": 0.5191398033541963,
+      "learning_rate": 1.574407356042005e-05,
+      "loss": 0.4035,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6529984623270118,
+      "grad_norm": 0.4796139010381861,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.3972,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6535110199897488,
+      "grad_norm": 0.5006167676884137,
+      "learning_rate": 1.573047361721552e-05,
+      "loss": 0.3862,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6540235776524859,
+      "grad_norm": 0.524776109060049,
+      "learning_rate": 1.572366771839335e-05,
+      "loss": 0.3545,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6545361353152229,
+      "grad_norm": 0.4712932579497906,
+      "learning_rate": 1.5716857874345258e-05,
+      "loss": 0.3296,
+      "step": 1277
+    },
+    {
+      "epoch": 0.65504869297796,
+      "grad_norm": 0.494790103715949,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.3654,
+      "step": 1278
+    },
+    {
+      "epoch": 0.655561250640697,
+      "grad_norm": 0.5084810287054835,
+      "learning_rate": 1.5703226369349642e-05,
+      "loss": 0.44,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6560738083034341,
+      "grad_norm": 0.5575623339665483,
+      "learning_rate": 1.5696404717798087e-05,
+      "loss": 0.3707,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6565863659661711,
+      "grad_norm": 0.49635504060865426,
+      "learning_rate": 1.5689579139812524e-05,
+      "loss": 0.4196,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6570989236289082,
+      "grad_norm": 0.48881691008692313,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.3796,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6576114812916453,
+      "grad_norm": 0.5668998848957568,
+      "learning_rate": 1.5675916223361095e-05,
+      "loss": 0.3699,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6581240389543823,
+      "grad_norm": 0.5188513300825194,
+      "learning_rate": 1.5669078894312848e-05,
+      "loss": 0.3915,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6586365966171194,
+      "grad_norm": 0.5103493842260141,
+      "learning_rate": 1.566223765766581e-05,
+      "loss": 0.3594,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6591491542798564,
+      "grad_norm": 0.5259413882031672,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 0.4229,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6596617119425935,
+      "grad_norm": 0.49197959457920964,
+      "learning_rate": 1.5648543480440264e-05,
+      "loss": 0.4277,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6601742696053307,
+      "grad_norm": 0.44892194428551985,
+      "learning_rate": 1.5641690549300913e-05,
+      "loss": 0.3727,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6606868272680677,
+      "grad_norm": 0.529632125237598,
+      "learning_rate": 1.5634833729441093e-05,
+      "loss": 0.3897,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6611993849308048,
+      "grad_norm": 0.4914908405491881,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.3794,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6617119425935418,
+      "grad_norm": 0.48524713514366796,
+      "learning_rate": 1.5621108442467877e-05,
+      "loss": 0.3717,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6622245002562789,
+      "grad_norm": 0.5304744653432029,
+      "learning_rate": 1.5614239984815084e-05,
+      "loss": 0.3441,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6627370579190159,
+      "grad_norm": 0.511543183661938,
+      "learning_rate": 1.5607367657363028e-05,
+      "loss": 0.3797,
+      "step": 1293
+    },
+    {
+      "epoch": 0.663249615581753,
+      "grad_norm": 0.496819030956105,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.3722,
+      "step": 1294
+    },
+    {
+      "epoch": 0.66376217324449,
+      "grad_norm": 0.5204017280289366,
+      "learning_rate": 1.5593611412011685e-05,
+      "loss": 0.3709,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6642747309072271,
+      "grad_norm": 0.4730599457372219,
+      "learning_rate": 1.5586727503594347e-05,
+      "loss": 0.3765,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6647872885699642,
+      "grad_norm": 0.43783471842967975,
+      "learning_rate": 1.557983974434162e-05,
+      "loss": 0.3542,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6652998462327012,
+      "grad_norm": 0.4869070093249163,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.3487,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6658124038954383,
+      "grad_norm": 0.5571518441886586,
+      "learning_rate": 1.5566052692323127e-05,
+      "loss": 0.3685,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6663249615581753,
+      "grad_norm": 0.4570112350470338,
+      "learning_rate": 1.5559153409060535e-05,
+      "loss": 0.3877,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6668375192209124,
+      "grad_norm": 0.5065690230304977,
+      "learning_rate": 1.5552250293968903e-05,
+      "loss": 0.3979,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6673500768836494,
+      "grad_norm": 0.48437941022626724,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.3633,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6678626345463865,
+      "grad_norm": 0.5323288257293891,
+      "learning_rate": 1.553843258733398e-05,
+      "loss": 0.3901,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6683751922091236,
+      "grad_norm": 0.5392570041614668,
+      "learning_rate": 1.5531518005314987e-05,
+      "loss": 0.3571,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6688877498718606,
+      "grad_norm": 0.5224403624495645,
+      "learning_rate": 1.552459961051557e-05,
+      "loss": 0.3586,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6694003075345977,
+      "grad_norm": 0.5217339307250121,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.4132,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6699128651973347,
+      "grad_norm": 0.4990829734040992,
+      "learning_rate": 1.551075140165301e-05,
+      "loss": 0.3474,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6704254228600718,
+      "grad_norm": 0.49818550275371315,
+      "learning_rate": 1.55038215971352e-05,
+      "loss": 0.3693,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6709379805228088,
+      "grad_norm": 0.4585221785371497,
+      "learning_rate": 1.549688799892762e-05,
+      "loss": 0.3424,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6714505381855459,
+      "grad_norm": 0.4816969591338806,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.3628,
+      "step": 1310
+    },
+    {
+      "epoch": 0.671963095848283,
+      "grad_norm": 0.4965185003771397,
+      "learning_rate": 1.5483009440562617e-05,
+      "loss": 0.3663,
+      "step": 1311
+    },
+    {
+      "epoch": 0.67247565351102,
+      "grad_norm": 0.486090579922493,
+      "learning_rate": 1.5476064489971444e-05,
+      "loss": 0.3767,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6729882111737571,
+      "grad_norm": 0.5054627944713437,
+      "learning_rate": 1.5469115764823e-05,
+      "loss": 0.4297,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6735007688364941,
+      "grad_norm": 0.4985181740156195,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.3493,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6740133264992312,
+      "grad_norm": 0.5518064647213198,
+      "learning_rate": 1.5455207010015456e-05,
+      "loss": 0.3971,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6745258841619682,
+      "grad_norm": 0.536683838788927,
+      "learning_rate": 1.544824698994342e-05,
+      "loss": 0.3835,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6750384418247053,
+      "grad_norm": 0.5067313177427939,
+      "learning_rate": 1.544128321448824e-05,
+      "loss": 0.3736,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6755509994874423,
+      "grad_norm": 0.563314483577618,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.4044,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6760635571501794,
+      "grad_norm": 0.5097097633753342,
+      "learning_rate": 1.5427344416631076e-05,
+      "loss": 0.3676,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6765761148129165,
+      "grad_norm": 0.4834697721375851,
+      "learning_rate": 1.5420369403836867e-05,
+      "loss": 0.3337,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6770886724756535,
+      "grad_norm": 0.5177539039827452,
+      "learning_rate": 1.541339065487506e-05,
+      "loss": 0.3837,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6776012301383906,
+      "grad_norm": 0.5062730915082158,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.338,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6781137878011276,
+      "grad_norm": 0.5133352134203722,
+      "learning_rate": 1.539942196769253e-05,
+      "loss": 0.3978,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6786263454638647,
+      "grad_norm": 0.5160788770504897,
+      "learning_rate": 1.5392432039100192e-05,
+      "loss": 0.3436,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6791389031266017,
+      "grad_norm": 0.4541779054417641,
+      "learning_rate": 1.5385438393596994e-05,
+      "loss": 0.376,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6796514607893388,
+      "grad_norm": 0.516479302058551,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.364,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6801640184520759,
+      "grad_norm": 0.5212718082884531,
+      "learning_rate": 1.5371439971142993e-05,
+      "loss": 0.4055,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6806765761148129,
+      "grad_norm": 0.5187825660397735,
+      "learning_rate": 1.5364435203841058e-05,
+      "loss": 0.3429,
+      "step": 1328
+    },
+    {
+      "epoch": 0.68118913377755,
+      "grad_norm": 0.5291924446380852,
+      "learning_rate": 1.5357426738926008e-05,
+      "loss": 0.362,
+      "step": 1329
+    },
+    {
+      "epoch": 0.681701691440287,
+      "grad_norm": 0.51453705616876,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4076,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6822142491030241,
+      "grad_norm": 0.5116833623265402,
+      "learning_rate": 1.5343398735582352e-05,
+      "loss": 0.3717,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6827268067657611,
+      "grad_norm": 0.5260935616326815,
+      "learning_rate": 1.5336379206823013e-05,
+      "loss": 0.3898,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6832393644284982,
+      "grad_norm": 0.4813713598417104,
+      "learning_rate": 1.5329355999789074e-05,
+      "loss": 0.3516,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6837519220912353,
+      "grad_norm": 0.5301669105439802,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.403,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6842644797539723,
+      "grad_norm": 0.522801747940241,
+      "learning_rate": 1.5315298570263827e-05,
+      "loss": 0.3894,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6847770374167094,
+      "grad_norm": 0.5416688544494416,
+      "learning_rate": 1.5308264357462068e-05,
+      "loss": 0.375,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6852895950794464,
+      "grad_norm": 0.5034282806922896,
+      "learning_rate": 1.530122648576479e-05,
+      "loss": 0.3696,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6858021527421835,
+      "grad_norm": 0.5459547437250816,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.4397,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6863147104049205,
+      "grad_norm": 0.5621843513063042,
+      "learning_rate": 1.5287139785090534e-05,
+      "loss": 0.4247,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6868272680676576,
+      "grad_norm": 0.5115941107308559,
+      "learning_rate": 1.528009096582328e-05,
+      "loss": 0.4059,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6873398257303946,
+      "grad_norm": 0.5240919171630058,
+      "learning_rate": 1.527303850707994e-05,
+      "loss": 0.4275,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6878523833931317,
+      "grad_norm": 0.46572337146402,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.3719,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6883649410558688,
+      "grad_norm": 0.4790009548598872,
+      "learning_rate": 1.5258922690612085e-05,
+      "loss": 0.3696,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6888774987186058,
+      "grad_norm": 0.4458910932056421,
+      "learning_rate": 1.5251859342617353e-05,
+      "loss": 0.3706,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6893900563813429,
+      "grad_norm": 0.4831176283259128,
+      "learning_rate": 1.524479237460611e-05,
+      "loss": 0.3413,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6899026140440799,
+      "grad_norm": 0.5003790236137974,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.3878,
+      "step": 1346
+    },
+    {
+      "epoch": 0.690415171706817,
+      "grad_norm": 0.46175750846735075,
+      "learning_rate": 1.5230647598021153e-05,
+      "loss": 0.2981,
+      "step": 1347
+    },
+    {
+      "epoch": 0.690927729369554,
+      "grad_norm": 0.48336755359389444,
+      "learning_rate": 1.5223569799197185e-05,
+      "loss": 0.3374,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6914402870322911,
+      "grad_norm": 0.4822996569778159,
+      "learning_rate": 1.5216488399856207e-05,
+      "loss": 0.3997,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6919528446950282,
+      "grad_norm": 0.5028494471732301,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.3658,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6924654023577652,
+      "grad_norm": 0.45951293406691945,
+      "learning_rate": 1.5202314819150044e-05,
+      "loss": 0.4117,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6929779600205023,
+      "grad_norm": 0.5400116570322007,
+      "learning_rate": 1.5195222647554459e-05,
+      "loss": 0.378,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6934905176832393,
+      "grad_norm": 0.4873228571855651,
+      "learning_rate": 1.5188126894981072e-05,
+      "loss": 0.3424,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6940030753459764,
+      "grad_norm": 0.5261296317361047,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.3706,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6945156330087134,
+      "grad_norm": 0.5164907289793792,
+      "learning_rate": 1.5173924666467259e-05,
+      "loss": 0.3526,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6950281906714505,
+      "grad_norm": 0.4760911775181999,
+      "learning_rate": 1.5166818200316192e-05,
+      "loss": 0.3931,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6955407483341876,
+      "grad_norm": 0.5000308811890761,
+      "learning_rate": 1.515970817276601e-05,
+      "loss": 0.3656,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6960533059969246,
+      "grad_norm": 0.5286174319739142,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.4056,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6965658636596617,
+      "grad_norm": 0.5001449837754821,
+      "learning_rate": 1.5145477453074056e-05,
+      "loss": 0.3865,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6970784213223987,
+      "grad_norm": 0.5231838322700835,
+      "learning_rate": 1.5138356770741274e-05,
+      "loss": 0.3765,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6975909789851358,
+      "grad_norm": 0.5093633816874403,
+      "learning_rate": 1.5131232546627355e-05,
+      "loss": 0.4125,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6981035366478728,
+      "grad_norm": 0.5442242148483528,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.391,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6986160943106099,
+      "grad_norm": 0.5015991662607845,
+      "learning_rate": 1.5116973492700977e-05,
+      "loss": 0.3807,
+      "step": 1363
+    },
+    {
+      "epoch": 0.699128651973347,
+      "grad_norm": 0.48247281948746146,
+      "learning_rate": 1.5109838672717034e-05,
+      "loss": 0.3843,
+      "step": 1364
+    },
+    {
+      "epoch": 0.699641209636084,
+      "grad_norm": 0.4911221743201765,
+      "learning_rate": 1.5102700330609e-05,
+      "loss": 0.402,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7001537672988212,
+      "grad_norm": 0.44076674835714363,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.3428,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7006663249615582,
+      "grad_norm": 0.4804116885035028,
+      "learning_rate": 1.5088413099704396e-05,
+      "loss": 0.399,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7011788826242953,
+      "grad_norm": 0.46020059469443764,
+      "learning_rate": 1.5081264220755775e-05,
+      "loss": 0.3564,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7016914402870323,
+      "grad_norm": 0.5352609685663365,
+      "learning_rate": 1.5074111839378938e-05,
+      "loss": 0.3818,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7022039979497694,
+      "grad_norm": 0.4580778376046765,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.3843,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7027165556125065,
+      "grad_norm": 0.5389540393593844,
+      "learning_rate": 1.505979658906307e-05,
+      "loss": 0.4036,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7032291132752435,
+      "grad_norm": 0.507636739604362,
+      "learning_rate": 1.5052633729991296e-05,
+      "loss": 0.4134,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7037416709379806,
+      "grad_norm": 0.49859282070808786,
+      "learning_rate": 1.5045467388225805e-05,
+      "loss": 0.3544,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7042542286007176,
+      "grad_norm": 0.538419730904198,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.3705,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7047667862634547,
+      "grad_norm": 0.5203424678996743,
+      "learning_rate": 1.5031124276374633e-05,
+      "loss": 0.4184,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7052793439261917,
+      "grad_norm": 0.4694442169522522,
+      "learning_rate": 1.502394751617541e-05,
+      "loss": 0.3423,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7057919015889288,
+      "grad_norm": 0.5032637258282101,
+      "learning_rate": 1.5016767293055388e-05,
+      "loss": 0.3632,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7063044592516659,
+      "grad_norm": 0.5310039676790624,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.3542,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7068170169144029,
+      "grad_norm": 0.5079432089518724,
+      "learning_rate": 1.5002396477852146e-05,
+      "loss": 0.385,
+      "step": 1379
+    },
+    {
+      "epoch": 0.70732957457714,
+      "grad_norm": 0.5187047993872902,
+      "learning_rate": 1.4995205895674489e-05,
+      "loss": 0.392,
+      "step": 1380
+    },
+    {
+      "epoch": 0.707842132239877,
+      "grad_norm": 0.4899687631240464,
+      "learning_rate": 1.4988011870387138e-05,
+      "loss": 0.3785,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7083546899026141,
+      "grad_norm": 0.5127407696229104,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.3509,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7088672475653511,
+      "grad_norm": 0.5400092959664277,
+      "learning_rate": 1.4973613510320595e-05,
+      "loss": 0.4327,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7093798052280882,
+      "grad_norm": 0.4925019972230039,
+      "learning_rate": 1.4966409185465949e-05,
+      "loss": 0.3886,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7098923628908252,
+      "grad_norm": 0.48047676629692443,
+      "learning_rate": 1.4959201437350688e-05,
+      "loss": 0.3835,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7104049205535623,
+      "grad_norm": 0.4566864934117825,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.3791,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7109174782162994,
+      "grad_norm": 0.47263739337022415,
+      "learning_rate": 1.494477569121339e-05,
+      "loss": 0.3469,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7114300358790364,
+      "grad_norm": 0.5397347475235895,
+      "learning_rate": 1.4937557703134767e-05,
+      "loss": 0.4258,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7119425935417735,
+      "grad_norm": 0.5267893350948304,
+      "learning_rate": 1.4930336311682365e-05,
+      "loss": 0.4083,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7124551512045105,
+      "grad_norm": 0.47279227926450207,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.3615,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7129677088672476,
+      "grad_norm": 0.4587811699138689,
+      "learning_rate": 1.4915883338568875e-05,
+      "loss": 0.3573,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7134802665299846,
+      "grad_norm": 0.4824921512393126,
+      "learning_rate": 1.4908651766869975e-05,
+      "loss": 0.3649,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7139928241927217,
+      "grad_norm": 0.5017287032657716,
+      "learning_rate": 1.4901416811721658e-05,
+      "loss": 0.3116,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7145053818554588,
+      "grad_norm": 0.47924654745299744,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.3669,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7150179395181958,
+      "grad_norm": 0.5038247197056689,
+      "learning_rate": 1.4886936771026826e-05,
+      "loss": 0.358,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7155304971809329,
+      "grad_norm": 0.4653424649841318,
+      "learning_rate": 1.4879691695461151e-05,
+      "loss": 0.3736,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7160430548436699,
+      "grad_norm": 0.458056683465254,
+      "learning_rate": 1.487244325640774e-05,
+      "loss": 0.378,
+      "step": 1397
+    },
+    {
+      "epoch": 0.716555612506407,
+      "grad_norm": 0.4653136785758063,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.375,
+      "step": 1398
+    },
+    {
+      "epoch": 0.717068170169144,
+      "grad_norm": 0.5093661207824752,
+      "learning_rate": 1.4857936307824923e-05,
+      "loss": 0.3508,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7175807278318811,
+      "grad_norm": 0.49594526347364515,
+      "learning_rate": 1.4850677808294902e-05,
+      "loss": 0.334,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7180932854946181,
+      "grad_norm": 0.4777976427957875,
+      "learning_rate": 1.484341596527592e-05,
+      "loss": 0.3842,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7186058431573552,
+      "grad_norm": 0.5073257062357591,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.3524,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7191184008200923,
+      "grad_norm": 0.4328640546131387,
+      "learning_rate": 1.4828882268795228e-05,
+      "loss": 0.3424,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7196309584828293,
+      "grad_norm": 0.49465196701965775,
+      "learning_rate": 1.4821610425351343e-05,
+      "loss": 0.396,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7201435161455664,
+      "grad_norm": 0.46614511551902715,
+      "learning_rate": 1.4814335258454144e-05,
+      "loss": 0.3111,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7206560738083034,
+      "grad_norm": 0.47018630544209067,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.3406,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7211686314710405,
+      "grad_norm": 0.489880485073502,
+      "learning_rate": 1.4799774974360679e-05,
+      "loss": 0.3518,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7216811891337775,
+      "grad_norm": 0.5214192570394324,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.3695,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7221937467965146,
+      "grad_norm": 0.5258309006627064,
+      "learning_rate": 1.4785201456659446e-05,
+      "loss": 0.3529,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7227063044592517,
+      "grad_norm": 0.49593553081857455,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.3664,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7232188621219887,
+      "grad_norm": 0.4718592077401747,
+      "learning_rate": 1.4770614745531538e-05,
+      "loss": 0.3536,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7237314197847258,
+      "grad_norm": 0.5168869967423495,
+      "learning_rate": 1.4763316454999121e-05,
+      "loss": 0.3769,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7242439774474628,
+      "grad_norm": 0.4983411223451127,
+      "learning_rate": 1.4756014881194422e-05,
+      "loss": 0.3847,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7247565351101999,
+      "grad_norm": 0.49805352688059246,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.3662,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7252690927729369,
+      "grad_norm": 0.49178190229340885,
+      "learning_rate": 1.4741401903901843e-05,
+      "loss": 0.3656,
+      "step": 1415
+    },
+    {
+      "epoch": 0.725781650435674,
+      "grad_norm": 0.4679442024448,
+      "learning_rate": 1.4734090510486435e-05,
+      "loss": 0.348,
+      "step": 1416
+    },
+    {
+      "epoch": 0.726294208098411,
+      "grad_norm": 0.49446215921078673,
+      "learning_rate": 1.4726775853943688e-05,
+      "loss": 0.3791,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7268067657611481,
+      "grad_norm": 0.5167513247117264,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.3624,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7273193234238852,
+      "grad_norm": 0.5054133866401934,
+      "learning_rate": 1.4712136771645891e-05,
+      "loss": 0.3554,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7278318810866222,
+      "grad_norm": 0.48825641217584487,
+      "learning_rate": 1.4704812355981308e-05,
+      "loss": 0.4127,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7283444387493593,
+      "grad_norm": 0.4666148807399284,
+      "learning_rate": 1.4697484697370317e-05,
+      "loss": 0.3805,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7288569964120963,
+      "grad_norm": 0.5140447853783096,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.3971,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7293695540748334,
+      "grad_norm": 0.5188673974542385,
+      "learning_rate": 1.4682819671514654e-05,
+      "loss": 0.3411,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7298821117375704,
+      "grad_norm": 0.4950261953014311,
+      "learning_rate": 1.4675482314378331e-05,
+      "loss": 0.3751,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7303946694003075,
+      "grad_norm": 0.486083324067171,
+      "learning_rate": 1.4668141734512294e-05,
+      "loss": 0.3539,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7309072270630446,
+      "grad_norm": 0.48370213820970764,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.3347,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7314197847257816,
+      "grad_norm": 0.4589930772925287,
+      "learning_rate": 1.4653450926832236e-05,
+      "loss": 0.3578,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7319323423885187,
+      "grad_norm": 0.522485016659336,
+      "learning_rate": 1.4646100709144335e-05,
+      "loss": 0.3928,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7324449000512557,
+      "grad_norm": 0.46880311404864566,
+      "learning_rate": 1.4638747288978954e-05,
+      "loss": 0.3785,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7329574577139928,
+      "grad_norm": 0.4832011945123479,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.3911,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7334700153767298,
+      "grad_norm": 0.4722585670856934,
+      "learning_rate": 1.4624030861492304e-05,
+      "loss": 0.3678,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7339825730394669,
+      "grad_norm": 0.5016443041043147,
+      "learning_rate": 1.461666786431482e-05,
+      "loss": 0.3503,
+      "step": 1432
+    },
+    {
+      "epoch": 0.734495130702204,
+      "grad_norm": 0.4629259616134305,
+      "learning_rate": 1.4609301684947403e-05,
+      "loss": 0.3225,
+      "step": 1433
+    },
+    {
+      "epoch": 0.735007688364941,
+      "grad_norm": 0.549438704910588,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.3724,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7355202460276781,
+      "grad_norm": 0.48656112039799243,
+      "learning_rate": 1.4594559799954523e-05,
+      "loss": 0.3922,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7360328036904151,
+      "grad_norm": 0.49356673450129357,
+      "learning_rate": 1.4587184104490377e-05,
+      "loss": 0.3712,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7365453613531522,
+      "grad_norm": 0.5116709998095905,
+      "learning_rate": 1.457980524715896e-05,
+      "loss": 0.3651,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7370579190158892,
+      "grad_norm": 0.5220454850851347,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.3645,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7375704766786263,
+      "grad_norm": 0.48386443119943445,
+      "learning_rate": 1.4565038067240956e-05,
+      "loss": 0.3747,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7380830343413634,
+      "grad_norm": 0.448858102739118,
+      "learning_rate": 1.455764975483313e-05,
+      "loss": 0.3594,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7385955920041004,
+      "grad_norm": 0.5276079097606935,
+      "learning_rate": 1.4550258300915551e-05,
+      "loss": 0.3733,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7391081496668375,
+      "grad_norm": 0.5111674482420316,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.3631,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7396207073295745,
+      "grad_norm": 0.48012509530755615,
+      "learning_rate": 1.4535465988932502e-05,
+      "loss": 0.3367,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7401332649923117,
+      "grad_norm": 0.46913970157146656,
+      "learning_rate": 1.4528065141063119e-05,
+      "loss": 0.4067,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7406458226550487,
+      "grad_norm": 0.5445257789064579,
+      "learning_rate": 1.4520661172076147e-05,
+      "loss": 0.4145,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7411583803177858,
+      "grad_norm": 0.5314090368984774,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.382,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7416709379805229,
+      "grad_norm": 0.4786715828712538,
+      "learning_rate": 1.4505843891165304e-05,
+      "loss": 0.396,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7421834956432599,
+      "grad_norm": 0.4862488099819625,
+      "learning_rate": 1.4498430589454726e-05,
+      "loss": 0.3775,
+      "step": 1448
+    },
+    {
+      "epoch": 0.742696053305997,
+      "grad_norm": 0.506867086382113,
+      "learning_rate": 1.449101418705315e-05,
+      "loss": 0.4073,
+      "step": 1449
+    },
+    {
+      "epoch": 0.743208610968734,
+      "grad_norm": 0.5346380734555378,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.3733,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7437211686314711,
+      "grad_norm": 0.49355213393741376,
+      "learning_rate": 1.4476172100627127e-05,
+      "loss": 0.3652,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7442337262942081,
+      "grad_norm": 0.4652909657628661,
+      "learning_rate": 1.4468746426833075e-05,
+      "loss": 0.3523,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7447462839569452,
+      "grad_norm": 0.5015163781505074,
+      "learning_rate": 1.4461317672808809e-05,
+      "loss": 0.3975,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7452588416196823,
+      "grad_norm": 0.49252009225494126,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.3832,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7457713992824193,
+      "grad_norm": 0.4794530495931546,
+      "learning_rate": 1.4446450944553793e-05,
+      "loss": 0.34,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7462839569451564,
+      "grad_norm": 0.5018145603529296,
+      "learning_rate": 1.4439012980570425e-05,
+      "loss": 0.3686,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7467965146078934,
+      "grad_norm": 0.47425303612131847,
+      "learning_rate": 1.44315719568516e-05,
+      "loss": 0.3498,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7473090722706305,
+      "grad_norm": 0.4886368457313323,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.3731,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7478216299333675,
+      "grad_norm": 0.497431835698092,
+      "learning_rate": 1.4416680750725542e-05,
+      "loss": 0.3682,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7483341875961046,
+      "grad_norm": 0.5218793931470161,
+      "learning_rate": 1.4409230578582566e-05,
+      "loss": 0.4047,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7488467452588417,
+      "grad_norm": 0.5154231354734998,
+      "learning_rate": 1.4401777367232625e-05,
+      "loss": 0.3736,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7493593029215787,
+      "grad_norm": 0.45719004863643137,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.3546,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7498718605843158,
+      "grad_norm": 0.49579922291124,
+      "learning_rate": 1.4386861847463429e-05,
+      "loss": 0.38,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7503844182470528,
+      "grad_norm": 0.49862274244977045,
+      "learning_rate": 1.4379399549325185e-05,
+      "loss": 0.3612,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7508969759097899,
+      "grad_norm": 0.456668590709913,
+      "learning_rate": 1.4371934232541991e-05,
+      "loss": 0.3466,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7514095335725269,
+      "grad_norm": 0.49872491354037923,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.4007,
+      "step": 1466
+    },
+    {
+      "epoch": 0.751922091235264,
+      "grad_norm": 0.4821353970384583,
+      "learning_rate": 1.4356994563625703e-05,
+      "loss": 0.3144,
+      "step": 1467
+    },
+    {
+      "epoch": 0.752434648898001,
+      "grad_norm": 0.4616909168631738,
+      "learning_rate": 1.4349520221790263e-05,
+      "loss": 0.3309,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7529472065607381,
+      "grad_norm": 0.44762120781142234,
+      "learning_rate": 1.4342042881905184e-05,
+      "loss": 0.3794,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7534597642234752,
+      "grad_norm": 0.4833383724988035,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.3627,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7539723218862122,
+      "grad_norm": 0.5038707500376719,
+      "learning_rate": 1.4327079228604177e-05,
+      "loss": 0.3317,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7544848795489493,
+      "grad_norm": 0.4512584602220797,
+      "learning_rate": 1.4319592925502443e-05,
+      "loss": 0.351,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7549974372116863,
+      "grad_norm": 0.48486188039618455,
+      "learning_rate": 1.4312103644979441e-05,
+      "loss": 0.3264,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7555099948744234,
+      "grad_norm": 0.4653409449986988,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.3508,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7560225525371604,
+      "grad_norm": 0.4989189965863489,
+      "learning_rate": 1.4297116172320606e-05,
+      "loss": 0.3409,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7565351101998975,
+      "grad_norm": 0.5188329382235826,
+      "learning_rate": 1.4289617990515383e-05,
+      "loss": 0.3695,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7570476678626346,
+      "grad_norm": 0.45583862239956235,
+      "learning_rate": 1.4282116851950103e-05,
+      "loss": 0.3416,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7575602255253716,
+      "grad_norm": 0.516848145653567,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.3642,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7580727831881087,
+      "grad_norm": 0.5259404696108063,
+      "learning_rate": 1.4267105725223023e-05,
+      "loss": 0.3902,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7585853408508457,
+      "grad_norm": 0.4867298667026852,
+      "learning_rate": 1.4259595747408133e-05,
+      "loss": 0.3645,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7590978985135828,
+      "grad_norm": 0.47750522307184357,
+      "learning_rate": 1.4252082833527003e-05,
+      "loss": 0.3786,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7596104561763198,
+      "grad_norm": 0.5352149492625023,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.4255,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7601230138390569,
+      "grad_norm": 0.46094040620125726,
+      "learning_rate": 1.4237048218282124e-05,
+      "loss": 0.3899,
+      "step": 1483
+    },
+    {
+      "epoch": 0.760635571501794,
+      "grad_norm": 0.4906167189513182,
+      "learning_rate": 1.422952652728148e-05,
+      "loss": 0.408,
+      "step": 1484
+    },
+    {
+      "epoch": 0.761148129164531,
+      "grad_norm": 0.4662223829478201,
+      "learning_rate": 1.4222001920940787e-05,
+      "loss": 0.3617,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7616606868272681,
+      "grad_norm": 0.4726229238281902,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.4032,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7621732444900051,
+      "grad_norm": 0.42314942534914685,
+      "learning_rate": 1.4206943982987604e-05,
+      "loss": 0.3251,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7626858021527422,
+      "grad_norm": 0.5582494082169284,
+      "learning_rate": 1.419941066175429e-05,
+      "loss": 0.4083,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7631983598154792,
+      "grad_norm": 0.4749934590180283,
+      "learning_rate": 1.4191874445939278e-05,
+      "loss": 0.3638,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7637109174782163,
+      "grad_norm": 0.4583574012055634,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.3848,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7642234751409533,
+      "grad_norm": 0.4833253220634529,
+      "learning_rate": 1.41767933513445e-05,
+      "loss": 0.3688,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7647360328036904,
+      "grad_norm": 0.49769719790450134,
+      "learning_rate": 1.4169248482959868e-05,
+      "loss": 0.3352,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7652485904664275,
+      "grad_norm": 0.509550277251258,
+      "learning_rate": 1.4161700740783815e-05,
+      "loss": 0.3684,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7657611481291645,
+      "grad_norm": 0.4846415538358595,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.3485,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7662737057919016,
+      "grad_norm": 0.5148677652520544,
+      "learning_rate": 1.4146596655869529e-05,
+      "loss": 0.3846,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7667862634546386,
+      "grad_norm": 0.46316811834133115,
+      "learning_rate": 1.4139040323542284e-05,
+      "loss": 0.3576,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7672988211173757,
+      "grad_norm": 0.4703383434659294,
+      "learning_rate": 1.4131481138245582e-05,
+      "loss": 0.3964,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7678113787801127,
+      "grad_norm": 0.48987636207114227,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.3572,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7683239364428498,
+      "grad_norm": 0.4901515672819091,
+      "learning_rate": 1.4116354229587432e-05,
+      "loss": 0.3816,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7688364941055869,
+      "grad_norm": 0.49122497598653386,
+      "learning_rate": 1.4108786516652704e-05,
+      "loss": 0.3839,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7693490517683239,
+      "grad_norm": 0.5044613311691365,
+      "learning_rate": 1.4101215971601945e-05,
+      "loss": 0.3402,
+      "step": 1501
+    },
+    {
+      "epoch": 0.769861609431061,
+      "grad_norm": 0.4991033317443066,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.3285,
+      "step": 1502
+    },
+    {
+      "epoch": 0.770374167093798,
+      "grad_norm": 0.4657681163127245,
+      "learning_rate": 1.4086066406027284e-05,
+      "loss": 0.2981,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7708867247565351,
+      "grad_norm": 0.5192607230192244,
+      "learning_rate": 1.4078487395945712e-05,
+      "loss": 0.3676,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7713992824192721,
+      "grad_norm": 0.5266072811364451,
+      "learning_rate": 1.4070905574632779e-05,
+      "loss": 0.4066,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7719118400820092,
+      "grad_norm": 0.5253946403862891,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.3795,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7724243977447462,
+      "grad_norm": 0.453496561070252,
+      "learning_rate": 1.4055733519218829e-05,
+      "loss": 0.336,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7729369554074833,
+      "grad_norm": 0.5052582388697123,
+      "learning_rate": 1.404814329557565e-05,
+      "loss": 0.3658,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7734495130702204,
+      "grad_norm": 0.48045480506613775,
+      "learning_rate": 1.4040550281616777e-05,
+      "loss": 0.3388,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7739620707329574,
+      "grad_norm": 0.46516349245393657,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.3394,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7744746283956945,
+      "grad_norm": 0.5080701034717423,
+      "learning_rate": 1.4025355903688794e-05,
+      "loss": 0.3812,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7749871860584315,
+      "grad_norm": 0.44786286461661007,
+      "learning_rate": 1.401775455019291e-05,
+      "loss": 0.3781,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7754997437211686,
+      "grad_norm": 0.4683044051286398,
+      "learning_rate": 1.401015042732777e-05,
+      "loss": 0.3386,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7760123013839056,
+      "grad_norm": 0.48297632634514337,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.3583,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7765248590466427,
+      "grad_norm": 0.5007325864452641,
+      "learning_rate": 1.3994933894457193e-05,
+      "loss": 0.3446,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7770374167093798,
+      "grad_norm": 0.5152087345650416,
+      "learning_rate": 1.398732149494025e-05,
+      "loss": 0.3588,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7775499743721168,
+      "grad_norm": 0.488228639915329,
+      "learning_rate": 1.3979706347031034e-05,
+      "loss": 0.3483,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7780625320348539,
+      "grad_norm": 0.4707466888623256,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.3599,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7785750896975909,
+      "grad_norm": 0.4984016764090868,
+      "learning_rate": 1.3964467827033637e-05,
+      "loss": 0.393,
+      "step": 1519
+    },
+    {
+      "epoch": 0.779087647360328,
+      "grad_norm": 0.5381823139245266,
+      "learning_rate": 1.3956844465449106e-05,
+      "loss": 0.3957,
+      "step": 1520
+    },
+    {
+      "epoch": 0.779600205023065,
+      "grad_norm": 0.516345913740466,
+      "learning_rate": 1.3949218376479593e-05,
+      "loss": 0.3622,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7801127626858022,
+      "grad_norm": 0.48358625823328344,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.3982,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7806253203485393,
+      "grad_norm": 0.4751669749363038,
+      "learning_rate": 1.3933958037413636e-05,
+      "loss": 0.3736,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7811378780112763,
+      "grad_norm": 0.4988493213701085,
+      "learning_rate": 1.3926323797835882e-05,
+      "loss": 0.3876,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7816504356740134,
+      "grad_norm": 0.4713765220287793,
+      "learning_rate": 1.3918686851910516e-05,
+      "loss": 0.3594,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7821629933367504,
+      "grad_norm": 0.4998596117655545,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.3624,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7826755509994875,
+      "grad_norm": 0.45257430389407277,
+      "learning_rate": 1.3903404862074891e-05,
+      "loss": 0.3561,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7831881086622245,
+      "grad_norm": 0.4745976295481429,
+      "learning_rate": 1.3895759828698243e-05,
+      "loss": 0.3846,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7837006663249616,
+      "grad_norm": 0.4788486967130179,
+      "learning_rate": 1.3888112110041207e-05,
+      "loss": 0.3312,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7842132239876987,
+      "grad_norm": 0.4770285032689454,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.332,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7847257816504357,
+      "grad_norm": 0.4825279288372994,
+      "learning_rate": 1.3872808637973583e-05,
+      "loss": 0.3573,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7852383393131728,
+      "grad_norm": 0.47191496409498224,
+      "learning_rate": 1.386515289511141e-05,
+      "loss": 0.4111,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7857508969759098,
+      "grad_norm": 0.5272373173385123,
+      "learning_rate": 1.3857494488065693e-05,
+      "loss": 0.365,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7862634546386469,
+      "grad_norm": 0.5172401080426673,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.3636,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7867760123013839,
+      "grad_norm": 0.5136003225708606,
+      "learning_rate": 1.384216970254065e-05,
+      "loss": 0.3973,
+      "step": 1535
+    },
+    {
+      "epoch": 0.787288569964121,
+      "grad_norm": 0.48353006845890945,
+      "learning_rate": 1.3834503334624443e-05,
+      "loss": 0.3491,
+      "step": 1536
+    },
+    {
+      "epoch": 0.787801127626858,
+      "grad_norm": 0.4672690655286873,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.3695,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7883136852895951,
+      "grad_norm": 0.5340310843619491,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.3755,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7888262429523322,
+      "grad_norm": 0.4894447170896454,
+      "learning_rate": 1.3811488393678085e-05,
+      "loss": 0.3419,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7893388006150692,
+      "grad_norm": 0.4367016797429956,
+      "learning_rate": 1.38038114852565e-05,
+      "loss": 0.3515,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7898513582778063,
+      "grad_norm": 0.4965507126774858,
+      "learning_rate": 1.379613195493294e-05,
+      "loss": 0.3258,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7903639159405433,
+      "grad_norm": 0.4907105237667023,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.3491,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7908764736032804,
+      "grad_norm": 0.5120815318721117,
+      "learning_rate": 1.3780765049755184e-05,
+      "loss": 0.3816,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7913890312660175,
+      "grad_norm": 0.5396410757240344,
+      "learning_rate": 1.3773077685493128e-05,
+      "loss": 0.3416,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7919015889287545,
+      "grad_norm": 0.5051247719598091,
+      "learning_rate": 1.376538772051338e-05,
+      "loss": 0.3748,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7924141465914916,
+      "grad_norm": 0.5051063627395481,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.3658,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7929267042542286,
+      "grad_norm": 0.4996879794321905,
+      "learning_rate": 1.3750000009604843e-05,
+      "loss": 0.3818,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7934392619169657,
+      "grad_norm": 0.4841675154092533,
+      "learning_rate": 1.3742302274282532e-05,
+      "loss": 0.333,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7939518195797027,
+      "grad_norm": 0.4558494330451396,
+      "learning_rate": 1.3734601959455495e-05,
+      "loss": 0.3552,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7944643772424398,
+      "grad_norm": 0.5021183958403396,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.3752,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7949769349051768,
+      "grad_norm": 0.463659797528096,
+      "learning_rate": 1.3719193612519789e-05,
+      "loss": 0.3785,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7954894925679139,
+      "grad_norm": 0.45224870770311254,
+      "learning_rate": 1.3711485591031821e-05,
+      "loss": 0.3175,
+      "step": 1552
+    },
+    {
+      "epoch": 0.796002050230651,
+      "grad_norm": 0.4803285034653485,
+      "learning_rate": 1.3703775011280538e-05,
+      "loss": 0.3414,
+      "step": 1553
+    },
+    {
+      "epoch": 0.796514607893388,
+      "grad_norm": 0.5441675486779248,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.3949,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7970271655561251,
+      "grad_norm": 0.47018300091574455,
+      "learning_rate": 1.3688346198248859e-05,
+      "loss": 0.3843,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7975397232188621,
+      "grad_norm": 0.5544200457498892,
+      "learning_rate": 1.3680627975603275e-05,
+      "loss": 0.3938,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7980522808815992,
+      "grad_norm": 0.528388172300649,
+      "learning_rate": 1.3672907215964e-05,
+      "loss": 0.3464,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7985648385443362,
+      "grad_norm": 0.44724235384198674,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.3394,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7990773962070733,
+      "grad_norm": 0.4694223298070473,
+      "learning_rate": 1.3657458106993257e-05,
+      "loss": 0.3559,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7995899538698104,
+      "grad_norm": 0.5003502257394156,
+      "learning_rate": 1.3649729768310598e-05,
+      "loss": 0.3608,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8001025115325474,
+      "grad_norm": 0.5103222288911216,
+      "learning_rate": 1.3641998913931849e-05,
+      "loss": 0.3851,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8006150691952845,
+      "grad_norm": 0.4788499220400211,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.3464,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8011276268580215,
+      "grad_norm": 0.5595722700992711,
+      "learning_rate": 1.362652967940279e-05,
+      "loss": 0.3537,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8016401845207586,
+      "grad_norm": 0.49583270331131696,
+      "learning_rate": 1.361879130991516e-05,
+      "loss": 0.3899,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8021527421834956,
+      "grad_norm": 0.4729735548100127,
+      "learning_rate": 1.3611050446056787e-05,
+      "loss": 0.3818,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8026652998462327,
+      "grad_norm": 0.4822222205734282,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.4205,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8031778575089698,
+      "grad_norm": 0.491134199383415,
+      "learning_rate": 1.3595561256572111e-05,
+      "loss": 0.369,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8036904151717068,
+      "grad_norm": 0.45549570646722004,
+      "learning_rate": 1.3587812941622238e-05,
+      "loss": 0.3694,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8042029728344439,
+      "grad_norm": 0.5181306296035669,
+      "learning_rate": 1.3580062153654484e-05,
+      "loss": 0.4131,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8047155304971809,
+      "grad_norm": 0.49198453856446295,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.3775,
+      "step": 1570
+    },
+    {
+      "epoch": 0.805228088159918,
+      "grad_norm": 0.5096834286631129,
+      "learning_rate": 1.3564553180036962e-05,
+      "loss": 0.3925,
+      "step": 1571
+    },
+    {
+      "epoch": 0.805740645822655,
+      "grad_norm": 0.5310978657524447,
+      "learning_rate": 1.3556795005077266e-05,
+      "loss": 0.3946,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8062532034853921,
+      "grad_norm": 0.5142794482687483,
+      "learning_rate": 1.3549034378479815e-05,
+      "loss": 0.3855,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8067657611481291,
+      "grad_norm": 0.45206131680860423,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.3261,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8072783188108662,
+      "grad_norm": 0.497437459502267,
+      "learning_rate": 1.3533505791770409e-05,
+      "loss": 0.3575,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8077908764736033,
+      "grad_norm": 0.5123506569666623,
+      "learning_rate": 1.3525737842362043e-05,
+      "loss": 0.3906,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8083034341363403,
+      "grad_norm": 0.5038398142280311,
+      "learning_rate": 1.3517967462723092e-05,
+      "loss": 0.3707,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8088159917990774,
+      "grad_norm": 0.43197961748311414,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.3502,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8093285494618144,
+      "grad_norm": 0.5208193508560429,
+      "learning_rate": 1.3502419434179068e-05,
+      "loss": 0.3642,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8098411071245515,
+      "grad_norm": 0.49361092122638256,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.3693,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8103536647872885,
+      "grad_norm": 0.4827302908741152,
+      "learning_rate": 1.3486861749006286e-05,
+      "loss": 0.3464,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8108662224500256,
+      "grad_norm": 0.5263941568429394,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.3608,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8113787801127627,
+      "grad_norm": 0.4867498997285664,
+      "learning_rate": 1.3471294450099327e-05,
+      "loss": 0.3533,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8118913377754997,
+      "grad_norm": 0.5066616502903957,
+      "learning_rate": 1.3463507208907332e-05,
+      "loss": 0.3748,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8124038954382368,
+      "grad_norm": 0.49917014932851567,
+      "learning_rate": 1.3455717580379266e-05,
+      "loss": 0.3856,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8129164531009738,
+      "grad_norm": 0.4838523040957045,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.3312,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8134290107637109,
+      "grad_norm": 0.4715788374030384,
+      "learning_rate": 1.3440131182793567e-05,
+      "loss": 0.3194,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8139415684264479,
+      "grad_norm": 0.44985296940239683,
+      "learning_rate": 1.3432334424479377e-05,
+      "loss": 0.3653,
+      "step": 1588
+    },
+    {
+      "epoch": 0.814454126089185,
+      "grad_norm": 0.4496239387502193,
+      "learning_rate": 1.3424535300315978e-05,
+      "loss": 0.3299,
+      "step": 1589
+    },
+    {
+      "epoch": 0.814966683751922,
+      "grad_norm": 0.5038103745007217,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.3957,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8154792414146591,
+      "grad_norm": 0.5054137558061381,
+      "learning_rate": 1.3408929975946379e-05,
+      "loss": 0.3894,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8159917990773962,
+      "grad_norm": 0.47693653520334184,
+      "learning_rate": 1.3401123786496664e-05,
+      "loss": 0.3518,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8165043567401332,
+      "grad_norm": 0.5052299601185096,
+      "learning_rate": 1.3393315252710695e-05,
+      "loss": 0.3623,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8170169144028703,
+      "grad_norm": 0.513723003581951,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.3255,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8175294720656073,
+      "grad_norm": 0.533631883386473,
+      "learning_rate": 1.3377691173660764e-05,
+      "loss": 0.3452,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8180420297283444,
+      "grad_norm": 0.5358195519401378,
+      "learning_rate": 1.3369875639166212e-05,
+      "loss": 0.3738,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8185545873910814,
+      "grad_norm": 0.45837028756255754,
+      "learning_rate": 1.3362057781874215e-05,
+      "loss": 0.4045,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8190671450538185,
+      "grad_norm": 0.5222483304901935,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.3441,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8195797027165556,
+      "grad_norm": 0.4357673666486136,
+      "learning_rate": 1.3346415120454354e-05,
+      "loss": 0.3243,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8200922603792927,
+      "grad_norm": 0.5252994499412944,
+      "learning_rate": 1.3338590327108709e-05,
+      "loss": 0.3854,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8206048180420298,
+      "grad_norm": 0.48094408816819517,
+      "learning_rate": 1.3330763232530048e-05,
+      "loss": 0.3721,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8211173757047668,
+      "grad_norm": 0.5121869935666045,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.4061,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8216299333675039,
+      "grad_norm": 0.49347500101736175,
+      "learning_rate": 1.3315102161255603e-05,
+      "loss": 0.3482,
+      "step": 1603
+    },
+    {
+      "epoch": 0.822142491030241,
+      "grad_norm": 0.506513166741962,
+      "learning_rate": 1.3307268195354718e-05,
+      "loss": 0.3921,
+      "step": 1604
+    },
+    {
+      "epoch": 0.822655048692978,
+      "grad_norm": 0.4550792524022784,
+      "learning_rate": 1.329943194981063e-05,
+      "loss": 0.3471,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8231676063557151,
+      "grad_norm": 0.47322652691278816,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4026,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8236801640184521,
+      "grad_norm": 0.4760578213923934,
+      "learning_rate": 1.3283752641399965e-05,
+      "loss": 0.3497,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8241927216811892,
+      "grad_norm": 0.4865318149350133,
+      "learning_rate": 1.3275909589340865e-05,
+      "loss": 0.3537,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8247052793439262,
+      "grad_norm": 0.44643422046289927,
+      "learning_rate": 1.3268064279253512e-05,
+      "loss": 0.3395,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8252178370066633,
+      "grad_norm": 0.46293658937683607,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.3407,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8257303946694003,
+      "grad_norm": 0.5236934792677871,
+      "learning_rate": 1.325236690662613e-05,
+      "loss": 0.3847,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8262429523321374,
+      "grad_norm": 0.5625794867507109,
+      "learning_rate": 1.3244514854906039e-05,
+      "loss": 0.3072,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8267555099948745,
+      "grad_norm": 0.5019321570359015,
+      "learning_rate": 1.3236660566797542e-05,
+      "loss": 0.3503,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8272680676576115,
+      "grad_norm": 0.5315371234499154,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.4159,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8277806253203486,
+      "grad_norm": 0.5247426867434019,
+      "learning_rate": 1.3220945303072172e-05,
+      "loss": 0.3913,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8282931829830856,
+      "grad_norm": 0.43074291440294993,
+      "learning_rate": 1.3213084338287563e-05,
+      "loss": 0.3436,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8288057406458227,
+      "grad_norm": 0.4749162375356055,
+      "learning_rate": 1.3205221158779063e-05,
+      "loss": 0.3785,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8293182983085597,
+      "grad_norm": 0.4975325981100459,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.4026,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8298308559712968,
+      "grad_norm": 0.4655212386130027,
+      "learning_rate": 1.3189488177271741e-05,
+      "loss": 0.355,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8303434136340339,
+      "grad_norm": 0.47660550944414287,
+      "learning_rate": 1.318161838611739e-05,
+      "loss": 0.3525,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8308559712967709,
+      "grad_norm": 0.4850312121109505,
+      "learning_rate": 1.3173746401928092e-05,
+      "loss": 0.3377,
+      "step": 1621
+    },
+    {
+      "epoch": 0.831368528959508,
+      "grad_norm": 0.4726408944063838,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.3544,
+      "step": 1622
+    },
+    {
+      "epoch": 0.831881086622245,
+      "grad_norm": 0.4925297311426149,
+      "learning_rate": 1.3157995876150252e-05,
+      "loss": 0.4056,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8323936442849821,
+      "grad_norm": 0.45736391660564446,
+      "learning_rate": 1.3150117345418277e-05,
+      "loss": 0.3418,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8329062019477191,
+      "grad_norm": 0.44046944876223126,
+      "learning_rate": 1.3142236643364481e-05,
+      "loss": 0.3812,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8334187596104562,
+      "grad_norm": 0.4736619049014531,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.3789,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8339313172731933,
+      "grad_norm": 0.5048014518671811,
+      "learning_rate": 1.3126468747021044e-05,
+      "loss": 0.3982,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8344438749359303,
+      "grad_norm": 0.4617308340286318,
+      "learning_rate": 1.3118581563599946e-05,
+      "loss": 0.3594,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8349564325986674,
+      "grad_norm": 0.4690004959147388,
+      "learning_rate": 1.31106922305941e-05,
+      "loss": 0.3418,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8354689902614044,
+      "grad_norm": 0.42267627794709234,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.3546,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8359815479241415,
+      "grad_norm": 0.45746998819670887,
+      "learning_rate": 1.3094907137581567e-05,
+      "loss": 0.3232,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8364941055868785,
+      "grad_norm": 0.47111366848355324,
+      "learning_rate": 1.3087011388455275e-05,
+      "loss": 0.3671,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8370066632496156,
+      "grad_norm": 0.4980450621429967,
+      "learning_rate": 1.307911351150501e-05,
+      "loss": 0.344,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8375192209123526,
+      "grad_norm": 0.44911815920692433,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.3569,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8380317785750897,
+      "grad_norm": 0.49551068210910626,
+      "learning_rate": 1.3063311395909531e-05,
+      "loss": 0.3459,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8385443362378268,
+      "grad_norm": 0.49056918032138747,
+      "learning_rate": 1.3055407168156438e-05,
+      "loss": 0.3703,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8390568939005638,
+      "grad_norm": 0.532210201057865,
+      "learning_rate": 1.3047500834363623e-05,
+      "loss": 0.3794,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8395694515633009,
+      "grad_norm": 0.46232465175855125,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.367,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8400820092260379,
+      "grad_norm": 0.5175977782705284,
+      "learning_rate": 1.3031681870459074e-05,
+      "loss": 0.3911,
+      "step": 1639
+    },
+    {
+      "epoch": 0.840594566888775,
+      "grad_norm": 0.5033204602990113,
+      "learning_rate": 1.3023769251251081e-05,
+      "loss": 0.3594,
+      "step": 1640
+    },
+    {
+      "epoch": 0.841107124551512,
+      "grad_norm": 0.4658343384811488,
+      "learning_rate": 1.301585454781084e-05,
+      "loss": 0.3661,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8416196822142491,
+      "grad_norm": 0.4636040377403396,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.4038,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8421322398769862,
+      "grad_norm": 0.4997818689029588,
+      "learning_rate": 1.3000018910056922e-05,
+      "loss": 0.3876,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8426447975397232,
+      "grad_norm": 0.4913728786774945,
+      "learning_rate": 1.2992097986658476e-05,
+      "loss": 0.3258,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8431573552024603,
+      "grad_norm": 0.45584892563056445,
+      "learning_rate": 1.2984175000858241e-05,
+      "loss": 0.4025,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8436699128651973,
+      "grad_norm": 0.49743151063881347,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.3673,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8441824705279344,
+      "grad_norm": 0.4972629907694658,
+      "learning_rate": 1.2968322863898533e-05,
+      "loss": 0.3669,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8446950281906714,
+      "grad_norm": 0.549334342638613,
+      "learning_rate": 1.2960393723665663e-05,
+      "loss": 0.3374,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8452075858534085,
+      "grad_norm": 0.5042401516018036,
+      "learning_rate": 1.2952462542884212e-05,
+      "loss": 0.3915,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8457201435161456,
+      "grad_norm": 0.5294943797732728,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.4052,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8462327011788826,
+      "grad_norm": 0.49315564790470895,
+      "learning_rate": 1.2936594081544258e-05,
+      "loss": 0.3695,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8467452588416197,
+      "grad_norm": 0.47220527606612267,
+      "learning_rate": 1.2928656811923621e-05,
+      "loss": 0.3824,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8472578165043567,
+      "grad_norm": 0.4600802608169791,
+      "learning_rate": 1.29207175236301e-05,
+      "loss": 0.3552,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8477703741670938,
+      "grad_norm": 0.5439737480211146,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.4041,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8482829318298308,
+      "grad_norm": 0.5041022507891273,
+      "learning_rate": 1.2904832912915481e-05,
+      "loss": 0.4008,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8487954894925679,
+      "grad_norm": 0.4927321179442074,
+      "learning_rate": 1.289688760144337e-05,
+      "loss": 0.3706,
+      "step": 1656
+    },
+    {
+      "epoch": 0.849308047155305,
+      "grad_norm": 0.47066293077723775,
+      "learning_rate": 1.2888940293196352e-05,
+      "loss": 0.3434,
+      "step": 1657
+    },
+    {
+      "epoch": 0.849820604818042,
+      "grad_norm": 0.49721116784413844,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.3706,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8503331624807791,
+      "grad_norm": 0.5253427302689918,
+      "learning_rate": 1.2873039708290753e-05,
+      "loss": 0.3534,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8508457201435161,
+      "grad_norm": 0.4615924652690999,
+      "learning_rate": 1.2865086442592165e-05,
+      "loss": 0.2998,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8513582778062532,
+      "grad_norm": 0.46356427216322754,
+      "learning_rate": 1.2857131202038668e-05,
+      "loss": 0.3732,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8518708354689902,
+      "grad_norm": 0.5233231161161653,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.4092,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8523833931317273,
+      "grad_norm": 0.4587025169578737,
+      "learning_rate": 1.2841214818301935e-05,
+      "loss": 0.3788,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8528959507944643,
+      "grad_norm": 0.47769819900901594,
+      "learning_rate": 1.2833253686089588e-05,
+      "loss": 0.3249,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8534085084572014,
+      "grad_norm": 0.49241960732109935,
+      "learning_rate": 1.2825290600964104e-05,
+      "loss": 0.3471,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8539210661199385,
+      "grad_norm": 0.5149341815103085,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.3233,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8544336237826755,
+      "grad_norm": 0.4673513098086092,
+      "learning_rate": 1.2809358593930332e-05,
+      "loss": 0.3546,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8549461814454126,
+      "grad_norm": 0.4902560474807241,
+      "learning_rate": 1.2801389683003701e-05,
+      "loss": 0.3503,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8554587391081496,
+      "grad_norm": 0.455212605296345,
+      "learning_rate": 1.2793418841127242e-05,
+      "loss": 0.3618,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8559712967708867,
+      "grad_norm": 0.4379285937641603,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.3119,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8564838544336237,
+      "grad_norm": 0.49159079957163815,
+      "learning_rate": 1.277747138650282e-05,
+      "loss": 0.3655,
+      "step": 1671
+    },
+    {
+      "epoch": 0.8569964120963608,
+      "grad_norm": 0.4739656489462452,
+      "learning_rate": 1.2769494784747164e-05,
+      "loss": 0.342,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8575089697590978,
+      "grad_norm": 0.525246026197525,
+      "learning_rate": 1.2761516274026289e-05,
+      "loss": 0.3808,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8580215274218349,
+      "grad_norm": 0.5346951954961683,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.4205,
+      "step": 1674
+    },
+    {
+      "epoch": 0.858534085084572,
+      "grad_norm": 0.45773200206666054,
+      "learning_rate": 1.2745553547687977e-05,
+      "loss": 0.3561,
+      "step": 1675
+    },
+    {
+      "epoch": 0.859046642747309,
+      "grad_norm": 0.49461224942627224,
+      "learning_rate": 1.2737569343073379e-05,
+      "loss": 0.4045,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8595592004100461,
+      "grad_norm": 0.48270270347095584,
+      "learning_rate": 1.2729583251499213e-05,
+      "loss": 0.3463,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8600717580727832,
+      "grad_norm": 0.49244140589943547,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.3875,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8605843157355203,
+      "grad_norm": 0.4936098645021005,
+      "learning_rate": 1.2713605429492193e-05,
+      "loss": 0.372,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8610968733982574,
+      "grad_norm": 0.46992093650307587,
+      "learning_rate": 1.2705613710072575e-05,
+      "loss": 0.3457,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8616094310609944,
+      "grad_norm": 0.485856663236743,
+      "learning_rate": 1.2697620125719862e-05,
+      "loss": 0.368,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8621219887237315,
+      "grad_norm": 0.5136985733606011,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.3479,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8626345463864685,
+      "grad_norm": 0.4759638745194652,
+      "learning_rate": 1.2681627384255796e-05,
+      "loss": 0.3906,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8631471040492056,
+      "grad_norm": 0.4957723817774408,
+      "learning_rate": 1.2673628238167965e-05,
+      "loss": 0.3656,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8636596617119426,
+      "grad_norm": 0.48616513737467726,
+      "learning_rate": 1.2665627249194078e-05,
+      "loss": 0.377,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8641722193746797,
+      "grad_norm": 0.5005361083294709,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.3697,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8646847770374168,
+      "grad_norm": 0.48580446224459717,
+      "learning_rate": 1.264961976464917e-05,
+      "loss": 0.3989,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8651973347001538,
+      "grad_norm": 0.510411229182461,
+      "learning_rate": 1.2641613280111837e-05,
+      "loss": 0.3507,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8657098923628909,
+      "grad_norm": 0.45939730298908515,
+      "learning_rate": 1.2633604974755807e-05,
+      "loss": 0.411,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8662224500256279,
+      "grad_norm": 0.48130840029102523,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.3897,
+      "step": 1690
+    },
+    {
+      "epoch": 0.866735007688365,
+      "grad_norm": 0.47768971681631955,
+      "learning_rate": 1.2617582923668855e-05,
+      "loss": 0.3482,
+      "step": 1691
+    },
+    {
+      "epoch": 0.867247565351102,
+      "grad_norm": 0.5205545256176649,
+      "learning_rate": 1.260956918898166e-05,
+      "loss": 0.4115,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8677601230138391,
+      "grad_norm": 0.4757777411284808,
+      "learning_rate": 1.2601553655563207e-05,
+      "loss": 0.3595,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8682726806765761,
+      "grad_norm": 0.5113058052151949,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.3709,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8687852383393132,
+      "grad_norm": 0.46279523466981726,
+      "learning_rate": 1.258551721463366e-05,
+      "loss": 0.3832,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8692977960020503,
+      "grad_norm": 0.523136626377708,
+      "learning_rate": 1.25774963181762e-05,
+      "loss": 0.3856,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8698103536647873,
+      "grad_norm": 0.47020879891319495,
+      "learning_rate": 1.2569473645094772e-05,
+      "loss": 0.3139,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8703229113275244,
+      "grad_norm": 0.4721726497801805,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.3969,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8708354689902614,
+      "grad_norm": 0.4789810619436148,
+      "learning_rate": 1.2553422991180766e-05,
+      "loss": 0.385,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8713480266529985,
+      "grad_norm": 0.49901817627630884,
+      "learning_rate": 1.254539502141163e-05,
+      "loss": 0.388,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8718605843157355,
+      "grad_norm": 0.4608258245165349,
+      "learning_rate": 1.2537365297145396e-05,
+      "loss": 0.3892,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8723731419784726,
+      "grad_norm": 0.4758484637329836,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.3563,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8728856996412097,
+      "grad_norm": 0.4563373288209085,
+      "learning_rate": 1.2521300607261826e-05,
+      "loss": 0.3762,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8733982573039467,
+      "grad_norm": 0.4590189612301811,
+      "learning_rate": 1.2513265652717613e-05,
+      "loss": 0.3559,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8739108149666838,
+      "grad_norm": 0.46759050027775556,
+      "learning_rate": 1.2505228965822511e-05,
+      "loss": 0.3678,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8744233726294208,
+      "grad_norm": 0.44129551920260235,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.324,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8749359302921579,
+      "grad_norm": 0.4592803130998646,
+      "learning_rate": 1.2489150417139067e-05,
+      "loss": 0.3534,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8754484879548949,
+      "grad_norm": 0.5119333173302258,
+      "learning_rate": 1.2481108566433388e-05,
+      "loss": 0.4153,
+      "step": 1708
+    },
+    {
+      "epoch": 0.875961045617632,
+      "grad_norm": 0.48485277370599195,
+      "learning_rate": 1.2473065005542155e-05,
+      "loss": 0.3584,
+      "step": 1709
+    },
+    {
+      "epoch": 0.876473603280369,
+      "grad_norm": 0.46076948911352483,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.3426,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8769861609431061,
+      "grad_norm": 0.48022957083490897,
+      "learning_rate": 1.2456972775381364e-05,
+      "loss": 0.3487,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8774987186058432,
+      "grad_norm": 0.455271770300195,
+      "learning_rate": 1.2448924117203904e-05,
+      "loss": 0.3504,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8780112762685802,
+      "grad_norm": 0.4838509970665582,
+      "learning_rate": 1.2440873771025079e-05,
+      "loss": 0.3133,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8785238339313173,
+      "grad_norm": 0.48731041042108597,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.3801,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8790363915940543,
+      "grad_norm": 0.4868457151072363,
+      "learning_rate": 1.2424768036860359e-05,
+      "loss": 0.3617,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8795489492567914,
+      "grad_norm": 0.4653625887608884,
+      "learning_rate": 1.2416712659975869e-05,
+      "loss": 0.3365,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8800615069195284,
+      "grad_norm": 0.466404373344266,
+      "learning_rate": 1.2408655617292822e-05,
+      "loss": 0.3746,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8805740645822655,
+      "grad_norm": 0.4921563298343448,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.3742,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8810866222450026,
+      "grad_norm": 0.49811509749663496,
+      "learning_rate": 1.2392536556746517e-05,
+      "loss": 0.4161,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8815991799077396,
+      "grad_norm": 0.4791822773011124,
+      "learning_rate": 1.238447454999385e-05,
+      "loss": 0.3449,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8821117375704767,
+      "grad_norm": 0.4653066381142429,
+      "learning_rate": 1.2376410899663802e-05,
+      "loss": 0.4053,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8826242952332137,
+      "grad_norm": 0.5351694000233692,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.3524,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8831368528959508,
+      "grad_norm": 0.481932074004,
+      "learning_rate": 1.236027869050523e-05,
+      "loss": 0.4065,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8836494105586878,
+      "grad_norm": 0.49324029614102327,
+      "learning_rate": 1.2352210142796359e-05,
+      "loss": 0.3902,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8841619682214249,
+      "grad_norm": 0.44210356662255273,
+      "learning_rate": 1.2344139973749406e-05,
+      "loss": 0.3465,
+      "step": 1725
+    },
+    {
+      "epoch": 0.884674525884162,
+      "grad_norm": 0.4861046433268693,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.3174,
+      "step": 1726
+    },
+    {
+      "epoch": 0.885187083546899,
+      "grad_norm": 0.4934440072160213,
+      "learning_rate": 1.2327994793892885e-05,
+      "loss": 0.3252,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8856996412096361,
+      "grad_norm": 0.4923372795437099,
+      "learning_rate": 1.231991979421192e-05,
+      "loss": 0.3669,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8862121988723731,
+      "grad_norm": 0.5654169148556434,
+      "learning_rate": 1.2311843195450046e-05,
+      "loss": 0.3976,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8867247565351102,
+      "grad_norm": 0.45075120521575807,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.2904,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8872373141978472,
+      "grad_norm": 0.47731801587066636,
+      "learning_rate": 1.2295685222952958e-05,
+      "loss": 0.3764,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8877498718605843,
+      "grad_norm": 0.4836144982694259,
+      "learning_rate": 1.228760386035515e-05,
+      "loss": 0.3457,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8882624295233214,
+      "grad_norm": 0.4518459047305948,
+      "learning_rate": 1.2279520920951262e-05,
+      "loss": 0.3751,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8887749871860584,
+      "grad_norm": 0.5078593098528896,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.3342,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8892875448487955,
+      "grad_norm": 0.5201564608811418,
+      "learning_rate": 1.2263350334012059e-05,
+      "loss": 0.3376,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8898001025115325,
+      "grad_norm": 0.45052811931586273,
+      "learning_rate": 1.2255262697622848e-05,
+      "loss": 0.3161,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8903126601742696,
+      "grad_norm": 0.5019598302757895,
+      "learning_rate": 1.2247173506719768e-05,
+      "loss": 0.3893,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8908252178370066,
+      "grad_norm": 0.49170209595822034,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.3622,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8913377754997437,
+      "grad_norm": 0.4793288403568117,
+      "learning_rate": 1.2230990483676024e-05,
+      "loss": 0.3411,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8918503331624807,
+      "grad_norm": 0.48267165655854066,
+      "learning_rate": 1.222289666269004e-05,
+      "loss": 0.3485,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8923628908252178,
+      "grad_norm": 0.47115692028256934,
+      "learning_rate": 1.2214801309499538e-05,
+      "loss": 0.3411,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8928754484879549,
+      "grad_norm": 0.4768162501753483,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.3646,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8933880061506919,
+      "grad_norm": 0.4773188849921369,
+      "learning_rate": 1.2198606028825978e-05,
+      "loss": 0.3581,
+      "step": 1743
+    },
+    {
+      "epoch": 0.893900563813429,
+      "grad_norm": 0.4343151750668318,
+      "learning_rate": 1.2190506112506054e-05,
+      "loss": 0.3215,
+      "step": 1744
+    },
+    {
+      "epoch": 0.894413121476166,
+      "grad_norm": 0.5044418407445775,
+      "learning_rate": 1.2182404686307857e-05,
+      "loss": 0.3824,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8949256791389031,
+      "grad_norm": 0.4114895414326818,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.3041,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8954382368016401,
+      "grad_norm": 0.4702440595088517,
+      "learning_rate": 1.2166197326614399e-05,
+      "loss": 0.3618,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8959507944643772,
+      "grad_norm": 0.5144740013164824,
+      "learning_rate": 1.2158091404290588e-05,
+      "loss": 0.3494,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8964633521271143,
+      "grad_norm": 0.505050705151979,
+      "learning_rate": 1.214998399443141e-05,
+      "loss": 0.3706,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8969759097898513,
+      "grad_norm": 0.4765860828696622,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.3665,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8974884674525884,
+      "grad_norm": 0.5011865131365494,
+      "learning_rate": 1.2133764734461173e-05,
+      "loss": 0.3247,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8980010251153254,
+      "grad_norm": 0.44850955009446974,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.3793,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8985135827780625,
+      "grad_norm": 0.442508038651308,
+      "learning_rate": 1.2117539591422308e-05,
+      "loss": 0.375,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8990261404407995,
+      "grad_norm": 0.477677730501286,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.3521,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8995386981035366,
+      "grad_norm": 0.4769476758147146,
+      "learning_rate": 1.2101308610049653e-05,
+      "loss": 0.3418,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9000512557662738,
+      "grad_norm": 0.48612983174160307,
+      "learning_rate": 1.2093190943972204e-05,
+      "loss": 0.3936,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9005638134290108,
+      "grad_norm": 0.4873404674227811,
+      "learning_rate": 1.208507183509416e-05,
+      "loss": 0.3364,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9010763710917479,
+      "grad_norm": 0.46497822828839724,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.3702,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9015889287544849,
+      "grad_norm": 0.5059548202520097,
+      "learning_rate": 1.2068829311322731e-05,
+      "loss": 0.3731,
+      "step": 1759
+    },
+    {
+      "epoch": 0.902101486417222,
+      "grad_norm": 0.526712008029892,
+      "learning_rate": 1.2060705907625042e-05,
+      "loss": 0.418,
+      "step": 1760
+    },
+    {
+      "epoch": 0.902614044079959,
+      "grad_norm": 0.46989009933406006,
+      "learning_rate": 1.2052581083518135e-05,
+      "loss": 0.3416,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9031266017426961,
+      "grad_norm": 0.45719646295959654,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.3438,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9036391594054332,
+      "grad_norm": 0.45082620872152634,
+      "learning_rate": 1.2036327196478866e-05,
+      "loss": 0.3667,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9041517170681702,
+      "grad_norm": 0.46081902789565976,
+      "learning_rate": 1.2028198144750023e-05,
+      "loss": 0.3666,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9046642747309073,
+      "grad_norm": 0.5188025431340871,
+      "learning_rate": 1.202006769501901e-05,
+      "loss": 0.3783,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9051768323936443,
+      "grad_norm": 0.45207967784513375,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.3587,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9056893900563814,
+      "grad_norm": 0.4858405364026096,
+      "learning_rate": 1.2003802623968144e-05,
+      "loss": 0.3872,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9062019477191184,
+      "grad_norm": 0.4414400298299832,
+      "learning_rate": 1.199566801385953e-05,
+      "loss": 0.3283,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9067145053818555,
+      "grad_norm": 0.5059156240702507,
+      "learning_rate": 1.19875320281712e-05,
+      "loss": 0.3831,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9072270630445926,
+      "grad_norm": 0.44823684648621226,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3343,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9077396207073296,
+      "grad_norm": 0.44635450939405025,
+      "learning_rate": 1.197125595248834e-05,
+      "loss": 0.3775,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9082521783700667,
+      "grad_norm": 0.46408384697351995,
+      "learning_rate": 1.196311587371262e-05,
+      "loss": 0.3609,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9087647360328037,
+      "grad_norm": 0.5124718347848816,
+      "learning_rate": 1.1954974441794827e-05,
+      "loss": 0.3768,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9092772936955408,
+      "grad_norm": 0.4893040692014562,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.3604,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9097898513582778,
+      "grad_norm": 0.4664360168550382,
+      "learning_rate": 1.1938687540980925e-05,
+      "loss": 0.3695,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9103024090210149,
+      "grad_norm": 0.4455614195645772,
+      "learning_rate": 1.1930542083311096e-05,
+      "loss": 0.3603,
+      "step": 1776
+    },
+    {
+      "epoch": 0.910814966683752,
+      "grad_norm": 0.5829123642467536,
+      "learning_rate": 1.1922395294951747e-05,
+      "loss": 0.35,
+      "step": 1777
+    },
+    {
+      "epoch": 0.911327524346489,
+      "grad_norm": 0.4879683223112704,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.3809,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9118400820092261,
+      "grad_norm": 0.4910955291014004,
+      "learning_rate": 1.190609774862715e-05,
+      "loss": 0.3588,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9123526396719631,
+      "grad_norm": 0.530668938103623,
+      "learning_rate": 1.1897947001895524e-05,
+      "loss": 0.3814,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9128651973347002,
+      "grad_norm": 0.4918163081400432,
+      "learning_rate": 1.1889794946941603e-05,
+      "loss": 0.3575,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9133777549974372,
+      "grad_norm": 0.45955171681091883,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.3692,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9138903126601743,
+      "grad_norm": 0.5069018469393888,
+      "learning_rate": 1.187348693484406e-05,
+      "loss": 0.3596,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9144028703229113,
+      "grad_norm": 0.504037585546518,
+      "learning_rate": 1.1865330988941271e-05,
+      "loss": 0.3778,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9149154279856484,
+      "grad_norm": 0.4467496879685425,
+      "learning_rate": 1.1857173757297843e-05,
+      "loss": 0.3677,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9154279856483855,
+      "grad_norm": 0.46467799130063575,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.3005,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9159405433111225,
+      "grad_norm": 0.4855932227917054,
+      "learning_rate": 1.1840855459280526e-05,
+      "loss": 0.3677,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9164531009738596,
+      "grad_norm": 0.4923625624669849,
+      "learning_rate": 1.1832694404154557e-05,
+      "loss": 0.3591,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9169656586365966,
+      "grad_norm": 0.44961270284250854,
+      "learning_rate": 1.1824532085783788e-05,
+      "loss": 0.3432,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9174782162993337,
+      "grad_norm": 0.491266322914333,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.3687,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9179907739620707,
+      "grad_norm": 0.49209823950683035,
+      "learning_rate": 1.1808203681813304e-05,
+      "loss": 0.354,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9185033316248078,
+      "grad_norm": 0.48719441938724395,
+      "learning_rate": 1.1800037607468477e-05,
+      "loss": 0.3705,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9190158892875449,
+      "grad_norm": 0.4841768735010603,
+      "learning_rate": 1.1791870292388624e-05,
+      "loss": 0.3819,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9195284469502819,
+      "grad_norm": 0.45330968664577,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3442,
+      "step": 1794
+    },
+    {
+      "epoch": 0.920041004613019,
+      "grad_norm": 0.48162693847804305,
+      "learning_rate": 1.1775531962543036e-05,
+      "loss": 0.3527,
+      "step": 1795
+    },
+    {
+      "epoch": 0.920553562275756,
+      "grad_norm": 0.4366344549611428,
+      "learning_rate": 1.1767360959039032e-05,
+      "loss": 0.3328,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9210661199384931,
+      "grad_norm": 0.45082468289790384,
+      "learning_rate": 1.1759188737323454e-05,
+      "loss": 0.3576,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9215786776012301,
+      "grad_norm": 0.4419470701162587,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.3054,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9220912352639672,
+      "grad_norm": 0.4810846365971059,
+      "learning_rate": 1.1742840661790294e-05,
+      "loss": 0.3976,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9226037929267042,
+      "grad_norm": 0.44623264435936594,
+      "learning_rate": 1.1734664819241159e-05,
+      "loss": 0.3322,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9231163505894413,
+      "grad_norm": 0.4674098090258592,
+      "learning_rate": 1.1726487781017337e-05,
+      "loss": 0.3712,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9236289082521784,
+      "grad_norm": 0.4545636669391432,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.3313,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9241414659149154,
+      "grad_norm": 0.49390836412518435,
+      "learning_rate": 1.1710130140091611e-05,
+      "loss": 0.3793,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9246540235776525,
+      "grad_norm": 0.46189473223712296,
+      "learning_rate": 1.1701949548664753e-05,
+      "loss": 0.3769,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9251665812403895,
+      "grad_norm": 0.47592566086591387,
+      "learning_rate": 1.1693767784113281e-05,
+      "loss": 0.38,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9256791389031266,
+      "grad_norm": 0.4953977193860679,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.3483,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9261916965658636,
+      "grad_norm": 0.4482992935260982,
+      "learning_rate": 1.1677400758195501e-05,
+      "loss": 0.319,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9267042542286007,
+      "grad_norm": 0.48959098694005376,
+      "learning_rate": 1.16692155081107e-05,
+      "loss": 0.3501,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9272168118913378,
+      "grad_norm": 0.504559967531381,
+      "learning_rate": 1.1661029107464298e-05,
+      "loss": 0.3373,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9277293695540748,
+      "grad_norm": 0.45201612743380687,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.3267,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9282419272168119,
+      "grad_norm": 0.5035552675449941,
+      "learning_rate": 1.1644652877058464e-05,
+      "loss": 0.3454,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9287544848795489,
+      "grad_norm": 0.46071392027421326,
+      "learning_rate": 1.1636463058586882e-05,
+      "loss": 0.3744,
+      "step": 1812
+    },
+    {
+      "epoch": 0.929267042542286,
+      "grad_norm": 0.4885859523913591,
+      "learning_rate": 1.1628272112129402e-05,
+      "loss": 0.353,
+      "step": 1813
+    },
+    {
+      "epoch": 0.929779600205023,
+      "grad_norm": 0.4911928137040222,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3552,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9302921578677601,
+      "grad_norm": 0.5208107196713049,
+      "learning_rate": 1.161188685784103e-05,
+      "loss": 0.3993,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9308047155304972,
+      "grad_norm": 0.47481907253015926,
+      "learning_rate": 1.1603692561304214e-05,
+      "loss": 0.3563,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9313172731932342,
+      "grad_norm": 0.4770933397790573,
+      "learning_rate": 1.1595497159369638e-05,
+      "loss": 0.3932,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9318298308559713,
+      "grad_norm": 0.46739731323151784,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3684,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9323423885187083,
+      "grad_norm": 0.45101309394910144,
+      "learning_rate": 1.1579103061903765e-05,
+      "loss": 0.2908,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9328549461814454,
+      "grad_norm": 0.5052199499693449,
+      "learning_rate": 1.1570904377672642e-05,
+      "loss": 0.3436,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9333675038441824,
+      "grad_norm": 0.47848185124641723,
+      "learning_rate": 1.156270461064409e-05,
+      "loss": 0.3386,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9338800615069195,
+      "grad_norm": 0.47118770233018586,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.3772,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9343926191696565,
+      "grad_norm": 0.5079756211875055,
+      "learning_rate": 1.1546301850803283e-05,
+      "loss": 0.3337,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9349051768323936,
+      "grad_norm": 0.45320969716931947,
+      "learning_rate": 1.153809886929717e-05,
+      "loss": 0.362,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9354177344951307,
+      "grad_norm": 0.4462573711792128,
+      "learning_rate": 1.1529894827605902e-05,
+      "loss": 0.3469,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9359302921578677,
+      "grad_norm": 0.43540724120632923,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.3591,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9364428498206048,
+      "grad_norm": 0.46813608598310336,
+      "learning_rate": 1.1513483586288262e-05,
+      "loss": 0.3346,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9369554074833418,
+      "grad_norm": 0.5197737191910624,
+      "learning_rate": 1.150527639797387e-05,
+      "loss": 0.378,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9374679651460789,
+      "grad_norm": 0.4767280606615702,
+      "learning_rate": 1.1497068172098295e-05,
+      "loss": 0.3745,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9379805228088159,
+      "grad_norm": 0.4551302301806354,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.3522,
+      "step": 1830
+    },
+    {
+      "epoch": 0.938493080471553,
+      "grad_norm": 0.4640152974312194,
+      "learning_rate": 1.148064863029545e-05,
+      "loss": 0.3155,
+      "step": 1831
+    },
+    {
+      "epoch": 0.93900563813429,
+      "grad_norm": 0.45070171061830544,
+      "learning_rate": 1.147243732568589e-05,
+      "loss": 0.3447,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9395181957970271,
+      "grad_norm": 0.4537746436410683,
+      "learning_rate": 1.1464225006150553e-05,
+      "loss": 0.3545,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9400307534597643,
+      "grad_norm": 0.4591881078673107,
+      "learning_rate": 1.1456011677350049e-05,
+      "loss": 0.3609,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9405433111225013,
+      "grad_norm": 0.4522354883140988,
+      "learning_rate": 1.1447797344945686e-05,
+      "loss": 0.3512,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9410558687852384,
+      "grad_norm": 0.44227969760952035,
+      "learning_rate": 1.1439582014599462e-05,
+      "loss": 0.3497,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9415684264479754,
+      "grad_norm": 0.44964153063704015,
+      "learning_rate": 1.1431365691974059e-05,
+      "loss": 0.3858,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9420809841107125,
+      "grad_norm": 0.4695904201794387,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3547,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9425935417734496,
+      "grad_norm": 0.49965348611520866,
+      "learning_rate": 1.1414930092539891e-05,
+      "loss": 0.3607,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9431060994361866,
+      "grad_norm": 0.4632882969922601,
+      "learning_rate": 1.14067108270599e-05,
+      "loss": 0.3702,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9436186570989237,
+      "grad_norm": 0.45398465207104105,
+      "learning_rate": 1.139849059195828e-05,
+      "loss": 0.3682,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9441312147616607,
+      "grad_norm": 0.4982723914095333,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.37,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9446437724243978,
+      "grad_norm": 0.5175006841608409,
+      "learning_rate": 1.1382047235555078e-05,
+      "loss": 0.2822,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9451563300871348,
+      "grad_norm": 0.45218804586177236,
+      "learning_rate": 1.1373824125587625e-05,
+      "loss": 0.3305,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9456688877498719,
+      "grad_norm": 0.4340810046841615,
+      "learning_rate": 1.1365600068666781e-05,
+      "loss": 0.3468,
+      "step": 1845
+    },
+    {
+      "epoch": 0.946181445412609,
+      "grad_norm": 0.45757420531014426,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.3278,
+      "step": 1846
+    },
+    {
+      "epoch": 0.946694003075346,
+      "grad_norm": 0.498043179779521,
+      "learning_rate": 1.1349149136640367e-05,
+      "loss": 0.3961,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9472065607380831,
+      "grad_norm": 0.4865732359890721,
+      "learning_rate": 1.1340922272874143e-05,
+      "loss": 0.3611,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9477191184008201,
+      "grad_norm": 0.46323741156704906,
+      "learning_rate": 1.1332694484833213e-05,
+      "loss": 0.3609,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9482316760635572,
+      "grad_norm": 0.4441251364031231,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.3344,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9487442337262942,
+      "grad_norm": 0.4508190467982966,
+      "learning_rate": 1.1316236158612952e-05,
+      "loss": 0.3823,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9492567913890313,
+      "grad_norm": 0.5203448555034247,
+      "learning_rate": 1.1308005631778061e-05,
+      "loss": 0.344,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9497693490517684,
+      "grad_norm": 0.4829487600186343,
+      "learning_rate": 1.129977420335734e-05,
+      "loss": 0.3462,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9502819067145054,
+      "grad_norm": 0.461204797550412,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.3394,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9507944643772425,
+      "grad_norm": 0.4839228198145935,
+      "learning_rate": 1.128330866445414e-05,
+      "loss": 0.355,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9513070220399795,
+      "grad_norm": 0.4823515652068172,
+      "learning_rate": 1.127507456532108e-05,
+      "loss": 0.3662,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9518195797027166,
+      "grad_norm": 0.4739562941849806,
+      "learning_rate": 1.1266839587301004e-05,
+      "loss": 0.374,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9523321373654536,
+      "grad_norm": 0.5019270476303439,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.3629,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9528446950281907,
+      "grad_norm": 0.5138303759994045,
+      "learning_rate": 1.1250367017305328e-05,
+      "loss": 0.3695,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9533572526909277,
+      "grad_norm": 0.43677114557180474,
+      "learning_rate": 1.1242129436683988e-05,
+      "loss": 0.3674,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9538698103536648,
+      "grad_norm": 0.4511049272324161,
+      "learning_rate": 1.1233890999884143e-05,
+      "loss": 0.3536,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9543823680164019,
+      "grad_norm": 0.44915543130334235,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.3061,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9548949256791389,
+      "grad_norm": 0.4727316103043516,
+      "learning_rate": 1.121741158046399e-05,
+      "loss": 0.3747,
+      "step": 1863
+    },
+    {
+      "epoch": 0.955407483341876,
+      "grad_norm": 0.49812547164303755,
+      "learning_rate": 1.1209170609202657e-05,
+      "loss": 0.3886,
+      "step": 1864
+    },
+    {
+      "epoch": 0.955920041004613,
+      "grad_norm": 0.4482348596286208,
+      "learning_rate": 1.120092880448078e-05,
+      "loss": 0.3289,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9564325986673501,
+      "grad_norm": 0.5013434728589455,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.3907,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9569451563300871,
+      "grad_norm": 0.4483746349991857,
+      "learning_rate": 1.1184442717379687e-05,
+      "loss": 0.3487,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9574577139928242,
+      "grad_norm": 0.45423612724481643,
+      "learning_rate": 1.1176198446364048e-05,
+      "loss": 0.3906,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9579702716555613,
+      "grad_norm": 0.4701034106051296,
+      "learning_rate": 1.1167953364615007e-05,
+      "loss": 0.3546,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9584828293182983,
+      "grad_norm": 0.4720419379094272,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.3196,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9589953869810354,
+      "grad_norm": 0.46189620084691935,
+      "learning_rate": 1.1151460791650046e-05,
+      "loss": 0.3326,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9595079446437724,
+      "grad_norm": 0.4733154473115124,
+      "learning_rate": 1.1143213311802177e-05,
+      "loss": 0.3806,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9600205023065095,
+      "grad_norm": 0.4683899603857252,
+      "learning_rate": 1.1134965043956987e-05,
+      "loss": 0.3507,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9605330599692465,
+      "grad_norm": 0.4885326937593014,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.3781,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9610456176319836,
+      "grad_norm": 0.49464420483541016,
+      "learning_rate": 1.1118466167016763e-05,
+      "loss": 0.3442,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9615581752947207,
+      "grad_norm": 0.4500772557599538,
+      "learning_rate": 1.1110215569294122e-05,
+      "loss": 0.3493,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9620707329574577,
+      "grad_norm": 0.42311834665336145,
+      "learning_rate": 1.1101964206318938e-05,
+      "loss": 0.3278,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9625832906201948,
+      "grad_norm": 0.5075989501334868,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.3543,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9630958482829318,
+      "grad_norm": 0.5041611460762169,
+      "learning_rate": 1.1085459207361579e-05,
+      "loss": 0.3583,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9636084059456689,
+      "grad_norm": 0.43746630432044575,
+      "learning_rate": 1.1077205582756015e-05,
+      "loss": 0.3093,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9641209636084059,
+      "grad_norm": 0.44007132124353404,
+      "learning_rate": 1.1068951215651132e-05,
+      "loss": 0.3454,
+      "step": 1881
+    },
+    {
+      "epoch": 0.964633521271143,
+      "grad_norm": 0.4545111362067691,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.349,
+      "step": 1882
+    },
+    {
+      "epoch": 0.96514607893388,
+      "grad_norm": 0.5007552454567603,
+      "learning_rate": 1.1052440276702276e-05,
+      "loss": 0.3852,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9656586365966171,
+      "grad_norm": 0.46257908600759434,
+      "learning_rate": 1.1044183716239019e-05,
+      "loss": 0.3418,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9661711942593542,
+      "grad_norm": 0.4429635047286092,
+      "learning_rate": 1.1035926436037845e-05,
+      "loss": 0.3285,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9666837519220912,
+      "grad_norm": 0.4815475586486571,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.3576,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9671963095848283,
+      "grad_norm": 0.5228740424952939,
+      "learning_rate": 1.1019409739188653e-05,
+      "loss": 0.3575,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9677088672475653,
+      "grad_norm": 0.4636063161253621,
+      "learning_rate": 1.101115033392531e-05,
+      "loss": 0.3653,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9682214249103024,
+      "grad_norm": 0.5471188292309686,
+      "learning_rate": 1.1002890231693395e-05,
+      "loss": 0.3987,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9687339825730394,
+      "grad_norm": 0.522057305860813,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3738,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9692465402357765,
+      "grad_norm": 0.5031642756429588,
+      "learning_rate": 1.0986367959098517e-05,
+      "loss": 0.3911,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9697590978985136,
+      "grad_norm": 0.47370805652469894,
+      "learning_rate": 1.0978105800124072e-05,
+      "loss": 0.3681,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9702716555612506,
+      "grad_norm": 0.5076319039454857,
+      "learning_rate": 1.0969842966958087e-05,
+      "loss": 0.3955,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9707842132239877,
+      "grad_norm": 0.4570901214457225,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.3049,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9712967708867247,
+      "grad_norm": 0.46992197733097607,
+      "learning_rate": 1.0953315300833665e-05,
+      "loss": 0.3845,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9718093285494618,
+      "grad_norm": 0.42641000484890357,
+      "learning_rate": 1.0945050479267472e-05,
+      "loss": 0.3826,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9723218862121988,
+      "grad_norm": 0.4517146637779568,
+      "learning_rate": 1.0936785006294197e-05,
+      "loss": 0.3302,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9728344438749359,
+      "grad_norm": 0.45411195332811294,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.3137,
+      "step": 1898
+    },
+    {
+      "epoch": 0.973347001537673,
+      "grad_norm": 0.5305795910325141,
+      "learning_rate": 1.092025212891587e-05,
+      "loss": 0.361,
+      "step": 1899
+    },
+    {
+      "epoch": 0.97385955920041,
+      "grad_norm": 0.4426795405762844,
+      "learning_rate": 1.0911984735906635e-05,
+      "loss": 0.3568,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9743721168631471,
+      "grad_norm": 0.4591907054579869,
+      "learning_rate": 1.090371671428198e-05,
+      "loss": 0.3338,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9748846745258841,
+      "grad_norm": 0.49478699384910746,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.3697,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9753972321886212,
+      "grad_norm": 0.49729630204226066,
+      "learning_rate": 1.0887178807982837e-05,
+      "loss": 0.3913,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9759097898513582,
+      "grad_norm": 0.47115409064859703,
+      "learning_rate": 1.0878908934707646e-05,
+      "loss": 0.3578,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9764223475140953,
+      "grad_norm": 0.4815145779920411,
+      "learning_rate": 1.0870638455615613e-05,
+      "loss": 0.3749,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9769349051768323,
+      "grad_norm": 0.47785096566388546,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.3849,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9774474628395694,
+      "grad_norm": 0.47303677952040757,
+      "learning_rate": 1.0854095702784222e-05,
+      "loss": 0.3496,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9779600205023065,
+      "grad_norm": 0.45425446696618815,
+      "learning_rate": 1.08458234404475e-05,
+      "loss": 0.3449,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9784725781650435,
+      "grad_norm": 0.5068426869749849,
+      "learning_rate": 1.08375505950992e-05,
+      "loss": 0.3345,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9789851358277806,
+      "grad_norm": 0.4686719623260925,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.3604,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9794976934905176,
+      "grad_norm": 0.5064816342901534,
+      "learning_rate": 1.0821003178177572e-05,
+      "loss": 0.3669,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9800102511532548,
+      "grad_norm": 0.43401452668855695,
+      "learning_rate": 1.0812728618010098e-05,
+      "loss": 0.3748,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9805228088159919,
+      "grad_norm": 0.4380014067068782,
+      "learning_rate": 1.0804453497642734e-05,
+      "loss": 0.3543,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9810353664787289,
+      "grad_norm": 0.4911600054186239,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.3238,
+      "step": 1914
+    },
+    {
+      "epoch": 0.981547924141466,
+      "grad_norm": 0.4452603037571246,
+      "learning_rate": 1.0787901599124322e-05,
+      "loss": 0.3607,
+      "step": 1915
+    },
+    {
+      "epoch": 0.982060481804203,
+      "grad_norm": 0.4543450075553829,
+      "learning_rate": 1.0779624832382211e-05,
+      "loss": 0.3954,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9825730394669401,
+      "grad_norm": 0.45434740541055135,
+      "learning_rate": 1.0771347528258084e-05,
+      "loss": 0.3146,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9830855971296771,
+      "grad_norm": 0.46695126993206415,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.3117,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9835981547924142,
+      "grad_norm": 0.4695484136630176,
+      "learning_rate": 1.0754791330685763e-05,
+      "loss": 0.3794,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9841107124551512,
+      "grad_norm": 0.48428104942896993,
+      "learning_rate": 1.0746512448649473e-05,
+      "loss": 0.3688,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9846232701178883,
+      "grad_norm": 0.44739118821705093,
+      "learning_rate": 1.073823305205496e-05,
+      "loss": 0.3548,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9851358277806254,
+      "grad_norm": 0.4661729380939632,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.3714,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9856483854433624,
+      "grad_norm": 0.4698060147715302,
+      "learning_rate": 1.0721672738019015e-05,
+      "loss": 0.3645,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9861609431060995,
+      "grad_norm": 0.44612072624981586,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.3416,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9866735007688365,
+      "grad_norm": 0.4767431145323526,
+      "learning_rate": 1.0705110434236888e-05,
+      "loss": 0.3432,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9871860584315736,
+      "grad_norm": 0.4863407358547148,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.3485,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9876986160943106,
+      "grad_norm": 0.4504412738034094,
+      "learning_rate": 1.0688546186373009e-05,
+      "loss": 0.3495,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9882111737570477,
+      "grad_norm": 0.471572956795188,
+      "learning_rate": 1.0680263347682021e-05,
+      "loss": 0.349,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9887237314197848,
+      "grad_norm": 0.48828829986067745,
+      "learning_rate": 1.067198004009718e-05,
+      "loss": 0.3877,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9892362890825218,
+      "grad_norm": 0.48544067031602056,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.3831,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9897488467452589,
+      "grad_norm": 0.4754453092286183,
+      "learning_rate": 1.0655412041084438e-05,
+      "loss": 0.3811,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9902614044079959,
+      "grad_norm": 0.5100165981877349,
+      "learning_rate": 1.0647127361076573e-05,
+      "loss": 0.3616,
+      "step": 1932
+    },
+    {
+      "epoch": 0.990773962070733,
+      "grad_norm": 0.44950914534172914,
+      "learning_rate": 1.0638842235014923e-05,
+      "loss": 0.3036,
+      "step": 1933
+    },
+    {
+      "epoch": 0.99128651973347,
+      "grad_norm": 0.44908258641054244,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.3636,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9917990773962071,
+      "grad_norm": 0.4732759391036496,
+      "learning_rate": 1.0622270667573757e-05,
+      "loss": 0.3482,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9923116350589442,
+      "grad_norm": 0.46264774692217975,
+      "learning_rate": 1.061398423761674e-05,
+      "loss": 0.3961,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9928241927216812,
+      "grad_norm": 0.4592759186758219,
+      "learning_rate": 1.0605697384450922e-05,
+      "loss": 0.318,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9933367503844183,
+      "grad_norm": 0.4894242146736082,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.3603,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9938493080471553,
+      "grad_norm": 0.4610663120779362,
+      "learning_rate": 1.0589122431341129e-05,
+      "loss": 0.3353,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9943618657098924,
+      "grad_norm": 0.4870536397582043,
+      "learning_rate": 1.0580834342821978e-05,
+      "loss": 0.3655,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9948744233726294,
+      "grad_norm": 0.44887226495907906,
+      "learning_rate": 1.0572545853943688e-05,
+      "loss": 0.3917,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9953869810353665,
+      "grad_norm": 0.4187460549165628,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.3398,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9958995386981035,
+      "grad_norm": 0.444250330330165,
+      "learning_rate": 1.0555967697962399e-05,
+      "loss": 0.3571,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9964120963608406,
+      "grad_norm": 0.467391000457538,
+      "learning_rate": 1.0547678042286435e-05,
+      "loss": 0.3898,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9969246540235777,
+      "grad_norm": 0.5058735919491707,
+      "learning_rate": 1.0539388009105404e-05,
+      "loss": 0.3793,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9974372116863147,
+      "grad_norm": 0.5105637637815359,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.3589,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9979497693490518,
+      "grad_norm": 0.4110528885572847,
+      "learning_rate": 1.0522806833085076e-05,
+      "loss": 0.3613,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9984623270117888,
+      "grad_norm": 0.43854331808185243,
+      "learning_rate": 1.0514515701674904e-05,
+      "loss": 0.338,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9989748846745259,
+      "grad_norm": 0.5586961665565003,
+      "learning_rate": 1.0506224215617887e-05,
+      "loss": 0.3685,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9994874423372629,
+      "grad_norm": 0.445974257133975,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.3386,
+      "step": 1950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4805913275587879,
+      "learning_rate": 1.0489640202424284e-05,
+      "loss": 0.3823,
+      "step": 1951
+    },
+    {
+      "epoch": 1.0005125576627372,
+      "grad_norm": 0.4760246116032915,
+      "learning_rate": 1.0481347686718775e-05,
+      "loss": 0.2636,
+      "step": 1952
+    },
+    {
+      "epoch": 1.0010251153254741,
+      "grad_norm": 0.4696885186397599,
+      "learning_rate": 1.0473054839228562e-05,
+      "loss": 0.2935,
+      "step": 1953
+    },
+    {
+      "epoch": 1.0015376729882113,
+      "grad_norm": 0.46536323376853733,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.2942,
+      "step": 1954
+    },
+    {
+      "epoch": 1.0020502306509482,
+      "grad_norm": 0.4586756497884205,
+      "learning_rate": 1.045646817175874e-05,
+      "loss": 0.2466,
+      "step": 1955
+    },
+    {
+      "epoch": 1.0025627883136854,
+      "grad_norm": 0.4620570894731852,
+      "learning_rate": 1.0448174363212029e-05,
+      "loss": 0.2962,
+      "step": 1956
+    },
+    {
+      "epoch": 1.0030753459764223,
+      "grad_norm": 0.42664430167404227,
+      "learning_rate": 1.0439880245746428e-05,
+      "loss": 0.2766,
+      "step": 1957
+    },
+    {
+      "epoch": 1.0035879036391595,
+      "grad_norm": 0.4344584313852051,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.2598,
+      "step": 1958
+    },
+    {
+      "epoch": 1.0041004613018965,
+      "grad_norm": 0.45343912209123516,
+      "learning_rate": 1.0423291106926704e-05,
+      "loss": 0.2533,
+      "step": 1959
+    },
+    {
+      "epoch": 1.0046130189646336,
+      "grad_norm": 0.5074335885693091,
+      "learning_rate": 1.0414996097007199e-05,
+      "loss": 0.2412,
+      "step": 1960
+    },
+    {
+      "epoch": 1.0051255766273706,
+      "grad_norm": 0.5047143617645681,
+      "learning_rate": 1.0406700801038006e-05,
+      "loss": 0.2489,
+      "step": 1961
+    },
+    {
+      "epoch": 1.0056381342901077,
+      "grad_norm": 0.49398814073049246,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.2435,
+      "step": 1962
+    },
+    {
+      "epoch": 1.0061506919528447,
+      "grad_norm": 0.4889969184639393,
+      "learning_rate": 1.039010937382197e-05,
+      "loss": 0.2518,
+      "step": 1963
+    },
+    {
+      "epoch": 1.0066632496155818,
+      "grad_norm": 0.5305743664523681,
+      "learning_rate": 1.0381813254011317e-05,
+      "loss": 0.2784,
+      "step": 1964
+    },
+    {
+      "epoch": 1.0071758072783188,
+      "grad_norm": 0.5094882677769771,
+      "learning_rate": 1.0373516871023336e-05,
+      "loss": 0.285,
+      "step": 1965
+    },
+    {
+      "epoch": 1.007688364941056,
+      "grad_norm": 0.5944681276504131,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.2817,
+      "step": 1966
+    },
+    {
+      "epoch": 1.008200922603793,
+      "grad_norm": 0.4980967864333201,
+      "learning_rate": 1.0356923338389807e-05,
+      "loss": 0.2506,
+      "step": 1967
+    },
+    {
+      "epoch": 1.00871348026653,
+      "grad_norm": 0.5323955472568954,
+      "learning_rate": 1.034862620018189e-05,
+      "loss": 0.2904,
+      "step": 1968
+    },
+    {
+      "epoch": 1.009226037929267,
+      "grad_norm": 0.44083514879902114,
+      "learning_rate": 1.0340328821671918e-05,
+      "loss": 0.2342,
+      "step": 1969
+    },
+    {
+      "epoch": 1.0097385955920042,
+      "grad_norm": 0.48295898796494213,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.248,
+      "step": 1970
+    },
+    {
+      "epoch": 1.0102511532547411,
+      "grad_norm": 0.5346318665430062,
+      "learning_rate": 1.032373336662293e-05,
+      "loss": 0.2921,
+      "step": 1971
+    },
+    {
+      "epoch": 1.0107637109174783,
+      "grad_norm": 0.4741703256485525,
+      "learning_rate": 1.0315435301522877e-05,
+      "loss": 0.2501,
+      "step": 1972
+    },
+    {
+      "epoch": 1.0112762685802152,
+      "grad_norm": 0.48379420517767985,
+      "learning_rate": 1.0307137018998683e-05,
+      "loss": 0.2364,
+      "step": 1973
+    },
+    {
+      "epoch": 1.0117888262429524,
+      "grad_norm": 0.4912484368960739,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.2633,
+      "step": 1974
+    },
+    {
+      "epoch": 1.0123013839056894,
+      "grad_norm": 0.4988848521924495,
+      "learning_rate": 1.0290539824557474e-05,
+      "loss": 0.2447,
+      "step": 1975
+    },
+    {
+      "epoch": 1.0128139415684265,
+      "grad_norm": 0.47240026121293444,
+      "learning_rate": 1.0282240924080625e-05,
+      "loss": 0.2546,
+      "step": 1976
+    },
+    {
+      "epoch": 1.0133264992311635,
+      "grad_norm": 0.5023105995368743,
+      "learning_rate": 1.027394182905995e-05,
+      "loss": 0.2794,
+      "step": 1977
+    },
+    {
+      "epoch": 1.0138390568939006,
+      "grad_norm": 0.510798123264994,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.2714,
+      "step": 1978
+    },
+    {
+      "epoch": 1.0143516145566376,
+      "grad_norm": 0.5168559009338509,
+      "learning_rate": 1.0257343078268948e-05,
+      "loss": 0.2689,
+      "step": 1979
+    },
+    {
+      "epoch": 1.0148641722193747,
+      "grad_norm": 0.49273251375842975,
+      "learning_rate": 1.0249043433939857e-05,
+      "loss": 0.2495,
+      "step": 1980
+    },
+    {
+      "epoch": 1.0153767298821117,
+      "grad_norm": 0.5089549478219513,
+      "learning_rate": 1.02407436179494e-05,
+      "loss": 0.2722,
+      "step": 1981
+    },
+    {
+      "epoch": 1.0158892875448489,
+      "grad_norm": 0.5036497985304708,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.2331,
+      "step": 1982
+    },
+    {
+      "epoch": 1.0164018452075858,
+      "grad_norm": 0.45967140354286373,
+      "learning_rate": 1.0224143493868191e-05,
+      "loss": 0.2314,
+      "step": 1983
+    },
+    {
+      "epoch": 1.016914402870323,
+      "grad_norm": 0.5546887240596567,
+      "learning_rate": 1.021584319721962e-05,
+      "loss": 0.29,
+      "step": 1984
+    },
+    {
+      "epoch": 1.01742696053306,
+      "grad_norm": 0.5167853946150652,
+      "learning_rate": 1.0207542751794035e-05,
+      "loss": 0.2333,
+      "step": 1985
+    },
+    {
+      "epoch": 1.017939518195797,
+      "grad_norm": 0.4946043556273633,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.2402,
+      "step": 1986
+    },
+    {
+      "epoch": 1.018452075858534,
+      "grad_norm": 0.5386496307161702,
+      "learning_rate": 1.019094143749735e-05,
+      "loss": 0.2908,
+      "step": 1987
+    },
+    {
+      "epoch": 1.0189646335212712,
+      "grad_norm": 0.5270223827865977,
+      "learning_rate": 1.0182640580069249e-05,
+      "loss": 0.2862,
+      "step": 1988
+    },
+    {
+      "epoch": 1.0194771911840081,
+      "grad_norm": 0.5026082325925341,
+      "learning_rate": 1.0174339596750131e-05,
+      "loss": 0.2272,
+      "step": 1989
+    },
+    {
+      "epoch": 1.0199897488467453,
+      "grad_norm": 0.5139093279521225,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.2449,
+      "step": 1990
+    },
+    {
+      "epoch": 1.0205023065094823,
+      "grad_norm": 0.4878096200983223,
+      "learning_rate": 1.0157737275325829e-05,
+      "loss": 0.2681,
+      "step": 1991
+    },
+    {
+      "epoch": 1.0210148641722194,
+      "grad_norm": 0.4999293303588626,
+      "learning_rate": 1.0149435948664338e-05,
+      "loss": 0.2599,
+      "step": 1992
+    },
+    {
+      "epoch": 1.0215274218349564,
+      "grad_norm": 0.4417032213467149,
+      "learning_rate": 1.0141134518999216e-05,
+      "loss": 0.2274,
+      "step": 1993
+    },
+    {
+      "epoch": 1.0220399794976935,
+      "grad_norm": 0.4672877825940755,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.2385,
+      "step": 1994
+    },
+    {
+      "epoch": 1.0225525371604305,
+      "grad_norm": 0.4880484993435755,
+      "learning_rate": 1.0124531373546262e-05,
+      "loss": 0.2577,
+      "step": 1995
+    },
+    {
+      "epoch": 1.0230650948231677,
+      "grad_norm": 0.5151884771750062,
+      "learning_rate": 1.0116229669202692e-05,
+      "loss": 0.2779,
+      "step": 1996
+    },
+    {
+      "epoch": 1.0235776524859046,
+      "grad_norm": 0.5146424276822807,
+      "learning_rate": 1.0107927884744014e-05,
+      "loss": 0.2699,
+      "step": 1997
+    },
+    {
+      "epoch": 1.0240902101486418,
+      "grad_norm": 0.46985361290038163,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.2523,
+      "step": 1998
+    },
+    {
+      "epoch": 1.0246027678113787,
+      "grad_norm": 0.5383366805353708,
+      "learning_rate": 1.009132409837046e-05,
+      "loss": 0.2416,
+      "step": 1999
+    },
+    {
+      "epoch": 1.0251153254741159,
+      "grad_norm": 0.5062367494390366,
+      "learning_rate": 1.0083022107900295e-05,
+      "loss": 0.279,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0256278831368528,
+      "grad_norm": 0.4717042574821369,
+      "learning_rate": 1.0074720060204417e-05,
+      "loss": 0.2592,
+      "step": 2001
+    },
+    {
+      "epoch": 1.02614044079959,
+      "grad_norm": 0.5195190178947321,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.2572,
+      "step": 2002
+    },
+    {
+      "epoch": 1.026652998462327,
+      "grad_norm": 0.45399480113588103,
+      "learning_rate": 1.0058115816025389e-05,
+      "loss": 0.2354,
+      "step": 2003
+    },
+    {
+      "epoch": 1.027165556125064,
+      "grad_norm": 0.5482570305565277,
+      "learning_rate": 1.0049813630987262e-05,
+      "loss": 0.2742,
+      "step": 2004
+    },
+    {
+      "epoch": 1.027678113787801,
+      "grad_norm": 0.5347842958780353,
+      "learning_rate": 1.0041511411613449e-05,
+      "loss": 0.2878,
+      "step": 2005
+    },
+    {
+      "epoch": 1.0281906714505382,
+      "grad_norm": 0.5045400406536991,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.2164,
+      "step": 2006
+    },
+    {
+      "epoch": 1.0287032291132752,
+      "grad_norm": 0.5216988091176072,
+      "learning_rate": 1.0024906892749115e-05,
+      "loss": 0.2482,
+      "step": 2007
+    },
+    {
+      "epoch": 1.0292157867760123,
+      "grad_norm": 0.49115113786158593,
+      "learning_rate": 1.0016604604703803e-05,
+      "loss": 0.2667,
+      "step": 2008
+    },
+    {
+      "epoch": 1.0297283444387493,
+      "grad_norm": 0.5461591253604524,
+      "learning_rate": 1.000830230521322e-05,
+      "loss": 0.2618,
+      "step": 2009
+    },
+    {
+      "epoch": 1.0302409021014864,
+      "grad_norm": 0.5476261884378741,
+      "learning_rate": 1e-05,
+      "loss": 0.2758,
+      "step": 2010
+    },
+    {
+      "epoch": 1.0307534597642234,
+      "grad_norm": 0.5149107374951961,
+      "learning_rate": 9.99169769478678e-06,
+      "loss": 0.2685,
+      "step": 2011
+    },
+    {
+      "epoch": 1.0312660174269606,
+      "grad_norm": 0.530697188941781,
+      "learning_rate": 9.9833953952962e-06,
+      "loss": 0.2673,
+      "step": 2012
+    },
+    {
+      "epoch": 1.0317785750896975,
+      "grad_norm": 0.49416936795738164,
+      "learning_rate": 9.975093107250888e-06,
+      "loss": 0.2497,
+      "step": 2013
+    },
+    {
+      "epoch": 1.0322911327524347,
+      "grad_norm": 0.5116016715310112,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.2756,
+      "step": 2014
+    },
+    {
+      "epoch": 1.0328036904151716,
+      "grad_norm": 0.5047762817773508,
+      "learning_rate": 9.958488588386551e-06,
+      "loss": 0.2761,
+      "step": 2015
+    },
+    {
+      "epoch": 1.0333162480779088,
+      "grad_norm": 0.48517280743378255,
+      "learning_rate": 9.950186369012744e-06,
+      "loss": 0.2582,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0338288057406457,
+      "grad_norm": 0.4669367052718487,
+      "learning_rate": 9.941884183974614e-06,
+      "loss": 0.2543,
+      "step": 2017
+    },
+    {
+      "epoch": 1.034341363403383,
+      "grad_norm": 0.4673706947289928,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.259,
+      "step": 2018
+    },
+    {
+      "epoch": 1.0348539210661198,
+      "grad_norm": 0.45201338967881016,
+      "learning_rate": 9.925279939795583e-06,
+      "loss": 0.2376,
+      "step": 2019
+    },
+    {
+      "epoch": 1.035366478728857,
+      "grad_norm": 0.5156407411251327,
+      "learning_rate": 9.916977892099707e-06,
+      "loss": 0.2596,
+      "step": 2020
+    },
+    {
+      "epoch": 1.035879036391594,
+      "grad_norm": 0.5562301229991968,
+      "learning_rate": 9.908675901629542e-06,
+      "loss": 0.2454,
+      "step": 2021
+    },
+    {
+      "epoch": 1.0363915940543311,
+      "grad_norm": 0.480362296064468,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.2462,
+      "step": 2022
+    },
+    {
+      "epoch": 1.036904151717068,
+      "grad_norm": 0.5736890964355176,
+      "learning_rate": 9.892072115255991e-06,
+      "loss": 0.2766,
+      "step": 2023
+    },
+    {
+      "epoch": 1.0374167093798052,
+      "grad_norm": 0.533129241615145,
+      "learning_rate": 9.88377033079731e-06,
+      "loss": 0.2514,
+      "step": 2024
+    },
+    {
+      "epoch": 1.0379292670425422,
+      "grad_norm": 0.5112487325021532,
+      "learning_rate": 9.875468626453741e-06,
+      "loss": 0.2778,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0384418247052793,
+      "grad_norm": 0.4815957076109414,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.2256,
+      "step": 2026
+    },
+    {
+      "epoch": 1.0389543823680163,
+      "grad_norm": 0.5298086070620445,
+      "learning_rate": 9.858865481000787e-06,
+      "loss": 0.2622,
+      "step": 2027
+    },
+    {
+      "epoch": 1.0394669400307535,
+      "grad_norm": 0.49358123154395134,
+      "learning_rate": 9.850564051335663e-06,
+      "loss": 0.2478,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0399794976934906,
+      "grad_norm": 0.5230824873897036,
+      "learning_rate": 9.842262724674173e-06,
+      "loss": 0.2538,
+      "step": 2029
+    },
+    {
+      "epoch": 1.0404920553562276,
+      "grad_norm": 0.510118160430244,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.2474,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0410046130189645,
+      "grad_norm": 0.48448550994170164,
+      "learning_rate": 9.82566040324987e-06,
+      "loss": 0.2472,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0415171706817017,
+      "grad_norm": 0.48083188901546914,
+      "learning_rate": 9.817359419930753e-06,
+      "loss": 0.2429,
+      "step": 2032
+    },
+    {
+      "epoch": 1.0420297283444389,
+      "grad_norm": 0.4881803907488733,
+      "learning_rate": 9.809058562502653e-06,
+      "loss": 0.2313,
+      "step": 2033
+    },
+    {
+      "epoch": 1.0425422860071758,
+      "grad_norm": 0.5072911342278233,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.2576,
+      "step": 2034
+    },
+    {
+      "epoch": 1.043054843669913,
+      "grad_norm": 0.5354053194610469,
+      "learning_rate": 9.792457248205967e-06,
+      "loss": 0.2461,
+      "step": 2035
+    },
+    {
+      "epoch": 1.04356740133265,
+      "grad_norm": 0.5417543648410824,
+      "learning_rate": 9.784156802780385e-06,
+      "loss": 0.2772,
+      "step": 2036
+    },
+    {
+      "epoch": 1.044079958995387,
+      "grad_norm": 0.5349659246158044,
+      "learning_rate": 9.77585650613181e-06,
+      "loss": 0.2862,
+      "step": 2037
+    },
+    {
+      "epoch": 1.044592516658124,
+      "grad_norm": 0.5102579774954631,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.2349,
+      "step": 2038
+    },
+    {
+      "epoch": 1.0451050743208612,
+      "grad_norm": 0.5197383766152786,
+      "learning_rate": 9.759256382050602e-06,
+      "loss": 0.2885,
+      "step": 2039
+    },
+    {
+      "epoch": 1.0456176319835981,
+      "grad_norm": 0.4993014540520361,
+      "learning_rate": 9.750956566060145e-06,
+      "loss": 0.2699,
+      "step": 2040
+    },
+    {
+      "epoch": 1.0461301896463353,
+      "grad_norm": 0.49949696267265614,
+      "learning_rate": 9.742656921731057e-06,
+      "loss": 0.2571,
+      "step": 2041
+    },
+    {
+      "epoch": 1.0466427473090723,
+      "grad_norm": 0.4773478578778226,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.2609,
+      "step": 2042
+    },
+    {
+      "epoch": 1.0471553049718094,
+      "grad_norm": 0.4957362465545343,
+      "learning_rate": 9.726058170940053e-06,
+      "loss": 0.2744,
+      "step": 2043
+    },
+    {
+      "epoch": 1.0476678626345464,
+      "grad_norm": 0.4918940873824536,
+      "learning_rate": 9.717759075919375e-06,
+      "loss": 0.2448,
+      "step": 2044
+    },
+    {
+      "epoch": 1.0481804202972835,
+      "grad_norm": 0.5635187353204976,
+      "learning_rate": 9.709460175442527e-06,
+      "loss": 0.2736,
+      "step": 2045
+    },
+    {
+      "epoch": 1.0486929779600205,
+      "grad_norm": 0.5281866980269353,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.3163,
+      "step": 2046
+    },
+    {
+      "epoch": 1.0492055356227576,
+      "grad_norm": 0.4801866834343016,
+      "learning_rate": 9.692862981001319e-06,
+      "loss": 0.2433,
+      "step": 2047
+    },
+    {
+      "epoch": 1.0497180932854946,
+      "grad_norm": 0.5071838949372861,
+      "learning_rate": 9.684564698477128e-06,
+      "loss": 0.258,
+      "step": 2048
+    },
+    {
+      "epoch": 1.0502306509482318,
+      "grad_norm": 0.5045926184273581,
+      "learning_rate": 9.676266633377073e-06,
+      "loss": 0.2761,
+      "step": 2049
+    },
+    {
+      "epoch": 1.0507432086109687,
+      "grad_norm": 0.46917133140630046,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.2487,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0512557662737059,
+      "grad_norm": 0.492447381019282,
+      "learning_rate": 9.659671178328083e-06,
+      "loss": 0.2349,
+      "step": 2051
+    },
+    {
+      "epoch": 1.0517683239364428,
+      "grad_norm": 0.546971364986402,
+      "learning_rate": 9.651373799818114e-06,
+      "loss": 0.2495,
+      "step": 2052
+    },
+    {
+      "epoch": 1.05228088159918,
+      "grad_norm": 0.5549451309228598,
+      "learning_rate": 9.643076661610197e-06,
+      "loss": 0.2337,
+      "step": 2053
+    },
+    {
+      "epoch": 1.052793439261917,
+      "grad_norm": 0.5099449379763369,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.2867,
+      "step": 2054
+    },
+    {
+      "epoch": 1.053305996924654,
+      "grad_norm": 0.5264791434980345,
+      "learning_rate": 9.626483128976665e-06,
+      "loss": 0.2607,
+      "step": 2055
+    },
+    {
+      "epoch": 1.053818554587391,
+      "grad_norm": 0.507721066525415,
+      "learning_rate": 9.618186745988688e-06,
+      "loss": 0.2569,
+      "step": 2056
+    },
+    {
+      "epoch": 1.0543311122501282,
+      "grad_norm": 0.5023651503735953,
+      "learning_rate": 9.60989062617803e-06,
+      "loss": 0.2571,
+      "step": 2057
+    },
+    {
+      "epoch": 1.0548436699128652,
+      "grad_norm": 0.5078757308651112,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.2446,
+      "step": 2058
+    },
+    {
+      "epoch": 1.0553562275756023,
+      "grad_norm": 0.48456979427561153,
+      "learning_rate": 9.593299198961999e-06,
+      "loss": 0.2211,
+      "step": 2059
+    },
+    {
+      "epoch": 1.0558687852383393,
+      "grad_norm": 0.602012634186767,
+      "learning_rate": 9.585003902992804e-06,
+      "loss": 0.2675,
+      "step": 2060
+    },
+    {
+      "epoch": 1.0563813429010764,
+      "grad_norm": 0.47413443290727353,
+      "learning_rate": 9.576708893073297e-06,
+      "loss": 0.2366,
+      "step": 2061
+    },
+    {
+      "epoch": 1.0568939005638134,
+      "grad_norm": 0.5356811305431864,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.2058,
+      "step": 2062
+    },
+    {
+      "epoch": 1.0574064582265505,
+      "grad_norm": 0.5088769378778013,
+      "learning_rate": 9.560119754253578e-06,
+      "loss": 0.2353,
+      "step": 2063
+    },
+    {
+      "epoch": 1.0579190158892875,
+      "grad_norm": 0.5054836598829702,
+      "learning_rate": 9.551825636787974e-06,
+      "loss": 0.2733,
+      "step": 2064
+    },
+    {
+      "epoch": 1.0584315735520247,
+      "grad_norm": 0.5040357307178984,
+      "learning_rate": 9.543531828241263e-06,
+      "loss": 0.2539,
+      "step": 2065
+    },
+    {
+      "epoch": 1.0589441312147616,
+      "grad_norm": 0.5122746263940327,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.2735,
+      "step": 2066
+    },
+    {
+      "epoch": 1.0594566888774988,
+      "grad_norm": 0.4867946679235616,
+      "learning_rate": 9.526945160771441e-06,
+      "loss": 0.2499,
+      "step": 2067
+    },
+    {
+      "epoch": 1.0599692465402357,
+      "grad_norm": 0.5001461968866207,
+      "learning_rate": 9.518652313281229e-06,
+      "loss": 0.2293,
+      "step": 2068
+    },
+    {
+      "epoch": 1.0604818042029729,
+      "grad_norm": 0.5276980779880595,
+      "learning_rate": 9.510359797575716e-06,
+      "loss": 0.2674,
+      "step": 2069
+    },
+    {
+      "epoch": 1.0609943618657098,
+      "grad_norm": 0.4926050350083189,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.2533,
+      "step": 2070
+    },
+    {
+      "epoch": 1.061506919528447,
+      "grad_norm": 0.5397500621917246,
+      "learning_rate": 9.493775784382115e-06,
+      "loss": 0.2624,
+      "step": 2071
+    },
+    {
+      "epoch": 1.062019477191184,
+      "grad_norm": 0.4802810474588019,
+      "learning_rate": 9.485484298325098e-06,
+      "loss": 0.225,
+      "step": 2072
+    },
+    {
+      "epoch": 1.0625320348539211,
+      "grad_norm": 0.5179171554732361,
+      "learning_rate": 9.477193166914926e-06,
+      "loss": 0.2374,
+      "step": 2073
+    },
+    {
+      "epoch": 1.063044592516658,
+      "grad_norm": 0.513050025407761,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.2523,
+      "step": 2074
+    },
+    {
+      "epoch": 1.0635571501793952,
+      "grad_norm": 0.5519195884286978,
+      "learning_rate": 9.460611990894599e-06,
+      "loss": 0.2357,
+      "step": 2075
+    },
+    {
+      "epoch": 1.0640697078421322,
+      "grad_norm": 0.5147927797385807,
+      "learning_rate": 9.452321957713563e-06,
+      "loss": 0.2634,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0645822655048693,
+      "grad_norm": 0.5012643266886696,
+      "learning_rate": 9.444032302037608e-06,
+      "loss": 0.2427,
+      "step": 2077
+    },
+    {
+      "epoch": 1.0650948231676063,
+      "grad_norm": 0.5046607815182187,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.2637,
+      "step": 2078
+    },
+    {
+      "epoch": 1.0656073808303435,
+      "grad_norm": 0.5460223507152464,
+      "learning_rate": 9.427454146056315e-06,
+      "loss": 0.2747,
+      "step": 2079
+    },
+    {
+      "epoch": 1.0661199384930804,
+      "grad_norm": 0.5162905768492041,
+      "learning_rate": 9.419165657178022e-06,
+      "loss": 0.2568,
+      "step": 2080
+    },
+    {
+      "epoch": 1.0666324961558176,
+      "grad_norm": 0.5367450864773977,
+      "learning_rate": 9.410877568658876e-06,
+      "loss": 0.2488,
+      "step": 2081
+    },
+    {
+      "epoch": 1.0671450538185545,
+      "grad_norm": 0.5293638750434786,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.2541,
+      "step": 2082
+    },
+    {
+      "epoch": 1.0676576114812917,
+      "grad_norm": 0.5593424291415136,
+      "learning_rate": 9.394302615549081e-06,
+      "loss": 0.2423,
+      "step": 2083
+    },
+    {
+      "epoch": 1.0681701691440286,
+      "grad_norm": 0.5187162909433227,
+      "learning_rate": 9.386015762383262e-06,
+      "loss": 0.2206,
+      "step": 2084
+    },
+    {
+      "epoch": 1.0686827268067658,
+      "grad_norm": 0.5781952714059188,
+      "learning_rate": 9.377729332426247e-06,
+      "loss": 0.2964,
+      "step": 2085
+    },
+    {
+      "epoch": 1.0691952844695027,
+      "grad_norm": 0.5041988017871358,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.2505,
+      "step": 2086
+    },
+    {
+      "epoch": 1.06970784213224,
+      "grad_norm": 0.4922604080536275,
+      "learning_rate": 9.361157764985079e-06,
+      "loss": 0.2465,
+      "step": 2087
+    },
+    {
+      "epoch": 1.0702203997949769,
+      "grad_norm": 0.5061916200677875,
+      "learning_rate": 9.35287263892343e-06,
+      "loss": 0.2431,
+      "step": 2088
+    },
+    {
+      "epoch": 1.070732957457714,
+      "grad_norm": 0.5239004218136527,
+      "learning_rate": 9.344587958915566e-06,
+      "loss": 0.2471,
+      "step": 2089
+    },
+    {
+      "epoch": 1.071245515120451,
+      "grad_norm": 0.5100110742676005,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.2539,
+      "step": 2090
+    },
+    {
+      "epoch": 1.0717580727831881,
+      "grad_norm": 0.4573273035321168,
+      "learning_rate": 9.328019959902821e-06,
+      "loss": 0.2532,
+      "step": 2091
+    },
+    {
+      "epoch": 1.072270630445925,
+      "grad_norm": 0.5117606881263284,
+      "learning_rate": 9.319736652317984e-06,
+      "loss": 0.2389,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0727831881086622,
+      "grad_norm": 0.5205373145748154,
+      "learning_rate": 9.311453813626993e-06,
+      "loss": 0.2691,
+      "step": 2093
+    },
+    {
+      "epoch": 1.0732957457713992,
+      "grad_norm": 0.48873925721824657,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.2479,
+      "step": 2094
+    },
+    {
+      "epoch": 1.0738083034341364,
+      "grad_norm": 0.5029530776343117,
+      "learning_rate": 9.294889565763119e-06,
+      "loss": 0.2184,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0743208610968733,
+      "grad_norm": 0.5035593409121861,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.2632,
+      "step": 2096
+    },
+    {
+      "epoch": 1.0748334187596105,
+      "grad_norm": 0.47372769741475174,
+      "learning_rate": 9.278327261980986e-06,
+      "loss": 0.2438,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0753459764223474,
+      "grad_norm": 0.4771210370695295,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.2253,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0758585340850846,
+      "grad_norm": 0.5164042683703862,
+      "learning_rate": 9.261766947945043e-06,
+      "loss": 0.2529,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0763710917478215,
+      "grad_norm": 0.5276352079858166,
+      "learning_rate": 9.25348755135053e-06,
+      "loss": 0.2607,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0768836494105587,
+      "grad_norm": 0.5216901470565397,
+      "learning_rate": 9.245208669314238e-06,
+      "loss": 0.2651,
+      "step": 2101
+    },
+    {
+      "epoch": 1.0773962070732956,
+      "grad_norm": 0.4822966435659085,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.2325,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0779087647360328,
+      "grad_norm": 0.48083967288723267,
+      "learning_rate": 9.228652471741917e-06,
+      "loss": 0.2307,
+      "step": 2103
+    },
+    {
+      "epoch": 1.07842132239877,
+      "grad_norm": 0.520944212469631,
+      "learning_rate": 9.22037516761779e-06,
+      "loss": 0.23,
+      "step": 2104
+    },
+    {
+      "epoch": 1.078933880061507,
+      "grad_norm": 0.4895892028719725,
+      "learning_rate": 9.212098400875678e-06,
+      "loss": 0.2247,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0794464377242439,
+      "grad_norm": 0.5640451705783466,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.2881,
+      "step": 2106
+    },
+    {
+      "epoch": 1.079958995386981,
+      "grad_norm": 0.5256059335510765,
+      "learning_rate": 9.195546502357268e-06,
+      "loss": 0.2313,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0804715530497182,
+      "grad_norm": 0.5594278531361538,
+      "learning_rate": 9.187271381989906e-06,
+      "loss": 0.3068,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0809841107124551,
+      "grad_norm": 0.5129881231302889,
+      "learning_rate": 9.17899682182243e-06,
+      "loss": 0.2302,
+      "step": 2109
+    },
+    {
+      "epoch": 1.081496668375192,
+      "grad_norm": 0.5313725069297646,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.2463,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0820092260379293,
+      "grad_norm": 0.4990801108429407,
+      "learning_rate": 9.162449404900803e-06,
+      "loss": 0.2355,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0825217837006664,
+      "grad_norm": 0.49597287565018455,
+      "learning_rate": 9.154176559552502e-06,
+      "loss": 0.2493,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0830343413634034,
+      "grad_norm": 0.5215722849643062,
+      "learning_rate": 9.145904297215785e-06,
+      "loss": 0.2586,
+      "step": 2113
+    },
+    {
+      "epoch": 1.0835468990261405,
+      "grad_norm": 0.5464419707799851,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.2832,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0840594566888775,
+      "grad_norm": 0.5419869984247184,
+      "learning_rate": 9.12936154438439e-06,
+      "loss": 0.2829,
+      "step": 2115
+    },
+    {
+      "epoch": 1.0845720143516147,
+      "grad_norm": 0.4969024264851095,
+      "learning_rate": 9.121091065292356e-06,
+      "loss": 0.2727,
+      "step": 2116
+    },
+    {
+      "epoch": 1.0850845720143516,
+      "grad_norm": 0.497781914427393,
+      "learning_rate": 9.112821192017168e-06,
+      "loss": 0.2304,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0855971296770888,
+      "grad_norm": 0.5229072246384241,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.2509,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0861096873398257,
+      "grad_norm": 0.47985853280565066,
+      "learning_rate": 9.096283285718025e-06,
+      "loss": 0.2529,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0866222450025629,
+      "grad_norm": 0.4988033398852772,
+      "learning_rate": 9.088015264093365e-06,
+      "loss": 0.2672,
+      "step": 2120
+    },
+    {
+      "epoch": 1.0871348026652998,
+      "grad_norm": 0.48326021569133504,
+      "learning_rate": 9.079747871084137e-06,
+      "loss": 0.2272,
+      "step": 2121
+    },
+    {
+      "epoch": 1.087647360328037,
+      "grad_norm": 0.48555670680500485,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.2559,
+      "step": 2122
+    },
+    {
+      "epoch": 1.088159917990774,
+      "grad_norm": 0.5308910319089534,
+      "learning_rate": 9.063214993705804e-06,
+      "loss": 0.2716,
+      "step": 2123
+    },
+    {
+      "epoch": 1.088672475653511,
+      "grad_norm": 0.5286261321948517,
+      "learning_rate": 9.054949520732535e-06,
+      "loss": 0.2658,
+      "step": 2124
+    },
+    {
+      "epoch": 1.089185033316248,
+      "grad_norm": 0.5149915259563753,
+      "learning_rate": 9.046684699166338e-06,
+      "loss": 0.2678,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0896975909789852,
+      "grad_norm": 0.5190282089699061,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.2664,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0902101486417222,
+      "grad_norm": 0.5173114527482684,
+      "learning_rate": 9.030157033041918e-06,
+      "loss": 0.2354,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0907227063044593,
+      "grad_norm": 0.5339466377029577,
+      "learning_rate": 9.021894199875934e-06,
+      "loss": 0.2813,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0912352639671963,
+      "grad_norm": 0.4618127309396129,
+      "learning_rate": 9.013632040901485e-06,
+      "loss": 0.2581,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0917478216299334,
+      "grad_norm": 0.5170777994874503,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.2868,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0922603792926704,
+      "grad_norm": 0.5186386626708287,
+      "learning_rate": 8.997109768306607e-06,
+      "loss": 0.2432,
+      "step": 2131
+    },
+    {
+      "epoch": 1.0927729369554076,
+      "grad_norm": 0.5097776046384366,
+      "learning_rate": 8.988849666074694e-06,
+      "loss": 0.2626,
+      "step": 2132
+    },
+    {
+      "epoch": 1.0932854946181445,
+      "grad_norm": 0.5319663406863578,
+      "learning_rate": 8.980590260811349e-06,
+      "loss": 0.2656,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0937980522808817,
+      "grad_norm": 0.4915727418006402,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.2359,
+      "step": 2134
+    },
+    {
+      "epoch": 1.0943106099436186,
+      "grad_norm": 0.4993995703852711,
+      "learning_rate": 8.964073563962158e-06,
+      "loss": 0.2435,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0948231676063558,
+      "grad_norm": 0.5154451214642152,
+      "learning_rate": 8.955816283760984e-06,
+      "loss": 0.2676,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0953357252690927,
+      "grad_norm": 0.5582765035408674,
+      "learning_rate": 8.947559723297726e-06,
+      "loss": 0.2492,
+      "step": 2137
+    },
+    {
+      "epoch": 1.09584828293183,
+      "grad_norm": 0.5207765517065268,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.2732,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0963608405945668,
+      "grad_norm": 0.4753814020875312,
+      "learning_rate": 8.931048784348875e-06,
+      "loss": 0.2176,
+      "step": 2139
+    },
+    {
+      "epoch": 1.096873398257304,
+      "grad_norm": 0.5269373944283438,
+      "learning_rate": 8.922794417243989e-06,
+      "loss": 0.2323,
+      "step": 2140
+    },
+    {
+      "epoch": 1.097385955920041,
+      "grad_norm": 0.500124768711827,
+      "learning_rate": 8.91454079263842e-06,
+      "loss": 0.2594,
+      "step": 2141
+    },
+    {
+      "epoch": 1.0978985135827781,
+      "grad_norm": 0.46241105991118486,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.2356,
+      "step": 2142
+    },
+    {
+      "epoch": 1.098411071245515,
+      "grad_norm": 0.49552474754532966,
+      "learning_rate": 8.898035793681064e-06,
+      "loss": 0.2428,
+      "step": 2143
+    },
+    {
+      "epoch": 1.0989236289082522,
+      "grad_norm": 0.49859162199872203,
+      "learning_rate": 8.889784430705881e-06,
+      "loss": 0.2423,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0994361865709892,
+      "grad_norm": 0.47153647428404544,
+      "learning_rate": 8.881533832983235e-06,
+      "loss": 0.2503,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0999487442337263,
+      "grad_norm": 0.5342360920460649,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.2648,
+      "step": 2146
+    },
+    {
+      "epoch": 1.1004613018964633,
+      "grad_norm": 0.5164610028293419,
+      "learning_rate": 8.865034956043017e-06,
+      "loss": 0.2432,
+      "step": 2147
+    },
+    {
+      "epoch": 1.1009738595592005,
+      "grad_norm": 0.5404466730340739,
+      "learning_rate": 8.856786688197827e-06,
+      "loss": 0.2872,
+      "step": 2148
+    },
+    {
+      "epoch": 1.1014864172219374,
+      "grad_norm": 0.5060501420191604,
+      "learning_rate": 8.848539208349956e-06,
+      "loss": 0.2639,
+      "step": 2149
+    },
+    {
+      "epoch": 1.1019989748846746,
+      "grad_norm": 0.5311773025964774,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.2692,
+      "step": 2150
+    },
+    {
+      "epoch": 1.1025115325474115,
+      "grad_norm": 0.4823843254431971,
+      "learning_rate": 8.832046635384996e-06,
+      "loss": 0.2528,
+      "step": 2151
+    },
+    {
+      "epoch": 1.1030240902101487,
+      "grad_norm": 0.48256793885296184,
+      "learning_rate": 8.823801553635955e-06,
+      "loss": 0.2588,
+      "step": 2152
+    },
+    {
+      "epoch": 1.1035366478728856,
+      "grad_norm": 0.5195394763623357,
+      "learning_rate": 8.81555728262032e-06,
+      "loss": 0.2477,
+      "step": 2153
+    },
+    {
+      "epoch": 1.1040492055356228,
+      "grad_norm": 0.499804547887182,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.2662,
+      "step": 2154
+    },
+    {
+      "epoch": 1.1045617631983597,
+      "grad_norm": 0.5228585416151428,
+      "learning_rate": 8.799071195519224e-06,
+      "loss": 0.2931,
+      "step": 2155
+    },
+    {
+      "epoch": 1.105074320861097,
+      "grad_norm": 0.5727017237420731,
+      "learning_rate": 8.790829390797342e-06,
+      "loss": 0.2646,
+      "step": 2156
+    },
+    {
+      "epoch": 1.1055868785238339,
+      "grad_norm": 0.48328142132017776,
+      "learning_rate": 8.782588419536015e-06,
+      "loss": 0.2434,
+      "step": 2157
+    },
+    {
+      "epoch": 1.106099436186571,
+      "grad_norm": 0.5394616323171039,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.2337,
+      "step": 2158
+    },
+    {
+      "epoch": 1.106611993849308,
+      "grad_norm": 0.5586315226907531,
+      "learning_rate": 8.766109000115857e-06,
+      "loss": 0.3067,
+      "step": 2159
+    },
+    {
+      "epoch": 1.1071245515120451,
+      "grad_norm": 0.5003159323495131,
+      "learning_rate": 8.757870563316018e-06,
+      "loss": 0.2244,
+      "step": 2160
+    },
+    {
+      "epoch": 1.107637109174782,
+      "grad_norm": 0.5337941770786216,
+      "learning_rate": 8.749632982694675e-06,
+      "loss": 0.2398,
+      "step": 2161
+    },
+    {
+      "epoch": 1.1081496668375193,
+      "grad_norm": 0.5277665507605129,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.2712,
+      "step": 2162
+    },
+    {
+      "epoch": 1.1086622245002562,
+      "grad_norm": 0.5208063595765086,
+      "learning_rate": 8.733160412698996e-06,
+      "loss": 0.2583,
+      "step": 2163
+    },
+    {
+      "epoch": 1.1091747821629934,
+      "grad_norm": 0.5095406903277916,
+      "learning_rate": 8.724925434678923e-06,
+      "loss": 0.2411,
+      "step": 2164
+    },
+    {
+      "epoch": 1.1096873398257303,
+      "grad_norm": 0.6072032008912487,
+      "learning_rate": 8.716691335545862e-06,
+      "loss": 0.2616,
+      "step": 2165
+    },
+    {
+      "epoch": 1.1101998974884675,
+      "grad_norm": 0.5045269912252148,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.265,
+      "step": 2166
+    },
+    {
+      "epoch": 1.1107124551512044,
+      "grad_norm": 0.5312342131084844,
+      "learning_rate": 8.700225796642663e-06,
+      "loss": 0.241,
+      "step": 2167
+    },
+    {
+      "epoch": 1.1112250128139416,
+      "grad_norm": 0.5309929979991208,
+      "learning_rate": 8.691994368221942e-06,
+      "loss": 0.2519,
+      "step": 2168
+    },
+    {
+      "epoch": 1.1117375704766785,
+      "grad_norm": 0.5071386350424731,
+      "learning_rate": 8.683763841387051e-06,
+      "loss": 0.2543,
+      "step": 2169
+    },
+    {
+      "epoch": 1.1122501281394157,
+      "grad_norm": 0.4710344772392592,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.2519,
+      "step": 2170
+    },
+    {
+      "epoch": 1.1127626858021527,
+      "grad_norm": 0.5137607937934529,
+      "learning_rate": 8.667305515166792e-06,
+      "loss": 0.2654,
+      "step": 2171
+    },
+    {
+      "epoch": 1.1132752434648898,
+      "grad_norm": 0.5022012393790797,
+      "learning_rate": 8.659077727125859e-06,
+      "loss": 0.2587,
+      "step": 2172
+    },
+    {
+      "epoch": 1.1137878011276268,
+      "grad_norm": 0.48794855926253095,
+      "learning_rate": 8.650850863359636e-06,
+      "loss": 0.2451,
+      "step": 2173
+    },
+    {
+      "epoch": 1.114300358790364,
+      "grad_norm": 0.5204149579649495,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.2688,
+      "step": 2174
+    },
+    {
+      "epoch": 1.1148129164531009,
+      "grad_norm": 0.5244238511162778,
+      "learning_rate": 8.634399931333226e-06,
+      "loss": 0.3016,
+      "step": 2175
+    },
+    {
+      "epoch": 1.115325474115838,
+      "grad_norm": 0.5512662220646297,
+      "learning_rate": 8.626175874412376e-06,
+      "loss": 0.2707,
+      "step": 2176
+    },
+    {
+      "epoch": 1.115838031778575,
+      "grad_norm": 0.4927262377389602,
+      "learning_rate": 8.617952764444922e-06,
+      "loss": 0.2306,
+      "step": 2177
+    },
+    {
+      "epoch": 1.1163505894413122,
+      "grad_norm": 0.5332703022445505,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.2718,
+      "step": 2178
+    },
+    {
+      "epoch": 1.116863147104049,
+      "grad_norm": 0.498644843007847,
+      "learning_rate": 8.601509408041722e-06,
+      "loss": 0.2559,
+      "step": 2179
+    },
+    {
+      "epoch": 1.1173757047667863,
+      "grad_norm": 0.5085530697558027,
+      "learning_rate": 8.593289172940102e-06,
+      "loss": 0.2679,
+      "step": 2180
+    },
+    {
+      "epoch": 1.1178882624295232,
+      "grad_norm": 0.5055868250967384,
+      "learning_rate": 8.58506990746011e-06,
+      "loss": 0.2588,
+      "step": 2181
+    },
+    {
+      "epoch": 1.1184008200922604,
+      "grad_norm": 0.5194465464193879,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.2558,
+      "step": 2182
+    },
+    {
+      "epoch": 1.1189133777549976,
+      "grad_norm": 0.49530375001645754,
+      "learning_rate": 8.568634308025945e-06,
+      "loss": 0.2324,
+      "step": 2183
+    },
+    {
+      "epoch": 1.1194259354177345,
+      "grad_norm": 0.5067480629382843,
+      "learning_rate": 8.560417985400541e-06,
+      "loss": 0.2512,
+      "step": 2184
+    },
+    {
+      "epoch": 1.1199384930804714,
+      "grad_norm": 0.5205161273858971,
+      "learning_rate": 8.552202655054315e-06,
+      "loss": 0.2487,
+      "step": 2185
+    },
+    {
+      "epoch": 1.1204510507432086,
+      "grad_norm": 0.486383702514792,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.2613,
+      "step": 2186
+    },
+    {
+      "epoch": 1.1209636084059458,
+      "grad_norm": 0.5316894061974605,
+      "learning_rate": 8.535774993849449e-06,
+      "loss": 0.2511,
+      "step": 2187
+    },
+    {
+      "epoch": 1.1214761660686827,
+      "grad_norm": 0.5411314954872105,
+      "learning_rate": 8.527562674314109e-06,
+      "loss": 0.2577,
+      "step": 2188
+    },
+    {
+      "epoch": 1.1219887237314197,
+      "grad_norm": 0.5480446593468252,
+      "learning_rate": 8.519351369704553e-06,
+      "loss": 0.2762,
+      "step": 2189
+    },
+    {
+      "epoch": 1.1225012813941568,
+      "grad_norm": 0.4850337756700158,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.2677,
+      "step": 2190
+    },
+    {
+      "epoch": 1.123013839056894,
+      "grad_norm": 0.4715738715843408,
+      "learning_rate": 8.502931827901706e-06,
+      "loss": 0.2244,
+      "step": 2191
+    },
+    {
+      "epoch": 1.123526396719631,
+      "grad_norm": 0.4598114991000119,
+      "learning_rate": 8.494723602026131e-06,
+      "loss": 0.2423,
+      "step": 2192
+    },
+    {
+      "epoch": 1.1240389543823681,
+      "grad_norm": 0.45919782425525246,
+      "learning_rate": 8.486516413711743e-06,
+      "loss": 0.2776,
+      "step": 2193
+    },
+    {
+      "epoch": 1.124551512045105,
+      "grad_norm": 0.5227532273392516,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.2119,
+      "step": 2194
+    },
+    {
+      "epoch": 1.1250640697078422,
+      "grad_norm": 0.5079847317918166,
+      "learning_rate": 8.4701051723941e-06,
+      "loss": 0.2453,
+      "step": 2195
+    },
+    {
+      "epoch": 1.1255766273705792,
+      "grad_norm": 0.48014114150356485,
+      "learning_rate": 8.461901130702836e-06,
+      "loss": 0.2647,
+      "step": 2196
+    },
+    {
+      "epoch": 1.1260891850333163,
+      "grad_norm": 0.5259641105553476,
+      "learning_rate": 8.45369814919672e-06,
+      "loss": 0.2263,
+      "step": 2197
+    },
+    {
+      "epoch": 1.1266017426960533,
+      "grad_norm": 0.513627052907793,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.2339,
+      "step": 2198
+    },
+    {
+      "epoch": 1.1271143003587905,
+      "grad_norm": 0.5225649785164845,
+      "learning_rate": 8.437295389355912e-06,
+      "loss": 0.2696,
+      "step": 2199
+    },
+    {
+      "epoch": 1.1276268580215274,
+      "grad_norm": 0.4981904331590521,
+      "learning_rate": 8.429095622327363e-06,
+      "loss": 0.2266,
+      "step": 2200
+    },
+    {
+      "epoch": 1.1281394156842646,
+      "grad_norm": 0.5838364029217351,
+      "learning_rate": 8.42089693809624e-06,
+      "loss": 0.2625,
+      "step": 2201
+    },
+    {
+      "epoch": 1.1286519733470015,
+      "grad_norm": 0.5266367998732105,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.2529,
+      "step": 2202
+    },
+    {
+      "epoch": 1.1291645310097387,
+      "grad_norm": 0.4990589711063256,
+      "learning_rate": 8.404502840630363e-06,
+      "loss": 0.2552,
+      "step": 2203
+    },
+    {
+      "epoch": 1.1296770886724756,
+      "grad_norm": 0.5189286343629017,
+      "learning_rate": 8.39630743869579e-06,
+      "loss": 0.2813,
+      "step": 2204
+    },
+    {
+      "epoch": 1.1301896463352128,
+      "grad_norm": 0.5244610722266267,
+      "learning_rate": 8.388113142158972e-06,
+      "loss": 0.2519,
+      "step": 2205
+    },
+    {
+      "epoch": 1.1307022039979497,
+      "grad_norm": 0.5509443628720357,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.2547,
+      "step": 2206
+    },
+    {
+      "epoch": 1.131214761660687,
+      "grad_norm": 0.5170963435066628,
+      "learning_rate": 8.3717278878706e-06,
+      "loss": 0.2379,
+      "step": 2207
+    },
+    {
+      "epoch": 1.1317273193234239,
+      "grad_norm": 0.4734085697792744,
+      "learning_rate": 8.363536941413121e-06,
+      "loss": 0.247,
+      "step": 2208
+    },
+    {
+      "epoch": 1.132239876986161,
+      "grad_norm": 0.5001529923372452,
+      "learning_rate": 8.355347122941538e-06,
+      "loss": 0.2464,
+      "step": 2209
+    },
+    {
+      "epoch": 1.132752434648898,
+      "grad_norm": 0.5237523383725793,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.2765,
+      "step": 2210
+    },
+    {
+      "epoch": 1.1332649923116351,
+      "grad_norm": 0.5393258354962868,
+      "learning_rate": 8.338970892535705e-06,
+      "loss": 0.2905,
+      "step": 2211
+    },
+    {
+      "epoch": 1.133777549974372,
+      "grad_norm": 0.4792991477306323,
+      "learning_rate": 8.330784491889303e-06,
+      "loss": 0.2462,
+      "step": 2212
+    },
+    {
+      "epoch": 1.1342901076371092,
+      "grad_norm": 0.5332679578668371,
+      "learning_rate": 8.322599241804502e-06,
+      "loss": 0.2606,
+      "step": 2213
+    },
+    {
+      "epoch": 1.1348026652998462,
+      "grad_norm": 0.5278404498863386,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.2636,
+      "step": 2214
+    },
+    {
+      "epoch": 1.1353152229625834,
+      "grad_norm": 0.4818291979445696,
+      "learning_rate": 8.306232215886722e-06,
+      "loss": 0.2167,
+      "step": 2215
+    },
+    {
+      "epoch": 1.1358277806253203,
+      "grad_norm": 0.4993340961249,
+      "learning_rate": 8.298050451335249e-06,
+      "loss": 0.2565,
+      "step": 2216
+    },
+    {
+      "epoch": 1.1363403382880575,
+      "grad_norm": 0.5221356900275397,
+      "learning_rate": 8.289869859908389e-06,
+      "loss": 0.2359,
+      "step": 2217
+    },
+    {
+      "epoch": 1.1368528959507944,
+      "grad_norm": 0.4833971736485918,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.2379,
+      "step": 2218
+    },
+    {
+      "epoch": 1.1373654536135316,
+      "grad_norm": 0.5335704181949393,
+      "learning_rate": 8.273512218982666e-06,
+      "loss": 0.2933,
+      "step": 2219
+    },
+    {
+      "epoch": 1.1378780112762685,
+      "grad_norm": 0.5151223492069769,
+      "learning_rate": 8.265335180758843e-06,
+      "loss": 0.2615,
+      "step": 2220
+    },
+    {
+      "epoch": 1.1383905689390057,
+      "grad_norm": 0.5269429034533429,
+      "learning_rate": 8.257159338209708e-06,
+      "loss": 0.2311,
+      "step": 2221
+    },
+    {
+      "epoch": 1.1389031266017426,
+      "grad_norm": 0.5302069440648126,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.2671,
+      "step": 2222
+    },
+    {
+      "epoch": 1.1394156842644798,
+      "grad_norm": 0.5562801557208159,
+      "learning_rate": 8.24081126267655e-06,
+      "loss": 0.3054,
+      "step": 2223
+    },
+    {
+      "epoch": 1.1399282419272168,
+      "grad_norm": 0.49113507738591067,
+      "learning_rate": 8.23263904096097e-06,
+      "loss": 0.2304,
+      "step": 2224
+    },
+    {
+      "epoch": 1.140440799589954,
+      "grad_norm": 0.5125609452913552,
+      "learning_rate": 8.224468037456969e-06,
+      "loss": 0.2433,
+      "step": 2225
+    },
+    {
+      "epoch": 1.1409533572526909,
+      "grad_norm": 0.48905428754377034,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.2414,
+      "step": 2226
+    },
+    {
+      "epoch": 1.141465914915428,
+      "grad_norm": 0.5241393463611118,
+      "learning_rate": 8.208129707611381e-06,
+      "loss": 0.2594,
+      "step": 2227
+    },
+    {
+      "epoch": 1.141978472578165,
+      "grad_norm": 0.5049300633242987,
+      "learning_rate": 8.199962392531526e-06,
+      "loss": 0.2533,
+      "step": 2228
+    },
+    {
+      "epoch": 1.1424910302409022,
+      "grad_norm": 0.5196382530486137,
+      "learning_rate": 8.191796318186701e-06,
+      "loss": 0.2519,
+      "step": 2229
+    },
+    {
+      "epoch": 1.143003587903639,
+      "grad_norm": 0.49624395659466264,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.2456,
+      "step": 2230
+    },
+    {
+      "epoch": 1.1435161455663763,
+      "grad_norm": 0.49721014679950304,
+      "learning_rate": 8.175467914216216e-06,
+      "loss": 0.2598,
+      "step": 2231
+    },
+    {
+      "epoch": 1.1440287032291132,
+      "grad_norm": 0.5080883471066374,
+      "learning_rate": 8.167305595845445e-06,
+      "loss": 0.2445,
+      "step": 2232
+    },
+    {
+      "epoch": 1.1445412608918504,
+      "grad_norm": 0.5164871870872808,
+      "learning_rate": 8.159144540719475e-06,
+      "loss": 0.2429,
+      "step": 2233
+    },
+    {
+      "epoch": 1.1450538185545873,
+      "grad_norm": 0.525659286120698,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.2233,
+      "step": 2234
+    },
+    {
+      "epoch": 1.1455663762173245,
+      "grad_norm": 0.5670195076280463,
+      "learning_rate": 8.142826242702157e-06,
+      "loss": 0.2714,
+      "step": 2235
+    },
+    {
+      "epoch": 1.1460789338800614,
+      "grad_norm": 0.47859321858234477,
+      "learning_rate": 8.134669011058734e-06,
+      "loss": 0.2558,
+      "step": 2236
+    },
+    {
+      "epoch": 1.1465914915427986,
+      "grad_norm": 0.486732526801992,
+      "learning_rate": 8.126513065155944e-06,
+      "loss": 0.2345,
+      "step": 2237
+    },
+    {
+      "epoch": 1.1471040492055355,
+      "grad_norm": 0.5353806206155548,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.2666,
+      "step": 2238
+    },
+    {
+      "epoch": 1.1476166068682727,
+      "grad_norm": 0.5198047059377432,
+      "learning_rate": 8.110205053058399e-06,
+      "loss": 0.2544,
+      "step": 2239
+    },
+    {
+      "epoch": 1.1481291645310097,
+      "grad_norm": 0.5375254800545902,
+      "learning_rate": 8.10205299810448e-06,
+      "loss": 0.2662,
+      "step": 2240
+    },
+    {
+      "epoch": 1.1486417221937468,
+      "grad_norm": 0.49480199767655414,
+      "learning_rate": 8.093902251372854e-06,
+      "loss": 0.2426,
+      "step": 2241
+    },
+    {
+      "epoch": 1.1491542798564838,
+      "grad_norm": 0.5009753269129115,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.2601,
+      "step": 2242
+    },
+    {
+      "epoch": 1.149666837519221,
+      "grad_norm": 0.500773720519238,
+      "learning_rate": 8.077604705048258e-06,
+      "loss": 0.225,
+      "step": 2243
+    },
+    {
+      "epoch": 1.1501793951819579,
+      "grad_norm": 0.45974565528074596,
+      "learning_rate": 8.069457916688907e-06,
+      "loss": 0.1978,
+      "step": 2244
+    },
+    {
+      "epoch": 1.150691952844695,
+      "grad_norm": 0.5238137797649489,
+      "learning_rate": 8.061312459019078e-06,
+      "loss": 0.239,
+      "step": 2245
+    },
+    {
+      "epoch": 1.151204510507432,
+      "grad_norm": 0.5090017229312772,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.26,
+      "step": 2246
+    },
+    {
+      "epoch": 1.1517170681701692,
+      "grad_norm": 0.5266439956425074,
+      "learning_rate": 8.04502555820518e-06,
+      "loss": 0.2633,
+      "step": 2247
+    },
+    {
+      "epoch": 1.1522296258329061,
+      "grad_norm": 0.6679499302065401,
+      "learning_rate": 8.036884126287382e-06,
+      "loss": 0.2949,
+      "step": 2248
+    },
+    {
+      "epoch": 1.1527421834956433,
+      "grad_norm": 0.505277262097061,
+      "learning_rate": 8.028744047511666e-06,
+      "loss": 0.2664,
+      "step": 2249
+    },
+    {
+      "epoch": 1.1532547411583802,
+      "grad_norm": 0.4994373023091951,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.2507,
+      "step": 2250
+    },
+    {
+      "epoch": 1.1537672988211174,
+      "grad_norm": 0.5095248324890755,
+      "learning_rate": 8.012467971828803e-06,
+      "loss": 0.2313,
+      "step": 2251
+    },
+    {
+      "epoch": 1.1542798564838543,
+      "grad_norm": 0.5101018617443275,
+      "learning_rate": 8.004331986140474e-06,
+      "loss": 0.2651,
+      "step": 2252
+    },
+    {
+      "epoch": 1.1547924141465915,
+      "grad_norm": 0.5288554100393075,
+      "learning_rate": 7.996197376031857e-06,
+      "loss": 0.2338,
+      "step": 2253
+    },
+    {
+      "epoch": 1.1553049718093287,
+      "grad_norm": 0.49951884946341923,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.2638,
+      "step": 2254
+    },
+    {
+      "epoch": 1.1558175294720656,
+      "grad_norm": 0.4911672026101427,
+      "learning_rate": 7.979932304980993e-06,
+      "loss": 0.2272,
+      "step": 2255
+    },
+    {
+      "epoch": 1.1563300871348026,
+      "grad_norm": 0.5457263106696684,
+      "learning_rate": 7.97180185524998e-06,
+      "loss": 0.243,
+      "step": 2256
+    },
+    {
+      "epoch": 1.1568426447975397,
+      "grad_norm": 0.4897018681180572,
+      "learning_rate": 7.963672803521136e-06,
+      "loss": 0.2601,
+      "step": 2257
+    },
+    {
+      "epoch": 1.157355202460277,
+      "grad_norm": 0.5050570233794145,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.247,
+      "step": 2258
+    },
+    {
+      "epoch": 1.1578677601230138,
+      "grad_norm": 0.5558433231056058,
+      "learning_rate": 7.947418916481867e-06,
+      "loss": 0.2672,
+      "step": 2259
+    },
+    {
+      "epoch": 1.1583803177857508,
+      "grad_norm": 0.5351591979961657,
+      "learning_rate": 7.939294092374961e-06,
+      "loss": 0.2515,
+      "step": 2260
+    },
+    {
+      "epoch": 1.158892875448488,
+      "grad_norm": 0.5385630725642093,
+      "learning_rate": 7.931170688677272e-06,
+      "loss": 0.2484,
+      "step": 2261
+    },
+    {
+      "epoch": 1.1594054331112251,
+      "grad_norm": 0.46478563614676255,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.2284,
+      "step": 2262
+    },
+    {
+      "epoch": 1.159917990773962,
+      "grad_norm": 0.5779642271128888,
+      "learning_rate": 7.914928164905844e-06,
+      "loss": 0.2675,
+      "step": 2263
+    },
+    {
+      "epoch": 1.160430548436699,
+      "grad_norm": 0.4920571253150662,
+      "learning_rate": 7.906809056027797e-06,
+      "loss": 0.2566,
+      "step": 2264
+    },
+    {
+      "epoch": 1.1609431060994362,
+      "grad_norm": 0.5476113641616556,
+      "learning_rate": 7.898691389950348e-06,
+      "loss": 0.2152,
+      "step": 2265
+    },
+    {
+      "epoch": 1.1614556637621734,
+      "grad_norm": 0.5227852867339912,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.242,
+      "step": 2266
+    },
+    {
+      "epoch": 1.1619682214249103,
+      "grad_norm": 0.4972574251944705,
+      "learning_rate": 7.882460408577696e-06,
+      "loss": 0.2342,
+      "step": 2267
+    },
+    {
+      "epoch": 1.1624807790876472,
+      "grad_norm": 0.520796748250538,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.2524,
+      "step": 2268
+    },
+    {
+      "epoch": 1.1629933367503844,
+      "grad_norm": 0.5026396286901406,
+      "learning_rate": 7.866235265538834e-06,
+      "loss": 0.2517,
+      "step": 2269
+    },
+    {
+      "epoch": 1.1635058944131216,
+      "grad_norm": 0.5400469542507097,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.281,
+      "step": 2270
+    },
+    {
+      "epoch": 1.1640184520758585,
+      "grad_norm": 0.5120953987659348,
+      "learning_rate": 7.850016005568593e-06,
+      "loss": 0.2889,
+      "step": 2271
+    },
+    {
+      "epoch": 1.1645310097385955,
+      "grad_norm": 0.5204780604042522,
+      "learning_rate": 7.841908595709416e-06,
+      "loss": 0.2484,
+      "step": 2272
+    },
+    {
+      "epoch": 1.1650435674013326,
+      "grad_norm": 0.5017571604298904,
+      "learning_rate": 7.833802673385604e-06,
+      "loss": 0.255,
+      "step": 2273
+    },
+    {
+      "epoch": 1.1655561250640698,
+      "grad_norm": 0.570284845625587,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2644,
+      "step": 2274
+    },
+    {
+      "epoch": 1.1660686827268067,
+      "grad_norm": 0.556732151352888,
+      "learning_rate": 7.817595313692143e-06,
+      "loss": 0.2604,
+      "step": 2275
+    },
+    {
+      "epoch": 1.166581240389544,
+      "grad_norm": 0.5295200998617984,
+      "learning_rate": 7.80949388749395e-06,
+      "loss": 0.2507,
+      "step": 2276
+    },
+    {
+      "epoch": 1.1670937980522809,
+      "grad_norm": 0.47832869605115014,
+      "learning_rate": 7.801393971174024e-06,
+      "loss": 0.2374,
+      "step": 2277
+    },
+    {
+      "epoch": 1.167606355715018,
+      "grad_norm": 0.5090891388300747,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.2522,
+      "step": 2278
+    },
+    {
+      "epoch": 1.168118913377755,
+      "grad_norm": 0.5117958478184721,
+      "learning_rate": 7.785198690500468e-06,
+      "loss": 0.2477,
+      "step": 2279
+    },
+    {
+      "epoch": 1.1686314710404921,
+      "grad_norm": 0.4876142743625617,
+      "learning_rate": 7.777103337309963e-06,
+      "loss": 0.2577,
+      "step": 2280
+    },
+    {
+      "epoch": 1.169144028703229,
+      "grad_norm": 0.516574818889365,
+      "learning_rate": 7.76900951632398e-06,
+      "loss": 0.2493,
+      "step": 2281
+    },
+    {
+      "epoch": 1.1696565863659663,
+      "grad_norm": 0.5332488636679411,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.2839,
+      "step": 2282
+    },
+    {
+      "epoch": 1.1701691440287032,
+      "grad_norm": 0.5421588897580533,
+      "learning_rate": 7.752826493280236e-06,
+      "loss": 0.2793,
+      "step": 2283
+    },
+    {
+      "epoch": 1.1706817016914404,
+      "grad_norm": 0.6015257157849299,
+      "learning_rate": 7.744737302377156e-06,
+      "loss": 0.2921,
+      "step": 2284
+    },
+    {
+      "epoch": 1.1711942593541773,
+      "grad_norm": 0.5322005829567272,
+      "learning_rate": 7.736649665987944e-06,
+      "loss": 0.2643,
+      "step": 2285
+    },
+    {
+      "epoch": 1.1717068170169145,
+      "grad_norm": 0.5321978664061342,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.2937,
+      "step": 2286
+    },
+    {
+      "epoch": 1.1722193746796514,
+      "grad_norm": 0.5331249069542525,
+      "learning_rate": 7.720479079048741e-06,
+      "loss": 0.2324,
+      "step": 2287
+    },
+    {
+      "epoch": 1.1727319323423886,
+      "grad_norm": 0.5634202438564926,
+      "learning_rate": 7.712396139644854e-06,
+      "loss": 0.2893,
+      "step": 2288
+    },
+    {
+      "epoch": 1.1732444900051255,
+      "grad_norm": 0.5589174037500428,
+      "learning_rate": 7.704314777047045e-06,
+      "loss": 0.2368,
+      "step": 2289
+    },
+    {
+      "epoch": 1.1737570476678627,
+      "grad_norm": 0.49657031310524785,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.2537,
+      "step": 2290
+    },
+    {
+      "epoch": 1.1742696053305997,
+      "grad_norm": 0.5683389062775803,
+      "learning_rate": 7.688156804549956e-06,
+      "loss": 0.2545,
+      "step": 2291
+    },
+    {
+      "epoch": 1.1747821629933368,
+      "grad_norm": 0.47759197565974437,
+      "learning_rate": 7.680080205788086e-06,
+      "loss": 0.2317,
+      "step": 2292
+    },
+    {
+      "epoch": 1.1752947206560738,
+      "grad_norm": 0.5129488064489816,
+      "learning_rate": 7.672005206107115e-06,
+      "loss": 0.2475,
+      "step": 2293
+    },
+    {
+      "epoch": 1.175807278318811,
+      "grad_norm": 0.46171034476881395,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.2298,
+      "step": 2294
+    },
+    {
+      "epoch": 1.1763198359815479,
+      "grad_norm": 0.5219266458517827,
+      "learning_rate": 7.655860026250596e-06,
+      "loss": 0.2547,
+      "step": 2295
+    },
+    {
+      "epoch": 1.176832393644285,
+      "grad_norm": 0.5598316677771373,
+      "learning_rate": 7.647789857203644e-06,
+      "loss": 0.2805,
+      "step": 2296
+    },
+    {
+      "epoch": 1.177344951307022,
+      "grad_norm": 0.532457667041288,
+      "learning_rate": 7.639721309494771e-06,
+      "loss": 0.2615,
+      "step": 2297
+    },
+    {
+      "epoch": 1.1778575089697592,
+      "grad_norm": 0.5173442863909762,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.2274,
+      "step": 2298
+    },
+    {
+      "epoch": 1.178370066632496,
+      "grad_norm": 0.49890740312538123,
+      "learning_rate": 7.623589100336199e-06,
+      "loss": 0.2344,
+      "step": 2299
+    },
+    {
+      "epoch": 1.1788826242952333,
+      "grad_norm": 0.48870637283089013,
+      "learning_rate": 7.6155254500061494e-06,
+      "loss": 0.2466,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1793951819579702,
+      "grad_norm": 0.4887045716993884,
+      "learning_rate": 7.607463443253487e-06,
+      "loss": 0.2565,
+      "step": 2301
+    },
+    {
+      "epoch": 1.1799077396207074,
+      "grad_norm": 0.5452857765364035,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.2701,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1804202972834443,
+      "grad_norm": 0.5344396543808454,
+      "learning_rate": 7.591344382707181e-06,
+      "loss": 0.2837,
+      "step": 2303
+    },
+    {
+      "epoch": 1.1809328549461815,
+      "grad_norm": 0.5920569405408033,
+      "learning_rate": 7.583287340024132e-06,
+      "loss": 0.2907,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1814454126089184,
+      "grad_norm": 0.5560719179321463,
+      "learning_rate": 7.575231963139645e-06,
+      "loss": 0.2681,
+      "step": 2305
+    },
+    {
+      "epoch": 1.1819579702716556,
+      "grad_norm": 0.5446971023776929,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.2552,
+      "step": 2306
+    },
+    {
+      "epoch": 1.1824705279343926,
+      "grad_norm": 0.4913893623632185,
+      "learning_rate": 7.559126228974921e-06,
+      "loss": 0.2636,
+      "step": 2307
+    },
+    {
+      "epoch": 1.1829830855971297,
+      "grad_norm": 0.52943961394468,
+      "learning_rate": 7.551075882796098e-06,
+      "loss": 0.2429,
+      "step": 2308
+    },
+    {
+      "epoch": 1.1834956432598667,
+      "grad_norm": 0.5266974108484638,
+      "learning_rate": 7.543027224618637e-06,
+      "loss": 0.241,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1840082009226038,
+      "grad_norm": 0.48881924399219023,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.2607,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1845207585853408,
+      "grad_norm": 0.5411757452046969,
+      "learning_rate": 7.5269349944578454e-06,
+      "loss": 0.2723,
+      "step": 2311
+    },
+    {
+      "epoch": 1.185033316248078,
+      "grad_norm": 0.5172420725487932,
+      "learning_rate": 7.518891433566617e-06,
+      "loss": 0.26,
+      "step": 2312
+    },
+    {
+      "epoch": 1.185545873910815,
+      "grad_norm": 0.48554250794049797,
+      "learning_rate": 7.510849582860936e-06,
+      "loss": 0.2663,
+      "step": 2313
+    },
+    {
+      "epoch": 1.186058431573552,
+      "grad_norm": 0.5161340678978056,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.2729,
+      "step": 2314
+    },
+    {
+      "epoch": 1.186570989236289,
+      "grad_norm": 0.5403417242069901,
+      "learning_rate": 7.49477103417749e-06,
+      "loss": 0.2505,
+      "step": 2315
+    },
+    {
+      "epoch": 1.1870835468990262,
+      "grad_norm": 0.4784489688912263,
+      "learning_rate": 7.486734347282394e-06,
+      "loss": 0.2385,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1875961045617631,
+      "grad_norm": 0.5121064126225968,
+      "learning_rate": 7.478699392738176e-06,
+      "loss": 0.2461,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1881086622245003,
+      "grad_norm": 0.49365870441600584,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.2679,
+      "step": 2318
+    },
+    {
+      "epoch": 1.1886212198872372,
+      "grad_norm": 0.49933463882824913,
+      "learning_rate": 7.462634702854607e-06,
+      "loss": 0.2471,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1891337775499744,
+      "grad_norm": 0.5226221243828247,
+      "learning_rate": 7.454604978588373e-06,
+      "loss": 0.2528,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1896463352127113,
+      "grad_norm": 0.5062059405251267,
+      "learning_rate": 7.446577008819238e-06,
+      "loss": 0.2627,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1901588928754485,
+      "grad_norm": 0.5451367648471334,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.2508,
+      "step": 2322
+    },
+    {
+      "epoch": 1.1906714505381855,
+      "grad_norm": 0.533618013277112,
+      "learning_rate": 7.430526354905232e-06,
+      "loss": 0.2561,
+      "step": 2323
+    },
+    {
+      "epoch": 1.1911840082009226,
+      "grad_norm": 0.5363874076993309,
+      "learning_rate": 7.4225036818238e-06,
+      "loss": 0.2669,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1916965658636596,
+      "grad_norm": 0.5287644678483414,
+      "learning_rate": 7.414482785366342e-06,
+      "loss": 0.2652,
+      "step": 2325
+    },
+    {
+      "epoch": 1.1922091235263967,
+      "grad_norm": 0.4859919247526991,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.2238,
+      "step": 2326
+    },
+    {
+      "epoch": 1.1927216811891337,
+      "grad_norm": 0.49742736878877586,
+      "learning_rate": 7.398446344436795e-06,
+      "loss": 0.2492,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1932342388518709,
+      "grad_norm": 0.5157293962040252,
+      "learning_rate": 7.390430811018343e-06,
+      "loss": 0.2546,
+      "step": 2328
+    },
+    {
+      "epoch": 1.1937467965146078,
+      "grad_norm": 0.5300032926559908,
+      "learning_rate": 7.382417076331148e-06,
+      "loss": 0.2336,
+      "step": 2329
+    },
+    {
+      "epoch": 1.194259354177345,
+      "grad_norm": 0.5661637378099852,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.2958,
+      "step": 2330
+    },
+    {
+      "epoch": 1.194771911840082,
+      "grad_norm": 0.5333851108575784,
+      "learning_rate": 7.3663950252441974e-06,
+      "loss": 0.2596,
+      "step": 2331
+    },
+    {
+      "epoch": 1.195284469502819,
+      "grad_norm": 0.5159539657914884,
+      "learning_rate": 7.358386719888166e-06,
+      "loss": 0.2599,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1957970271655562,
+      "grad_norm": 0.504063892302749,
+      "learning_rate": 7.35038023535083e-06,
+      "loss": 0.241,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1963095848282932,
+      "grad_norm": 0.4969094113583747,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.2309,
+      "step": 2334
+    },
+    {
+      "epoch": 1.1968221424910301,
+      "grad_norm": 0.5796627200585714,
+      "learning_rate": 7.334372750805926e-06,
+      "loss": 0.2533,
+      "step": 2335
+    },
+    {
+      "epoch": 1.1973347001537673,
+      "grad_norm": 0.519054328452811,
+      "learning_rate": 7.326371761832035e-06,
+      "loss": 0.2592,
+      "step": 2336
+    },
+    {
+      "epoch": 1.1978472578165045,
+      "grad_norm": 0.5214867347529252,
+      "learning_rate": 7.318372615744207e-06,
+      "loss": 0.2493,
+      "step": 2337
+    },
+    {
+      "epoch": 1.1983598154792414,
+      "grad_norm": 0.5030796981849853,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.2372,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1988723731419784,
+      "grad_norm": 0.46537517125057487,
+      "learning_rate": 7.30237987428014e-06,
+      "loss": 0.2184,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1993849308047155,
+      "grad_norm": 0.473494339859272,
+      "learning_rate": 7.294386289927425e-06,
+      "loss": 0.2453,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1998974884674527,
+      "grad_norm": 0.518373077127871,
+      "learning_rate": 7.286394570507811e-06,
+      "loss": 0.2615,
+      "step": 2341
+    },
+    {
+      "epoch": 1.2004100461301896,
+      "grad_norm": 0.5224753080690295,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.2522,
+      "step": 2342
+    },
+    {
+      "epoch": 1.2009226037929266,
+      "grad_norm": 0.5166275908859789,
+      "learning_rate": 7.270416748500788e-06,
+      "loss": 0.2485,
+      "step": 2343
+    },
+    {
+      "epoch": 1.2014351614556638,
+      "grad_norm": 0.5092184206963953,
+      "learning_rate": 7.2624306569266265e-06,
+      "loss": 0.259,
+      "step": 2344
+    },
+    {
+      "epoch": 1.201947719118401,
+      "grad_norm": 0.479320968304076,
+      "learning_rate": 7.254446452312027e-06,
+      "loss": 0.2489,
+      "step": 2345
+    },
+    {
+      "epoch": 1.2024602767811379,
+      "grad_norm": 0.5195952361317394,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.2525,
+      "step": 2346
+    },
+    {
+      "epoch": 1.2029728344438748,
+      "grad_norm": 0.5097743047484631,
+      "learning_rate": 7.238483725973715e-06,
+      "loss": 0.2286,
+      "step": 2347
+    },
+    {
+      "epoch": 1.203485392106612,
+      "grad_norm": 0.5850907469723176,
+      "learning_rate": 7.230505215252839e-06,
+      "loss": 0.2772,
+      "step": 2348
+    },
+    {
+      "epoch": 1.2039979497693492,
+      "grad_norm": 0.4905661263098126,
+      "learning_rate": 7.2225286134971815e-06,
+      "loss": 0.224,
+      "step": 2349
+    },
+    {
+      "epoch": 1.204510507432086,
+      "grad_norm": 0.5239480036603439,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.261,
+      "step": 2350
+    },
+    {
+      "epoch": 1.2050230650948233,
+      "grad_norm": 0.5405130864192055,
+      "learning_rate": 7.206581158872761e-06,
+      "loss": 0.262,
+      "step": 2351
+    },
+    {
+      "epoch": 1.2055356227575602,
+      "grad_norm": 0.5093059276246136,
+      "learning_rate": 7.198610316996303e-06,
+      "loss": 0.254,
+      "step": 2352
+    },
+    {
+      "epoch": 1.2060481804202974,
+      "grad_norm": 0.538818188138369,
+      "learning_rate": 7.190641406069672e-06,
+      "loss": 0.2465,
+      "step": 2353
+    },
+    {
+      "epoch": 1.2065607380830343,
+      "grad_norm": 0.498615317654422,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.233,
+      "step": 2354
+    },
+    {
+      "epoch": 1.2070732957457715,
+      "grad_norm": 0.5057681570982077,
+      "learning_rate": 7.1747093990359e-06,
+      "loss": 0.2719,
+      "step": 2355
+    },
+    {
+      "epoch": 1.2075858534085084,
+      "grad_norm": 0.5523898970490984,
+      "learning_rate": 7.166746313910413e-06,
+      "loss": 0.2398,
+      "step": 2356
+    },
+    {
+      "epoch": 1.2080984110712456,
+      "grad_norm": 0.5320047749987078,
+      "learning_rate": 7.158785181698069e-06,
+      "loss": 0.258,
+      "step": 2357
+    },
+    {
+      "epoch": 1.2086109687339825,
+      "grad_norm": 0.4713813790961183,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.2287,
+      "step": 2358
+    },
+    {
+      "epoch": 1.2091235263967197,
+      "grad_norm": 0.5118152678457204,
+      "learning_rate": 7.142868797961335e-06,
+      "loss": 0.2671,
+      "step": 2359
+    },
+    {
+      "epoch": 1.2096360840594567,
+      "grad_norm": 0.5034747673366126,
+      "learning_rate": 7.134913557407838e-06,
+      "loss": 0.2421,
+      "step": 2360
+    },
+    {
+      "epoch": 1.2101486417221938,
+      "grad_norm": 0.501013257482637,
+      "learning_rate": 7.126960291709248e-06,
+      "loss": 0.2262,
+      "step": 2361
+    },
+    {
+      "epoch": 1.2106611993849308,
+      "grad_norm": 0.5519210231400711,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.2718,
+      "step": 2362
+    },
+    {
+      "epoch": 1.211173757047668,
+      "grad_norm": 0.5881324942913527,
+      "learning_rate": 7.111059706803649e-06,
+      "loss": 0.2934,
+      "step": 2363
+    },
+    {
+      "epoch": 1.2116863147104049,
+      "grad_norm": 0.5020395695857044,
+      "learning_rate": 7.103112398556631e-06,
+      "loss": 0.2653,
+      "step": 2364
+    },
+    {
+      "epoch": 1.212198872373142,
+      "grad_norm": 0.5218274199964807,
+      "learning_rate": 7.095167087084518e-06,
+      "loss": 0.2613,
+      "step": 2365
+    },
+    {
+      "epoch": 1.212711430035879,
+      "grad_norm": 0.5076411847585143,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.2523,
+      "step": 2366
+    },
+    {
+      "epoch": 1.2132239876986162,
+      "grad_norm": 0.514753996459874,
+      "learning_rate": 7.079282476369902e-06,
+      "loss": 0.2739,
+      "step": 2367
+    },
+    {
+      "epoch": 1.2137365453613531,
+      "grad_norm": 0.48829678107264646,
+      "learning_rate": 7.071343188076383e-06,
+      "loss": 0.2364,
+      "step": 2368
+    },
+    {
+      "epoch": 1.2142491030240903,
+      "grad_norm": 0.5360828599014714,
+      "learning_rate": 7.0634059184557415e-06,
+      "loss": 0.228,
+      "step": 2369
+    },
+    {
+      "epoch": 1.2147616606868272,
+      "grad_norm": 0.5402018670849565,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.2684,
+      "step": 2370
+    },
+    {
+      "epoch": 1.2152742183495644,
+      "grad_norm": 0.5352594735130318,
+      "learning_rate": 7.04753745711579e-06,
+      "loss": 0.2363,
+      "step": 2371
+    },
+    {
+      "epoch": 1.2157867760123013,
+      "grad_norm": 0.4982542623730721,
+      "learning_rate": 7.039606276334336e-06,
+      "loss": 0.2578,
+      "step": 2372
+    },
+    {
+      "epoch": 1.2162993336750385,
+      "grad_norm": 0.5269347147918944,
+      "learning_rate": 7.031677136101471e-06,
+      "loss": 0.2441,
+      "step": 2373
+    },
+    {
+      "epoch": 1.2168118913377755,
+      "grad_norm": 0.5316300441542314,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.2673,
+      "step": 2374
+    },
+    {
+      "epoch": 1.2173244490005126,
+      "grad_norm": 0.4632000964225764,
+      "learning_rate": 7.015824999141762e-06,
+      "loss": 0.2502,
+      "step": 2375
+    },
+    {
+      "epoch": 1.2178370066632496,
+      "grad_norm": 0.5537675690275924,
+      "learning_rate": 7.007902013341525e-06,
+      "loss": 0.2531,
+      "step": 2376
+    },
+    {
+      "epoch": 1.2183495643259867,
+      "grad_norm": 0.502871663018061,
+      "learning_rate": 6.999981089943083e-06,
+      "loss": 0.2432,
+      "step": 2377
+    },
+    {
+      "epoch": 1.2188621219887237,
+      "grad_norm": 0.496016022536357,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.253,
+      "step": 2378
+    },
+    {
+      "epoch": 1.2193746796514608,
+      "grad_norm": 0.5208586105219468,
+      "learning_rate": 6.984145452189162e-06,
+      "loss": 0.2761,
+      "step": 2379
+    },
+    {
+      "epoch": 1.2198872373141978,
+      "grad_norm": 0.5568823878861947,
+      "learning_rate": 6.976230748748919e-06,
+      "loss": 0.2706,
+      "step": 2380
+    },
+    {
+      "epoch": 1.220399794976935,
+      "grad_norm": 0.5515297128111026,
+      "learning_rate": 6.968318129540929e-06,
+      "loss": 0.2658,
+      "step": 2381
+    },
+    {
+      "epoch": 1.220912352639672,
+      "grad_norm": 0.515443607999139,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.2604,
+      "step": 2382
+    },
+    {
+      "epoch": 1.221424910302409,
+      "grad_norm": 0.5468265073406282,
+      "learning_rate": 6.952499165636379e-06,
+      "loss": 0.2531,
+      "step": 2383
+    },
+    {
+      "epoch": 1.221937467965146,
+      "grad_norm": 0.5304026103743049,
+      "learning_rate": 6.944592831843566e-06,
+      "loss": 0.2908,
+      "step": 2384
+    },
+    {
+      "epoch": 1.2224500256278832,
+      "grad_norm": 0.47579905827514046,
+      "learning_rate": 6.936688604090472e-06,
+      "loss": 0.2352,
+      "step": 2385
+    },
+    {
+      "epoch": 1.2229625832906201,
+      "grad_norm": 0.46856086376819306,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.2407,
+      "step": 2386
+    },
+    {
+      "epoch": 1.2234751409533573,
+      "grad_norm": 0.5321524840008518,
+      "learning_rate": 6.920886488494989e-06,
+      "loss": 0.2549,
+      "step": 2387
+    },
+    {
+      "epoch": 1.2239876986160942,
+      "grad_norm": 0.5146608422831295,
+      "learning_rate": 6.9129886115447294e-06,
+      "loss": 0.2519,
+      "step": 2388
+    },
+    {
+      "epoch": 1.2245002562788314,
+      "grad_norm": 0.5515254863333425,
+      "learning_rate": 6.905092862418434e-06,
+      "loss": 0.3027,
+      "step": 2389
+    },
+    {
+      "epoch": 1.2250128139415684,
+      "grad_norm": 0.493567870698942,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.2395,
+      "step": 2390
+    },
+    {
+      "epoch": 1.2255253716043055,
+      "grad_norm": 0.5587149745340853,
+      "learning_rate": 6.889307769405904e-06,
+      "loss": 0.3046,
+      "step": 2391
+    },
+    {
+      "epoch": 1.2260379292670425,
+      "grad_norm": 0.49354803993223084,
+      "learning_rate": 6.881418436400056e-06,
+      "loss": 0.2882,
+      "step": 2392
+    },
+    {
+      "epoch": 1.2265504869297796,
+      "grad_norm": 0.5259978931103384,
+      "learning_rate": 6.873531252978959e-06,
+      "loss": 0.2765,
+      "step": 2393
+    },
+    {
+      "epoch": 1.2270630445925166,
+      "grad_norm": 0.567568927605565,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.2967,
+      "step": 2394
+    },
+    {
+      "epoch": 1.2275756022552538,
+      "grad_norm": 0.5205461514540273,
+      "learning_rate": 6.857763356635525e-06,
+      "loss": 0.2707,
+      "step": 2395
+    },
+    {
+      "epoch": 1.2280881599179907,
+      "grad_norm": 0.5040883099422797,
+      "learning_rate": 6.849882654581726e-06,
+      "loss": 0.231,
+      "step": 2396
+    },
+    {
+      "epoch": 1.2286007175807279,
+      "grad_norm": 0.5405457400033858,
+      "learning_rate": 6.8420041238497525e-06,
+      "loss": 0.2422,
+      "step": 2397
+    },
+    {
+      "epoch": 1.2291132752434648,
+      "grad_norm": 0.4834935806395893,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.2524,
+      "step": 2398
+    },
+    {
+      "epoch": 1.229625832906202,
+      "grad_norm": 0.48084184221094844,
+      "learning_rate": 6.826253598071911e-06,
+      "loss": 0.2367,
+      "step": 2399
+    },
+    {
+      "epoch": 1.230138390568939,
+      "grad_norm": 0.49239932907371403,
+      "learning_rate": 6.818381613882612e-06,
+      "loss": 0.2255,
+      "step": 2400
+    },
+    {
+      "epoch": 1.230650948231676,
+      "grad_norm": 0.5113966647567281,
+      "learning_rate": 6.8105118227282605e-06,
+      "loss": 0.2446,
+      "step": 2401
+    },
+    {
+      "epoch": 1.231163505894413,
+      "grad_norm": 0.558348256575361,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.2561,
+      "step": 2402
+    },
+    {
+      "epoch": 1.2316760635571502,
+      "grad_norm": 0.5024163712543215,
+      "learning_rate": 6.79477884122094e-06,
+      "loss": 0.2706,
+      "step": 2403
+    },
+    {
+      "epoch": 1.2321886212198871,
+      "grad_norm": 0.5682913678877385,
+      "learning_rate": 6.786915661712439e-06,
+      "loss": 0.2445,
+      "step": 2404
+    },
+    {
+      "epoch": 1.2327011788826243,
+      "grad_norm": 0.5332002713603435,
+      "learning_rate": 6.779054696927831e-06,
+      "loss": 0.2359,
+      "step": 2405
+    },
+    {
+      "epoch": 1.2332137365453613,
+      "grad_norm": 0.5112479961594727,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.23,
+      "step": 2406
+    },
+    {
+      "epoch": 1.2337262942080984,
+      "grad_norm": 0.5006277954258033,
+      "learning_rate": 6.763339433202462e-06,
+      "loss": 0.2457,
+      "step": 2407
+    },
+    {
+      "epoch": 1.2342388518708354,
+      "grad_norm": 0.5064999563605932,
+      "learning_rate": 6.755485145093965e-06,
+      "loss": 0.255,
+      "step": 2408
+    },
+    {
+      "epoch": 1.2347514095335725,
+      "grad_norm": 0.5280728669085032,
+      "learning_rate": 6.747633093373873e-06,
+      "loss": 0.2435,
+      "step": 2409
+    },
+    {
+      "epoch": 1.2352639671963095,
+      "grad_norm": 0.5499075357554034,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.2727,
+      "step": 2410
+    },
+    {
+      "epoch": 1.2357765248590467,
+      "grad_norm": 0.5240265139313158,
+      "learning_rate": 6.731935720746492e-06,
+      "loss": 0.2488,
+      "step": 2411
+    },
+    {
+      "epoch": 1.2362890825217838,
+      "grad_norm": 0.5131754098530877,
+      "learning_rate": 6.724090410659137e-06,
+      "loss": 0.253,
+      "step": 2412
+    },
+    {
+      "epoch": 1.2368016401845208,
+      "grad_norm": 0.4831065829545866,
+      "learning_rate": 6.716247358600038e-06,
+      "loss": 0.2552,
+      "step": 2413
+    },
+    {
+      "epoch": 1.2373141978472577,
+      "grad_norm": 0.4960064957601512,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.2729,
+      "step": 2414
+    },
+    {
+      "epoch": 1.2378267555099949,
+      "grad_norm": 0.5070643795937531,
+      "learning_rate": 6.700568050189371e-06,
+      "loss": 0.2631,
+      "step": 2415
+    },
+    {
+      "epoch": 1.238339313172732,
+      "grad_norm": 0.5273220201426861,
+      "learning_rate": 6.6927318046452825e-06,
+      "loss": 0.2549,
+      "step": 2416
+    },
+    {
+      "epoch": 1.238851870835469,
+      "grad_norm": 0.5049836515446374,
+      "learning_rate": 6.684897838744403e-06,
+      "loss": 0.2392,
+      "step": 2417
+    },
+    {
+      "epoch": 1.239364428498206,
+      "grad_norm": 0.529224386325416,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.2287,
+      "step": 2418
+    },
+    {
+      "epoch": 1.239876986160943,
+      "grad_norm": 0.4885528052058858,
+      "learning_rate": 6.669236767469952e-06,
+      "loss": 0.2624,
+      "step": 2419
+    },
+    {
+      "epoch": 1.2403895438236803,
+      "grad_norm": 0.5201744165603445,
+      "learning_rate": 6.6614096728912945e-06,
+      "loss": 0.2686,
+      "step": 2420
+    },
+    {
+      "epoch": 1.2409021014864172,
+      "grad_norm": 0.49542003080536245,
+      "learning_rate": 6.6535848795456485e-06,
+      "loss": 0.2599,
+      "step": 2421
+    },
+    {
+      "epoch": 1.2414146591491542,
+      "grad_norm": 0.5086693362345428,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.2977,
+      "step": 2422
+    },
+    {
+      "epoch": 1.2419272168118913,
+      "grad_norm": 0.5059009902952912,
+      "learning_rate": 6.637942218125786e-06,
+      "loss": 0.2513,
+      "step": 2423
+    },
+    {
+      "epoch": 1.2424397744746285,
+      "grad_norm": 0.5082974865636213,
+      "learning_rate": 6.630124360833794e-06,
+      "loss": 0.2495,
+      "step": 2424
+    },
+    {
+      "epoch": 1.2429523321373654,
+      "grad_norm": 0.508405142732196,
+      "learning_rate": 6.622308826339237e-06,
+      "loss": 0.2569,
+      "step": 2425
+    },
+    {
+      "epoch": 1.2434648898001024,
+      "grad_norm": 0.4823892433737954,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.2394,
+      "step": 2426
+    },
+    {
+      "epoch": 1.2439774474628396,
+      "grad_norm": 0.5529852632081494,
+      "learning_rate": 6.6066847472893046e-06,
+      "loss": 0.2495,
+      "step": 2427
+    },
+    {
+      "epoch": 1.2444900051255767,
+      "grad_norm": 0.5769712511420587,
+      "learning_rate": 6.5988762135033405e-06,
+      "loss": 0.2641,
+      "step": 2428
+    },
+    {
+      "epoch": 1.2450025627883137,
+      "grad_norm": 0.537076755864153,
+      "learning_rate": 6.591070024053623e-06,
+      "loss": 0.2381,
+      "step": 2429
+    },
+    {
+      "epoch": 1.2455151204510508,
+      "grad_norm": 0.47560819105935054,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.2192,
+      "step": 2430
+    },
+    {
+      "epoch": 1.2460276781137878,
+      "grad_norm": 0.48883981084850336,
+      "learning_rate": 6.575464699684029e-06,
+      "loss": 0.2493,
+      "step": 2431
+    },
+    {
+      "epoch": 1.246540235776525,
+      "grad_norm": 0.5027658900266005,
+      "learning_rate": 6.567665575520625e-06,
+      "loss": 0.2313,
+      "step": 2432
+    },
+    {
+      "epoch": 1.247052793439262,
+      "grad_norm": 0.505843185679179,
+      "learning_rate": 6.559868817206435e-06,
+      "loss": 0.2184,
+      "step": 2433
+    },
+    {
+      "epoch": 1.247565351101999,
+      "grad_norm": 0.5382249571682779,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.2322,
+      "step": 2434
+    },
+    {
+      "epoch": 1.248077908764736,
+      "grad_norm": 0.578947980012249,
+      "learning_rate": 6.54428241962074e-06,
+      "loss": 0.2633,
+      "step": 2435
+    },
+    {
+      "epoch": 1.2485904664274732,
+      "grad_norm": 0.6151125849195237,
+      "learning_rate": 6.53649279109267e-06,
+      "loss": 0.2638,
+      "step": 2436
+    },
+    {
+      "epoch": 1.2491030240902101,
+      "grad_norm": 0.5154895473064298,
+      "learning_rate": 6.528705549900674e-06,
+      "loss": 0.2642,
+      "step": 2437
+    },
+    {
+      "epoch": 1.2496155817529473,
+      "grad_norm": 0.5169596549109796,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.2304,
+      "step": 2438
+    },
+    {
+      "epoch": 1.2501281394156842,
+      "grad_norm": 0.49546854248282124,
+      "learning_rate": 6.513138250993716e-06,
+      "loss": 0.2418,
+      "step": 2439
+    },
+    {
+      "epoch": 1.2506406970784214,
+      "grad_norm": 0.49800438948615494,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.2364,
+      "step": 2440
+    },
+    {
+      "epoch": 1.2511532547411583,
+      "grad_norm": 0.5060157071270097,
+      "learning_rate": 6.4975805658209324e-06,
+      "loss": 0.2512,
+      "step": 2441
+    },
+    {
+      "epoch": 1.2516658124038955,
+      "grad_norm": 0.5136097766321261,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.2474,
+      "step": 2442
+    },
+    {
+      "epoch": 1.2521783700666325,
+      "grad_norm": 0.5204463056684469,
+      "learning_rate": 6.4820325372769126e-06,
+      "loss": 0.2199,
+      "step": 2443
+    },
+    {
+      "epoch": 1.2526909277293696,
+      "grad_norm": 0.4879776288898382,
+      "learning_rate": 6.474262157637958e-06,
+      "loss": 0.2386,
+      "step": 2444
+    },
+    {
+      "epoch": 1.2532034853921066,
+      "grad_norm": 0.5157492723532902,
+      "learning_rate": 6.4664942082295925e-06,
+      "loss": 0.2666,
+      "step": 2445
+    },
+    {
+      "epoch": 1.2537160430548437,
+      "grad_norm": 0.5512669542263444,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.2821,
+      "step": 2446
+    },
+    {
+      "epoch": 1.2542286007175807,
+      "grad_norm": 0.4979294288440422,
+      "learning_rate": 6.450965621520187e-06,
+      "loss": 0.2432,
+      "step": 2447
+    },
+    {
+      "epoch": 1.2547411583803179,
+      "grad_norm": 0.5169134782704756,
+      "learning_rate": 6.443204994922738e-06,
+      "loss": 0.2179,
+      "step": 2448
+    },
+    {
+      "epoch": 1.2552537160430548,
+      "grad_norm": 0.5640419285638771,
+      "learning_rate": 6.4354468199630425e-06,
+      "loss": 0.2588,
+      "step": 2449
+    },
+    {
+      "epoch": 1.255766273705792,
+      "grad_norm": 0.5459458510224707,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.2657,
+      "step": 2450
+    },
+    {
+      "epoch": 1.256278831368529,
+      "grad_norm": 0.5084056091418061,
+      "learning_rate": 6.419937846345519e-06,
+      "loss": 0.2328,
+      "step": 2451
+    },
+    {
+      "epoch": 1.256791389031266,
+      "grad_norm": 0.5167194152736742,
+      "learning_rate": 6.412187058377762e-06,
+      "loss": 0.2547,
+      "step": 2452
+    },
+    {
+      "epoch": 1.257303946694003,
+      "grad_norm": 0.4814327192682674,
+      "learning_rate": 6.404438743427894e-06,
+      "loss": 0.2516,
+      "step": 2453
+    },
+    {
+      "epoch": 1.2578165043567402,
+      "grad_norm": 0.499192542061777,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.2324,
+      "step": 2454
+    },
+    {
+      "epoch": 1.2583290620194771,
+      "grad_norm": 0.4979418933080457,
+      "learning_rate": 6.388949553943216e-06,
+      "loss": 0.2271,
+      "step": 2455
+    },
+    {
+      "epoch": 1.2588416196822143,
+      "grad_norm": 0.5169680952749267,
+      "learning_rate": 6.3812086900848456e-06,
+      "loss": 0.2372,
+      "step": 2456
+    },
+    {
+      "epoch": 1.2593541773449513,
+      "grad_norm": 0.5141199547900289,
+      "learning_rate": 6.373470320597212e-06,
+      "loss": 0.2567,
+      "step": 2457
+    },
+    {
+      "epoch": 1.2598667350076884,
+      "grad_norm": 0.4866024393163067,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.2371,
+      "step": 2458
+    },
+    {
+      "epoch": 1.2603792926704254,
+      "grad_norm": 0.47955956191675686,
+      "learning_rate": 6.358001086068153e-06,
+      "loss": 0.2532,
+      "step": 2459
+    },
+    {
+      "epoch": 1.2608918503331625,
+      "grad_norm": 0.5228507832415459,
+      "learning_rate": 6.350270231689405e-06,
+      "loss": 0.2335,
+      "step": 2460
+    },
+    {
+      "epoch": 1.2614044079958995,
+      "grad_norm": 0.5086423809544871,
+      "learning_rate": 6.342541893006746e-06,
+      "loss": 0.2402,
+      "step": 2461
+    },
+    {
+      "epoch": 1.2619169656586366,
+      "grad_norm": 0.5161778519709852,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2608,
+      "step": 2462
+    },
+    {
+      "epoch": 1.2624295233213736,
+      "grad_norm": 0.5977926500158645,
+      "learning_rate": 6.327092784036002e-06,
+      "loss": 0.2625,
+      "step": 2463
+    },
+    {
+      "epoch": 1.2629420809841108,
+      "grad_norm": 0.5012209828135894,
+      "learning_rate": 6.319372024396728e-06,
+      "loss": 0.2355,
+      "step": 2464
+    },
+    {
+      "epoch": 1.2634546386468477,
+      "grad_norm": 0.5023251433281644,
+      "learning_rate": 6.311653801751145e-06,
+      "loss": 0.2369,
+      "step": 2465
+    },
+    {
+      "epoch": 1.2639671963095849,
+      "grad_norm": 0.5276329696276642,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.2517,
+      "step": 2466
+    },
+    {
+      "epoch": 1.2644797539723218,
+      "grad_norm": 0.5029173406813046,
+      "learning_rate": 6.296224988719466e-06,
+      "loss": 0.2413,
+      "step": 2467
+    },
+    {
+      "epoch": 1.264992311635059,
+      "grad_norm": 0.5390195927524575,
+      "learning_rate": 6.288514408968183e-06,
+      "loss": 0.2495,
+      "step": 2468
+    },
+    {
+      "epoch": 1.265504869297796,
+      "grad_norm": 0.5401009373762207,
+      "learning_rate": 6.2808063874802165e-06,
+      "loss": 0.2826,
+      "step": 2469
+    },
+    {
+      "epoch": 1.266017426960533,
+      "grad_norm": 0.553194738892275,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.2557,
+      "step": 2470
+    },
+    {
+      "epoch": 1.26652998462327,
+      "grad_norm": 0.5107584897188191,
+      "learning_rate": 6.265398040544508e-06,
+      "loss": 0.2622,
+      "step": 2471
+    },
+    {
+      "epoch": 1.2670425422860072,
+      "grad_norm": 0.5199763030600897,
+      "learning_rate": 6.257697725717469e-06,
+      "loss": 0.2791,
+      "step": 2472
+    },
+    {
+      "epoch": 1.2675550999487442,
+      "grad_norm": 0.49388295757723866,
+      "learning_rate": 6.249999990395162e-06,
+      "loss": 0.2447,
+      "step": 2473
+    },
+    {
+      "epoch": 1.2680676576114813,
+      "grad_norm": 0.4876242701422739,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.2548,
+      "step": 2474
+    },
+    {
+      "epoch": 1.2685802152742183,
+      "grad_norm": 0.544575472015716,
+      "learning_rate": 6.23461227948662e-06,
+      "loss": 0.2634,
+      "step": 2475
+    },
+    {
+      "epoch": 1.2690927729369554,
+      "grad_norm": 0.5483278985094436,
+      "learning_rate": 6.226922314506874e-06,
+      "loss": 0.2615,
+      "step": 2476
+    },
+    {
+      "epoch": 1.2696053305996924,
+      "grad_norm": 0.4756221716207978,
+      "learning_rate": 6.219234950244816e-06,
+      "loss": 0.2183,
+      "step": 2477
+    },
+    {
+      "epoch": 1.2701178882624296,
+      "grad_norm": 0.5272354894163076,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.2712,
+      "step": 2478
+    },
+    {
+      "epoch": 1.2706304459251667,
+      "grad_norm": 0.5246532809427572,
+      "learning_rate": 6.2038680450670605e-06,
+      "loss": 0.2719,
+      "step": 2479
+    },
+    {
+      "epoch": 1.2711430035879037,
+      "grad_norm": 0.48845882979859845,
+      "learning_rate": 6.1961885147435015e-06,
+      "loss": 0.2385,
+      "step": 2480
+    },
+    {
+      "epoch": 1.2716555612506406,
+      "grad_norm": 0.5004852384504955,
+      "learning_rate": 6.188511606321918e-06,
+      "loss": 0.2178,
+      "step": 2481
+    },
+    {
+      "epoch": 1.2721681189133778,
+      "grad_norm": 0.5047695590283983,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.2367,
+      "step": 2482
+    },
+    {
+      "epoch": 1.272680676576115,
+      "grad_norm": 0.4775919485356019,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.2233,
+      "step": 2483
+    },
+    {
+      "epoch": 1.273193234238852,
+      "grad_norm": 0.5217436991119405,
+      "learning_rate": 6.165496665375556e-06,
+      "loss": 0.2618,
+      "step": 2484
+    },
+    {
+      "epoch": 1.2737057919015888,
+      "grad_norm": 0.5061297750115425,
+      "learning_rate": 6.1578302974593526e-06,
+      "loss": 0.2445,
+      "step": 2485
+    },
+    {
+      "epoch": 1.274218349564326,
+      "grad_norm": 0.4554652565465731,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.2323,
+      "step": 2486
+    },
+    {
+      "epoch": 1.2747309072270632,
+      "grad_norm": 0.49158013576862986,
+      "learning_rate": 6.1425055119343115e-06,
+      "loss": 0.2228,
+      "step": 2487
+    },
+    {
+      "epoch": 1.2752434648898001,
+      "grad_norm": 0.4856568616523521,
+      "learning_rate": 6.134847104888588e-06,
+      "loss": 0.2598,
+      "step": 2488
+    },
+    {
+      "epoch": 1.275756022552537,
+      "grad_norm": 0.5427258901898124,
+      "learning_rate": 6.1271913620264225e-06,
+      "loss": 0.2621,
+      "step": 2489
+    },
+    {
+      "epoch": 1.2762685802152742,
+      "grad_norm": 0.5193017721596687,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.2599,
+      "step": 2490
+    },
+    {
+      "epoch": 1.2767811378780114,
+      "grad_norm": 0.4999837263387376,
+      "learning_rate": 6.111887889958795e-06,
+      "loss": 0.2665,
+      "step": 2491
+    },
+    {
+      "epoch": 1.2772936955407483,
+      "grad_norm": 0.5140937300805926,
+      "learning_rate": 6.104240171301756e-06,
+      "loss": 0.2701,
+      "step": 2492
+    },
+    {
+      "epoch": 1.2778062532034853,
+      "grad_norm": 0.5172413580492691,
+      "learning_rate": 6.096595137925111e-06,
+      "loss": 0.2657,
+      "step": 2493
+    },
+    {
+      "epoch": 1.2783188108662225,
+      "grad_norm": 0.49098998402994726,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.2407,
+      "step": 2494
+    },
+    {
+      "epoch": 1.2788313685289596,
+      "grad_norm": 0.5027336229896262,
+      "learning_rate": 6.081313148089485e-06,
+      "loss": 0.2235,
+      "step": 2495
+    },
+    {
+      "epoch": 1.2793439261916966,
+      "grad_norm": 0.5464178110403574,
+      "learning_rate": 6.073676202164123e-06,
+      "loss": 0.2423,
+      "step": 2496
+    },
+    {
+      "epoch": 1.2798564838544335,
+      "grad_norm": 0.5031983216879574,
+      "learning_rate": 6.066041962586367e-06,
+      "loss": 0.2423,
+      "step": 2497
+    },
+    {
+      "epoch": 1.2803690415171707,
+      "grad_norm": 0.536329583541824,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.2607,
+      "step": 2498
+    },
+    {
+      "epoch": 1.2808815991799078,
+      "grad_norm": 0.4950177761813918,
+      "learning_rate": 6.0507816235204065e-06,
+      "loss": 0.231,
+      "step": 2499
+    },
+    {
+      "epoch": 1.2813941568426448,
+      "grad_norm": 0.4988369220702212,
+      "learning_rate": 6.043155534550897e-06,
+      "loss": 0.2331,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2819067145053817,
+      "grad_norm": 0.5213335171039187,
+      "learning_rate": 6.035532172966367e-06,
+      "loss": 0.2651,
+      "step": 2501
+    },
+    {
+      "epoch": 1.282419272168119,
+      "grad_norm": 0.5175508572681506,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.2465,
+      "step": 2502
+    },
+    {
+      "epoch": 1.282931829830856,
+      "grad_norm": 0.4965438861580493,
+      "learning_rate": 6.020293652968969e-06,
+      "loss": 0.2331,
+      "step": 2503
+    },
+    {
+      "epoch": 1.283444387493593,
+      "grad_norm": 0.49265040090751033,
+      "learning_rate": 6.012678505059755e-06,
+      "loss": 0.2394,
+      "step": 2504
+    },
+    {
+      "epoch": 1.28395694515633,
+      "grad_norm": 0.520292667627535,
+      "learning_rate": 6.00506610554281e-06,
+      "loss": 0.2577,
+      "step": 2505
+    },
+    {
+      "epoch": 1.2844695028190671,
+      "grad_norm": 0.5415825161118798,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.2612,
+      "step": 2506
+    },
+    {
+      "epoch": 1.2849820604818043,
+      "grad_norm": 0.5618686760177504,
+      "learning_rate": 5.989849572672235e-06,
+      "loss": 0.2625,
+      "step": 2507
+    },
+    {
+      "epoch": 1.2854946181445412,
+      "grad_norm": 0.5017835026089567,
+      "learning_rate": 5.982245449807094e-06,
+      "loss": 0.2263,
+      "step": 2508
+    },
+    {
+      "epoch": 1.2860071758072782,
+      "grad_norm": 0.48080987125739877,
+      "learning_rate": 5.9746440963112095e-06,
+      "loss": 0.206,
+      "step": 2509
+    },
+    {
+      "epoch": 1.2865197334700154,
+      "grad_norm": 0.5227478810045003,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.2703,
+      "step": 2510
+    },
+    {
+      "epoch": 1.2870322911327525,
+      "grad_norm": 0.4899609873652123,
+      "learning_rate": 5.9594497183832276e-06,
+      "loss": 0.2569,
+      "step": 2511
+    },
+    {
+      "epoch": 1.2875448487954895,
+      "grad_norm": 0.530344749874674,
+      "learning_rate": 5.951856704424353e-06,
+      "loss": 0.2576,
+      "step": 2512
+    },
+    {
+      "epoch": 1.2880574064582264,
+      "grad_norm": 0.46890105028550605,
+      "learning_rate": 5.944266480781173e-06,
+      "loss": 0.2219,
+      "step": 2513
+    },
+    {
+      "epoch": 1.2885699641209636,
+      "grad_norm": 0.49960185170318494,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.2295,
+      "step": 2514
+    },
+    {
+      "epoch": 1.2890825217837008,
+      "grad_norm": 0.48235543384664786,
+      "learning_rate": 5.929094425367224e-06,
+      "loss": 0.2471,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2895950794464377,
+      "grad_norm": 0.5449379550582383,
+      "learning_rate": 5.921512604054289e-06,
+      "loss": 0.2643,
+      "step": 2516
+    },
+    {
+      "epoch": 1.2901076371091749,
+      "grad_norm": 0.46358030714263926,
+      "learning_rate": 5.913933593972719e-06,
+      "loss": 0.2285,
+      "step": 2517
+    },
+    {
+      "epoch": 1.2906201947719118,
+      "grad_norm": 0.5107788486373495,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.2555,
+      "step": 2518
+    },
+    {
+      "epoch": 1.291132752434649,
+      "grad_norm": 0.5400366476299389,
+      "learning_rate": 5.898784028398058e-06,
+      "loss": 0.2466,
+      "step": 2519
+    },
+    {
+      "epoch": 1.291645310097386,
+      "grad_norm": 0.5513526760177174,
+      "learning_rate": 5.8912134833473e-06,
+      "loss": 0.2727,
+      "step": 2520
+    },
+    {
+      "epoch": 1.292157867760123,
+      "grad_norm": 0.5035523280351851,
+      "learning_rate": 5.883645770412574e-06,
+      "loss": 0.2487,
+      "step": 2521
+    },
+    {
+      "epoch": 1.29267042542286,
+      "grad_norm": 0.5135611373699764,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.1984,
+      "step": 2522
+    },
+    {
+      "epoch": 1.2931829830855972,
+      "grad_norm": 0.46541054849248326,
+      "learning_rate": 5.868518861754422e-06,
+      "loss": 0.2402,
+      "step": 2523
+    },
+    {
+      "epoch": 1.2936955407483341,
+      "grad_norm": 0.5046688077604533,
+      "learning_rate": 5.860959676457718e-06,
+      "loss": 0.2326,
+      "step": 2524
+    },
+    {
+      "epoch": 1.2942080984110713,
+      "grad_norm": 0.4593509337800865,
+      "learning_rate": 5.853403344130476e-06,
+      "loss": 0.2215,
+      "step": 2525
+    },
+    {
+      "epoch": 1.2947206560738083,
+      "grad_norm": 0.4885417202544483,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.2066,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2952332137365454,
+      "grad_norm": 0.5847444603865174,
+      "learning_rate": 5.838299259216187e-06,
+      "loss": 0.2666,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2957457713992824,
+      "grad_norm": 0.4934529554718069,
+      "learning_rate": 5.830751517040133e-06,
+      "loss": 0.2197,
+      "step": 2528
+    },
+    {
+      "epoch": 1.2962583290620195,
+      "grad_norm": 0.5629191052644454,
+      "learning_rate": 5.823206648655503e-06,
+      "loss": 0.2554,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2967708867247565,
+      "grad_norm": 0.5110997546426863,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.233,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2972834443874937,
+      "grad_norm": 0.5404468204051014,
+      "learning_rate": 5.808125554060723e-06,
+      "loss": 0.2698,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2977960020502306,
+      "grad_norm": 0.5444348878703112,
+      "learning_rate": 5.8005893382457125e-06,
+      "loss": 0.2495,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2983085597129678,
+      "grad_norm": 0.5086868840461163,
+      "learning_rate": 5.7930560170123985e-06,
+      "loss": 0.2406,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2988211173757047,
+      "grad_norm": 0.447972392081748,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.2167,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2993336750384419,
+      "grad_norm": 0.5045517123247597,
+      "learning_rate": 5.777998079059215e-06,
+      "loss": 0.274,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2998462327011788,
+      "grad_norm": 0.48345230327090943,
+      "learning_rate": 5.770473472718525e-06,
+      "loss": 0.2303,
+      "step": 2536
+    },
+    {
+      "epoch": 1.300358790363916,
+      "grad_norm": 0.5331802158682861,
+      "learning_rate": 5.762951781717879e-06,
+      "loss": 0.2405,
+      "step": 2537
+    },
+    {
+      "epoch": 1.300871348026653,
+      "grad_norm": 0.5095406614781585,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.2305,
+      "step": 2538
+    },
+    {
+      "epoch": 1.30138390568939,
+      "grad_norm": 0.5041279968621207,
+      "learning_rate": 5.747917166473e-06,
+      "loss": 0.2339,
+      "step": 2539
+    },
+    {
+      "epoch": 1.301896463352127,
+      "grad_norm": 0.48775102266840614,
+      "learning_rate": 5.74040425259187e-06,
+      "loss": 0.243,
+      "step": 2540
+    },
+    {
+      "epoch": 1.3024090210148642,
+      "grad_norm": 0.5439389373363818,
+      "learning_rate": 5.7328942747769824e-06,
+      "loss": 0.2427,
+      "step": 2541
+    },
+    {
+      "epoch": 1.3029215786776012,
+      "grad_norm": 0.5318963166721099,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.2449,
+      "step": 2542
+    },
+    {
+      "epoch": 1.3034341363403383,
+      "grad_norm": 0.5685013961749393,
+      "learning_rate": 5.717883148049901e-06,
+      "loss": 0.2315,
+      "step": 2543
+    },
+    {
+      "epoch": 1.3039466940030753,
+      "grad_norm": 0.5382664807129798,
+      "learning_rate": 5.710382009484616e-06,
+      "loss": 0.2678,
+      "step": 2544
+    },
+    {
+      "epoch": 1.3044592516658124,
+      "grad_norm": 0.5059203783722912,
+      "learning_rate": 5.702883827679397e-06,
+      "loss": 0.255,
+      "step": 2545
+    },
+    {
+      "epoch": 1.3049718093285494,
+      "grad_norm": 0.5115110681414926,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.2382,
+      "step": 2546
+    },
+    {
+      "epoch": 1.3054843669912866,
+      "grad_norm": 0.5639459895415478,
+      "learning_rate": 5.687896355020563e-06,
+      "loss": 0.2521,
+      "step": 2547
+    },
+    {
+      "epoch": 1.3059969246540235,
+      "grad_norm": 0.4963408340284299,
+      "learning_rate": 5.680407074497558e-06,
+      "loss": 0.2461,
+      "step": 2548
+    },
+    {
+      "epoch": 1.3065094823167607,
+      "grad_norm": 0.4779186605011712,
+      "learning_rate": 5.672920771395822e-06,
+      "loss": 0.2381,
+      "step": 2549
+    },
+    {
+      "epoch": 1.3070220399794976,
+      "grad_norm": 0.5075922738596074,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.2693,
+      "step": 2550
+    },
+    {
+      "epoch": 1.3075345976422348,
+      "grad_norm": 0.5065143776586216,
+      "learning_rate": 5.657957118094818e-06,
+      "loss": 0.2632,
+      "step": 2551
+    },
+    {
+      "epoch": 1.3080471553049717,
+      "grad_norm": 0.45048114380198667,
+      "learning_rate": 5.6504797782097386e-06,
+      "loss": 0.2419,
+      "step": 2552
+    },
+    {
+      "epoch": 1.308559712967709,
+      "grad_norm": 0.481297221307459,
+      "learning_rate": 5.6430054363742995e-06,
+      "loss": 0.2364,
+      "step": 2553
+    },
+    {
+      "epoch": 1.3090722706304458,
+      "grad_norm": 0.5775014557776591,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.3077,
+      "step": 2554
+    },
+    {
+      "epoch": 1.309584828293183,
+      "grad_norm": 0.4785775159732644,
+      "learning_rate": 5.628065767458011e-06,
+      "loss": 0.2213,
+      "step": 2555
+    },
+    {
+      "epoch": 1.31009738595592,
+      "grad_norm": 0.503642235836788,
+      "learning_rate": 5.620600450674819e-06,
+      "loss": 0.2405,
+      "step": 2556
+    },
+    {
+      "epoch": 1.3106099436186571,
+      "grad_norm": 0.5274200675020765,
+      "learning_rate": 5.61313815253657e-06,
+      "loss": 0.2525,
+      "step": 2557
+    },
+    {
+      "epoch": 1.3111225012813943,
+      "grad_norm": 0.5117152245683873,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.2456,
+      "step": 2558
+    },
+    {
+      "epoch": 1.3116350589441312,
+      "grad_norm": 0.535167379782528,
+      "learning_rate": 5.598222632767374e-06,
+      "loss": 0.2428,
+      "step": 2559
+    },
+    {
+      "epoch": 1.3121476166068682,
+      "grad_norm": 0.5502114648229134,
+      "learning_rate": 5.590769421417435e-06,
+      "loss": 0.2428,
+      "step": 2560
+    },
+    {
+      "epoch": 1.3126601742696054,
+      "grad_norm": 0.5599930165156427,
+      "learning_rate": 5.5833192492744635e-06,
+      "loss": 0.2387,
+      "step": 2561
+    },
+    {
+      "epoch": 1.3131727319323425,
+      "grad_norm": 0.4444107923928092,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.1873,
+      "step": 2562
+    },
+    {
+      "epoch": 1.3136852895950795,
+      "grad_norm": 0.4670926310998341,
+      "learning_rate": 5.568428043148402e-06,
+      "loss": 0.2331,
+      "step": 2563
+    },
+    {
+      "epoch": 1.3141978472578164,
+      "grad_norm": 0.5201980281082624,
+      "learning_rate": 5.560987019429575e-06,
+      "loss": 0.2433,
+      "step": 2564
+    },
+    {
+      "epoch": 1.3147104049205536,
+      "grad_norm": 0.4770590402974806,
+      "learning_rate": 5.553549055446209e-06,
+      "loss": 0.237,
+      "step": 2565
+    },
+    {
+      "epoch": 1.3152229625832907,
+      "grad_norm": 0.48069286063347283,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.2311,
+      "step": 2566
+    },
+    {
+      "epoch": 1.3157355202460277,
+      "grad_norm": 0.4890333490248736,
+      "learning_rate": 5.538682327191195e-06,
+      "loss": 0.2298,
+      "step": 2567
+    },
+    {
+      "epoch": 1.3162480779087646,
+      "grad_norm": 0.5105379001471538,
+      "learning_rate": 5.531253573166928e-06,
+      "loss": 0.2316,
+      "step": 2568
+    },
+    {
+      "epoch": 1.3167606355715018,
+      "grad_norm": 0.5324574255663623,
+      "learning_rate": 5.523827899372876e-06,
+      "loss": 0.233,
+      "step": 2569
+    },
+    {
+      "epoch": 1.317273193234239,
+      "grad_norm": 0.4882654822572307,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.2503,
+      "step": 2570
+    },
+    {
+      "epoch": 1.317785750896976,
+      "grad_norm": 0.4611423936261431,
+      "learning_rate": 5.50898581294685e-06,
+      "loss": 0.2245,
+      "step": 2571
+    },
+    {
+      "epoch": 1.3182983085597129,
+      "grad_norm": 0.49289216419119725,
+      "learning_rate": 5.501569410545278e-06,
+      "loss": 0.2638,
+      "step": 2572
+    },
+    {
+      "epoch": 1.31881086622245,
+      "grad_norm": 0.4468255787682782,
+      "learning_rate": 5.494156108834703e-06,
+      "loss": 0.216,
+      "step": 2573
+    },
+    {
+      "epoch": 1.3193234238851872,
+      "grad_norm": 0.4900237302326073,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.2202,
+      "step": 2574
+    },
+    {
+      "epoch": 1.3198359815479241,
+      "grad_norm": 0.5222614402690201,
+      "learning_rate": 5.479338827923854e-06,
+      "loss": 0.2338,
+      "step": 2575
+    },
+    {
+      "epoch": 1.320348539210661,
+      "grad_norm": 0.5612469577311083,
+      "learning_rate": 5.471934858936887e-06,
+      "loss": 0.2819,
+      "step": 2576
+    },
+    {
+      "epoch": 1.3208610968733983,
+      "grad_norm": 0.573581697027328,
+      "learning_rate": 5.464534011067499e-06,
+      "loss": 0.2696,
+      "step": 2577
+    },
+    {
+      "epoch": 1.3213736545361354,
+      "grad_norm": 0.5379264872216687,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.2499,
+      "step": 2578
+    },
+    {
+      "epoch": 1.3218862121988724,
+      "grad_norm": 0.5184055633791488,
+      "learning_rate": 5.4497416990844545e-06,
+      "loss": 0.2618,
+      "step": 2579
+    },
+    {
+      "epoch": 1.3223987698616093,
+      "grad_norm": 0.522960669586445,
+      "learning_rate": 5.442350245166872e-06,
+      "loss": 0.2614,
+      "step": 2580
+    },
+    {
+      "epoch": 1.3229113275243465,
+      "grad_norm": 0.5557945228928917,
+      "learning_rate": 5.4349619327590474e-06,
+      "loss": 0.2708,
+      "step": 2581
+    },
+    {
+      "epoch": 1.3234238851870836,
+      "grad_norm": 0.4723617507823411,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.2562,
+      "step": 2582
+    },
+    {
+      "epoch": 1.3239364428498206,
+      "grad_norm": 0.48971317026242983,
+      "learning_rate": 5.4201947528410425e-06,
+      "loss": 0.2267,
+      "step": 2583
+    },
+    {
+      "epoch": 1.3244490005125575,
+      "grad_norm": 0.5304269817599963,
+      "learning_rate": 5.412815895509624e-06,
+      "loss": 0.2689,
+      "step": 2584
+    },
+    {
+      "epoch": 1.3249615581752947,
+      "grad_norm": 0.5528473730287224,
+      "learning_rate": 5.405440200045482e-06,
+      "loss": 0.2426,
+      "step": 2585
+    },
+    {
+      "epoch": 1.3254741158380319,
+      "grad_norm": 0.560753869455338,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.2584,
+      "step": 2586
+    },
+    {
+      "epoch": 1.3259866735007688,
+      "grad_norm": 0.4773028999858124,
+      "learning_rate": 5.390698315052598e-06,
+      "loss": 0.2379,
+      "step": 2587
+    },
+    {
+      "epoch": 1.3264992311635058,
+      "grad_norm": 0.46862229545422235,
+      "learning_rate": 5.383332135685186e-06,
+      "loss": 0.2296,
+      "step": 2588
+    },
+    {
+      "epoch": 1.327011788826243,
+      "grad_norm": 0.4988844549972113,
+      "learning_rate": 5.375969138507695e-06,
+      "loss": 0.2319,
+      "step": 2589
+    },
+    {
+      "epoch": 1.32752434648898,
+      "grad_norm": 0.49819117854915335,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.2234,
+      "step": 2590
+    },
+    {
+      "epoch": 1.328036904151717,
+      "grad_norm": 0.5463144690628838,
+      "learning_rate": 5.3612527110210495e-06,
+      "loss": 0.2524,
+      "step": 2591
+    },
+    {
+      "epoch": 1.328549461814454,
+      "grad_norm": 0.4946224537373504,
+      "learning_rate": 5.353899290855666e-06,
+      "loss": 0.2258,
+      "step": 2592
+    },
+    {
+      "epoch": 1.3290620194771912,
+      "grad_norm": 0.5397006908618118,
+      "learning_rate": 5.346549073167766e-06,
+      "loss": 0.2321,
+      "step": 2593
+    },
+    {
+      "epoch": 1.3295745771399283,
+      "grad_norm": 0.5529293772081201,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.2657,
+      "step": 2594
+    },
+    {
+      "epoch": 1.3300871348026653,
+      "grad_norm": 0.5166478233984764,
+      "learning_rate": 5.3318582654877074e-06,
+      "loss": 0.2551,
+      "step": 2595
+    },
+    {
+      "epoch": 1.3305996924654024,
+      "grad_norm": 0.4801002025061306,
+      "learning_rate": 5.324517685621671e-06,
+      "loss": 0.2489,
+      "step": 2596
+    },
+    {
+      "epoch": 1.3311122501281394,
+      "grad_norm": 0.5063937110357613,
+      "learning_rate": 5.3171803284853544e-06,
+      "loss": 0.2248,
+      "step": 2597
+    },
+    {
+      "epoch": 1.3316248077908766,
+      "grad_norm": 0.5271503722030908,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.2377,
+      "step": 2598
+    },
+    {
+      "epoch": 1.3321373654536135,
+      "grad_norm": 0.5446183206834085,
+      "learning_rate": 5.302515302629686e-06,
+      "loss": 0.2325,
+      "step": 2599
+    },
+    {
+      "epoch": 1.3326499231163507,
+      "grad_norm": 0.4718374057543461,
+      "learning_rate": 5.295187644018694e-06,
+      "loss": 0.2445,
+      "step": 2600
+    },
+    {
+      "epoch": 1.3331624807790876,
+      "grad_norm": 0.515771148642771,
+      "learning_rate": 5.287863228354113e-06,
+      "loss": 0.2742,
+      "step": 2601
+    },
+    {
+      "epoch": 1.3336750384418248,
+      "grad_norm": 0.48971612353407923,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.235,
+      "step": 2602
+    },
+    {
+      "epoch": 1.3341875961045617,
+      "grad_norm": 0.5503978825740152,
+      "learning_rate": 5.273224146056315e-06,
+      "loss": 0.2992,
+      "step": 2603
+    },
+    {
+      "epoch": 1.334700153767299,
+      "grad_norm": 0.5137353409514249,
+      "learning_rate": 5.265909489513568e-06,
+      "loss": 0.2413,
+      "step": 2604
+    },
+    {
+      "epoch": 1.3352127114300358,
+      "grad_norm": 0.5244996972755682,
+      "learning_rate": 5.25859809609816e-06,
+      "loss": 0.2493,
+      "step": 2605
+    },
+    {
+      "epoch": 1.335725269092773,
+      "grad_norm": 0.4924328442003221,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.2549,
+      "step": 2606
+    },
+    {
+      "epoch": 1.33623782675551,
+      "grad_norm": 0.5507412017644685,
+      "learning_rate": 5.243985118805577e-06,
+      "loss": 0.2896,
+      "step": 2607
+    },
+    {
+      "epoch": 1.3367503844182471,
+      "grad_norm": 0.5521799326127118,
+      "learning_rate": 5.2366835450008836e-06,
+      "loss": 0.2477,
+      "step": 2608
+    },
+    {
+      "epoch": 1.337262942080984,
+      "grad_norm": 0.5226445139930481,
+      "learning_rate": 5.229385254468469e-06,
+      "loss": 0.244,
+      "step": 2609
+    },
+    {
+      "epoch": 1.3377754997437212,
+      "grad_norm": 0.5253510477195242,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.2392,
+      "step": 2610
+    },
+    {
+      "epoch": 1.3382880574064582,
+      "grad_norm": 0.5116589379619284,
+      "learning_rate": 5.2147985433405555e-06,
+      "loss": 0.2794,
+      "step": 2611
+    },
+    {
+      "epoch": 1.3388006150691953,
+      "grad_norm": 0.4883023137768304,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.2527,
+      "step": 2612
+    },
+    {
+      "epoch": 1.3393131727319323,
+      "grad_norm": 0.4969511071174131,
+      "learning_rate": 5.20022502563932e-06,
+      "loss": 0.257,
+      "step": 2613
+    },
+    {
+      "epoch": 1.3398257303946695,
+      "grad_norm": 0.49702570046583694,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.2569,
+      "step": 2614
+    },
+    {
+      "epoch": 1.3403382880574064,
+      "grad_norm": 0.4893689904892226,
+      "learning_rate": 5.185664741545862e-06,
+      "loss": 0.2305,
+      "step": 2615
+    },
+    {
+      "epoch": 1.3408508457201436,
+      "grad_norm": 0.4768870139439399,
+      "learning_rate": 5.178389574648659e-06,
+      "loss": 0.2379,
+      "step": 2616
+    },
+    {
+      "epoch": 1.3413634033828805,
+      "grad_norm": 0.5128270749434748,
+      "learning_rate": 5.171117731204775e-06,
+      "loss": 0.2528,
+      "step": 2617
+    },
+    {
+      "epoch": 1.3418759610456177,
+      "grad_norm": 0.5053603479632709,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.2464,
+      "step": 2618
+    },
+    {
+      "epoch": 1.3423885187083546,
+      "grad_norm": 0.5482264521653295,
+      "learning_rate": 5.156584034724081e-06,
+      "loss": 0.2893,
+      "step": 2619
+    },
+    {
+      "epoch": 1.3429010763710918,
+      "grad_norm": 0.47643203208857626,
+      "learning_rate": 5.149322191705102e-06,
+      "loss": 0.2278,
+      "step": 2620
+    },
+    {
+      "epoch": 1.3434136340338287,
+      "grad_norm": 0.5006794570539145,
+      "learning_rate": 5.142063692175084e-06,
+      "loss": 0.258,
+      "step": 2621
+    },
+    {
+      "epoch": 1.343926191696566,
+      "grad_norm": 0.5360880674427793,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.2626,
+      "step": 2622
+    },
+    {
+      "epoch": 1.3444387493593029,
+      "grad_norm": 0.5209712669361808,
+      "learning_rate": 5.127556743592263e-06,
+      "loss": 0.2448,
+      "step": 2623
+    },
+    {
+      "epoch": 1.34495130702204,
+      "grad_norm": 0.5494530757681839,
+      "learning_rate": 5.120308304538852e-06,
+      "loss": 0.2847,
+      "step": 2624
+    },
+    {
+      "epoch": 1.345463864684777,
+      "grad_norm": 0.525033399650878,
+      "learning_rate": 5.113063228973174e-06,
+      "loss": 0.2534,
+      "step": 2625
+    },
+    {
+      "epoch": 1.3459764223475141,
+      "grad_norm": 0.5261349415504945,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.2656,
+      "step": 2626
+    },
+    {
+      "epoch": 1.346488980010251,
+      "grad_norm": 0.5747039451998944,
+      "learning_rate": 5.098583188278345e-06,
+      "loss": 0.2428,
+      "step": 2627
+    },
+    {
+      "epoch": 1.3470015376729882,
+      "grad_norm": 0.524862384901596,
+      "learning_rate": 5.0913482331300244e-06,
+      "loss": 0.2843,
+      "step": 2628
+    },
+    {
+      "epoch": 1.3475140953357252,
+      "grad_norm": 0.5223588305189625,
+      "learning_rate": 5.0841166614311266e-06,
+      "loss": 0.2525,
+      "step": 2629
+    },
+    {
+      "epoch": 1.3480266529984624,
+      "grad_norm": 0.5346845843318712,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.2541,
+      "step": 2630
+    },
+    {
+      "epoch": 1.3485392106611993,
+      "grad_norm": 0.49810781929382614,
+      "learning_rate": 5.069663688317638e-06,
+      "loss": 0.2315,
+      "step": 2631
+    },
+    {
+      "epoch": 1.3490517683239365,
+      "grad_norm": 0.4604071198432951,
+      "learning_rate": 5.062442296865234e-06,
+      "loss": 0.218,
+      "step": 2632
+    },
+    {
+      "epoch": 1.3495643259866734,
+      "grad_norm": 0.547431599955885,
+      "learning_rate": 5.055224308786614e-06,
+      "loss": 0.2463,
+      "step": 2633
+    },
+    {
+      "epoch": 1.3500768836494106,
+      "grad_norm": 0.47559831031057787,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.2482,
+      "step": 2634
+    },
+    {
+      "epoch": 1.3505894413121475,
+      "grad_norm": 0.5095112135347194,
+      "learning_rate": 5.040798562649314e-06,
+      "loss": 0.2816,
+      "step": 2635
+    },
+    {
+      "epoch": 1.3511019989748847,
+      "grad_norm": 0.49635495285953235,
+      "learning_rate": 5.033590814534054e-06,
+      "loss": 0.2748,
+      "step": 2636
+    },
+    {
+      "epoch": 1.3516145566376219,
+      "grad_norm": 0.46760214430417196,
+      "learning_rate": 5.026386489679408e-06,
+      "loss": 0.2413,
+      "step": 2637
+    },
+    {
+      "epoch": 1.3521271143003588,
+      "grad_norm": 0.5330299817426193,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.2546,
+      "step": 2638
+    },
+    {
+      "epoch": 1.3526396719630958,
+      "grad_norm": 0.525619477717961,
+      "learning_rate": 5.011988129612868e-06,
+      "loss": 0.2486,
+      "step": 2639
+    },
+    {
+      "epoch": 1.353152229625833,
+      "grad_norm": 0.5024769505314648,
+      "learning_rate": 5.004794104325512e-06,
+      "loss": 0.2381,
+      "step": 2640
+    },
+    {
+      "epoch": 1.35366478728857,
+      "grad_norm": 0.5275177944406891,
+      "learning_rate": 4.997603522147856e-06,
+      "loss": 0.2492,
+      "step": 2641
+    },
+    {
+      "epoch": 1.354177344951307,
+      "grad_norm": 0.517887120871458,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.2524,
+      "step": 2642
+    },
+    {
+      "epoch": 1.354689902614044,
+      "grad_norm": 0.4833132742586356,
+      "learning_rate": 4.983232706944613e-06,
+      "loss": 0.2329,
+      "step": 2643
+    },
+    {
+      "epoch": 1.3552024602767812,
+      "grad_norm": 0.5027483696606287,
+      "learning_rate": 4.976052483824595e-06,
+      "loss": 0.2214,
+      "step": 2644
+    },
+    {
+      "epoch": 1.3557150179395183,
+      "grad_norm": 0.5205672972494466,
+      "learning_rate": 4.968875723625371e-06,
+      "loss": 0.233,
+      "step": 2645
+    },
+    {
+      "epoch": 1.3562275756022553,
+      "grad_norm": 0.5350710151946503,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.2688,
+      "step": 2646
+    },
+    {
+      "epoch": 1.3567401332649922,
+      "grad_norm": 0.5070073894022253,
+      "learning_rate": 4.954532611774195e-06,
+      "loss": 0.2561,
+      "step": 2647
+    },
+    {
+      "epoch": 1.3572526909277294,
+      "grad_norm": 0.5695324644826026,
+      "learning_rate": 4.947366270008708e-06,
+      "loss": 0.283,
+      "step": 2648
+    },
+    {
+      "epoch": 1.3577652485904665,
+      "grad_norm": 0.4883245423055449,
+      "learning_rate": 4.940203410936934e-06,
+      "loss": 0.2481,
+      "step": 2649
+    },
+    {
+      "epoch": 1.3582778062532035,
+      "grad_norm": 0.5009040767957541,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.2628,
+      "step": 2650
+    },
+    {
+      "epoch": 1.3587903639159404,
+      "grad_norm": 0.5314520322638253,
+      "learning_rate": 4.925888160621064e-06,
+      "loss": 0.2511,
+      "step": 2651
+    },
+    {
+      "epoch": 1.3593029215786776,
+      "grad_norm": 0.5390517800811301,
+      "learning_rate": 4.918735779244226e-06,
+      "loss": 0.269,
+      "step": 2652
+    },
+    {
+      "epoch": 1.3598154792414148,
+      "grad_norm": 0.4996517705863301,
+      "learning_rate": 4.911586900295606e-06,
+      "loss": 0.2305,
+      "step": 2653
+    },
+    {
+      "epoch": 1.3603280369041517,
+      "grad_norm": 0.507673949284833,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.2367,
+      "step": 2654
+    },
+    {
+      "epoch": 1.3608405945668887,
+      "grad_norm": 0.5261206964376673,
+      "learning_rate": 4.897299669391006e-06,
+      "loss": 0.266,
+      "step": 2655
+    },
+    {
+      "epoch": 1.3613531522296258,
+      "grad_norm": 0.5587178001196389,
+      "learning_rate": 4.890161327282968e-06,
+      "loss": 0.2674,
+      "step": 2656
+    },
+    {
+      "epoch": 1.361865709892363,
+      "grad_norm": 0.5106635392683897,
+      "learning_rate": 4.883026507299028e-06,
+      "loss": 0.2486,
+      "step": 2657
+    },
+    {
+      "epoch": 1.3623782675551,
+      "grad_norm": 0.5399304583813702,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.2858,
+      "step": 2658
+    },
+    {
+      "epoch": 1.3628908252178369,
+      "grad_norm": 0.5222846805271688,
+      "learning_rate": 4.868767453372649e-06,
+      "loss": 0.2462,
+      "step": 2659
+    },
+    {
+      "epoch": 1.363403382880574,
+      "grad_norm": 0.5325328380563498,
+      "learning_rate": 4.8616432292587315e-06,
+      "loss": 0.2465,
+      "step": 2660
+    },
+    {
+      "epoch": 1.3639159405433112,
+      "grad_norm": 0.5111742426416055,
+      "learning_rate": 4.8545225469259435e-06,
+      "loss": 0.2454,
+      "step": 2661
+    },
+    {
+      "epoch": 1.3644284982060482,
+      "grad_norm": 0.44058529384478917,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.2123,
+      "step": 2662
+    },
+    {
+      "epoch": 1.3649410558687851,
+      "grad_norm": 0.4986949643571755,
+      "learning_rate": 4.8402918272339905e-06,
+      "loss": 0.2719,
+      "step": 2663
+    },
+    {
+      "epoch": 1.3654536135315223,
+      "grad_norm": 0.5138797867084516,
+      "learning_rate": 4.833181799683812e-06,
+      "loss": 0.2375,
+      "step": 2664
+    },
+    {
+      "epoch": 1.3659661711942594,
+      "grad_norm": 0.46330404377889445,
+      "learning_rate": 4.826075333532742e-06,
+      "loss": 0.211,
+      "step": 2665
+    },
+    {
+      "epoch": 1.3664787288569964,
+      "grad_norm": 0.5231835384017229,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.2618,
+      "step": 2666
+    },
+    {
+      "epoch": 1.3669912865197333,
+      "grad_norm": 0.47107968232280106,
+      "learning_rate": 4.81187310501893e-06,
+      "loss": 0.2447,
+      "step": 2667
+    },
+    {
+      "epoch": 1.3675038441824705,
+      "grad_norm": 0.5799186953592933,
+      "learning_rate": 4.804777352445542e-06,
+      "loss": 0.274,
+      "step": 2668
+    },
+    {
+      "epoch": 1.3680164018452077,
+      "grad_norm": 0.5952888550992259,
+      "learning_rate": 4.79768518084996e-06,
+      "loss": 0.2554,
+      "step": 2669
+    },
+    {
+      "epoch": 1.3685289595079446,
+      "grad_norm": 0.5064364699106617,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.2632,
+      "step": 2670
+    },
+    {
+      "epoch": 1.3690415171706816,
+      "grad_norm": 0.5036427547264193,
+      "learning_rate": 4.783511600143796e-06,
+      "loss": 0.2495,
+      "step": 2671
+    },
+    {
+      "epoch": 1.3695540748334187,
+      "grad_norm": 0.48109497035328375,
+      "learning_rate": 4.776430200802819e-06,
+      "loss": 0.243,
+      "step": 2672
+    },
+    {
+      "epoch": 1.370066632496156,
+      "grad_norm": 0.4865662405694858,
+      "learning_rate": 4.7693524019788525e-06,
+      "loss": 0.251,
+      "step": 2673
+    },
+    {
+      "epoch": 1.3705791901588928,
+      "grad_norm": 0.5007569430380457,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.2434,
+      "step": 2674
+    },
+    {
+      "epoch": 1.37109174782163,
+      "grad_norm": 0.5440940339623456,
+      "learning_rate": 4.7552076253938905e-06,
+      "loss": 0.2488,
+      "step": 2675
+    },
+    {
+      "epoch": 1.371604305484367,
+      "grad_norm": 0.47905514400050686,
+      "learning_rate": 4.7481406573826496e-06,
+      "loss": 0.2176,
+      "step": 2676
+    },
+    {
+      "epoch": 1.3721168631471041,
+      "grad_norm": 0.48830298429038194,
+      "learning_rate": 4.741077309387921e-06,
+      "loss": 0.242,
+      "step": 2677
+    },
+    {
+      "epoch": 1.372629420809841,
+      "grad_norm": 0.5185289847934935,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2456,
+      "step": 2678
+    },
+    {
+      "epoch": 1.3731419784725782,
+      "grad_norm": 0.5305414868398456,
+      "learning_rate": 4.726961492920057e-06,
+      "loss": 0.2606,
+      "step": 2679
+    },
+    {
+      "epoch": 1.3736545361353152,
+      "grad_norm": 0.4970249301519707,
+      "learning_rate": 4.719909034176724e-06,
+      "loss": 0.2563,
+      "step": 2680
+    },
+    {
+      "epoch": 1.3741670937980524,
+      "grad_norm": 0.5246046929656653,
+      "learning_rate": 4.712860214909466e-06,
+      "loss": 0.2634,
+      "step": 2681
+    },
+    {
+      "epoch": 1.3746796514607893,
+      "grad_norm": 0.47260464800162516,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.2256,
+      "step": 2682
+    },
+    {
+      "epoch": 1.3751922091235265,
+      "grad_norm": 0.478870465995926,
+      "learning_rate": 4.698773514235212e-06,
+      "loss": 0.2662,
+      "step": 2683
+    },
+    {
+      "epoch": 1.3757047667862634,
+      "grad_norm": 0.5267585911036742,
+      "learning_rate": 4.691735642537934e-06,
+      "loss": 0.2376,
+      "step": 2684
+    },
+    {
+      "epoch": 1.3762173244490006,
+      "grad_norm": 0.4718598053178981,
+      "learning_rate": 4.6847014297361735e-06,
+      "loss": 0.2309,
+      "step": 2685
+    },
+    {
+      "epoch": 1.3767298821117375,
+      "grad_norm": 0.5920303188167082,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.2928,
+      "step": 2686
+    },
+    {
+      "epoch": 1.3772424397744747,
+      "grad_norm": 0.48833552970915284,
+      "learning_rate": 4.6706440002109275e-06,
+      "loss": 0.244,
+      "step": 2687
+    },
+    {
+      "epoch": 1.3777549974372116,
+      "grad_norm": 0.5344804779912966,
+      "learning_rate": 4.6636207931769884e-06,
+      "loss": 0.2393,
+      "step": 2688
+    },
+    {
+      "epoch": 1.3782675550999488,
+      "grad_norm": 0.5008201192698263,
+      "learning_rate": 4.65660126441765e-06,
+      "loss": 0.2466,
+      "step": 2689
+    },
+    {
+      "epoch": 1.3787801127626858,
+      "grad_norm": 0.5482851577916646,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.2711,
+      "step": 2690
+    },
+    {
+      "epoch": 1.379292670425423,
+      "grad_norm": 0.5309566867245732,
+      "learning_rate": 4.642573261073996e-06,
+      "loss": 0.2578,
+      "step": 2691
+    },
+    {
+      "epoch": 1.3798052280881599,
+      "grad_norm": 0.47820297901671777,
+      "learning_rate": 4.635564796158946e-06,
+      "loss": 0.2115,
+      "step": 2692
+    },
+    {
+      "epoch": 1.380317785750897,
+      "grad_norm": 0.5351378992578941,
+      "learning_rate": 4.628560028857008e-06,
+      "loss": 0.2383,
+      "step": 2693
+    },
+    {
+      "epoch": 1.380830343413634,
+      "grad_norm": 0.5151531725070799,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.2373,
+      "step": 2694
+    },
+    {
+      "epoch": 1.3813429010763711,
+      "grad_norm": 0.4927828803138208,
+      "learning_rate": 4.6145616064030095e-06,
+      "loss": 0.2194,
+      "step": 2695
+    },
+    {
+      "epoch": 1.381855458739108,
+      "grad_norm": 0.5373474547544395,
+      "learning_rate": 4.60756796089981e-06,
+      "loss": 0.2446,
+      "step": 2696
+    },
+    {
+      "epoch": 1.3823680164018453,
+      "grad_norm": 0.5021742366505709,
+      "learning_rate": 4.60057803230747e-06,
+      "loss": 0.2269,
+      "step": 2697
+    },
+    {
+      "epoch": 1.3828805740645822,
+      "grad_norm": 0.5626995402854468,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2361,
+      "step": 2698
+    },
+    {
+      "epoch": 1.3833931317273194,
+      "grad_norm": 0.5276226490036021,
+      "learning_rate": 4.586609345124944e-06,
+      "loss": 0.2541,
+      "step": 2699
+    },
+    {
+      "epoch": 1.3839056893900563,
+      "grad_norm": 0.5183403343758763,
+      "learning_rate": 4.579630596163134e-06,
+      "loss": 0.2534,
+      "step": 2700
+    },
+    {
+      "epoch": 1.3844182470527935,
+      "grad_norm": 0.5031341296644977,
+      "learning_rate": 4.572655583368927e-06,
+      "loss": 0.2119,
+      "step": 2701
+    },
+    {
+      "epoch": 1.3849308047155304,
+      "grad_norm": 0.5108210051960552,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.2639,
+      "step": 2702
+    },
+    {
+      "epoch": 1.3854433623782676,
+      "grad_norm": 0.5474108944582746,
+      "learning_rate": 4.558716785511764e-06,
+      "loss": 0.2988,
+      "step": 2703
+    },
+    {
+      "epoch": 1.3859559200410045,
+      "grad_norm": 0.4834724139553934,
+      "learning_rate": 4.551753010056584e-06,
+      "loss": 0.2184,
+      "step": 2704
+    },
+    {
+      "epoch": 1.3864684777037417,
+      "grad_norm": 0.5541261904252728,
+      "learning_rate": 4.544792989984544e-06,
+      "loss": 0.2401,
+      "step": 2705
+    },
+    {
+      "epoch": 1.3869810353664787,
+      "grad_norm": 0.5407147508523791,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.2466,
+      "step": 2706
+    },
+    {
+      "epoch": 1.3874935930292158,
+      "grad_norm": 0.4771087871847018,
+      "learning_rate": 4.5308842351770026e-06,
+      "loss": 0.2254,
+      "step": 2707
+    },
+    {
+      "epoch": 1.3880061506919528,
+      "grad_norm": 0.5404126745470428,
+      "learning_rate": 4.523935510028556e-06,
+      "loss": 0.2593,
+      "step": 2708
+    },
+    {
+      "epoch": 1.38851870835469,
+      "grad_norm": 0.5589818116993125,
+      "learning_rate": 4.516990559437388e-06,
+      "loss": 0.2698,
+      "step": 2709
+    },
+    {
+      "epoch": 1.3890312660174269,
+      "grad_norm": 0.5350549793667265,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.2583,
+      "step": 2710
+    },
+    {
+      "epoch": 1.389543823680164,
+      "grad_norm": 0.5362162442789421,
+      "learning_rate": 4.503112001072381e-06,
+      "loss": 0.2708,
+      "step": 2711
+    },
+    {
+      "epoch": 1.3900563813429012,
+      "grad_norm": 0.5463499404184535,
+      "learning_rate": 4.4961784028648025e-06,
+      "loss": 0.2679,
+      "step": 2712
+    },
+    {
+      "epoch": 1.3905689390056382,
+      "grad_norm": 0.5572908915461321,
+      "learning_rate": 4.489248598346994e-06,
+      "loss": 0.2695,
+      "step": 2713
+    },
+    {
+      "epoch": 1.391081496668375,
+      "grad_norm": 0.49696436267347754,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.2282,
+      "step": 2714
+    },
+    {
+      "epoch": 1.3915940543311123,
+      "grad_norm": 0.5353881432861487,
+      "learning_rate": 4.47540038948443e-06,
+      "loss": 0.2632,
+      "step": 2715
+    },
+    {
+      "epoch": 1.3921066119938494,
+      "grad_norm": 0.5511870354957058,
+      "learning_rate": 4.468481994685018e-06,
+      "loss": 0.2753,
+      "step": 2716
+    },
+    {
+      "epoch": 1.3926191696565864,
+      "grad_norm": 0.4694785410596922,
+      "learning_rate": 4.4615674126660226e-06,
+      "loss": 0.2135,
+      "step": 2717
+    },
+    {
+      "epoch": 1.3931317273193233,
+      "grad_norm": 0.5530124786607367,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.2433,
+      "step": 2718
+    },
+    {
+      "epoch": 1.3936442849820605,
+      "grad_norm": 0.5119532138323598,
+      "learning_rate": 4.447749706031096e-06,
+      "loss": 0.2607,
+      "step": 2719
+    },
+    {
+      "epoch": 1.3941568426447977,
+      "grad_norm": 0.48141189121627387,
+      "learning_rate": 4.440846590939469e-06,
+      "loss": 0.2306,
+      "step": 2720
+    },
+    {
+      "epoch": 1.3946694003075346,
+      "grad_norm": 0.5160891115152177,
+      "learning_rate": 4.4339473076768765e-06,
+      "loss": 0.2689,
+      "step": 2721
+    },
+    {
+      "epoch": 1.3951819579702716,
+      "grad_norm": 0.4942015656801272,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.2288,
+      "step": 2722
+    },
+    {
+      "epoch": 1.3956945156330087,
+      "grad_norm": 0.516008720915471,
+      "learning_rate": 4.4201602556583775e-06,
+      "loss": 0.2457,
+      "step": 2723
+    },
+    {
+      "epoch": 1.396207073295746,
+      "grad_norm": 0.5212331653258999,
+      "learning_rate": 4.4132724964056576e-06,
+      "loss": 0.2599,
+      "step": 2724
+    },
+    {
+      "epoch": 1.3967196309584828,
+      "grad_norm": 0.5249084349377573,
+      "learning_rate": 4.406388587988318e-06,
+      "loss": 0.2328,
+      "step": 2725
+    },
+    {
+      "epoch": 1.3972321886212198,
+      "grad_norm": 0.5244178386388963,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.2567,
+      "step": 2726
+    },
+    {
+      "epoch": 1.397744746283957,
+      "grad_norm": 0.5032719955535463,
+      "learning_rate": 4.392632342636978e-06,
+      "loss": 0.2359,
+      "step": 2727
+    },
+    {
+      "epoch": 1.3982573039466941,
+      "grad_norm": 0.5447097201648587,
+      "learning_rate": 4.385760015184919e-06,
+      "loss": 0.2852,
+      "step": 2728
+    },
+    {
+      "epoch": 1.398769861609431,
+      "grad_norm": 0.501866821808872,
+      "learning_rate": 4.378891557532123e-06,
+      "loss": 0.2335,
+      "step": 2729
+    },
+    {
+      "epoch": 1.399282419272168,
+      "grad_norm": 0.5302533289807867,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.2297,
+      "step": 2730
+    },
+    {
+      "epoch": 1.3997949769349052,
+      "grad_norm": 0.5281886297107157,
+      "learning_rate": 4.3651662705589104e-06,
+      "loss": 0.2453,
+      "step": 2731
+    },
+    {
+      "epoch": 1.4003075345976423,
+      "grad_norm": 0.5379473159097063,
+      "learning_rate": 4.358309450699087e-06,
+      "loss": 0.2245,
+      "step": 2732
+    },
+    {
+      "epoch": 1.4008200922603793,
+      "grad_norm": 0.5815823566799437,
+      "learning_rate": 4.3514565195597356e-06,
+      "loss": 0.2863,
+      "step": 2733
+    },
+    {
+      "epoch": 1.4013326499231162,
+      "grad_norm": 0.4841396444704307,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.2351,
+      "step": 2734
+    },
+    {
+      "epoch": 1.4018452075858534,
+      "grad_norm": 0.5356363837431538,
+      "learning_rate": 4.337762342334189e-06,
+      "loss": 0.2326,
+      "step": 2735
+    },
+    {
+      "epoch": 1.4023577652485906,
+      "grad_norm": 0.5238393100602086,
+      "learning_rate": 4.330921105687155e-06,
+      "loss": 0.243,
+      "step": 2736
+    },
+    {
+      "epoch": 1.4028703229113275,
+      "grad_norm": 0.5652289445122705,
+      "learning_rate": 4.324083776638905e-06,
+      "loss": 0.2702,
+      "step": 2737
+    },
+    {
+      "epoch": 1.4033828805740645,
+      "grad_norm": 0.528765665354688,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.2611,
+      "step": 2738
+    },
+    {
+      "epoch": 1.4038954382368016,
+      "grad_norm": 0.5010536146031103,
+      "learning_rate": 4.31042086018748e-06,
+      "loss": 0.2352,
+      "step": 2739
+    },
+    {
+      "epoch": 1.4044079958995388,
+      "grad_norm": 0.5250173122759618,
+      "learning_rate": 4.303595282201918e-06,
+      "loss": 0.2457,
+      "step": 2740
+    },
+    {
+      "epoch": 1.4049205535622757,
+      "grad_norm": 0.49527498374103085,
+      "learning_rate": 4.296773630650358e-06,
+      "loss": 0.2588,
+      "step": 2741
+    },
+    {
+      "epoch": 1.4054331112250127,
+      "grad_norm": 0.511998338110813,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.2343,
+      "step": 2742
+    },
+    {
+      "epoch": 1.4059456688877499,
+      "grad_norm": 0.516589613733181,
+      "learning_rate": 4.283142125654748e-06,
+      "loss": 0.235,
+      "step": 2743
+    },
+    {
+      "epoch": 1.406458226550487,
+      "grad_norm": 0.5332070562179088,
+      "learning_rate": 4.27633228160665e-06,
+      "loss": 0.2386,
+      "step": 2744
+    },
+    {
+      "epoch": 1.406970784213224,
+      "grad_norm": 0.5415608839341663,
+      "learning_rate": 4.2695263827844835e-06,
+      "loss": 0.2291,
+      "step": 2745
+    },
+    {
+      "epoch": 1.407483341875961,
+      "grad_norm": 0.5102756972093746,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.2328,
+      "step": 2746
+    },
+    {
+      "epoch": 1.407995899538698,
+      "grad_norm": 0.4759583367951981,
+      "learning_rate": 4.255926439579948e-06,
+      "loss": 0.238,
+      "step": 2747
+    },
+    {
+      "epoch": 1.4085084572014352,
+      "grad_norm": 0.5704398668670948,
+      "learning_rate": 4.249132404571794e-06,
+      "loss": 0.2794,
+      "step": 2748
+    },
+    {
+      "epoch": 1.4090210148641722,
+      "grad_norm": 0.49642764068657724,
+      "learning_rate": 4.242342333537978e-06,
+      "loss": 0.2514,
+      "step": 2749
+    },
+    {
+      "epoch": 1.4095335725269091,
+      "grad_norm": 0.5431578103710577,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.2513,
+      "step": 2750
+    },
+    {
+      "epoch": 1.4100461301896463,
+      "grad_norm": 0.4851552214102709,
+      "learning_rate": 4.228774102111714e-06,
+      "loss": 0.2315,
+      "step": 2751
+    },
+    {
+      "epoch": 1.4105586878523835,
+      "grad_norm": 0.5363838183939623,
+      "learning_rate": 4.221995951071623e-06,
+      "loss": 0.2437,
+      "step": 2752
+    },
+    {
+      "epoch": 1.4110712455151204,
+      "grad_norm": 0.49779761877376294,
+      "learning_rate": 4.215221782710559e-06,
+      "loss": 0.2488,
+      "step": 2753
+    },
+    {
+      "epoch": 1.4115838031778576,
+      "grad_norm": 0.5199995542111651,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.2429,
+      "step": 2754
+    },
+    {
+      "epoch": 1.4120963608405945,
+      "grad_norm": 0.557017126215738,
+      "learning_rate": 4.201685412700026e-06,
+      "loss": 0.2523,
+      "step": 2755
+    },
+    {
+      "epoch": 1.4126089185033317,
+      "grad_norm": 0.5041328496842157,
+      "learning_rate": 4.194923220380947e-06,
+      "loss": 0.2227,
+      "step": 2756
+    },
+    {
+      "epoch": 1.4131214761660686,
+      "grad_norm": 0.47220936755752324,
+      "learning_rate": 4.1881650294016606e-06,
+      "loss": 0.2288,
+      "step": 2757
+    },
+    {
+      "epoch": 1.4136340338288058,
+      "grad_norm": 0.44452232226048,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.2144,
+      "step": 2758
+    },
+    {
+      "epoch": 1.4141465914915428,
+      "grad_norm": 0.523560078305763,
+      "learning_rate": 4.1746606700929245e-06,
+      "loss": 0.2386,
+      "step": 2759
+    },
+    {
+      "epoch": 1.41465914915428,
+      "grad_norm": 0.5134750175136422,
+      "learning_rate": 4.167914511071805e-06,
+      "loss": 0.2566,
+      "step": 2760
+    },
+    {
+      "epoch": 1.4151717068170169,
+      "grad_norm": 0.5432335067996515,
+      "learning_rate": 4.161172372007115e-06,
+      "loss": 0.2628,
+      "step": 2761
+    },
+    {
+      "epoch": 1.415684264479754,
+      "grad_norm": 0.4993627802340633,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.2655,
+      "step": 2762
+    },
+    {
+      "epoch": 1.416196822142491,
+      "grad_norm": 0.4974063269274859,
+      "learning_rate": 4.147700172333222e-06,
+      "loss": 0.2292,
+      "step": 2763
+    },
+    {
+      "epoch": 1.4167093798052282,
+      "grad_norm": 0.4830330954548037,
+      "learning_rate": 4.140970121010177e-06,
+      "loss": 0.253,
+      "step": 2764
+    },
+    {
+      "epoch": 1.417221937467965,
+      "grad_norm": 0.5226941121101316,
+      "learning_rate": 4.134244108215867e-06,
+      "loss": 0.2375,
+      "step": 2765
+    },
+    {
+      "epoch": 1.4177344951307023,
+      "grad_norm": 0.5226990114021675,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2199,
+      "step": 2766
+    },
+    {
+      "epoch": 1.4182470527934392,
+      "grad_norm": 0.5261622317476922,
+      "learning_rate": 4.120804216755183e-06,
+      "loss": 0.2322,
+      "step": 2767
+    },
+    {
+      "epoch": 1.4187596104561764,
+      "grad_norm": 0.5665274272849649,
+      "learning_rate": 4.1140903473526935e-06,
+      "loss": 0.2756,
+      "step": 2768
+    },
+    {
+      "epoch": 1.4192721681189133,
+      "grad_norm": 0.5368558281541221,
+      "learning_rate": 4.107380535006709e-06,
+      "loss": 0.2349,
+      "step": 2769
+    },
+    {
+      "epoch": 1.4197847257816505,
+      "grad_norm": 0.5011993155736864,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.2574,
+      "step": 2770
+    },
+    {
+      "epoch": 1.4202972834443874,
+      "grad_norm": 0.5147783878375501,
+      "learning_rate": 4.093973099981292e-06,
+      "loss": 0.2672,
+      "step": 2771
+    },
+    {
+      "epoch": 1.4208098411071246,
+      "grad_norm": 0.475278275217271,
+      "learning_rate": 4.087275486543374e-06,
+      "loss": 0.2431,
+      "step": 2772
+    },
+    {
+      "epoch": 1.4213223987698616,
+      "grad_norm": 0.5691088640369176,
+      "learning_rate": 4.080581948644985e-06,
+      "loss": 0.2663,
+      "step": 2773
+    },
+    {
+      "epoch": 1.4218349564325987,
+      "grad_norm": 0.5269462050667394,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.2597,
+      "step": 2774
+    },
+    {
+      "epoch": 1.4223475140953357,
+      "grad_norm": 0.5327960391207287,
+      "learning_rate": 4.067207117918944e-06,
+      "loss": 0.271,
+      "step": 2775
+    },
+    {
+      "epoch": 1.4228600717580728,
+      "grad_norm": 0.49038932942348373,
+      "learning_rate": 4.0605258343103315e-06,
+      "loss": 0.2303,
+      "step": 2776
+    },
+    {
+      "epoch": 1.4233726294208098,
+      "grad_norm": 0.49812237677488147,
+      "learning_rate": 4.053848644679322e-06,
+      "loss": 0.2206,
+      "step": 2777
+    },
+    {
+      "epoch": 1.423885187083547,
+      "grad_norm": 0.5023512275623443,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.2401,
+      "step": 2778
+    },
+    {
+      "epoch": 1.424397744746284,
+      "grad_norm": 0.5508950527990639,
+      "learning_rate": 4.04050656575719e-06,
+      "loss": 0.2443,
+      "step": 2779
+    },
+    {
+      "epoch": 1.424910302409021,
+      "grad_norm": 0.5212584657925546,
+      "learning_rate": 4.033841685662529e-06,
+      "loss": 0.2404,
+      "step": 2780
+    },
+    {
+      "epoch": 1.425422860071758,
+      "grad_norm": 0.4781692693357959,
+      "learning_rate": 4.0271809179384035e-06,
+      "loss": 0.239,
+      "step": 2781
+    },
+    {
+      "epoch": 1.4259354177344952,
+      "grad_norm": 0.5371561546475093,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.2429,
+      "step": 2782
+    },
+    {
+      "epoch": 1.4264479753972321,
+      "grad_norm": 0.497176648754111,
+      "learning_rate": 4.013871737963506e-06,
+      "loss": 0.2216,
+      "step": 2783
+    },
+    {
+      "epoch": 1.4269605330599693,
+      "grad_norm": 0.5308025783149986,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.2585,
+      "step": 2784
+    },
+    {
+      "epoch": 1.4274730907227062,
+      "grad_norm": 0.5671513278948714,
+      "learning_rate": 4.00057906252766e-06,
+      "loss": 0.2631,
+      "step": 2785
+    },
+    {
+      "epoch": 1.4279856483854434,
+      "grad_norm": 0.49865883803009475,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2566,
+      "step": 2786
+    },
+    {
+      "epoch": 1.4284982060481803,
+      "grad_norm": 0.4983342832085773,
+      "learning_rate": 3.987302928280507e-06,
+      "loss": 0.2266,
+      "step": 2787
+    },
+    {
+      "epoch": 1.4290107637109175,
+      "grad_norm": 0.5230941494977537,
+      "learning_rate": 3.980671075543238e-06,
+      "loss": 0.2477,
+      "step": 2788
+    },
+    {
+      "epoch": 1.4295233213736545,
+      "grad_norm": 0.5356732742336032,
+      "learning_rate": 3.974043371826087e-06,
+      "loss": 0.2441,
+      "step": 2789
+    },
+    {
+      "epoch": 1.4300358790363916,
+      "grad_norm": 0.5395301259689879,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.2448,
+      "step": 2790
+    },
+    {
+      "epoch": 1.4305484366991288,
+      "grad_norm": 0.5828651074053851,
+      "learning_rate": 3.960800429722734e-06,
+      "loss": 0.2601,
+      "step": 2791
+    },
+    {
+      "epoch": 1.4310609943618657,
+      "grad_norm": 0.5151125818169313,
+      "learning_rate": 3.954185200464663e-06,
+      "loss": 0.2516,
+      "step": 2792
+    },
+    {
+      "epoch": 1.4315735520246027,
+      "grad_norm": 0.5807940115743717,
+      "learning_rate": 3.9475741384829715e-06,
+      "loss": 0.2723,
+      "step": 2793
+    },
+    {
+      "epoch": 1.4320861096873398,
+      "grad_norm": 0.5293191391072701,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.257,
+      "step": 2794
+    },
+    {
+      "epoch": 1.432598667350077,
+      "grad_norm": 0.5130401119898735,
+      "learning_rate": 3.93436453457341e-06,
+      "loss": 0.2572,
+      "step": 2795
+    },
+    {
+      "epoch": 1.433111225012814,
+      "grad_norm": 0.4858656219143425,
+      "learning_rate": 3.927766001750702e-06,
+      "loss": 0.2173,
+      "step": 2796
+    },
+    {
+      "epoch": 1.433623782675551,
+      "grad_norm": 0.5375963076771593,
+      "learning_rate": 3.921171654414667e-06,
+      "loss": 0.2484,
+      "step": 2797
+    },
+    {
+      "epoch": 1.434136340338288,
+      "grad_norm": 0.5521974213197726,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2663,
+      "step": 2798
+    },
+    {
+      "epoch": 1.4346488980010252,
+      "grad_norm": 0.49500086075264405,
+      "learning_rate": 3.907995534381237e-06,
+      "loss": 0.2299,
+      "step": 2799
+    },
+    {
+      "epoch": 1.4351614556637622,
+      "grad_norm": 0.5270601696383518,
+      "learning_rate": 3.901413770765905e-06,
+      "loss": 0.2647,
+      "step": 2800
+    },
+    {
+      "epoch": 1.4356740133264991,
+      "grad_norm": 0.5139646451375087,
+      "learning_rate": 3.894836210801397e-06,
+      "loss": 0.2441,
+      "step": 2801
+    },
+    {
+      "epoch": 1.4361865709892363,
+      "grad_norm": 0.5506515863074938,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2326,
+      "step": 2802
+    },
+    {
+      "epoch": 1.4366991286519735,
+      "grad_norm": 0.5048688804903922,
+      "learning_rate": 3.881693719957136e-06,
+      "loss": 0.2374,
+      "step": 2803
+    },
+    {
+      "epoch": 1.4372116863147104,
+      "grad_norm": 0.49026873278834754,
+      "learning_rate": 3.875128798136276e-06,
+      "loss": 0.2545,
+      "step": 2804
+    },
+    {
+      "epoch": 1.4377242439774474,
+      "grad_norm": 0.5090683598820657,
+      "learning_rate": 3.868568098084017e-06,
+      "loss": 0.2506,
+      "step": 2805
+    },
+    {
+      "epoch": 1.4382368016401845,
+      "grad_norm": 0.4978106461983983,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2496,
+      "step": 2806
+    },
+    {
+      "epoch": 1.4387493593029217,
+      "grad_norm": 0.5957346936888235,
+      "learning_rate": 3.8554593813710975e-06,
+      "loss": 0.2598,
+      "step": 2807
+    },
+    {
+      "epoch": 1.4392619169656586,
+      "grad_norm": 0.4711757877555722,
+      "learning_rate": 3.848911373746051e-06,
+      "loss": 0.2259,
+      "step": 2808
+    },
+    {
+      "epoch": 1.4397744746283956,
+      "grad_norm": 0.5180207923004838,
+      "learning_rate": 3.842367605960821e-06,
+      "loss": 0.2635,
+      "step": 2809
+    },
+    {
+      "epoch": 1.4402870322911328,
+      "grad_norm": 0.5088152075558492,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.272,
+      "step": 2810
+    },
+    {
+      "epoch": 1.44079958995387,
+      "grad_norm": 0.49980500175530707,
+      "learning_rate": 3.829292807948937e-06,
+      "loss": 0.2424,
+      "step": 2811
+    },
+    {
+      "epoch": 1.4413121476166069,
+      "grad_norm": 0.494729713088461,
+      "learning_rate": 3.822761786734506e-06,
+      "loss": 0.2255,
+      "step": 2812
+    },
+    {
+      "epoch": 1.4418247052793438,
+      "grad_norm": 0.5004081833893824,
+      "learning_rate": 3.81623502338436e-06,
+      "loss": 0.2511,
+      "step": 2813
+    },
+    {
+      "epoch": 1.442337262942081,
+      "grad_norm": 0.5601369037473904,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.2508,
+      "step": 2814
+    },
+    {
+      "epoch": 1.4428498206048181,
+      "grad_norm": 0.5292394059396741,
+      "learning_rate": 3.803194288269122e-06,
+      "loss": 0.2397,
+      "step": 2815
+    },
+    {
+      "epoch": 1.443362378267555,
+      "grad_norm": 0.5464596783675346,
+      "learning_rate": 3.7966803254927852e-06,
+      "loss": 0.2551,
+      "step": 2816
+    },
+    {
+      "epoch": 1.443874935930292,
+      "grad_norm": 0.513007135903009,
+      "learning_rate": 3.7901706385582372e-06,
+      "loss": 0.2282,
+      "step": 2817
+    },
+    {
+      "epoch": 1.4443874935930292,
+      "grad_norm": 0.5432726293686879,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2479,
+      "step": 2818
+    },
+    {
+      "epoch": 1.4449000512557664,
+      "grad_norm": 0.5281463275344115,
+      "learning_rate": 3.7771641101596e-06,
+      "loss": 0.2535,
+      "step": 2819
+    },
+    {
+      "epoch": 1.4454126089185033,
+      "grad_norm": 0.534958435234811,
+      "learning_rate": 3.770667277660696e-06,
+      "loss": 0.24,
+      "step": 2820
+    },
+    {
+      "epoch": 1.4459251665812403,
+      "grad_norm": 0.5340017110423951,
+      "learning_rate": 3.7641747389339232e-06,
+      "loss": 0.2811,
+      "step": 2821
+    },
+    {
+      "epoch": 1.4464377242439774,
+      "grad_norm": 0.54044224850304,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.2444,
+      "step": 2822
+    },
+    {
+      "epoch": 1.4469502819067146,
+      "grad_norm": 0.5064508338143706,
+      "learning_rate": 3.7512025606946e-06,
+      "loss": 0.26,
+      "step": 2823
+    },
+    {
+      "epoch": 1.4474628395694515,
+      "grad_norm": 0.5729264678094134,
+      "learning_rate": 3.7447229301235443e-06,
+      "loss": 0.2574,
+      "step": 2824
+    },
+    {
+      "epoch": 1.4479753972321885,
+      "grad_norm": 0.5312220424539259,
+      "learning_rate": 3.7382476112076227e-06,
+      "loss": 0.2819,
+      "step": 2825
+    },
+    {
+      "epoch": 1.4484879548949257,
+      "grad_norm": 0.5339343851548345,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.2717,
+      "step": 2826
+    },
+    {
+      "epoch": 1.4490005125576628,
+      "grad_norm": 0.4897458860061051,
+      "learning_rate": 3.7253099261914794e-06,
+      "loss": 0.2278,
+      "step": 2827
+    },
+    {
+      "epoch": 1.4495130702203998,
+      "grad_norm": 0.4986974005089407,
+      "learning_rate": 3.718847569008992e-06,
+      "loss": 0.2374,
+      "step": 2828
+    },
+    {
+      "epoch": 1.450025627883137,
+      "grad_norm": 0.5794075968592284,
+      "learning_rate": 3.7123895413170706e-06,
+      "loss": 0.2665,
+      "step": 2829
+    },
+    {
+      "epoch": 1.4505381855458739,
+      "grad_norm": 0.49363222104316007,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.2534,
+      "step": 2830
+    },
+    {
+      "epoch": 1.451050743208611,
+      "grad_norm": 0.5188061835566682,
+      "learning_rate": 3.6994864922075756e-06,
+      "loss": 0.2401,
+      "step": 2831
+    },
+    {
+      "epoch": 1.451563300871348,
+      "grad_norm": 0.5177462172747275,
+      "learning_rate": 3.6930414796838587e-06,
+      "loss": 0.2484,
+      "step": 2832
+    },
+    {
+      "epoch": 1.4520758585340852,
+      "grad_norm": 0.4922866356572482,
+      "learning_rate": 3.686600814438399e-06,
+      "loss": 0.2201,
+      "step": 2833
+    },
+    {
+      "epoch": 1.452588416196822,
+      "grad_norm": 0.48567826188068813,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.2295,
+      "step": 2834
+    },
+    {
+      "epoch": 1.4531009738595593,
+      "grad_norm": 0.5160387872644587,
+      "learning_rate": 3.6737325435370376e-06,
+      "loss": 0.24,
+      "step": 2835
+    },
+    {
+      "epoch": 1.4536135315222962,
+      "grad_norm": 0.4716896258760731,
+      "learning_rate": 3.667304946751011e-06,
+      "loss": 0.2236,
+      "step": 2836
+    },
+    {
+      "epoch": 1.4541260891850334,
+      "grad_norm": 0.49557801115295586,
+      "learning_rate": 3.660881714982999e-06,
+      "loss": 0.2107,
+      "step": 2837
+    },
+    {
+      "epoch": 1.4546386468477703,
+      "grad_norm": 0.53196773175651,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.2738,
+      "step": 2838
+    },
+    {
+      "epoch": 1.4551512045105075,
+      "grad_norm": 0.47092453176612287,
+      "learning_rate": 3.648048364207698e-06,
+      "loss": 0.2144,
+      "step": 2839
+    },
+    {
+      "epoch": 1.4556637621732444,
+      "grad_norm": 0.4626062244787271,
+      "learning_rate": 3.641638254046218e-06,
+      "loss": 0.2335,
+      "step": 2840
+    },
+    {
+      "epoch": 1.4561763198359816,
+      "grad_norm": 0.5357922216723315,
+      "learning_rate": 3.6352325265943634e-06,
+      "loss": 0.2464,
+      "step": 2841
+    },
+    {
+      "epoch": 1.4566888774987186,
+      "grad_norm": 0.47937209047739454,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.2105,
+      "step": 2842
+    },
+    {
+      "epoch": 1.4572014351614557,
+      "grad_norm": 0.5398446209566874,
+      "learning_rate": 3.6224342374779365e-06,
+      "loss": 0.2638,
+      "step": 2843
+    },
+    {
+      "epoch": 1.4577139928241927,
+      "grad_norm": 0.5003469827738948,
+      "learning_rate": 3.6160416846350056e-06,
+      "loss": 0.2273,
+      "step": 2844
+    },
+    {
+      "epoch": 1.4582265504869298,
+      "grad_norm": 0.4971785576566374,
+      "learning_rate": 3.609653532144972e-06,
+      "loss": 0.2439,
+      "step": 2845
+    },
+    {
+      "epoch": 1.4587391081496668,
+      "grad_norm": 0.4887149425345939,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.2305,
+      "step": 2846
+    },
+    {
+      "epoch": 1.459251665812404,
+      "grad_norm": 0.5182355350564947,
+      "learning_rate": 3.5968904458335573e-06,
+      "loss": 0.2398,
+      "step": 2847
+    },
+    {
+      "epoch": 1.459764223475141,
+      "grad_norm": 0.5145555234043517,
+      "learning_rate": 3.590515520809542e-06,
+      "loss": 0.2409,
+      "step": 2848
+    },
+    {
+      "epoch": 1.460276781137878,
+      "grad_norm": 0.4711681778431692,
+      "learning_rate": 3.584145013733178e-06,
+      "loss": 0.2348,
+      "step": 2849
+    },
+    {
+      "epoch": 1.460789338800615,
+      "grad_norm": 0.5251343450772937,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.2598,
+      "step": 2850
+    },
+    {
+      "epoch": 1.4613018964633522,
+      "grad_norm": 0.526570294525289,
+      "learning_rate": 3.5714172709846686e-06,
+      "loss": 0.2294,
+      "step": 2851
+    },
+    {
+      "epoch": 1.4618144541260891,
+      "grad_norm": 0.5298732299623796,
+      "learning_rate": 3.5650600440855375e-06,
+      "loss": 0.2253,
+      "step": 2852
+    },
+    {
+      "epoch": 1.4623270117888263,
+      "grad_norm": 0.592468751445862,
+      "learning_rate": 3.558707252680078e-06,
+      "loss": 0.2874,
+      "step": 2853
+    },
+    {
+      "epoch": 1.4628395694515632,
+      "grad_norm": 0.5462812328968311,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2814,
+      "step": 2854
+    },
+    {
+      "epoch": 1.4633521271143004,
+      "grad_norm": 0.566363001649573,
+      "learning_rate": 3.546014993862591e-06,
+      "loss": 0.2668,
+      "step": 2855
+    },
+    {
+      "epoch": 1.4638646847770374,
+      "grad_norm": 0.49616119794546226,
+      "learning_rate": 3.5396755351991197e-06,
+      "loss": 0.2505,
+      "step": 2856
+    },
+    {
+      "epoch": 1.4643772424397745,
+      "grad_norm": 0.5516540394568447,
+      "learning_rate": 3.5333405295264255e-06,
+      "loss": 0.2521,
+      "step": 2857
+    },
+    {
+      "epoch": 1.4648898001025115,
+      "grad_norm": 0.5239711021556687,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.2629,
+      "step": 2858
+    },
+    {
+      "epoch": 1.4654023577652486,
+      "grad_norm": 0.5170927730070536,
+      "learning_rate": 3.520683894616739e-06,
+      "loss": 0.2451,
+      "step": 2859
+    },
+    {
+      "epoch": 1.4659149154279856,
+      "grad_norm": 0.5178063155570998,
+      "learning_rate": 3.5143622741037432e-06,
+      "loss": 0.255,
+      "step": 2860
+    },
+    {
+      "epoch": 1.4664274730907227,
+      "grad_norm": 0.5177331091518852,
+      "learning_rate": 3.5080451240295288e-06,
+      "loss": 0.2336,
+      "step": 2861
+    },
+    {
+      "epoch": 1.4669400307534597,
+      "grad_norm": 0.5194904460202592,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.2341,
+      "step": 2862
+    },
+    {
+      "epoch": 1.4674525884161969,
+      "grad_norm": 0.4831386833726277,
+      "learning_rate": 3.495424252611538e-06,
+      "loss": 0.2319,
+      "step": 2863
+    },
+    {
+      "epoch": 1.4679651460789338,
+      "grad_norm": 0.4752449676503757,
+      "learning_rate": 3.4891205399671226e-06,
+      "loss": 0.2296,
+      "step": 2864
+    },
+    {
+      "epoch": 1.468477703741671,
+      "grad_norm": 0.546241980021249,
+      "learning_rate": 3.4828213151601695e-06,
+      "loss": 0.2542,
+      "step": 2865
+    },
+    {
+      "epoch": 1.468990261404408,
+      "grad_norm": 0.5285657211980681,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.2408,
+      "step": 2866
+    },
+    {
+      "epoch": 1.469502819067145,
+      "grad_norm": 0.5673692609859431,
+      "learning_rate": 3.4702363464233604e-06,
+      "loss": 0.2476,
+      "step": 2867
+    },
+    {
+      "epoch": 1.470015376729882,
+      "grad_norm": 0.5360972429640449,
+      "learning_rate": 3.463950611168111e-06,
+      "loss": 0.2426,
+      "step": 2868
+    },
+    {
+      "epoch": 1.4705279343926192,
+      "grad_norm": 0.5389401847814864,
+      "learning_rate": 3.457669381099523e-06,
+      "loss": 0.2311,
+      "step": 2869
+    },
+    {
+      "epoch": 1.4710404920553564,
+      "grad_norm": 0.4868873686351302,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2333,
+      "step": 2870
+    },
+    {
+      "epoch": 1.4715530497180933,
+      "grad_norm": 0.5303463848474584,
+      "learning_rate": 3.445120453837424e-06,
+      "loss": 0.2487,
+      "step": 2871
+    },
+    {
+      "epoch": 1.4720656073808303,
+      "grad_norm": 0.5184604916364336,
+      "learning_rate": 3.438852765293671e-06,
+      "loss": 0.263,
+      "step": 2872
+    },
+    {
+      "epoch": 1.4725781650435674,
+      "grad_norm": 0.5323429845762316,
+      "learning_rate": 3.432589599236099e-06,
+      "loss": 0.237,
+      "step": 2873
+    },
+    {
+      "epoch": 1.4730907227063046,
+      "grad_norm": 0.5497807275807922,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.2582,
+      "step": 2874
+    },
+    {
+      "epoch": 1.4736032803690415,
+      "grad_norm": 0.5248687681303138,
+      "learning_rate": 3.4200768518447524e-06,
+      "loss": 0.2367,
+      "step": 2875
+    },
+    {
+      "epoch": 1.4741158380317785,
+      "grad_norm": 0.5384373018300335,
+      "learning_rate": 3.4138272791357995e-06,
+      "loss": 0.2436,
+      "step": 2876
+    },
+    {
+      "epoch": 1.4746283956945156,
+      "grad_norm": 0.5059168436497485,
+      "learning_rate": 3.4075822461626663e-06,
+      "loss": 0.2131,
+      "step": 2877
+    },
+    {
+      "epoch": 1.4751409533572528,
+      "grad_norm": 0.5155013231422648,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.2257,
+      "step": 2878
+    },
+    {
+      "epoch": 1.4756535110199898,
+      "grad_norm": 0.45494407186489655,
+      "learning_rate": 3.395105816639106e-06,
+      "loss": 0.2002,
+      "step": 2879
+    },
+    {
+      "epoch": 1.4761660686827267,
+      "grad_norm": 0.5489681314248084,
+      "learning_rate": 3.3888744286884657e-06,
+      "loss": 0.2715,
+      "step": 2880
+    },
+    {
+      "epoch": 1.4766786263454639,
+      "grad_norm": 0.5288047931737466,
+      "learning_rate": 3.3826475976732144e-06,
+      "loss": 0.2367,
+      "step": 2881
+    },
+    {
+      "epoch": 1.477191184008201,
+      "grad_norm": 0.5339256240864259,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.2578,
+      "step": 2882
+    },
+    {
+      "epoch": 1.477703741670938,
+      "grad_norm": 0.5388234585554258,
+      "learning_rate": 3.3702076236139402e-06,
+      "loss": 0.2248,
+      "step": 2883
+    },
+    {
+      "epoch": 1.478216299333675,
+      "grad_norm": 0.5266646464833434,
+      "learning_rate": 3.363994489144575e-06,
+      "loss": 0.2332,
+      "step": 2884
+    },
+    {
+      "epoch": 1.478728856996412,
+      "grad_norm": 0.5181863956352016,
+      "learning_rate": 3.357785928759917e-06,
+      "loss": 0.2291,
+      "step": 2885
+    },
+    {
+      "epoch": 1.4792414146591493,
+      "grad_norm": 0.5399797708389331,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.2499,
+      "step": 2886
+    },
+    {
+      "epoch": 1.4797539723218862,
+      "grad_norm": 0.7910171039268231,
+      "learning_rate": 3.3453825473593805e-06,
+      "loss": 0.2473,
+      "step": 2887
+    },
+    {
+      "epoch": 1.4802665299846232,
+      "grad_norm": 0.5195591549248073,
+      "learning_rate": 3.339187734892941e-06,
+      "loss": 0.2506,
+      "step": 2888
+    },
+    {
+      "epoch": 1.4807790876473603,
+      "grad_norm": 0.497654938956051,
+      "learning_rate": 3.3329975136100787e-06,
+      "loss": 0.2083,
+      "step": 2889
+    },
+    {
+      "epoch": 1.4812916453100975,
+      "grad_norm": 0.5293004705270719,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.2434,
+      "step": 2890
+    },
+    {
+      "epoch": 1.4818042029728344,
+      "grad_norm": 0.5288888243723955,
+      "learning_rate": 3.320630861659171e-06,
+      "loss": 0.2207,
+      "step": 2891
+    },
+    {
+      "epoch": 1.4823167606355714,
+      "grad_norm": 0.4880516507845217,
+      "learning_rate": 3.3144544395152457e-06,
+      "loss": 0.2637,
+      "step": 2892
+    },
+    {
+      "epoch": 1.4828293182983086,
+      "grad_norm": 0.5055941589811903,
+      "learning_rate": 3.308282625603134e-06,
+      "loss": 0.2561,
+      "step": 2893
+    },
+    {
+      "epoch": 1.4833418759610457,
+      "grad_norm": 0.5345438450217431,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2649,
+      "step": 2894
+    },
+    {
+      "epoch": 1.4838544336237827,
+      "grad_norm": 0.5144166951223942,
+      "learning_rate": 3.2959528394876703e-06,
+      "loss": 0.2431,
+      "step": 2895
+    },
+    {
+      "epoch": 1.4843669912865196,
+      "grad_norm": 0.49814383081479346,
+      "learning_rate": 3.2897948757830256e-06,
+      "loss": 0.2306,
+      "step": 2896
+    },
+    {
+      "epoch": 1.4848795489492568,
+      "grad_norm": 0.46446755175962295,
+      "learning_rate": 3.283641537307615e-06,
+      "loss": 0.222,
+      "step": 2897
+    },
+    {
+      "epoch": 1.485392106611994,
+      "grad_norm": 0.4551619707399669,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.2073,
+      "step": 2898
+    },
+    {
+      "epoch": 1.485904664274731,
+      "grad_norm": 0.5195944159769776,
+      "learning_rate": 3.2713487530068356e-06,
+      "loss": 0.2204,
+      "step": 2899
+    },
+    {
+      "epoch": 1.4864172219374678,
+      "grad_norm": 0.5045694430685503,
+      "learning_rate": 3.2652093156546795e-06,
+      "loss": 0.2667,
+      "step": 2900
+    },
+    {
+      "epoch": 1.486929779600205,
+      "grad_norm": 0.5078109216499845,
+      "learning_rate": 3.2590745204781537e-06,
+      "loss": 0.2489,
+      "step": 2901
+    },
+    {
+      "epoch": 1.4874423372629422,
+      "grad_norm": 0.5513249125681956,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.258,
+      "step": 2902
+    },
+    {
+      "epoch": 1.4879548949256791,
+      "grad_norm": 0.5106317135058229,
+      "learning_rate": 3.246818873563232e-06,
+      "loss": 0.2425,
+      "step": 2903
+    },
+    {
+      "epoch": 1.488467452588416,
+      "grad_norm": 0.5231930231454878,
+      "learning_rate": 3.2406980302724422e-06,
+      "loss": 0.2604,
+      "step": 2904
+    },
+    {
+      "epoch": 1.4889800102511532,
+      "grad_norm": 0.49027343063084594,
+      "learning_rate": 3.2345818460524927e-06,
+      "loss": 0.2271,
+      "step": 2905
+    },
+    {
+      "epoch": 1.4894925679138904,
+      "grad_norm": 0.535356631591648,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.2505,
+      "step": 2906
+    },
+    {
+      "epoch": 1.4900051255766273,
+      "grad_norm": 1.610279716286711,
+      "learning_rate": 3.2223634716850227e-06,
+      "loss": 0.212,
+      "step": 2907
+    },
+    {
+      "epoch": 1.4905176832393645,
+      "grad_norm": 0.5195632922013577,
+      "learning_rate": 3.216261289959417e-06,
+      "loss": 0.2186,
+      "step": 2908
+    },
+    {
+      "epoch": 1.4910302409021015,
+      "grad_norm": 0.5002917597369569,
+      "learning_rate": 3.210163784148478e-06,
+      "loss": 0.258,
+      "step": 2909
+    },
+    {
+      "epoch": 1.4915427985648386,
+      "grad_norm": 0.5253859041471729,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2578,
+      "step": 2910
+    },
+    {
+      "epoch": 1.4920553562275756,
+      "grad_norm": 0.5471618987996401,
+      "learning_rate": 3.197982817078994e-06,
+      "loss": 0.2464,
+      "step": 2911
+    },
+    {
+      "epoch": 1.4925679138903127,
+      "grad_norm": 0.530283748618024,
+      "learning_rate": 3.191899364216581e-06,
+      "loss": 0.3001,
+      "step": 2912
+    },
+    {
+      "epoch": 1.4930804715530497,
+      "grad_norm": 0.4981550956957725,
+      "learning_rate": 3.1858206040610883e-06,
+      "loss": 0.2447,
+      "step": 2913
+    },
+    {
+      "epoch": 1.4935930292157868,
+      "grad_norm": 0.5456089190802053,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.2632,
+      "step": 2914
+    },
+    {
+      "epoch": 1.4941055868785238,
+      "grad_norm": 0.5203608791233832,
+      "learning_rate": 3.1736771786275834e-06,
+      "loss": 0.2608,
+      "step": 2915
+    },
+    {
+      "epoch": 1.494618144541261,
+      "grad_norm": 0.5862185381131446,
+      "learning_rate": 3.1676125217198175e-06,
+      "loss": 0.2653,
+      "step": 2916
+    },
+    {
+      "epoch": 1.495130702203998,
+      "grad_norm": 0.49239907463963095,
+      "learning_rate": 3.16155257425948e-06,
+      "loss": 0.2424,
+      "step": 2917
+    },
+    {
+      "epoch": 1.495643259866735,
+      "grad_norm": 0.475777566570927,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2388,
+      "step": 2918
+    },
+    {
+      "epoch": 1.496155817529472,
+      "grad_norm": 0.5158120340461063,
+      "learning_rate": 3.149446824385909e-06,
+      "loss": 0.2524,
+      "step": 2919
+    },
+    {
+      "epoch": 1.4966683751922092,
+      "grad_norm": 0.45749873948233283,
+      "learning_rate": 3.1434010303169593e-06,
+      "loss": 0.2199,
+      "step": 2920
+    },
+    {
+      "epoch": 1.4971809328549461,
+      "grad_norm": 0.4980825133605471,
+      "learning_rate": 3.1373599623840024e-06,
+      "loss": 0.2284,
+      "step": 2921
+    },
+    {
+      "epoch": 1.4976934905176833,
+      "grad_norm": 0.5413983056350928,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.252,
+      "step": 2922
+    },
+    {
+      "epoch": 1.4982060481804202,
+      "grad_norm": 0.5259194015414607,
+      "learning_rate": 3.125292021578822e-06,
+      "loss": 0.2514,
+      "step": 2923
+    },
+    {
+      "epoch": 1.4987186058431574,
+      "grad_norm": 0.4910978974320567,
+      "learning_rate": 3.1192651570248232e-06,
+      "loss": 0.2345,
+      "step": 2924
+    },
+    {
+      "epoch": 1.4992311635058944,
+      "grad_norm": 0.5775068971674365,
+      "learning_rate": 3.11324303524326e-06,
+      "loss": 0.272,
+      "step": 2925
+    },
+    {
+      "epoch": 1.4997437211686315,
+      "grad_norm": 0.5350406978853839,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.2261,
+      "step": 2926
+    },
+    {
+      "epoch": 1.5002562788313685,
+      "grad_norm": 0.5211453861329499,
+      "learning_rate": 3.1012130365979476e-06,
+      "loss": 0.2267,
+      "step": 2927
+    },
+    {
+      "epoch": 1.5007688364941056,
+      "grad_norm": 0.5541745154104105,
+      "learning_rate": 3.0952051680262675e-06,
+      "loss": 0.2669,
+      "step": 2928
+    },
+    {
+      "epoch": 1.5012813941568428,
+      "grad_norm": 0.5127896142093221,
+      "learning_rate": 3.089202058811166e-06,
+      "loss": 0.2262,
+      "step": 2929
+    },
+    {
+      "epoch": 1.5017939518195798,
+      "grad_norm": 0.47054522412323246,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.1884,
+      "step": 2930
+    },
+    {
+      "epoch": 1.5023065094823167,
+      "grad_norm": 0.5332083426847334,
+      "learning_rate": 3.0772101349987506e-06,
+      "loss": 0.2522,
+      "step": 2931
+    },
+    {
+      "epoch": 1.5028190671450539,
+      "grad_norm": 0.5156325087532021,
+      "learning_rate": 3.0712213286672686e-06,
+      "loss": 0.2465,
+      "step": 2932
+    },
+    {
+      "epoch": 1.503331624807791,
+      "grad_norm": 0.5654505509985084,
+      "learning_rate": 3.065237298224013e-06,
+      "loss": 0.2719,
+      "step": 2933
+    },
+    {
+      "epoch": 1.503844182470528,
+      "grad_norm": 0.5415059991051329,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.244,
+      "step": 2934
+    },
+    {
+      "epoch": 1.504356740133265,
+      "grad_norm": 0.4923511264324253,
+      "learning_rate": 3.0532835814976193e-06,
+      "loss": 0.2144,
+      "step": 2935
+    },
+    {
+      "epoch": 1.504869297796002,
+      "grad_norm": 0.5079539377143977,
+      "learning_rate": 3.047313903453979e-06,
+      "loss": 0.2325,
+      "step": 2936
+    },
+    {
+      "epoch": 1.5053818554587393,
+      "grad_norm": 0.4936750739066036,
+      "learning_rate": 3.0413490177775397e-06,
+      "loss": 0.2257,
+      "step": 2937
+    },
+    {
+      "epoch": 1.5058944131214762,
+      "grad_norm": 0.519667715773605,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.2319,
+      "step": 2938
+    },
+    {
+      "epoch": 1.5064069707842132,
+      "grad_norm": 0.48782339791617846,
+      "learning_rate": 3.029433639968925e-06,
+      "loss": 0.2276,
+      "step": 2939
+    },
+    {
+      "epoch": 1.5069195284469503,
+      "grad_norm": 0.5268431585342307,
+      "learning_rate": 3.023483156049817e-06,
+      "loss": 0.2406,
+      "step": 2940
+    },
+    {
+      "epoch": 1.5074320861096875,
+      "grad_norm": 0.5457843738236537,
+      "learning_rate": 3.017537480924032e-06,
+      "loss": 0.2664,
+      "step": 2941
+    },
+    {
+      "epoch": 1.5079446437724244,
+      "grad_norm": 0.5686980044016422,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.2718,
+      "step": 2942
+    },
+    {
+      "epoch": 1.5084572014351614,
+      "grad_norm": 0.509006264331281,
+      "learning_rate": 3.0056605734421253e-06,
+      "loss": 0.245,
+      "step": 2943
+    },
+    {
+      "epoch": 1.5089697590978985,
+      "grad_norm": 0.5048654246005035,
+      "learning_rate": 2.9997293492725577e-06,
+      "loss": 0.2549,
+      "step": 2944
+    },
+    {
+      "epoch": 1.5094823167606357,
+      "grad_norm": 0.4995200352212472,
+      "learning_rate": 2.993802950269402e-06,
+      "loss": 0.2286,
+      "step": 2945
+    },
+    {
+      "epoch": 1.5099948744233727,
+      "grad_norm": 0.4953620054798623,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2414,
+      "step": 2946
+    },
+    {
+      "epoch": 1.5105074320861096,
+      "grad_norm": 0.5478612649236791,
+      "learning_rate": 2.9819646440988713e-06,
+      "loss": 0.2712,
+      "step": 2947
+    },
+    {
+      "epoch": 1.5110199897488468,
+      "grad_norm": 0.48414015435846186,
+      "learning_rate": 2.976052745091438e-06,
+      "loss": 0.2421,
+      "step": 2948
+    },
+    {
+      "epoch": 1.511532547411584,
+      "grad_norm": 0.5021518356190368,
+      "learning_rate": 2.9701456875702937e-06,
+      "loss": 0.2442,
+      "step": 2949
+    },
+    {
+      "epoch": 1.5120451050743209,
+      "grad_norm": 0.521063929615499,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2442,
+      "step": 2950
+    },
+    {
+      "epoch": 1.5125576627370578,
+      "grad_norm": 0.558551152355406,
+      "learning_rate": 2.9583461132700875e-06,
+      "loss": 0.265,
+      "step": 2951
+    },
+    {
+      "epoch": 1.513070220399795,
+      "grad_norm": 0.478838582524175,
+      "learning_rate": 2.9524536046242603e-06,
+      "loss": 0.2303,
+      "step": 2952
+    },
+    {
+      "epoch": 1.5135827780625322,
+      "grad_norm": 0.5001794884084084,
+      "learning_rate": 2.946565953731211e-06,
+      "loss": 0.222,
+      "step": 2953
+    },
+    {
+      "epoch": 1.514095335725269,
+      "grad_norm": 0.472435973023686,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.2409,
+      "step": 2954
+    },
+    {
+      "epoch": 1.514607893388006,
+      "grad_norm": 0.5695788605146566,
+      "learning_rate": 2.934805241433115e-06,
+      "loss": 0.2528,
+      "step": 2955
+    },
+    {
+      "epoch": 1.5151204510507432,
+      "grad_norm": 0.47296896365564817,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.2147,
+      "step": 2956
+    },
+    {
+      "epoch": 1.5156330087134804,
+      "grad_norm": 0.5445980326270042,
+      "learning_rate": 2.9230640088016204e-06,
+      "loss": 0.2689,
+      "step": 2957
+    },
+    {
+      "epoch": 1.5161455663762173,
+      "grad_norm": 0.49732538465672355,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2419,
+      "step": 2958
+    },
+    {
+      "epoch": 1.5166581240389543,
+      "grad_norm": 0.4220869612104664,
+      "learning_rate": 2.9113422882088403e-06,
+      "loss": 0.1987,
+      "step": 2959
+    },
+    {
+      "epoch": 1.5171706817016914,
+      "grad_norm": 0.5754011640422099,
+      "learning_rate": 2.9054887550285483e-06,
+      "loss": 0.235,
+      "step": 2960
+    },
+    {
+      "epoch": 1.5176832393644286,
+      "grad_norm": 0.5218718923683728,
+      "learning_rate": 2.8996401119730923e-06,
+      "loss": 0.2587,
+      "step": 2961
+    },
+    {
+      "epoch": 1.5181957970271656,
+      "grad_norm": 0.5577339142640889,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.2767,
+      "step": 2962
+    },
+    {
+      "epoch": 1.5187083546899025,
+      "grad_norm": 0.5038881317118271,
+      "learning_rate": 2.887957512358813e-06,
+      "loss": 0.2423,
+      "step": 2963
+    },
+    {
+      "epoch": 1.5192209123526397,
+      "grad_norm": 0.5364602002539419,
+      "learning_rate": 2.882123563852598e-06,
+      "loss": 0.2727,
+      "step": 2964
+    },
+    {
+      "epoch": 1.5197334700153768,
+      "grad_norm": 0.4738081089427845,
+      "learning_rate": 2.87629452157645e-06,
+      "loss": 0.1917,
+      "step": 2965
+    },
+    {
+      "epoch": 1.5202460276781138,
+      "grad_norm": 0.5043929659868738,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.2185,
+      "step": 2966
+    },
+    {
+      "epoch": 1.5207585853408507,
+      "grad_norm": 0.5159517607065054,
+      "learning_rate": 2.864651171782402e-06,
+      "loss": 0.24,
+      "step": 2967
+    },
+    {
+      "epoch": 1.521271143003588,
+      "grad_norm": 0.5255200364432794,
+      "learning_rate": 2.858836872290064e-06,
+      "loss": 0.2555,
+      "step": 2968
+    },
+    {
+      "epoch": 1.521783700666325,
+      "grad_norm": 0.532727074217409,
+      "learning_rate": 2.8530274950789095e-06,
+      "loss": 0.2745,
+      "step": 2969
+    },
+    {
+      "epoch": 1.522296258329062,
+      "grad_norm": 0.5444718827171022,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.2307,
+      "step": 2970
+    },
+    {
+      "epoch": 1.522808815991799,
+      "grad_norm": 0.5677704917935464,
+      "learning_rate": 2.841423523513961e-06,
+      "loss": 0.2647,
+      "step": 2971
+    },
+    {
+      "epoch": 1.5233213736545361,
+      "grad_norm": 0.5883458071167541,
+      "learning_rate": 2.835628937158591e-06,
+      "loss": 0.2546,
+      "step": 2972
+    },
+    {
+      "epoch": 1.5238339313172733,
+      "grad_norm": 0.511919533929984,
+      "learning_rate": 2.8298392890812344e-06,
+      "loss": 0.2517,
+      "step": 2973
+    },
+    {
+      "epoch": 1.5243464889800102,
+      "grad_norm": 0.5398581017767609,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.2273,
+      "step": 2974
+    },
+    {
+      "epoch": 1.5248590466427472,
+      "grad_norm": 0.5510022382778469,
+      "learning_rate": 2.8182748237199755e-06,
+      "loss": 0.2661,
+      "step": 2975
+    },
+    {
+      "epoch": 1.5253716043054844,
+      "grad_norm": 0.5329025822483537,
+      "learning_rate": 2.8125000144072602e-06,
+      "loss": 0.2579,
+      "step": 2976
+    },
+    {
+      "epoch": 1.5258841619682215,
+      "grad_norm": 0.5568208319726613,
+      "learning_rate": 2.806730159314929e-06,
+      "loss": 0.2339,
+      "step": 2977
+    },
+    {
+      "epoch": 1.5263967196309585,
+      "grad_norm": 0.5281004400012169,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.2522,
+      "step": 2978
+    },
+    {
+      "epoch": 1.5269092772936954,
+      "grad_norm": 0.49474205689193,
+      "learning_rate": 2.7952053276962444e-06,
+      "loss": 0.2328,
+      "step": 2979
+    },
+    {
+      "epoch": 1.5274218349564326,
+      "grad_norm": 0.4894745043653836,
+      "learning_rate": 2.7894503591137656e-06,
+      "loss": 0.2526,
+      "step": 2980
+    },
+    {
+      "epoch": 1.5279343926191697,
+      "grad_norm": 0.5402455940139439,
+      "learning_rate": 2.7837003606393987e-06,
+      "loss": 0.2785,
+      "step": 2981
+    },
+    {
+      "epoch": 1.5284469502819067,
+      "grad_norm": 0.4813196235211439,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.2542,
+      "step": 2982
+    },
+    {
+      "epoch": 1.5289595079446436,
+      "grad_norm": 0.539472769603119,
+      "learning_rate": 2.7722152898650788e-06,
+      "loss": 0.2645,
+      "step": 2983
+    },
+    {
+      "epoch": 1.5294720656073808,
+      "grad_norm": 0.49756295463525524,
+      "learning_rate": 2.766480225481595e-06,
+      "loss": 0.2239,
+      "step": 2984
+    },
+    {
+      "epoch": 1.529984623270118,
+      "grad_norm": 0.5090112859645092,
+      "learning_rate": 2.7607501470391384e-06,
+      "loss": 0.2435,
+      "step": 2985
+    },
+    {
+      "epoch": 1.530497180932855,
+      "grad_norm": 0.49564787691287676,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.2301,
+      "step": 2986
+    },
+    {
+      "epoch": 1.5310097385955919,
+      "grad_norm": 0.5446251726924476,
+      "learning_rate": 2.7493049637724724e-06,
+      "loss": 0.2778,
+      "step": 2987
+    },
+    {
+      "epoch": 1.531522296258329,
+      "grad_norm": 0.4824881769626571,
+      "learning_rate": 2.7435898668372296e-06,
+      "loss": 0.2351,
+      "step": 2988
+    },
+    {
+      "epoch": 1.5320348539210662,
+      "grad_norm": 0.5479960137743843,
+      "learning_rate": 2.7378797716209506e-06,
+      "loss": 0.2484,
+      "step": 2989
+    },
+    {
+      "epoch": 1.5325474115838031,
+      "grad_norm": 0.5461798950567729,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.2487,
+      "step": 2990
+    },
+    {
+      "epoch": 1.53305996924654,
+      "grad_norm": 0.48919435246382253,
+      "learning_rate": 2.7264746020853217e-06,
+      "loss": 0.2273,
+      "step": 2991
+    },
+    {
+      "epoch": 1.5335725269092773,
+      "grad_norm": 0.5322064595848072,
+      "learning_rate": 2.720779535627359e-06,
+      "loss": 0.239,
+      "step": 2992
+    },
+    {
+      "epoch": 1.5340850845720144,
+      "grad_norm": 0.5133965753321528,
+      "learning_rate": 2.715089486611132e-06,
+      "loss": 0.2412,
+      "step": 2993
+    },
+    {
+      "epoch": 1.5345976422347514,
+      "grad_norm": 0.5098632532982835,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.2559,
+      "step": 2994
+    },
+    {
+      "epoch": 1.5351101998974883,
+      "grad_norm": 0.5434447664773335,
+      "learning_rate": 2.7037244565886335e-06,
+      "loss": 0.2456,
+      "step": 2995
+    },
+    {
+      "epoch": 1.5356227575602255,
+      "grad_norm": 0.4806584643840181,
+      "learning_rate": 2.698049483416083e-06,
+      "loss": 0.2526,
+      "step": 2996
+    },
+    {
+      "epoch": 1.5361353152229626,
+      "grad_norm": 0.559337766722489,
+      "learning_rate": 2.692379543352699e-06,
+      "loss": 0.2807,
+      "step": 2997
+    },
+    {
+      "epoch": 1.5366478728856996,
+      "grad_norm": 0.516316630851891,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2643,
+      "step": 2998
+    },
+    {
+      "epoch": 1.5371604305484365,
+      "grad_norm": 0.5773809374164198,
+      "learning_rate": 2.681054778182748e-06,
+      "loss": 0.2609,
+      "step": 2999
+    },
+    {
+      "epoch": 1.5376729882111737,
+      "grad_norm": 0.4974285132360085,
+      "learning_rate": 2.675399960882138e-06,
+      "loss": 0.254,
+      "step": 3000
+    },
+    {
+      "epoch": 1.5381855458739109,
+      "grad_norm": 0.538724587609134,
+      "learning_rate": 2.669750192302628e-06,
+      "loss": 0.255,
+      "step": 3001
+    },
+    {
+      "epoch": 1.5386981035366478,
+      "grad_norm": 0.5069900134119818,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.2596,
+      "step": 3002
+    },
+    {
+      "epoch": 1.5392106611993848,
+      "grad_norm": 0.5472833255198619,
+      "learning_rate": 2.658465816880572e-06,
+      "loss": 0.2238,
+      "step": 3003
+    },
+    {
+      "epoch": 1.539723218862122,
+      "grad_norm": 0.5310864821004946,
+      "learning_rate": 2.652831217816151e-06,
+      "loss": 0.2492,
+      "step": 3004
+    },
+    {
+      "epoch": 1.540235776524859,
+      "grad_norm": 0.5535706219705497,
+      "learning_rate": 2.647201683029075e-06,
+      "loss": 0.2345,
+      "step": 3005
+    },
+    {
+      "epoch": 1.540748334187596,
+      "grad_norm": 0.5473707005440255,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.2507,
+      "step": 3006
+    },
+    {
+      "epoch": 1.5412608918503332,
+      "grad_norm": 0.4955192539334103,
+      "learning_rate": 2.6359578218048287e-06,
+      "loss": 0.2276,
+      "step": 3007
+    },
+    {
+      "epoch": 1.5417734495130704,
+      "grad_norm": 0.49971808717512584,
+      "learning_rate": 2.630343503117859e-06,
+      "loss": 0.2376,
+      "step": 3008
+    },
+    {
+      "epoch": 1.5422860071758073,
+      "grad_norm": 0.5010464229579439,
+      "learning_rate": 2.6247342642086304e-06,
+      "loss": 0.238,
+      "step": 3009
+    },
+    {
+      "epoch": 1.5427985648385443,
+      "grad_norm": 0.5062179862152687,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.2439,
+      "step": 3010
+    },
+    {
+      "epoch": 1.5433111225012814,
+      "grad_norm": 0.5019229600390536,
+      "learning_rate": 2.6135310411852977e-06,
+      "loss": 0.2445,
+      "step": 3011
+    },
+    {
+      "epoch": 1.5438236801640186,
+      "grad_norm": 0.5678853874114322,
+      "learning_rate": 2.6079370647933834e-06,
+      "loss": 0.2646,
+      "step": 3012
+    },
+    {
+      "epoch": 1.5443362378267556,
+      "grad_norm": 0.5758775150515856,
+      "learning_rate": 2.602348183623582e-06,
+      "loss": 0.2748,
+      "step": 3013
+    },
+    {
+      "epoch": 1.5448487954894925,
+      "grad_norm": 0.48888119334784863,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.2236,
+      "step": 3014
+    },
+    {
+      "epoch": 1.5453613531522297,
+      "grad_norm": 0.5067785140768815,
+      "learning_rate": 2.591185722356082e-06,
+      "loss": 0.2515,
+      "step": 3015
+    },
+    {
+      "epoch": 1.5458739108149668,
+      "grad_norm": 0.4759674707311523,
+      "learning_rate": 2.5856121499524832e-06,
+      "loss": 0.2634,
+      "step": 3016
+    },
+    {
+      "epoch": 1.5463864684777038,
+      "grad_norm": 0.5872927009087061,
+      "learning_rate": 2.580043688159172e-06,
+      "loss": 0.2698,
+      "step": 3017
+    },
+    {
+      "epoch": 1.5468990261404407,
+      "grad_norm": 0.5213053358929479,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2384,
+      "step": 3018
+    },
+    {
+      "epoch": 1.547411583803178,
+      "grad_norm": 0.5836867610936175,
+      "learning_rate": 2.568922111752893e-06,
+      "loss": 0.2563,
+      "step": 3019
+    },
+    {
+      "epoch": 1.547924141465915,
+      "grad_norm": 0.5375456039418391,
+      "learning_rate": 2.5633690048058346e-06,
+      "loss": 0.2441,
+      "step": 3020
+    },
+    {
+      "epoch": 1.548436699128652,
+      "grad_norm": 0.5124884121096905,
+      "learning_rate": 2.5578210238008883e-06,
+      "loss": 0.2255,
+      "step": 3021
+    },
+    {
+      "epoch": 1.548949256791389,
+      "grad_norm": 0.5436731151904414,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.259,
+      "step": 3022
+    },
+    {
+      "epoch": 1.5494618144541261,
+      "grad_norm": 0.5043493344068016,
+      "learning_rate": 2.5467404549103094e-06,
+      "loss": 0.2294,
+      "step": 3023
+    },
+    {
+      "epoch": 1.5499743721168633,
+      "grad_norm": 0.5160278382226638,
+      "learning_rate": 2.5412078746623225e-06,
+      "loss": 0.2288,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5504869297796002,
+      "grad_norm": 0.5267759119953673,
+      "learning_rate": 2.5356804356317345e-06,
+      "loss": 0.2298,
+      "step": 3025
+    },
+    {
+      "epoch": 1.5509994874423372,
+      "grad_norm": 0.49728022460972693,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.2111,
+      "step": 3026
+    },
+    {
+      "epoch": 1.5515120451050743,
+      "grad_norm": 0.5369321309226305,
+      "learning_rate": 2.5246409964590855e-06,
+      "loss": 0.249,
+      "step": 3027
+    },
+    {
+      "epoch": 1.5520246027678115,
+      "grad_norm": 0.5266943837257894,
+      "learning_rate": 2.5191290039263203e-06,
+      "loss": 0.2412,
+      "step": 3028
+    },
+    {
+      "epoch": 1.5525371604305485,
+      "grad_norm": 0.5181892743135796,
+      "learning_rate": 2.5136221678295413e-06,
+      "loss": 0.268,
+      "step": 3029
+    },
+    {
+      "epoch": 1.5530497180932854,
+      "grad_norm": 0.5329813433249545,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.2476,
+      "step": 3030
+    },
+    {
+      "epoch": 1.5535622757560226,
+      "grad_norm": 0.5247521021129103,
+      "learning_rate": 2.502623980123453e-06,
+      "loss": 0.2232,
+      "step": 3031
+    },
+    {
+      "epoch": 1.5540748334187597,
+      "grad_norm": 0.49974756871557385,
+      "learning_rate": 2.4971326360950032e-06,
+      "loss": 0.2473,
+      "step": 3032
+    },
+    {
+      "epoch": 1.5545873910814967,
+      "grad_norm": 0.48455636473457214,
+      "learning_rate": 2.491646463664261e-06,
+      "loss": 0.2495,
+      "step": 3033
+    },
+    {
+      "epoch": 1.5550999487442336,
+      "grad_norm": 0.4995109825941713,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.2239,
+      "step": 3034
+    },
+    {
+      "epoch": 1.5556125064069708,
+      "grad_norm": 0.5133001352044422,
+      "learning_rate": 2.4806896487184207e-06,
+      "loss": 0.2381,
+      "step": 3035
+    },
+    {
+      "epoch": 1.556125064069708,
+      "grad_norm": 0.5218820889826726,
+      "learning_rate": 2.4752190137556676e-06,
+      "loss": 0.2685,
+      "step": 3036
+    },
+    {
+      "epoch": 1.556637621732445,
+      "grad_norm": 0.531184400236219,
+      "learning_rate": 2.469753565495303e-06,
+      "loss": 0.2317,
+      "step": 3037
+    },
+    {
+      "epoch": 1.5571501793951819,
+      "grad_norm": 0.5102107422461594,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.2367,
+      "step": 3038
+    },
+    {
+      "epoch": 1.557662737057919,
+      "grad_norm": 0.48862849875767733,
+      "learning_rate": 2.4588382441471193e-06,
+      "loss": 0.2277,
+      "step": 3039
+    },
+    {
+      "epoch": 1.5581752947206562,
+      "grad_norm": 0.4963162438053482,
+      "learning_rate": 2.4533883785830438e-06,
+      "loss": 0.2417,
+      "step": 3040
+    },
+    {
+      "epoch": 1.5586878523833931,
+      "grad_norm": 0.507359173596799,
+      "learning_rate": 2.447943714768839e-06,
+      "loss": 0.2438,
+      "step": 3041
+    },
+    {
+      "epoch": 1.55920041004613,
+      "grad_norm": 0.5199262027772462,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2439,
+      "step": 3042
+    },
+    {
+      "epoch": 1.5597129677088672,
+      "grad_norm": 0.4769516962413674,
+      "learning_rate": 2.437070007398107e-06,
+      "loss": 0.2279,
+      "step": 3043
+    },
+    {
+      "epoch": 1.5602255253716044,
+      "grad_norm": 0.489520558891542,
+      "learning_rate": 2.4316409713366353e-06,
+      "loss": 0.2461,
+      "step": 3044
+    },
+    {
+      "epoch": 1.5607380830343414,
+      "grad_norm": 0.564414345554938,
+      "learning_rate": 2.4262171520151557e-06,
+      "loss": 0.2529,
+      "step": 3045
+    },
+    {
+      "epoch": 1.5612506406970783,
+      "grad_norm": 0.4878955965170003,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.2401,
+      "step": 3046
+    },
+    {
+      "epoch": 1.5617631983598155,
+      "grad_norm": 0.5291189552300684,
+      "learning_rate": 2.415385178542725e-06,
+      "loss": 0.249,
+      "step": 3047
+    },
+    {
+      "epoch": 1.5622757560225526,
+      "grad_norm": 0.5624096596924621,
+      "learning_rate": 2.409977031858075e-06,
+      "loss": 0.2369,
+      "step": 3048
+    },
+    {
+      "epoch": 1.5627883136852896,
+      "grad_norm": 0.5190907441974739,
+      "learning_rate": 2.4045741168459944e-06,
+      "loss": 0.25,
+      "step": 3049
+    },
+    {
+      "epoch": 1.5633008713480265,
+      "grad_norm": 0.5114146104054029,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.234,
+      "step": 3050
+    },
+    {
+      "epoch": 1.5638134290107637,
+      "grad_norm": 0.5024505167631466,
+      "learning_rate": 2.393783996732462e-06,
+      "loss": 0.2496,
+      "step": 3051
+    },
+    {
+      "epoch": 1.5643259866735009,
+      "grad_norm": 0.5244576323029445,
+      "learning_rate": 2.388396799068463e-06,
+      "loss": 0.2475,
+      "step": 3052
+    },
+    {
+      "epoch": 1.5648385443362378,
+      "grad_norm": 0.575713854590954,
+      "learning_rate": 2.38301484795191e-06,
+      "loss": 0.276,
+      "step": 3053
+    },
+    {
+      "epoch": 1.5653511019989748,
+      "grad_norm": 0.4937918958844317,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.2621,
+      "step": 3054
+    },
+    {
+      "epoch": 1.565863659661712,
+      "grad_norm": 0.45454785102405065,
+      "learning_rate": 2.3722667001962898e-06,
+      "loss": 0.1915,
+      "step": 3055
+    },
+    {
+      "epoch": 1.566376217324449,
+      "grad_norm": 0.5377144694937355,
+      "learning_rate": 2.366900510965733e-06,
+      "loss": 0.2482,
+      "step": 3056
+    },
+    {
+      "epoch": 1.566888774987186,
+      "grad_norm": 0.6140187477898489,
+      "learning_rate": 2.361539583099649e-06,
+      "loss": 0.2754,
+      "step": 3057
+    },
+    {
+      "epoch": 1.567401332649923,
+      "grad_norm": 0.542119830242135,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.2551,
+      "step": 3058
+    },
+    {
+      "epoch": 1.5679138903126602,
+      "grad_norm": 0.5316306594823743,
+      "learning_rate": 2.3508335262380542e-06,
+      "loss": 0.2369,
+      "step": 3059
+    },
+    {
+      "epoch": 1.5684264479753973,
+      "grad_norm": 0.5468930494870622,
+      "learning_rate": 2.3454884046220463e-06,
+      "loss": 0.2572,
+      "step": 3060
+    },
+    {
+      "epoch": 1.5689390056381343,
+      "grad_norm": 0.5187416102460544,
+      "learning_rate": 2.340148559129507e-06,
+      "loss": 0.2405,
+      "step": 3061
+    },
+    {
+      "epoch": 1.5694515633008712,
+      "grad_norm": 0.531664386242341,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.232,
+      "step": 3062
+    },
+    {
+      "epoch": 1.5699641209636084,
+      "grad_norm": 0.5334091403125983,
+      "learning_rate": 2.329484711233858e-06,
+      "loss": 0.2604,
+      "step": 3063
+    },
+    {
+      "epoch": 1.5704766786263455,
+      "grad_norm": 0.523484536381414,
+      "learning_rate": 2.3241607161811553e-06,
+      "loss": 0.2431,
+      "step": 3064
+    },
+    {
+      "epoch": 1.5709892362890825,
+      "grad_norm": 0.5425416545305555,
+      "learning_rate": 2.31884201195273e-06,
+      "loss": 0.2632,
+      "step": 3065
+    },
+    {
+      "epoch": 1.5715017939518194,
+      "grad_norm": 0.6413947760427886,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.22,
+      "step": 3066
+    },
+    {
+      "epoch": 1.5720143516145566,
+      "grad_norm": 0.528725811844811,
+      "learning_rate": 2.3082204906294448e-06,
+      "loss": 0.2374,
+      "step": 3067
+    },
+    {
+      "epoch": 1.5725269092772938,
+      "grad_norm": 0.4554220863947333,
+      "learning_rate": 2.3029176808558097e-06,
+      "loss": 0.2274,
+      "step": 3068
+    },
+    {
+      "epoch": 1.5730394669400307,
+      "grad_norm": 0.47650262222257267,
+      "learning_rate": 2.297620176548916e-06,
+      "loss": 0.2024,
+      "step": 3069
+    },
+    {
+      "epoch": 1.5735520246027677,
+      "grad_norm": 0.5470092094688457,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.2263,
+      "step": 3070
+    },
+    {
+      "epoch": 1.5740645822655048,
+      "grad_norm": 0.5153767584631598,
+      "learning_rate": 2.2870410989376067e-06,
+      "loss": 0.2514,
+      "step": 3071
+    },
+    {
+      "epoch": 1.574577139928242,
+      "grad_norm": 0.4979482048711437,
+      "learning_rate": 2.2817595329251663e-06,
+      "loss": 0.2454,
+      "step": 3072
+    },
+    {
+      "epoch": 1.575089697590979,
+      "grad_norm": 0.5419381739737,
+      "learning_rate": 2.2764832869634134e-06,
+      "loss": 0.2539,
+      "step": 3073
+    },
+    {
+      "epoch": 1.5756022552537159,
+      "grad_norm": 0.4929771862488478,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.2246,
+      "step": 3074
+    },
+    {
+      "epoch": 1.576114812916453,
+      "grad_norm": 0.5541030500591361,
+      "learning_rate": 2.2659467697356074e-06,
+      "loss": 0.2663,
+      "step": 3075
+    },
+    {
+      "epoch": 1.5766273705791902,
+      "grad_norm": 0.49901620924134704,
+      "learning_rate": 2.2606865057321938e-06,
+      "loss": 0.2402,
+      "step": 3076
+    },
+    {
+      "epoch": 1.5771399282419272,
+      "grad_norm": 0.4771482117691628,
+      "learning_rate": 2.255431576304744e-06,
+      "loss": 0.2021,
+      "step": 3077
+    },
+    {
+      "epoch": 1.5776524859046641,
+      "grad_norm": 0.5616534907073568,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.267,
+      "step": 3078
+    },
+    {
+      "epoch": 1.5781650435674013,
+      "grad_norm": 0.5455341640523875,
+      "learning_rate": 2.244937735662589e-06,
+      "loss": 0.2105,
+      "step": 3079
+    },
+    {
+      "epoch": 1.5786776012301384,
+      "grad_norm": 0.5290625343337305,
+      "learning_rate": 2.239698831681105e-06,
+      "loss": 0.2592,
+      "step": 3080
+    },
+    {
+      "epoch": 1.5791901588928754,
+      "grad_norm": 0.5164753403650651,
+      "learning_rate": 2.2344652767420337e-06,
+      "loss": 0.2484,
+      "step": 3081
+    },
+    {
+      "epoch": 1.5797027165556123,
+      "grad_norm": 0.5249480156594774,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.2427,
+      "step": 3082
+    },
+    {
+      "epoch": 1.5802152742183497,
+      "grad_norm": 0.5133350471795772,
+      "learning_rate": 2.2240142284170165e-06,
+      "loss": 0.2424,
+      "step": 3083
+    },
+    {
+      "epoch": 1.5807278318810867,
+      "grad_norm": 0.49371334587132737,
+      "learning_rate": 2.2187967422348023e-06,
+      "loss": 0.2412,
+      "step": 3084
+    },
+    {
+      "epoch": 1.5812403895438236,
+      "grad_norm": 0.5615722452703445,
+      "learning_rate": 2.213584619502451e-06,
+      "loss": 0.2395,
+      "step": 3085
+    },
+    {
+      "epoch": 1.5817529472065608,
+      "grad_norm": 0.5049148329233929,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.2407,
+      "step": 3086
+    },
+    {
+      "epoch": 1.582265504869298,
+      "grad_norm": 0.4770069442409838,
+      "learning_rate": 2.2031764787541244e-06,
+      "loss": 0.2357,
+      "step": 3087
+    },
+    {
+      "epoch": 1.582778062532035,
+      "grad_norm": 0.5859143566238632,
+      "learning_rate": 2.1979804679123108e-06,
+      "loss": 0.27,
+      "step": 3088
+    },
+    {
+      "epoch": 1.5832906201947718,
+      "grad_norm": 0.5251332844240465,
+      "learning_rate": 2.1927898348686515e-06,
+      "loss": 0.2445,
+      "step": 3089
+    },
+    {
+      "epoch": 1.583803177857509,
+      "grad_norm": 0.5182356396589252,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.2417,
+      "step": 3090
+    },
+    {
+      "epoch": 1.5843157355202462,
+      "grad_norm": 0.49498427546926144,
+      "learning_rate": 2.1824247164833656e-06,
+      "loss": 0.2385,
+      "step": 3091
+    },
+    {
+      "epoch": 1.5848282931829831,
+      "grad_norm": 0.5954105980060099,
+      "learning_rate": 2.177250238286235e-06,
+      "loss": 0.2439,
+      "step": 3092
+    },
+    {
+      "epoch": 1.58534085084572,
+      "grad_norm": 0.5533345242315244,
+      "learning_rate": 2.1720811521762554e-06,
+      "loss": 0.2782,
+      "step": 3093
+    },
+    {
+      "epoch": 1.5858534085084572,
+      "grad_norm": 0.49846828437424007,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.2227,
+      "step": 3094
+    },
+    {
+      "epoch": 1.5863659661711944,
+      "grad_norm": 0.5237519910205727,
+      "learning_rate": 2.161759170465878e-06,
+      "loss": 0.2444,
+      "step": 3095
+    },
+    {
+      "epoch": 1.5868785238339314,
+      "grad_norm": 0.5463318471667173,
+      "learning_rate": 2.15660628198025e-06,
+      "loss": 0.279,
+      "step": 3096
+    },
+    {
+      "epoch": 1.5873910814966683,
+      "grad_norm": 0.5779438746804688,
+      "learning_rate": 2.1514587998112956e-06,
+      "loss": 0.2588,
+      "step": 3097
+    },
+    {
+      "epoch": 1.5879036391594055,
+      "grad_norm": 0.5693514938510426,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2412,
+      "step": 3098
+    },
+    {
+      "epoch": 1.5884161968221426,
+      "grad_norm": 0.5296832487239268,
+      "learning_rate": 2.141180068611971e-06,
+      "loss": 0.2607,
+      "step": 3099
+    },
+    {
+      "epoch": 1.5889287544848796,
+      "grad_norm": 0.5527640986133651,
+      "learning_rate": 2.1360488266665534e-06,
+      "loss": 0.2443,
+      "step": 3100
+    },
+    {
+      "epoch": 1.5894413121476165,
+      "grad_norm": 0.4866725753847905,
+      "learning_rate": 2.13092300520771e-06,
+      "loss": 0.2449,
+      "step": 3101
+    },
+    {
+      "epoch": 1.5899538698103537,
+      "grad_norm": 0.5499955299349529,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2516,
+      "step": 3102
+    },
+    {
+      "epoch": 1.5904664274730909,
+      "grad_norm": 0.5639681384142792,
+      "learning_rate": 2.1206876378785877e-06,
+      "loss": 0.2408,
+      "step": 3103
+    },
+    {
+      "epoch": 1.5909789851358278,
+      "grad_norm": 0.4851855470711491,
+      "learning_rate": 2.1155780990633644e-06,
+      "loss": 0.2335,
+      "step": 3104
+    },
+    {
+      "epoch": 1.5914915427985648,
+      "grad_norm": 0.5108296275022218,
+      "learning_rate": 2.1104739948448415e-06,
+      "loss": 0.2493,
+      "step": 3105
+    },
+    {
+      "epoch": 1.592004100461302,
+      "grad_norm": 0.5491982963373662,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2318,
+      "step": 3106
+    },
+    {
+      "epoch": 1.592516658124039,
+      "grad_norm": 0.48736461259596114,
+      "learning_rate": 2.100282104266829e-06,
+      "loss": 0.2292,
+      "step": 3107
+    },
+    {
+      "epoch": 1.593029215786776,
+      "grad_norm": 0.4961409917374532,
+      "learning_rate": 2.0951943249324334e-06,
+      "loss": 0.2253,
+      "step": 3108
+    },
+    {
+      "epoch": 1.593541773449513,
+      "grad_norm": 0.48867565743622887,
+      "learning_rate": 2.090111994244923e-06,
+      "loss": 0.2273,
+      "step": 3109
+    },
+    {
+      "epoch": 1.5940543311122501,
+      "grad_norm": 0.551808861383016,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2674,
+      "step": 3110
+    },
+    {
+      "epoch": 1.5945668887749873,
+      "grad_norm": 0.5312778105396738,
+      "learning_rate": 2.07996369281945e-06,
+      "loss": 0.2522,
+      "step": 3111
+    },
+    {
+      "epoch": 1.5950794464377243,
+      "grad_norm": 0.5413055772152011,
+      "learning_rate": 2.074897729076536e-06,
+      "loss": 0.2223,
+      "step": 3112
+    },
+    {
+      "epoch": 1.5955920041004612,
+      "grad_norm": 0.5263472017990082,
+      "learning_rate": 2.069837227970599e-06,
+      "loss": 0.2514,
+      "step": 3113
+    },
+    {
+      "epoch": 1.5961045617631984,
+      "grad_norm": 0.5338307463493753,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.2451,
+      "step": 3114
+    },
+    {
+      "epoch": 1.5966171194259355,
+      "grad_norm": 0.5313613487087695,
+      "learning_rate": 2.0597326276183716e-06,
+      "loss": 0.2256,
+      "step": 3115
+    },
+    {
+      "epoch": 1.5971296770886725,
+      "grad_norm": 0.5594807149975337,
+      "learning_rate": 2.0546885353369984e-06,
+      "loss": 0.2647,
+      "step": 3116
+    },
+    {
+      "epoch": 1.5976422347514094,
+      "grad_norm": 0.5053277592538636,
+      "learning_rate": 2.0496499196224605e-06,
+      "loss": 0.2321,
+      "step": 3117
+    },
+    {
+      "epoch": 1.5981547924141466,
+      "grad_norm": 0.4970290690114874,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.2308,
+      "step": 3118
+    },
+    {
+      "epoch": 1.5986673500768838,
+      "grad_norm": 0.5247576427586187,
+      "learning_rate": 2.0395891317822146e-06,
+      "loss": 0.2785,
+      "step": 3119
+    },
+    {
+      "epoch": 1.5991799077396207,
+      "grad_norm": 0.49505044233317685,
+      "learning_rate": 2.0345669665912402e-06,
+      "loss": 0.2117,
+      "step": 3120
+    },
+    {
+      "epoch": 1.5996924654023577,
+      "grad_norm": 0.543995061941087,
+      "learning_rate": 2.0295502918365473e-06,
+      "loss": 0.2387,
+      "step": 3121
+    },
+    {
+      "epoch": 1.6002050230650948,
+      "grad_norm": 0.5177078022256205,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.2557,
+      "step": 3122
+    },
+    {
+      "epoch": 1.600717580727832,
+      "grad_norm": 0.5060777707856164,
+      "learning_rate": 2.019533427463851e-06,
+      "loss": 0.2077,
+      "step": 3123
+    },
+    {
+      "epoch": 1.601230138390569,
+      "grad_norm": 0.5479810829238886,
+      "learning_rate": 2.0145332447502995e-06,
+      "loss": 0.2672,
+      "step": 3124
+    },
+    {
+      "epoch": 1.6017426960533059,
+      "grad_norm": 0.5527091027976196,
+      "learning_rate": 2.009538566281931e-06,
+      "loss": 0.252,
+      "step": 3125
+    },
+    {
+      "epoch": 1.602255253716043,
+      "grad_norm": 0.5201146637681893,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2391,
+      "step": 3126
+    },
+    {
+      "epoch": 1.6027678113787802,
+      "grad_norm": 0.5062492475341338,
+      "learning_rate": 1.9995657358479293e-06,
+      "loss": 0.2174,
+      "step": 3127
+    },
+    {
+      "epoch": 1.6032803690415172,
+      "grad_norm": 0.675013611994724,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.2812,
+      "step": 3128
+    },
+    {
+      "epoch": 1.603792926704254,
+      "grad_norm": 0.5161004350103808,
+      "learning_rate": 1.989614963658244e-06,
+      "loss": 0.2322,
+      "step": 3129
+    },
+    {
+      "epoch": 1.6043054843669913,
+      "grad_norm": 0.49941292948763405,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.2455,
+      "step": 3130
+    },
+    {
+      "epoch": 1.6048180420297284,
+      "grad_norm": 0.5208660624096583,
+      "learning_rate": 1.9796862771484526e-06,
+      "loss": 0.255,
+      "step": 3131
+    },
+    {
+      "epoch": 1.6053305996924654,
+      "grad_norm": 0.481576723664151,
+      "learning_rate": 1.9747302245804944e-06,
+      "loss": 0.2213,
+      "step": 3132
+    },
+    {
+      "epoch": 1.6058431573552023,
+      "grad_norm": 0.5552454196849025,
+      "learning_rate": 1.969779703693253e-06,
+      "loss": 0.2494,
+      "step": 3133
+    },
+    {
+      "epoch": 1.6063557150179395,
+      "grad_norm": 0.504886881154331,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.2455,
+      "step": 3134
+    },
+    {
+      "epoch": 1.6068682726806767,
+      "grad_norm": 0.5102917395266561,
+      "learning_rate": 1.959895270606356e-06,
+      "loss": 0.2559,
+      "step": 3135
+    },
+    {
+      "epoch": 1.6073808303434136,
+      "grad_norm": 0.5161641945032281,
+      "learning_rate": 1.9549613652198663e-06,
+      "loss": 0.241,
+      "step": 3136
+    },
+    {
+      "epoch": 1.6078933880061506,
+      "grad_norm": 0.5273590255322894,
+      "learning_rate": 1.950033005140436e-06,
+      "loss": 0.2334,
+      "step": 3137
+    },
+    {
+      "epoch": 1.6084059456688877,
+      "grad_norm": 0.5361312962641317,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.2398,
+      "step": 3138
+    },
+    {
+      "epoch": 1.608918503331625,
+      "grad_norm": 0.5120633278478968,
+      "learning_rate": 1.9401929344870563e-06,
+      "loss": 0.2462,
+      "step": 3139
+    },
+    {
+      "epoch": 1.6094310609943618,
+      "grad_norm": 0.4428402673777169,
+      "learning_rate": 1.9352812306956993e-06,
+      "loss": 0.1861,
+      "step": 3140
+    },
+    {
+      "epoch": 1.6099436186570988,
+      "grad_norm": 0.4715205377609909,
+      "learning_rate": 1.9303750857765767e-06,
+      "loss": 0.226,
+      "step": 3141
+    },
+    {
+      "epoch": 1.610456176319836,
+      "grad_norm": 0.5987196665204105,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.2541,
+      "step": 3142
+    },
+    {
+      "epoch": 1.6109687339825731,
+      "grad_norm": 0.5439638425496803,
+      "learning_rate": 1.920579486078091e-06,
+      "loss": 0.2524,
+      "step": 3143
+    },
+    {
+      "epoch": 1.61148129164531,
+      "grad_norm": 0.5458816924272684,
+      "learning_rate": 1.915690038050666e-06,
+      "loss": 0.2336,
+      "step": 3144
+    },
+    {
+      "epoch": 1.611993849308047,
+      "grad_norm": 0.53937246511486,
+      "learning_rate": 1.910806162399349e-06,
+      "loss": 0.2372,
+      "step": 3145
+    },
+    {
+      "epoch": 1.6125064069707842,
+      "grad_norm": 0.5377435604500345,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.2221,
+      "step": 3146
+    },
+    {
+      "epoch": 1.6130189646335213,
+      "grad_norm": 0.5076950863971811,
+      "learning_rate": 1.901055141686684e-06,
+      "loss": 0.2592,
+      "step": 3147
+    },
+    {
+      "epoch": 1.6135315222962583,
+      "grad_norm": 0.5610919966086337,
+      "learning_rate": 1.896188003346544e-06,
+      "loss": 0.2631,
+      "step": 3148
+    },
+    {
+      "epoch": 1.6140440799589952,
+      "grad_norm": 0.5172093074439709,
+      "learning_rate": 1.891326450824933e-06,
+      "loss": 0.2391,
+      "step": 3149
+    },
+    {
+      "epoch": 1.6145566376217324,
+      "grad_norm": 0.512591586211109,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.2522,
+      "step": 3150
+    },
+    {
+      "epoch": 1.6150691952844696,
+      "grad_norm": 0.5308466969263923,
+      "learning_rate": 1.881620116637375e-06,
+      "loss": 0.2313,
+      "step": 3151
+    },
+    {
+      "epoch": 1.6155817529472065,
+      "grad_norm": 0.4615822668926308,
+      "learning_rate": 1.8767753416618372e-06,
+      "loss": 0.2148,
+      "step": 3152
+    },
+    {
+      "epoch": 1.6160943106099435,
+      "grad_norm": 0.5458222295580524,
+      "learning_rate": 1.871936165885644e-06,
+      "loss": 0.2519,
+      "step": 3153
+    },
+    {
+      "epoch": 1.6166068682726806,
+      "grad_norm": 0.485945404900315,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.228,
+      "step": 3154
+    },
+    {
+      "epoch": 1.6171194259354178,
+      "grad_norm": 0.5394881112149802,
+      "learning_rate": 1.8622746252696522e-06,
+      "loss": 0.2629,
+      "step": 3155
+    },
+    {
+      "epoch": 1.6176319835981547,
+      "grad_norm": 0.5037093396003983,
+      "learning_rate": 1.8574522670893935e-06,
+      "loss": 0.2393,
+      "step": 3156
+    },
+    {
+      "epoch": 1.6181445412608917,
+      "grad_norm": 0.5137337450958432,
+      "learning_rate": 1.8526355214275382e-06,
+      "loss": 0.2129,
+      "step": 3157
+    },
+    {
+      "epoch": 1.6186570989236289,
+      "grad_norm": 0.508695158212496,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.2195,
+      "step": 3158
+    },
+    {
+      "epoch": 1.619169656586366,
+      "grad_norm": 0.5539565395202343,
+      "learning_rate": 1.843018880935572e-06,
+      "loss": 0.2443,
+      "step": 3159
+    },
+    {
+      "epoch": 1.619682214249103,
+      "grad_norm": 0.4814818562175274,
+      "learning_rate": 1.8382189927340454e-06,
+      "loss": 0.2241,
+      "step": 3160
+    },
+    {
+      "epoch": 1.6201947719118401,
+      "grad_norm": 0.5341182515572347,
+      "learning_rate": 1.8334247303080888e-06,
+      "loss": 0.2547,
+      "step": 3161
+    },
+    {
+      "epoch": 1.6207073295745773,
+      "grad_norm": 0.5131554119961185,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2553,
+      "step": 3162
+    },
+    {
+      "epoch": 1.6212198872373142,
+      "grad_norm": 0.4927066706139979,
+      "learning_rate": 1.8238530959974143e-06,
+      "loss": 0.246,
+      "step": 3163
+    },
+    {
+      "epoch": 1.6217324449000512,
+      "grad_norm": 0.5378624001302635,
+      "learning_rate": 1.8190757307102646e-06,
+      "loss": 0.2524,
+      "step": 3164
+    },
+    {
+      "epoch": 1.6222450025627884,
+      "grad_norm": 0.5510933669338351,
+      "learning_rate": 1.8143040043938054e-06,
+      "loss": 0.2409,
+      "step": 3165
+    },
+    {
+      "epoch": 1.6227575602255255,
+      "grad_norm": 0.530552743425026,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.2274,
+      "step": 3166
+    },
+    {
+      "epoch": 1.6232701178882625,
+      "grad_norm": 0.5003623721721365,
+      "learning_rate": 1.8047774818253483e-06,
+      "loss": 0.2321,
+      "step": 3167
+    },
+    {
+      "epoch": 1.6237826755509994,
+      "grad_norm": 0.49389465747679084,
+      "learning_rate": 1.8000226921398234e-06,
+      "loss": 0.251,
+      "step": 3168
+    },
+    {
+      "epoch": 1.6242952332137366,
+      "grad_norm": 0.47919568398277984,
+      "learning_rate": 1.7952735545579181e-06,
+      "loss": 0.2129,
+      "step": 3169
+    },
+    {
+      "epoch": 1.6248077908764738,
+      "grad_norm": 0.5047134998578975,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.223,
+      "step": 3170
+    },
+    {
+      "epoch": 1.6253203485392107,
+      "grad_norm": 0.5034916247916313,
+      "learning_rate": 1.7857922487950873e-06,
+      "loss": 0.2329,
+      "step": 3171
+    },
+    {
+      "epoch": 1.6258329062019476,
+      "grad_norm": 0.5152449897496599,
+      "learning_rate": 1.7810600871494555e-06,
+      "loss": 0.2388,
+      "step": 3172
+    },
+    {
+      "epoch": 1.6263454638646848,
+      "grad_norm": 0.5712774612488613,
+      "learning_rate": 1.7763335906780487e-06,
+      "loss": 0.2474,
+      "step": 3173
+    },
+    {
+      "epoch": 1.626858021527422,
+      "grad_norm": 0.5210441783830833,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.2499,
+      "step": 3174
+    },
+    {
+      "epoch": 1.627370579190159,
+      "grad_norm": 0.5197135698582757,
+      "learning_rate": 1.766897606285568e-06,
+      "loss": 0.2459,
+      "step": 3175
+    },
+    {
+      "epoch": 1.6278831368528959,
+      "grad_norm": 0.5568779037074354,
+      "learning_rate": 1.7621881248685569e-06,
+      "loss": 0.2346,
+      "step": 3176
+    },
+    {
+      "epoch": 1.628395694515633,
+      "grad_norm": 0.5288944611402787,
+      "learning_rate": 1.7574843216338878e-06,
+      "loss": 0.2457,
+      "step": 3177
+    },
+    {
+      "epoch": 1.6289082521783702,
+      "grad_norm": 0.5382478857303544,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.2717,
+      "step": 3178
+    },
+    {
+      "epoch": 1.6294208098411072,
+      "grad_norm": 0.49875471351228157,
+      "learning_rate": 1.7480937626766648e-06,
+      "loss": 0.2356,
+      "step": 3179
+    },
+    {
+      "epoch": 1.629933367503844,
+      "grad_norm": 0.4910422255986165,
+      "learning_rate": 1.743407013426861e-06,
+      "loss": 0.2226,
+      "step": 3180
+    },
+    {
+      "epoch": 1.6304459251665813,
+      "grad_norm": 0.529931070046108,
+      "learning_rate": 1.7387259553048963e-06,
+      "loss": 0.2616,
+      "step": 3181
+    },
+    {
+      "epoch": 1.6309584828293184,
+      "grad_norm": 0.4870687952203956,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.2266,
+      "step": 3182
+    },
+    {
+      "epoch": 1.6314710404920554,
+      "grad_norm": 0.515448112619337,
+      "learning_rate": 1.7293809253468607e-06,
+      "loss": 0.2278,
+      "step": 3183
+    },
+    {
+      "epoch": 1.6319835981547923,
+      "grad_norm": 0.533581161127086,
+      "learning_rate": 1.7247169599521529e-06,
+      "loss": 0.2478,
+      "step": 3184
+    },
+    {
+      "epoch": 1.6324961558175295,
+      "grad_norm": 0.5397930060587866,
+      "learning_rate": 1.72005869856802e-06,
+      "loss": 0.2398,
+      "step": 3185
+    },
+    {
+      "epoch": 1.6330087134802667,
+      "grad_norm": 0.5688828173190857,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.263,
+      "step": 3186
+    },
+    {
+      "epoch": 1.6335212711430036,
+      "grad_norm": 0.5025173967991595,
+      "learning_rate": 1.7107593006709799e-06,
+      "loss": 0.2205,
+      "step": 3187
+    },
+    {
+      "epoch": 1.6340338288057406,
+      "grad_norm": 0.5216802405835953,
+      "learning_rate": 1.7061181705679874e-06,
+      "loss": 0.2397,
+      "step": 3188
+    },
+    {
+      "epoch": 1.6345463864684777,
+      "grad_norm": 0.5056983821105027,
+      "learning_rate": 1.7014827572953974e-06,
+      "loss": 0.2318,
+      "step": 3189
+    },
+    {
+      "epoch": 1.6350589441312149,
+      "grad_norm": 0.487387404599789,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2412,
+      "step": 3190
+    },
+    {
+      "epoch": 1.6355715017939518,
+      "grad_norm": 0.5110121548421106,
+      "learning_rate": 1.6922290940179098e-06,
+      "loss": 0.2267,
+      "step": 3191
+    },
+    {
+      "epoch": 1.6360840594566888,
+      "grad_norm": 0.5197977367663441,
+      "learning_rate": 1.687610850391408e-06,
+      "loss": 0.2605,
+      "step": 3192
+    },
+    {
+      "epoch": 1.636596617119426,
+      "grad_norm": 0.5180408410029227,
+      "learning_rate": 1.6829983363520842e-06,
+      "loss": 0.237,
+      "step": 3193
+    },
+    {
+      "epoch": 1.6371091747821631,
+      "grad_norm": 0.5335580180022514,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.2396,
+      "step": 3194
+    },
+    {
+      "epoch": 1.6376217324449,
+      "grad_norm": 0.5606157310684317,
+      "learning_rate": 1.6737905097483254e-06,
+      "loss": 0.243,
+      "step": 3195
+    },
+    {
+      "epoch": 1.638134290107637,
+      "grad_norm": 0.505330367236862,
+      "learning_rate": 1.6691952035306857e-06,
+      "loss": 0.2251,
+      "step": 3196
+    },
+    {
+      "epoch": 1.6386468477703742,
+      "grad_norm": 0.49778573563180034,
+      "learning_rate": 1.664605639593818e-06,
+      "loss": 0.2538,
+      "step": 3197
+    },
+    {
+      "epoch": 1.6391594054331113,
+      "grad_norm": 0.5244446257425001,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2445,
+      "step": 3198
+    },
+    {
+      "epoch": 1.6396719630958483,
+      "grad_norm": 0.5753818789916271,
+      "learning_rate": 1.6554437512124465e-06,
+      "loss": 0.2218,
+      "step": 3199
+    },
+    {
+      "epoch": 1.6401845207585852,
+      "grad_norm": 0.51920145766535,
+      "learning_rate": 1.6508714330830823e-06,
+      "loss": 0.232,
+      "step": 3200
+    },
+    {
+      "epoch": 1.6406970784213224,
+      "grad_norm": 0.4778010599766316,
+      "learning_rate": 1.6463048698647444e-06,
+      "loss": 0.2013,
+      "step": 3201
+    },
+    {
+      "epoch": 1.6412096360840596,
+      "grad_norm": 0.5157881474300374,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.231,
+      "step": 3202
+    },
+    {
+      "epoch": 1.6417221937467965,
+      "grad_norm": 0.5114929970591788,
+      "learning_rate": 1.6371890207477937e-06,
+      "loss": 0.2516,
+      "step": 3203
+    },
+    {
+      "epoch": 1.6422347514095335,
+      "grad_norm": 0.48739984110275514,
+      "learning_rate": 1.6326397411325845e-06,
+      "loss": 0.2323,
+      "step": 3204
+    },
+    {
+      "epoch": 1.6427473090722706,
+      "grad_norm": 0.5633303797647968,
+      "learning_rate": 1.6280962289951907e-06,
+      "loss": 0.2685,
+      "step": 3205
+    },
+    {
+      "epoch": 1.6432598667350078,
+      "grad_norm": 0.5346271636052836,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.2386,
+      "step": 3206
+    },
+    {
+      "epoch": 1.6437724243977447,
+      "grad_norm": 0.5495561128433274,
+      "learning_rate": 1.6190265196769539e-06,
+      "loss": 0.2582,
+      "step": 3207
+    },
+    {
+      "epoch": 1.6442849820604817,
+      "grad_norm": 0.496998995711455,
+      "learning_rate": 1.6145003287476968e-06,
+      "loss": 0.2337,
+      "step": 3208
+    },
+    {
+      "epoch": 1.6447975397232188,
+      "grad_norm": 0.5172639410079406,
+      "learning_rate": 1.6099799177994491e-06,
+      "loss": 0.2465,
+      "step": 3209
+    },
+    {
+      "epoch": 1.645310097385956,
+      "grad_norm": 0.5335075183349058,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.2473,
+      "step": 3210
+    },
+    {
+      "epoch": 1.645822655048693,
+      "grad_norm": 0.5454504363154796,
+      "learning_rate": 1.6009564483053496e-06,
+      "loss": 0.2615,
+      "step": 3211
+    },
+    {
+      "epoch": 1.64633521271143,
+      "grad_norm": 0.5364076968514258,
+      "learning_rate": 1.5964533959792217e-06,
+      "loss": 0.2697,
+      "step": 3212
+    },
+    {
+      "epoch": 1.646847770374167,
+      "grad_norm": 0.5305191419974998,
+      "learning_rate": 1.5919561360735403e-06,
+      "loss": 0.2491,
+      "step": 3213
+    },
+    {
+      "epoch": 1.6473603280369042,
+      "grad_norm": 0.5340300533887727,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2224,
+      "step": 3214
+    },
+    {
+      "epoch": 1.6478728856996412,
+      "grad_norm": 0.5351935598840123,
+      "learning_rate": 1.5829790059190586e-06,
+      "loss": 0.2026,
+      "step": 3215
+    },
+    {
+      "epoch": 1.6483854433623781,
+      "grad_norm": 0.5100094762414473,
+      "learning_rate": 1.578499141858041e-06,
+      "loss": 0.2254,
+      "step": 3216
+    },
+    {
+      "epoch": 1.6488980010251153,
+      "grad_norm": 0.4760816346643629,
+      "learning_rate": 1.574025082593027e-06,
+      "loss": 0.2377,
+      "step": 3217
+    },
+    {
+      "epoch": 1.6494105586878525,
+      "grad_norm": 0.4790750583772664,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.2189,
+      "step": 3218
+    },
+    {
+      "epoch": 1.6499231163505894,
+      "grad_norm": 0.505507910088423,
+      "learning_rate": 1.5650943907825888e-06,
+      "loss": 0.2334,
+      "step": 3219
+    },
+    {
+      "epoch": 1.6504356740133264,
+      "grad_norm": 0.5283318615513791,
+      "learning_rate": 1.5606377643929305e-06,
+      "loss": 0.252,
+      "step": 3220
+    },
+    {
+      "epoch": 1.6509482316760635,
+      "grad_norm": 0.5322562038780867,
+      "learning_rate": 1.5561869551108212e-06,
+      "loss": 0.242,
+      "step": 3221
+    },
+    {
+      "epoch": 1.6514607893388007,
+      "grad_norm": 0.527859923448585,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.254,
+      "step": 3222
+    },
+    {
+      "epoch": 1.6519733470015376,
+      "grad_norm": 0.5635662125918758,
+      "learning_rate": 1.5473028001366973e-06,
+      "loss": 0.2638,
+      "step": 3223
+    },
+    {
+      "epoch": 1.6524859046642746,
+      "grad_norm": 0.5206507311973239,
+      "learning_rate": 1.542869460568379e-06,
+      "loss": 0.2524,
+      "step": 3224
+    },
+    {
+      "epoch": 1.6529984623270118,
+      "grad_norm": 0.4906338720823249,
+      "learning_rate": 1.5384419503549941e-06,
+      "loss": 0.236,
+      "step": 3225
+    },
+    {
+      "epoch": 1.653511019989749,
+      "grad_norm": 0.4975849464395455,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2489,
+      "step": 3226
+    },
+    {
+      "epoch": 1.6540235776524859,
+      "grad_norm": 0.5550790303614296,
+      "learning_rate": 1.529604430196232e-06,
+      "loss": 0.2525,
+      "step": 3227
+    },
+    {
+      "epoch": 1.6545361353152228,
+      "grad_norm": 0.5720025573784405,
+      "learning_rate": 1.5251944263424056e-06,
+      "loss": 0.2575,
+      "step": 3228
+    },
+    {
+      "epoch": 1.65504869297796,
+      "grad_norm": 0.5110925448752148,
+      "learning_rate": 1.5207902640266114e-06,
+      "loss": 0.2295,
+      "step": 3229
+    },
+    {
+      "epoch": 1.6555612506406971,
+      "grad_norm": 0.5395004750053991,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.2524,
+      "step": 3230
+    },
+    {
+      "epoch": 1.656073808303434,
+      "grad_norm": 0.5054457926943821,
+      "learning_rate": 1.5119994761479429e-06,
+      "loss": 0.2317,
+      "step": 3231
+    },
+    {
+      "epoch": 1.656586365966171,
+      "grad_norm": 0.5848652658667447,
+      "learning_rate": 1.507612856644405e-06,
+      "loss": 0.2589,
+      "step": 3232
+    },
+    {
+      "epoch": 1.6570989236289082,
+      "grad_norm": 0.5131028813957551,
+      "learning_rate": 1.5032320907975772e-06,
+      "loss": 0.2369,
+      "step": 3233
+    },
+    {
+      "epoch": 1.6576114812916454,
+      "grad_norm": 0.5349585557721194,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.2293,
+      "step": 3234
+    },
+    {
+      "epoch": 1.6581240389543823,
+      "grad_norm": 0.46627282688050714,
+      "learning_rate": 1.4944881321483428e-06,
+      "loss": 0.2469,
+      "step": 3235
+    },
+    {
+      "epoch": 1.6586365966171193,
+      "grad_norm": 0.5482512296378098,
+      "learning_rate": 1.4901249453730038e-06,
+      "loss": 0.2654,
+      "step": 3236
+    },
+    {
+      "epoch": 1.6591491542798564,
+      "grad_norm": 0.5104594694290455,
+      "learning_rate": 1.485767624308485e-06,
+      "loss": 0.2457,
+      "step": 3237
+    },
+    {
+      "epoch": 1.6596617119425936,
+      "grad_norm": 0.5382601139190156,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.237,
+      "step": 3238
+    },
+    {
+      "epoch": 1.6601742696053305,
+      "grad_norm": 0.5449162822420923,
+      "learning_rate": 1.4770705913215743e-06,
+      "loss": 0.243,
+      "step": 3239
+    },
+    {
+      "epoch": 1.6606868272680677,
+      "grad_norm": 0.4683044255425816,
+      "learning_rate": 1.4727308853939004e-06,
+      "loss": 0.2194,
+      "step": 3240
+    },
+    {
+      "epoch": 1.6611993849308049,
+      "grad_norm": 0.531219066340037,
+      "learning_rate": 1.4683970571664763e-06,
+      "loss": 0.2373,
+      "step": 3241
+    },
+    {
+      "epoch": 1.6617119425935418,
+      "grad_norm": 0.5127385996209569,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.2338,
+      "step": 3242
+    },
+    {
+      "epoch": 1.6622245002562788,
+      "grad_norm": 0.5104777326531248,
+      "learning_rate": 1.4597470457572583e-06,
+      "loss": 0.2378,
+      "step": 3243
+    },
+    {
+      "epoch": 1.662737057919016,
+      "grad_norm": 0.49531451227412604,
+      "learning_rate": 1.4554308685377694e-06,
+      "loss": 0.2085,
+      "step": 3244
+    },
+    {
+      "epoch": 1.663249615581753,
+      "grad_norm": 0.5169470455665702,
+      "learning_rate": 1.451120580943134e-06,
+      "loss": 0.2138,
+      "step": 3245
+    },
+    {
+      "epoch": 1.66376217324449,
+      "grad_norm": 0.5329604205602907,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.2745,
+      "step": 3246
+    },
+    {
+      "epoch": 1.664274730907227,
+      "grad_norm": 0.5023521543081089,
+      "learning_rate": 1.4425176865083946e-06,
+      "loss": 0.2224,
+      "step": 3247
+    },
+    {
+      "epoch": 1.6647872885699642,
+      "grad_norm": 0.5131457609973199,
+      "learning_rate": 1.4382250855981172e-06,
+      "loss": 0.2386,
+      "step": 3248
+    },
+    {
+      "epoch": 1.6652998462327013,
+      "grad_norm": 0.593795129152365,
+      "learning_rate": 1.4339383861723465e-06,
+      "loss": 0.3015,
+      "step": 3249
+    },
+    {
+      "epoch": 1.6658124038954383,
+      "grad_norm": 0.5030573272744473,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2305,
+      "step": 3250
+    },
+    {
+      "epoch": 1.6663249615581752,
+      "grad_norm": 0.5399565115137506,
+      "learning_rate": 1.4253827035892442e-06,
+      "loss": 0.247,
+      "step": 3251
+    },
+    {
+      "epoch": 1.6668375192209124,
+      "grad_norm": 0.5293563751908983,
+      "learning_rate": 1.4211137263291952e-06,
+      "loss": 0.252,
+      "step": 3252
+    },
+    {
+      "epoch": 1.6673500768836496,
+      "grad_norm": 0.5429077208820838,
+      "learning_rate": 1.4168506623482202e-06,
+      "loss": 0.2394,
+      "step": 3253
+    },
+    {
+      "epoch": 1.6678626345463865,
+      "grad_norm": 0.5106442494007915,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2403,
+      "step": 3254
+    },
+    {
+      "epoch": 1.6683751922091234,
+      "grad_norm": 0.5779687890295951,
+      "learning_rate": 1.4083422859732365e-06,
+      "loss": 0.2562,
+      "step": 3255
+    },
+    {
+      "epoch": 1.6688877498718606,
+      "grad_norm": 0.5092930422748942,
+      "learning_rate": 1.4040969794439041e-06,
+      "loss": 0.2403,
+      "step": 3256
+    },
+    {
+      "epoch": 1.6694003075345978,
+      "grad_norm": 0.5460900759217091,
+      "learning_rate": 1.3998575979229944e-06,
+      "loss": 0.2514,
+      "step": 3257
+    },
+    {
+      "epoch": 1.6699128651973347,
+      "grad_norm": 0.47989733259827555,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.2018,
+      "step": 3258
+    },
+    {
+      "epoch": 1.6704254228600717,
+      "grad_norm": 0.49272485132022953,
+      "learning_rate": 1.3913966215908926e-06,
+      "loss": 0.2401,
+      "step": 3259
+    },
+    {
+      "epoch": 1.6709379805228088,
+      "grad_norm": 0.545779972912956,
+      "learning_rate": 1.387175032611705e-06,
+      "loss": 0.2431,
+      "step": 3260
+    },
+    {
+      "epoch": 1.671450538185546,
+      "grad_norm": 0.4625724744269356,
+      "learning_rate": 1.382959380304948e-06,
+      "loss": 0.2028,
+      "step": 3261
+    },
+    {
+      "epoch": 1.671963095848283,
+      "grad_norm": 0.5291040082046965,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.235,
+      "step": 3262
+    },
+    {
+      "epoch": 1.67247565351102,
+      "grad_norm": 0.5014860959002262,
+      "learning_rate": 1.3745458973277392e-06,
+      "loss": 0.1999,
+      "step": 3263
+    },
+    {
+      "epoch": 1.672988211173757,
+      "grad_norm": 0.5110390015610551,
+      "learning_rate": 1.3703480724565577e-06,
+      "loss": 0.2247,
+      "step": 3264
+    },
+    {
+      "epoch": 1.6735007688364942,
+      "grad_norm": 0.4937128748265562,
+      "learning_rate": 1.3661561958563407e-06,
+      "loss": 0.2136,
+      "step": 3265
+    },
+    {
+      "epoch": 1.6740133264992312,
+      "grad_norm": 0.5505039723549386,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.2708,
+      "step": 3266
+    },
+    {
+      "epoch": 1.6745258841619681,
+      "grad_norm": 0.529598228646798,
+      "learning_rate": 1.3577902990222547e-06,
+      "loss": 0.2756,
+      "step": 3267
+    },
+    {
+      "epoch": 1.6750384418247053,
+      "grad_norm": 0.5212092463322514,
+      "learning_rate": 1.3536162845548572e-06,
+      "loss": 0.2281,
+      "step": 3268
+    },
+    {
+      "epoch": 1.6755509994874425,
+      "grad_norm": 0.5444650086596087,
+      "learning_rate": 1.3494482298913614e-06,
+      "loss": 0.2481,
+      "step": 3269
+    },
+    {
+      "epoch": 1.6760635571501794,
+      "grad_norm": 0.5449520230204529,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.2335,
+      "step": 3270
+    },
+    {
+      "epoch": 1.6765761148129164,
+      "grad_norm": 0.5818434558339727,
+      "learning_rate": 1.341130011463826e-06,
+      "loss": 0.2881,
+      "step": 3271
+    },
+    {
+      "epoch": 1.6770886724756535,
+      "grad_norm": 0.5746055214826089,
+      "learning_rate": 1.336979853433399e-06,
+      "loss": 0.2424,
+      "step": 3272
+    },
+    {
+      "epoch": 1.6776012301383907,
+      "grad_norm": 0.554661294582167,
+      "learning_rate": 1.3328356666740739e-06,
+      "loss": 0.2345,
+      "step": 3273
+    },
+    {
+      "epoch": 1.6781137878011276,
+      "grad_norm": 0.5489594571696332,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2756,
+      "step": 3274
+    },
+    {
+      "epoch": 1.6786263454638646,
+      "grad_norm": 0.48943771580143397,
+      "learning_rate": 1.3245652183906965e-06,
+      "loss": 0.2398,
+      "step": 3275
+    },
+    {
+      "epoch": 1.6791389031266017,
+      "grad_norm": 0.5371124664346122,
+      "learning_rate": 1.3204389625673208e-06,
+      "loss": 0.2237,
+      "step": 3276
+    },
+    {
+      "epoch": 1.679651460789339,
+      "grad_norm": 0.5569170136163664,
+      "learning_rate": 1.3163186894164026e-06,
+      "loss": 0.277,
+      "step": 3277
+    },
+    {
+      "epoch": 1.6801640184520759,
+      "grad_norm": 0.5750203420838248,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.2601,
+      "step": 3278
+    },
+    {
+      "epoch": 1.6806765761148128,
+      "grad_norm": 0.4748619954596225,
+      "learning_rate": 1.3080961024879501e-06,
+      "loss": 0.2205,
+      "step": 3279
+    },
+    {
+      "epoch": 1.68118913377755,
+      "grad_norm": 0.5191693879117384,
+      "learning_rate": 1.3039937943781046e-06,
+      "loss": 0.2381,
+      "step": 3280
+    },
+    {
+      "epoch": 1.6817016914402871,
+      "grad_norm": 0.5267513186699142,
+      "learning_rate": 1.29989748027609e-06,
+      "loss": 0.2264,
+      "step": 3281
+    },
+    {
+      "epoch": 1.682214249103024,
+      "grad_norm": 0.521262374143951,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.2504,
+      "step": 3282
+    },
+    {
+      "epoch": 1.682726806765761,
+      "grad_norm": 0.4879678277222229,
+      "learning_rate": 1.2917228453854936e-06,
+      "loss": 0.2232,
+      "step": 3283
+    },
+    {
+      "epoch": 1.6832393644284982,
+      "grad_norm": 0.5223174600572797,
+      "learning_rate": 1.2876445302315489e-06,
+      "loss": 0.2276,
+      "step": 3284
+    },
+    {
+      "epoch": 1.6837519220912354,
+      "grad_norm": 0.5859403508623747,
+      "learning_rate": 1.2835722203546996e-06,
+      "loss": 0.2593,
+      "step": 3285
+    },
+    {
+      "epoch": 1.6842644797539723,
+      "grad_norm": 0.4803442302728703,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.2113,
+      "step": 3286
+    },
+    {
+      "epoch": 1.6847770374167093,
+      "grad_norm": 0.4876640726138578,
+      "learning_rate": 1.2754456276560534e-06,
+      "loss": 0.2036,
+      "step": 3287
+    },
+    {
+      "epoch": 1.6852895950794464,
+      "grad_norm": 0.5649389835651036,
+      "learning_rate": 1.27139135043577e-06,
+      "loss": 0.2402,
+      "step": 3288
+    },
+    {
+      "epoch": 1.6858021527421836,
+      "grad_norm": 0.5737938012311957,
+      "learning_rate": 1.2673430896956252e-06,
+      "loss": 0.256,
+      "step": 3289
+    },
+    {
+      "epoch": 1.6863147104049205,
+      "grad_norm": 0.5013885057720785,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.2109,
+      "step": 3290
+    },
+    {
+      "epoch": 1.6868272680676575,
+      "grad_norm": 0.5442580330191681,
+      "learning_rate": 1.2592646288131804e-06,
+      "loss": 0.2492,
+      "step": 3291
+    },
+    {
+      "epoch": 1.6873398257303946,
+      "grad_norm": 0.5358969257962213,
+      "learning_rate": 1.2552344342392242e-06,
+      "loss": 0.2698,
+      "step": 3292
+    },
+    {
+      "epoch": 1.6878523833931318,
+      "grad_norm": 0.5093953042115243,
+      "learning_rate": 1.2512102672820902e-06,
+      "loss": 0.2202,
+      "step": 3293
+    },
+    {
+      "epoch": 1.6883649410558688,
+      "grad_norm": 0.49272923056179146,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2145,
+      "step": 3294
+    },
+    {
+      "epoch": 1.6888774987186057,
+      "grad_norm": 0.4976413293109661,
+      "learning_rate": 1.2431800273092832e-06,
+      "loss": 0.2313,
+      "step": 3295
+    },
+    {
+      "epoch": 1.6893900563813429,
+      "grad_norm": 0.5544855187037272,
+      "learning_rate": 1.2391739598287167e-06,
+      "loss": 0.2482,
+      "step": 3296
+    },
+    {
+      "epoch": 1.68990261404408,
+      "grad_norm": 0.5555022592159052,
+      "learning_rate": 1.2351739310351796e-06,
+      "loss": 0.2299,
+      "step": 3297
+    },
+    {
+      "epoch": 1.690415171706817,
+      "grad_norm": 0.530135479734727,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.2427,
+      "step": 3298
+    },
+    {
+      "epoch": 1.690927729369554,
+      "grad_norm": 0.49918619882861104,
+      "learning_rate": 1.2271920005336425e-06,
+      "loss": 0.2223,
+      "step": 3299
+    },
+    {
+      "epoch": 1.691440287032291,
+      "grad_norm": 0.4967234427041997,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.2291,
+      "step": 3300
+    },
+    {
+      "epoch": 1.6919528446950283,
+      "grad_norm": 0.525746176290551,
+      "learning_rate": 1.21923425781189e-06,
+      "loss": 0.2443,
+      "step": 3301
+    },
+    {
+      "epoch": 1.6924654023577652,
+      "grad_norm": 0.4830706213682881,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.2072,
+      "step": 3302
+    },
+    {
+      "epoch": 1.6929779600205022,
+      "grad_norm": 0.4896959101319103,
+      "learning_rate": 1.2113007248104625e-06,
+      "loss": 0.2424,
+      "step": 3303
+    },
+    {
+      "epoch": 1.6934905176832393,
+      "grad_norm": 0.559992561389048,
+      "learning_rate": 1.2073430437930412e-06,
+      "loss": 0.2591,
+      "step": 3304
+    },
+    {
+      "epoch": 1.6940030753459765,
+      "grad_norm": 0.47634581140154936,
+      "learning_rate": 1.2033914234031552e-06,
+      "loss": 0.217,
+      "step": 3305
+    },
+    {
+      "epoch": 1.6945156330087134,
+      "grad_norm": 0.5307553725677446,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.2598,
+      "step": 3306
+    },
+    {
+      "epoch": 1.6950281906714504,
+      "grad_norm": 0.5577007463120576,
+      "learning_rate": 1.1955063753969365e-06,
+      "loss": 0.2721,
+      "step": 3307
+    },
+    {
+      "epoch": 1.6955407483341876,
+      "grad_norm": 0.4699388457116617,
+      "learning_rate": 1.1915729532156372e-06,
+      "loss": 0.2176,
+      "step": 3308
+    },
+    {
+      "epoch": 1.6960533059969247,
+      "grad_norm": 0.5177251636272668,
+      "learning_rate": 1.187645602531925e-06,
+      "loss": 0.2374,
+      "step": 3309
+    },
+    {
+      "epoch": 1.6965658636596617,
+      "grad_norm": 0.5583422797605423,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.2347,
+      "step": 3310
+    },
+    {
+      "epoch": 1.6970784213223986,
+      "grad_norm": 0.529959976458408,
+      "learning_rate": 1.179809126481295e-06,
+      "loss": 0.243,
+      "step": 3311
+    },
+    {
+      "epoch": 1.6975909789851358,
+      "grad_norm": 0.4847256790007376,
+      "learning_rate": 1.1759000065159276e-06,
+      "loss": 0.2039,
+      "step": 3312
+    },
+    {
+      "epoch": 1.698103536647873,
+      "grad_norm": 0.5495398278112248,
+      "learning_rate": 1.1719969688512389e-06,
+      "loss": 0.2434,
+      "step": 3313
+    },
+    {
+      "epoch": 1.69861609431061,
+      "grad_norm": 0.4907774809817668,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.217,
+      "step": 3314
+    },
+    {
+      "epoch": 1.6991286519733468,
+      "grad_norm": 0.5092073745204287,
+      "learning_rate": 1.164209151180895e-06,
+      "loss": 0.2345,
+      "step": 3315
+    },
+    {
+      "epoch": 1.699641209636084,
+      "grad_norm": 0.5328820948816221,
+      "learning_rate": 1.1603243765432527e-06,
+      "loss": 0.2337,
+      "step": 3316
+    },
+    {
+      "epoch": 1.7001537672988212,
+      "grad_norm": 0.5335396152015577,
+      "learning_rate": 1.156445694942302e-06,
+      "loss": 0.241,
+      "step": 3317
+    },
+    {
+      "epoch": 1.7006663249615581,
+      "grad_norm": 0.5658757942223469,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2502,
+      "step": 3318
+    },
+    {
+      "epoch": 1.7011788826242953,
+      "grad_norm": 0.4907415509211336,
+      "learning_rate": 1.1487066215403186e-06,
+      "loss": 0.2345,
+      "step": 3319
+    },
+    {
+      "epoch": 1.7016914402870325,
+      "grad_norm": 0.4898078890800497,
+      "learning_rate": 1.1448462350736956e-06,
+      "loss": 0.2305,
+      "step": 3320
+    },
+    {
+      "epoch": 1.7022039979497694,
+      "grad_norm": 0.5092993192739954,
+      "learning_rate": 1.1409919523125823e-06,
+      "loss": 0.244,
+      "step": 3321
+    },
+    {
+      "epoch": 1.7027165556125063,
+      "grad_norm": 0.5148485869112024,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.224,
+      "step": 3322
+    },
+    {
+      "epoch": 1.7032291132752435,
+      "grad_norm": 0.5539948259555254,
+      "learning_rate": 1.1333017085294528e-06,
+      "loss": 0.2564,
+      "step": 3323
+    },
+    {
+      "epoch": 1.7037416709379807,
+      "grad_norm": 0.51585904140871,
+      "learning_rate": 1.1294657528081842e-06,
+      "loss": 0.2118,
+      "step": 3324
+    },
+    {
+      "epoch": 1.7042542286007176,
+      "grad_norm": 0.5214801153976563,
+      "learning_rate": 1.1256359113939309e-06,
+      "loss": 0.2439,
+      "step": 3325
+    },
+    {
+      "epoch": 1.7047667862634546,
+      "grad_norm": 0.5050392694139995,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.2254,
+      "step": 3326
+    },
+    {
+      "epoch": 1.7052793439261917,
+      "grad_norm": 0.5114913725383123,
+      "learning_rate": 1.1179945820416282e-06,
+      "loss": 0.2201,
+      "step": 3327
+    },
+    {
+      "epoch": 1.705791901588929,
+      "grad_norm": 0.5185552332341519,
+      "learning_rate": 1.1141830993706149e-06,
+      "loss": 0.248,
+      "step": 3328
+    },
+    {
+      "epoch": 1.7063044592516659,
+      "grad_norm": 0.544383870976034,
+      "learning_rate": 1.110377741540687e-06,
+      "loss": 0.2434,
+      "step": 3329
+    },
+    {
+      "epoch": 1.7068170169144028,
+      "grad_norm": 0.48306222137929783,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.2365,
+      "step": 3330
+    },
+    {
+      "epoch": 1.70732957457714,
+      "grad_norm": 0.48623255209966865,
+      "learning_rate": 1.1027854108917335e-06,
+      "loss": 0.2319,
+      "step": 3331
+    },
+    {
+      "epoch": 1.7078421322398771,
+      "grad_norm": 0.512049960084452,
+      "learning_rate": 1.098998443305972e-06,
+      "loss": 0.2281,
+      "step": 3332
+    },
+    {
+      "epoch": 1.708354689902614,
+      "grad_norm": 0.525232727640963,
+      "learning_rate": 1.0952176110278167e-06,
+      "loss": 0.2507,
+      "step": 3333
+    },
+    {
+      "epoch": 1.708867247565351,
+      "grad_norm": 0.5249429284916607,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.2013,
+      "step": 3334
+    },
+    {
+      "epoch": 1.7093798052280882,
+      "grad_norm": 0.5507899894614532,
+      "learning_rate": 1.087674362814355e-06,
+      "loss": 0.2291,
+      "step": 3335
+    },
+    {
+      "epoch": 1.7098923628908254,
+      "grad_norm": 0.5400254086404921,
+      "learning_rate": 1.0839119520784758e-06,
+      "loss": 0.2666,
+      "step": 3336
+    },
+    {
+      "epoch": 1.7104049205535623,
+      "grad_norm": 0.5069065218716144,
+      "learning_rate": 1.080155687049067e-06,
+      "loss": 0.2254,
+      "step": 3337
+    },
+    {
+      "epoch": 1.7109174782162992,
+      "grad_norm": 0.5087295959096043,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.2543,
+      "step": 3338
+    },
+    {
+      "epoch": 1.7114300358790364,
+      "grad_norm": 0.5206351210062521,
+      "learning_rate": 1.0726616044619243e-06,
+      "loss": 0.2275,
+      "step": 3339
+    },
+    {
+      "epoch": 1.7119425935417736,
+      "grad_norm": 0.5512106795229935,
+      "learning_rate": 1.068923792069736e-06,
+      "loss": 0.2441,
+      "step": 3340
+    },
+    {
+      "epoch": 1.7124551512045105,
+      "grad_norm": 0.5033075181619387,
+      "learning_rate": 1.0651921357150997e-06,
+      "loss": 0.203,
+      "step": 3341
+    },
+    {
+      "epoch": 1.7129677088672475,
+      "grad_norm": 0.5309487806398077,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.2348,
+      "step": 3342
+    },
+    {
+      "epoch": 1.7134802665299846,
+      "grad_norm": 0.5335028140025285,
+      "learning_rate": 1.0577473014028872e-06,
+      "loss": 0.2363,
+      "step": 3343
+    },
+    {
+      "epoch": 1.7139928241927218,
+      "grad_norm": 0.5016933507429788,
+      "learning_rate": 1.054034128576913e-06,
+      "loss": 0.2388,
+      "step": 3344
+    },
+    {
+      "epoch": 1.7145053818554588,
+      "grad_norm": 0.49529937013687053,
+      "learning_rate": 1.0503271220516775e-06,
+      "loss": 0.2121,
+      "step": 3345
+    },
+    {
+      "epoch": 1.7150179395181957,
+      "grad_norm": 0.484300842904864,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.2285,
+      "step": 3346
+    },
+    {
+      "epoch": 1.7155304971809329,
+      "grad_norm": 0.5151207842788802,
+      "learning_rate": 1.0429316181198735e-06,
+      "loss": 0.2216,
+      "step": 3347
+    },
+    {
+      "epoch": 1.71604305484367,
+      "grad_norm": 0.5341557034328228,
+      "learning_rate": 1.039243125810898e-06,
+      "loss": 0.2237,
+      "step": 3348
+    },
+    {
+      "epoch": 1.716555612506407,
+      "grad_norm": 0.4784071950256765,
+      "learning_rate": 1.0355608099978454e-06,
+      "loss": 0.1887,
+      "step": 3349
+    },
+    {
+      "epoch": 1.717068170169144,
+      "grad_norm": 0.5168218858126105,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.2232,
+      "step": 3350
+    },
+    {
+      "epoch": 1.717580727831881,
+      "grad_norm": 0.5036812367584662,
+      "learning_rate": 1.0282147180078761e-06,
+      "loss": 0.2389,
+      "step": 3351
+    },
+    {
+      "epoch": 1.7180932854946183,
+      "grad_norm": 0.5520952298285793,
+      "learning_rate": 1.0245509468944992e-06,
+      "loss": 0.2082,
+      "step": 3352
+    },
+    {
+      "epoch": 1.7186058431573552,
+      "grad_norm": 0.5202662629830659,
+      "learning_rate": 1.0208933624041085e-06,
+      "loss": 0.2432,
+      "step": 3353
+    },
+    {
+      "epoch": 1.7191184008200922,
+      "grad_norm": 0.5965831004255551,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2741,
+      "step": 3354
+    },
+    {
+      "epoch": 1.7196309584828293,
+      "grad_norm": 0.53146165865676,
+      "learning_rate": 1.0135967633724708e-06,
+      "loss": 0.2396,
+      "step": 3355
+    },
+    {
+      "epoch": 1.7201435161455665,
+      "grad_norm": 0.5510796954553456,
+      "learning_rate": 1.009957753860642e-06,
+      "loss": 0.265,
+      "step": 3356
+    },
+    {
+      "epoch": 1.7206560738083034,
+      "grad_norm": 0.5685478013575435,
+      "learning_rate": 1.006324941030643e-06,
+      "loss": 0.2531,
+      "step": 3357
+    },
+    {
+      "epoch": 1.7211686314710404,
+      "grad_norm": 0.6027870662655229,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.2521,
+      "step": 3358
+    },
+    {
+      "epoch": 1.7216811891337775,
+      "grad_norm": 0.5190093927816598,
+      "learning_rate": 9.99077915427994e-07,
+      "loss": 0.2295,
+      "step": 3359
+    },
+    {
+      "epoch": 1.7221937467965147,
+      "grad_norm": 0.5357084873140564,
+      "learning_rate": 9.954637076505946e-07,
+      "loss": 0.2339,
+      "step": 3360
+    },
+    {
+      "epoch": 1.7227063044592517,
+      "grad_norm": 0.7215548656356966,
+      "learning_rate": 9.918557065455193e-07,
+      "loss": 0.2402,
+      "step": 3361
+    },
+    {
+      "epoch": 1.7232188621219886,
+      "grad_norm": 0.56645576151543,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2706,
+      "step": 3362
+    },
+    {
+      "epoch": 1.7237314197847258,
+      "grad_norm": 0.516925874964003,
+      "learning_rate": 9.84658334295796e-07,
+      "loss": 0.2403,
+      "step": 3363
+    },
+    {
+      "epoch": 1.724243977447463,
+      "grad_norm": 0.4969620302907086,
+      "learning_rate": 9.81068968112172e-07,
+      "loss": 0.2204,
+      "step": 3364
+    },
+    {
+      "epoch": 1.7247565351101999,
+      "grad_norm": 0.5779781190922875,
+      "learning_rate": 9.774858185229198e-07,
+      "loss": 0.2456,
+      "step": 3365
+    },
+    {
+      "epoch": 1.7252690927729368,
+      "grad_norm": 0.4801503882238692,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.228,
+      "step": 3366
+    },
+    {
+      "epoch": 1.725781650435674,
+      "grad_norm": 0.5434670503929323,
+      "learning_rate": 9.703381790024547e-07,
+      "loss": 0.2364,
+      "step": 3367
+    },
+    {
+      "epoch": 1.7262942080984112,
+      "grad_norm": 0.5238090842425055,
+      "learning_rate": 9.667736939979888e-07,
+      "loss": 0.2412,
+      "step": 3368
+    },
+    {
+      "epoch": 1.726806765761148,
+      "grad_norm": 0.5433946435151384,
+      "learning_rate": 9.63215435441378e-07,
+      "loss": 0.2478,
+      "step": 3369
+    },
+    {
+      "epoch": 1.727319323423885,
+      "grad_norm": 0.5793976499313717,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2482,
+      "step": 3370
+    },
+    {
+      "epoch": 1.7278318810866222,
+      "grad_norm": 0.4982229949107075,
+      "learning_rate": 9.561176074780299e-07,
+      "loss": 0.2252,
+      "step": 3371
+    },
+    {
+      "epoch": 1.7283444387493594,
+      "grad_norm": 0.547574423121492,
+      "learning_rate": 9.525780429636999e-07,
+      "loss": 0.2458,
+      "step": 3372
+    },
+    {
+      "epoch": 1.7288569964120963,
+      "grad_norm": 0.5360621364961248,
+      "learning_rate": 9.49044714682047e-07,
+      "loss": 0.2618,
+      "step": 3373
+    },
+    {
+      "epoch": 1.7293695540748333,
+      "grad_norm": 0.5082869660457797,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.2495,
+      "step": 3374
+    },
+    {
+      "epoch": 1.7298821117375704,
+      "grad_norm": 0.4972044772453906,
+      "learning_rate": 9.419967765543225e-07,
+      "loss": 0.2271,
+      "step": 3375
+    },
+    {
+      "epoch": 1.7303946694003076,
+      "grad_norm": 0.5136022812636324,
+      "learning_rate": 9.384821715662729e-07,
+      "loss": 0.2141,
+      "step": 3376
+    },
+    {
+      "epoch": 1.7309072270630446,
+      "grad_norm": 0.5298378941586892,
+      "learning_rate": 9.349738125269425e-07,
+      "loss": 0.2522,
+      "step": 3377
+    },
+    {
+      "epoch": 1.7314197847257815,
+      "grad_norm": 0.5170415481433326,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2427,
+      "step": 3378
+    },
+    {
+      "epoch": 1.7319323423885187,
+      "grad_norm": 0.4864708649448814,
+      "learning_rate": 9.279758419631402e-07,
+      "loss": 0.2226,
+      "step": 3379
+    },
+    {
+      "epoch": 1.7324449000512558,
+      "grad_norm": 0.5269948197390086,
+      "learning_rate": 9.244862352622485e-07,
+      "loss": 0.2223,
+      "step": 3380
+    },
+    {
+      "epoch": 1.7329574577139928,
+      "grad_norm": 0.48696330562288226,
+      "learning_rate": 9.210028841572338e-07,
+      "loss": 0.2083,
+      "step": 3381
+    },
+    {
+      "epoch": 1.7334700153767297,
+      "grad_norm": 0.5312557165019651,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2256,
+      "step": 3382
+    },
+    {
+      "epoch": 1.733982573039467,
+      "grad_norm": 0.5237404683452935,
+      "learning_rate": 9.140549583345826e-07,
+      "loss": 0.2458,
+      "step": 3383
+    },
+    {
+      "epoch": 1.734495130702204,
+      "grad_norm": 0.5184189745149902,
+      "learning_rate": 9.105903884060307e-07,
+      "loss": 0.2297,
+      "step": 3384
+    },
+    {
+      "epoch": 1.735007688364941,
+      "grad_norm": 0.5824441520721818,
+      "learning_rate": 9.071320836515263e-07,
+      "loss": 0.2687,
+      "step": 3385
+    },
+    {
+      "epoch": 1.735520246027678,
+      "grad_norm": 0.5356174885724618,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.2413,
+      "step": 3386
+    },
+    {
+      "epoch": 1.7360328036904151,
+      "grad_norm": 0.5599900605769957,
+      "learning_rate": 9.002342791953345e-07,
+      "loss": 0.2335,
+      "step": 3387
+    },
+    {
+      "epoch": 1.7365453613531523,
+      "grad_norm": 0.5168095285528398,
+      "learning_rate": 8.967947842481894e-07,
+      "loss": 0.2385,
+      "step": 3388
+    },
+    {
+      "epoch": 1.7370579190158892,
+      "grad_norm": 0.4942997489554866,
+      "learning_rate": 8.933615639841608e-07,
+      "loss": 0.2422,
+      "step": 3389
+    },
+    {
+      "epoch": 1.7375704766786262,
+      "grad_norm": 0.515167124977177,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2542,
+      "step": 3390
+    },
+    {
+      "epoch": 1.7380830343413634,
+      "grad_norm": 0.5056786469862952,
+      "learning_rate": 8.865139569669778e-07,
+      "loss": 0.2132,
+      "step": 3391
+    },
+    {
+      "epoch": 1.7385955920041005,
+      "grad_norm": 0.5380083784850581,
+      "learning_rate": 8.830995749337612e-07,
+      "loss": 0.2335,
+      "step": 3392
+    },
+    {
+      "epoch": 1.7391081496668375,
+      "grad_norm": 0.5577843511900098,
+      "learning_rate": 8.796914770235365e-07,
+      "loss": 0.2367,
+      "step": 3393
+    },
+    {
+      "epoch": 1.7396207073295744,
+      "grad_norm": 0.547600102615428,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.247,
+      "step": 3394
+    },
+    {
+      "epoch": 1.7401332649923118,
+      "grad_norm": 0.48974199164920185,
+      "learning_rate": 8.728941429643045e-07,
+      "loss": 0.2427,
+      "step": 3395
+    },
+    {
+      "epoch": 1.7406458226550487,
+      "grad_norm": 0.490466802224485,
+      "learning_rate": 8.695049115005838e-07,
+      "loss": 0.2447,
+      "step": 3396
+    },
+    {
+      "epoch": 1.7411583803177857,
+      "grad_norm": 0.5540153133830357,
+      "learning_rate": 8.661219735304238e-07,
+      "loss": 0.2508,
+      "step": 3397
+    },
+    {
+      "epoch": 1.7416709379805229,
+      "grad_norm": 0.508547052792991,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.2418,
+      "step": 3398
+    },
+    {
+      "epoch": 1.74218349564326,
+      "grad_norm": 0.5102682699466877,
+      "learning_rate": 8.593749873936497e-07,
+      "loss": 0.2397,
+      "step": 3399
+    },
+    {
+      "epoch": 1.742696053305997,
+      "grad_norm": 0.5177288935843186,
+      "learning_rate": 8.560109438776176e-07,
+      "loss": 0.2515,
+      "step": 3400
+    },
+    {
+      "epoch": 1.743208610968734,
+      "grad_norm": 0.501979462421677,
+      "learning_rate": 8.526532031563051e-07,
+      "loss": 0.2062,
+      "step": 3401
+    },
+    {
+      "epoch": 1.743721168631471,
+      "grad_norm": 0.5238978603180079,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.2446,
+      "step": 3402
+    },
+    {
+      "epoch": 1.7442337262942083,
+      "grad_norm": 0.5287496372862599,
+      "learning_rate": 8.459566393512331e-07,
+      "loss": 0.2151,
+      "step": 3403
+    },
+    {
+      "epoch": 1.7447462839569452,
+      "grad_norm": 0.46596760668185316,
+      "learning_rate": 8.426178208832958e-07,
+      "loss": 0.2289,
+      "step": 3404
+    },
+    {
+      "epoch": 1.7452588416196821,
+      "grad_norm": 0.47609421177485434,
+      "learning_rate": 8.392853144417312e-07,
+      "loss": 0.2288,
+      "step": 3405
+    },
+    {
+      "epoch": 1.7457713992824193,
+      "grad_norm": 0.5431491181348291,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.2489,
+      "step": 3406
+    },
+    {
+      "epoch": 1.7462839569451565,
+      "grad_norm": 0.5382969059451627,
+      "learning_rate": 8.326392468215206e-07,
+      "loss": 0.2503,
+      "step": 3407
+    },
+    {
+      "epoch": 1.7467965146078934,
+      "grad_norm": 0.5042223357186406,
+      "learning_rate": 8.293256902238933e-07,
+      "loss": 0.2291,
+      "step": 3408
+    },
+    {
+      "epoch": 1.7473090722706304,
+      "grad_norm": 0.50784328619509,
+      "learning_rate": 8.260184548146755e-07,
+      "loss": 0.2076,
+      "step": 3409
+    },
+    {
+      "epoch": 1.7478216299333675,
+      "grad_norm": 0.5426608390810376,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.2598,
+      "step": 3410
+    },
+    {
+      "epoch": 1.7483341875961047,
+      "grad_norm": 0.5477829494263995,
+      "learning_rate": 8.194229566755885e-07,
+      "loss": 0.2724,
+      "step": 3411
+    },
+    {
+      "epoch": 1.7488467452588417,
+      "grad_norm": 0.5524256220370549,
+      "learning_rate": 8.161346984918827e-07,
+      "loss": 0.2545,
+      "step": 3412
+    },
+    {
+      "epoch": 1.7493593029215786,
+      "grad_norm": 0.5147573516674522,
+      "learning_rate": 8.128527705889088e-07,
+      "loss": 0.2157,
+      "step": 3413
+    },
+    {
+      "epoch": 1.7498718605843158,
+      "grad_norm": 0.5026341775340237,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.2472,
+      "step": 3414
+    },
+    {
+      "epoch": 1.750384418247053,
+      "grad_norm": 0.5298504897874496,
+      "learning_rate": 8.063079146695019e-07,
+      "loss": 0.2436,
+      "step": 3415
+    },
+    {
+      "epoch": 1.7508969759097899,
+      "grad_norm": 0.5251452164567743,
+      "learning_rate": 8.030449911643223e-07,
+      "loss": 0.2457,
+      "step": 3416
+    },
+    {
+      "epoch": 1.7514095335725268,
+      "grad_norm": 0.5162467904050223,
+      "learning_rate": 7.997884069623907e-07,
+      "loss": 0.2374,
+      "step": 3417
+    },
+    {
+      "epoch": 1.751922091235264,
+      "grad_norm": 0.5415921166376173,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.2501,
+      "step": 3418
+    },
+    {
+      "epoch": 1.7524346488980012,
+      "grad_norm": 0.5698384269193828,
+      "learning_rate": 7.932942654427089e-07,
+      "loss": 0.2584,
+      "step": 3419
+    },
+    {
+      "epoch": 1.752947206560738,
+      "grad_norm": 0.4990438920357137,
+      "learning_rate": 7.900567126012648e-07,
+      "loss": 0.2072,
+      "step": 3420
+    },
+    {
+      "epoch": 1.753459764223475,
+      "grad_norm": 0.5188568483642277,
+      "learning_rate": 7.868255080156606e-07,
+      "loss": 0.2265,
+      "step": 3421
+    },
+    {
+      "epoch": 1.7539723218862122,
+      "grad_norm": 0.530388571290447,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.268,
+      "step": 3422
+    },
+    {
+      "epoch": 1.7544848795489494,
+      "grad_norm": 0.5299798934091111,
+      "learning_rate": 7.80382152516449e-07,
+      "loss": 0.2805,
+      "step": 3423
+    },
+    {
+      "epoch": 1.7549974372116863,
+      "grad_norm": 0.5034812743212649,
+      "learning_rate": 7.7717000604414e-07,
+      "loss": 0.2412,
+      "step": 3424
+    },
+    {
+      "epoch": 1.7555099948744233,
+      "grad_norm": 0.5141186863169478,
+      "learning_rate": 7.739642167102546e-07,
+      "loss": 0.2533,
+      "step": 3425
+    },
+    {
+      "epoch": 1.7560225525371604,
+      "grad_norm": 0.5128517859044259,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2492,
+      "step": 3426
+    },
+    {
+      "epoch": 1.7565351101998976,
+      "grad_norm": 0.5284121096713792,
+      "learning_rate": 7.675717182921649e-07,
+      "loss": 0.2461,
+      "step": 3427
+    },
+    {
+      "epoch": 1.7570476678626346,
+      "grad_norm": 0.4955605476938147,
+      "learning_rate": 7.643850136141972e-07,
+      "loss": 0.2587,
+      "step": 3428
+    },
+    {
+      "epoch": 1.7575602255253715,
+      "grad_norm": 0.5530863799763457,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.2427,
+      "step": 3429
+    },
+    {
+      "epoch": 1.7580727831881087,
+      "grad_norm": 0.4988060640943368,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.2328,
+      "step": 3430
+    },
+    {
+      "epoch": 1.7585853408508458,
+      "grad_norm": 0.5472699470650839,
+      "learning_rate": 7.548631040499321e-07,
+      "loss": 0.2441,
+      "step": 3431
+    },
+    {
+      "epoch": 1.7590978985135828,
+      "grad_norm": 0.4717749207348705,
+      "learning_rate": 7.517018763109318e-07,
+      "loss": 0.2244,
+      "step": 3432
+    },
+    {
+      "epoch": 1.7596104561763197,
+      "grad_norm": 0.5479189046219579,
+      "learning_rate": 7.485470232651027e-07,
+      "loss": 0.2499,
+      "step": 3433
+    },
+    {
+      "epoch": 1.760123013839057,
+      "grad_norm": 0.5438244082235693,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.2466,
+      "step": 3434
+    },
+    {
+      "epoch": 1.760635571501794,
+      "grad_norm": 0.48851955818467396,
+      "learning_rate": 7.422564499469065e-07,
+      "loss": 0.2209,
+      "step": 3435
+    },
+    {
+      "epoch": 1.761148129164531,
+      "grad_norm": 0.49587915863034526,
+      "learning_rate": 7.391207340105233e-07,
+      "loss": 0.2107,
+      "step": 3436
+    },
+    {
+      "epoch": 1.761660686827268,
+      "grad_norm": 0.47232194035766817,
+      "learning_rate": 7.359914014392744e-07,
+      "loss": 0.2419,
+      "step": 3437
+    },
+    {
+      "epoch": 1.7621732444900051,
+      "grad_norm": 0.49410632287735995,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.2327,
+      "step": 3438
+    },
+    {
+      "epoch": 1.7626858021527423,
+      "grad_norm": 0.5408467266536386,
+      "learning_rate": 7.297518950157712e-07,
+      "loss": 0.2319,
+      "step": 3439
+    },
+    {
+      "epoch": 1.7631983598154792,
+      "grad_norm": 0.5077209672908062,
+      "learning_rate": 7.266417254642966e-07,
+      "loss": 0.2445,
+      "step": 3440
+    },
+    {
+      "epoch": 1.7637109174782162,
+      "grad_norm": 0.5102744838556093,
+      "learning_rate": 7.235379478795268e-07,
+      "loss": 0.232,
+      "step": 3441
+    },
+    {
+      "epoch": 1.7642234751409533,
+      "grad_norm": 0.5240645437164064,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.2183,
+      "step": 3442
+    },
+    {
+      "epoch": 1.7647360328036905,
+      "grad_norm": 0.4791700680907802,
+      "learning_rate": 7.1734957716321e-07,
+      "loss": 0.2061,
+      "step": 3443
+    },
+    {
+      "epoch": 1.7652485904664275,
+      "grad_norm": 0.5308755540729109,
+      "learning_rate": 7.142649882972008e-07,
+      "loss": 0.2526,
+      "step": 3444
+    },
+    {
+      "epoch": 1.7657611481291644,
+      "grad_norm": 0.5670228203045444,
+      "learning_rate": 7.111867999289679e-07,
+      "loss": 0.2582,
+      "step": 3445
+    },
+    {
+      "epoch": 1.7662737057919016,
+      "grad_norm": 0.5435774578237708,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2281,
+      "step": 3446
+    },
+    {
+      "epoch": 1.7667862634546387,
+      "grad_norm": 0.5075992447726339,
+      "learning_rate": 7.050496331683843e-07,
+      "loss": 0.2419,
+      "step": 3447
+    },
+    {
+      "epoch": 1.7672988211173757,
+      "grad_norm": 0.5463709883119504,
+      "learning_rate": 7.019906590062764e-07,
+      "loss": 0.262,
+      "step": 3448
+    },
+    {
+      "epoch": 1.7678113787801126,
+      "grad_norm": 0.5449129150748541,
+      "learning_rate": 6.989380938024293e-07,
+      "loss": 0.2411,
+      "step": 3449
+    },
+    {
+      "epoch": 1.7683239364428498,
+      "grad_norm": 0.5503506214926179,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.2436,
+      "step": 3450
+    },
+    {
+      "epoch": 1.768836494105587,
+      "grad_norm": 0.4812719433853579,
+      "learning_rate": 6.928521986814196e-07,
+      "loss": 0.2203,
+      "step": 3451
+    },
+    {
+      "epoch": 1.769349051768324,
+      "grad_norm": 0.5054490962667714,
+      "learning_rate": 6.898188729591582e-07,
+      "loss": 0.2286,
+      "step": 3452
+    },
+    {
+      "epoch": 1.7698616094310609,
+      "grad_norm": 0.5419394005650775,
+      "learning_rate": 6.867919645849619e-07,
+      "loss": 0.2558,
+      "step": 3453
+    },
+    {
+      "epoch": 1.770374167093798,
+      "grad_norm": 0.5345676613701156,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.2516,
+      "step": 3454
+    },
+    {
+      "epoch": 1.7708867247565352,
+      "grad_norm": 0.5007132516476915,
+      "learning_rate": 6.807574082219148e-07,
+      "loss": 0.2314,
+      "step": 3455
+    },
+    {
+      "epoch": 1.7713992824192721,
+      "grad_norm": 0.49401721097812984,
+      "learning_rate": 6.777497643925834e-07,
+      "loss": 0.2174,
+      "step": 3456
+    },
+    {
+      "epoch": 1.771911840082009,
+      "grad_norm": 0.5046317186245495,
+      "learning_rate": 6.747485462303449e-07,
+      "loss": 0.2294,
+      "step": 3457
+    },
+    {
+      "epoch": 1.7724243977447462,
+      "grad_norm": 0.5474581060146247,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2788,
+      "step": 3458
+    },
+    {
+      "epoch": 1.7729369554074834,
+      "grad_norm": 0.528276056815066,
+      "learning_rate": 6.687653951774642e-07,
+      "loss": 0.2328,
+      "step": 3459
+    },
+    {
+      "epoch": 1.7734495130702204,
+      "grad_norm": 0.5444092711670563,
+      "learning_rate": 6.657834664109075e-07,
+      "loss": 0.263,
+      "step": 3460
+    },
+    {
+      "epoch": 1.7739620707329573,
+      "grad_norm": 0.5387700254433981,
+      "learning_rate": 6.62807971559607e-07,
+      "loss": 0.2353,
+      "step": 3461
+    },
+    {
+      "epoch": 1.7744746283956945,
+      "grad_norm": 0.5383662608063817,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.254,
+      "step": 3462
+    },
+    {
+      "epoch": 1.7749871860584316,
+      "grad_norm": 0.5028408595694304,
+      "learning_rate": 6.568762918021688e-07,
+      "loss": 0.1975,
+      "step": 3463
+    },
+    {
+      "epoch": 1.7754997437211686,
+      "grad_norm": 0.5673211348454336,
+      "learning_rate": 6.539201109846372e-07,
+      "loss": 0.2525,
+      "step": 3464
+    },
+    {
+      "epoch": 1.7760123013839055,
+      "grad_norm": 0.5346700751637131,
+      "learning_rate": 6.509703722595684e-07,
+      "loss": 0.2523,
+      "step": 3465
+    },
+    {
+      "epoch": 1.7765248590466427,
+      "grad_norm": 0.5224161604190463,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2383,
+      "step": 3466
+    },
+    {
+      "epoch": 1.7770374167093799,
+      "grad_norm": 0.5574890305583541,
+      "learning_rate": 6.450902292151973e-07,
+      "loss": 0.2404,
+      "step": 3467
+    },
+    {
+      "epoch": 1.7775499743721168,
+      "grad_norm": 0.5275953193163245,
+      "learning_rate": 6.421598289489772e-07,
+      "loss": 0.2276,
+      "step": 3468
+    },
+    {
+      "epoch": 1.7780625320348538,
+      "grad_norm": 0.5338929588515102,
+      "learning_rate": 6.392358788813802e-07,
+      "loss": 0.2562,
+      "step": 3469
+    },
+    {
+      "epoch": 1.778575089697591,
+      "grad_norm": 0.494134436922555,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2305,
+      "step": 3470
+    },
+    {
+      "epoch": 1.779087647360328,
+      "grad_norm": 0.5312612666046042,
+      "learning_rate": 6.334073373993266e-07,
+      "loss": 0.2329,
+      "step": 3471
+    },
+    {
+      "epoch": 1.779600205023065,
+      "grad_norm": 0.49349505844960045,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.2258,
+      "step": 3472
+    },
+    {
+      "epoch": 1.7801127626858022,
+      "grad_norm": 0.5073376769667722,
+      "learning_rate": 6.276046208390873e-07,
+      "loss": 0.228,
+      "step": 3473
+    },
+    {
+      "epoch": 1.7806253203485394,
+      "grad_norm": 0.532397022604236,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2445,
+      "step": 3474
+    },
+    {
+      "epoch": 1.7811378780112763,
+      "grad_norm": 0.5272747384825595,
+      "learning_rate": 6.218277451995147e-07,
+      "loss": 0.2603,
+      "step": 3475
+    },
+    {
+      "epoch": 1.7816504356740133,
+      "grad_norm": 0.521783717630854,
+      "learning_rate": 6.18949002705137e-07,
+      "loss": 0.2446,
+      "step": 3476
+    },
+    {
+      "epoch": 1.7821629933367504,
+      "grad_norm": 0.5521151745050742,
+      "learning_rate": 6.160767264082079e-07,
+      "loss": 0.2539,
+      "step": 3477
+    },
+    {
+      "epoch": 1.7826755509994876,
+      "grad_norm": 0.4838520882557958,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2049,
+      "step": 3478
+    },
+    {
+      "epoch": 1.7831881086622245,
+      "grad_norm": 0.5324933864423501,
+      "learning_rate": 6.103515803214799e-07,
+      "loss": 0.2284,
+      "step": 3479
+    },
+    {
+      "epoch": 1.7837006663249615,
+      "grad_norm": 0.5282225996995394,
+      "learning_rate": 6.074987144779254e-07,
+      "loss": 0.2206,
+      "step": 3480
+    },
+    {
+      "epoch": 1.7842132239876987,
+      "grad_norm": 0.5656229146098024,
+      "learning_rate": 6.046523227243062e-07,
+      "loss": 0.244,
+      "step": 3481
+    },
+    {
+      "epoch": 1.7847257816504358,
+      "grad_norm": 0.5165413943477596,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2228,
+      "step": 3482
+    },
+    {
+      "epoch": 1.7852383393131728,
+      "grad_norm": 0.5327206765134981,
+      "learning_rate": 5.989789693302872e-07,
+      "loss": 0.2509,
+      "step": 3483
+    },
+    {
+      "epoch": 1.7857508969759097,
+      "grad_norm": 0.5320236923541599,
+      "learning_rate": 5.961520116004326e-07,
+      "loss": 0.2436,
+      "step": 3484
+    },
+    {
+      "epoch": 1.7862634546386469,
+      "grad_norm": 0.5026904597974902,
+      "learning_rate": 5.933315357816005e-07,
+      "loss": 0.2468,
+      "step": 3485
+    },
+    {
+      "epoch": 1.786776012301384,
+      "grad_norm": 0.4859907557456066,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2169,
+      "step": 3486
+    },
+    {
+      "epoch": 1.787288569964121,
+      "grad_norm": 0.5396860649274772,
+      "learning_rate": 5.877100376489597e-07,
+      "loss": 0.2551,
+      "step": 3487
+    },
+    {
+      "epoch": 1.787801127626858,
+      "grad_norm": 0.4965025696754931,
+      "learning_rate": 5.849090192099505e-07,
+      "loss": 0.2564,
+      "step": 3488
+    },
+    {
+      "epoch": 1.7883136852895951,
+      "grad_norm": 0.4832523048817021,
+      "learning_rate": 5.821144904315701e-07,
+      "loss": 0.2237,
+      "step": 3489
+    },
+    {
+      "epoch": 1.7888262429523323,
+      "grad_norm": 0.5185540281366555,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.2307,
+      "step": 3490
+    },
+    {
+      "epoch": 1.7893388006150692,
+      "grad_norm": 0.47125639035319494,
+      "learning_rate": 5.765449095570863e-07,
+      "loss": 0.1973,
+      "step": 3491
+    },
+    {
+      "epoch": 1.7898513582778062,
+      "grad_norm": 0.5429472376897574,
+      "learning_rate": 5.737698613000031e-07,
+      "loss": 0.2491,
+      "step": 3492
+    },
+    {
+      "epoch": 1.7903639159405433,
+      "grad_norm": 0.545616735676098,
+      "learning_rate": 5.710013103815748e-07,
+      "loss": 0.2488,
+      "step": 3493
+    },
+    {
+      "epoch": 1.7908764736032805,
+      "grad_norm": 0.5380376171953217,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.2545,
+      "step": 3494
+    },
+    {
+      "epoch": 1.7913890312660175,
+      "grad_norm": 0.532129639904329,
+      "learning_rate": 5.654837081894626e-07,
+      "loss": 0.2351,
+      "step": 3495
+    },
+    {
+      "epoch": 1.7919015889287544,
+      "grad_norm": 0.5584870586517574,
+      "learning_rate": 5.627346607189665e-07,
+      "loss": 0.2293,
+      "step": 3496
+    },
+    {
+      "epoch": 1.7924141465914916,
+      "grad_norm": 0.5320906461030782,
+      "learning_rate": 5.599921181935009e-07,
+      "loss": 0.2192,
+      "step": 3497
+    },
+    {
+      "epoch": 1.7929267042542287,
+      "grad_norm": 0.5198764914795407,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2434,
+      "step": 3498
+    },
+    {
+      "epoch": 1.7934392619169657,
+      "grad_norm": 0.5228470981461295,
+      "learning_rate": 5.545265555347235e-07,
+      "loss": 0.2305,
+      "step": 3499
+    },
+    {
+      "epoch": 1.7939518195797026,
+      "grad_norm": 0.5313577952616376,
+      "learning_rate": 5.518035391687293e-07,
+      "loss": 0.2628,
+      "step": 3500
+    },
+    {
+      "epoch": 1.7944643772424398,
+      "grad_norm": 0.49763735581291424,
+      "learning_rate": 5.490870352823996e-07,
+      "loss": 0.2421,
+      "step": 3501
+    },
+    {
+      "epoch": 1.794976934905177,
+      "grad_norm": 0.4927668558324164,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.2209,
+      "step": 3502
+    },
+    {
+      "epoch": 1.795489492567914,
+      "grad_norm": 0.5014147836422022,
+      "learning_rate": 5.436735724339981e-07,
+      "loss": 0.2128,
+      "step": 3503
+    },
+    {
+      "epoch": 1.7960020502306508,
+      "grad_norm": 0.5127427556656917,
+      "learning_rate": 5.40976617203336e-07,
+      "loss": 0.2321,
+      "step": 3504
+    },
+    {
+      "epoch": 1.796514607893388,
+      "grad_norm": 0.52225812484839,
+      "learning_rate": 5.382861819151475e-07,
+      "loss": 0.2117,
+      "step": 3505
+    },
+    {
+      "epoch": 1.7970271655561252,
+      "grad_norm": 0.47793947268772186,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.2343,
+      "step": 3506
+    },
+    {
+      "epoch": 1.7975397232188621,
+      "grad_norm": 0.485041218584336,
+      "learning_rate": 5.329248785795882e-07,
+      "loss": 0.2156,
+      "step": 3507
+    },
+    {
+      "epoch": 1.798052280881599,
+      "grad_norm": 0.5472042842079988,
+      "learning_rate": 5.302540142276679e-07,
+      "loss": 0.2577,
+      "step": 3508
+    },
+    {
+      "epoch": 1.7985648385443362,
+      "grad_norm": 0.5041933700447829,
+      "learning_rate": 5.275896772091316e-07,
+      "loss": 0.2312,
+      "step": 3509
+    },
+    {
+      "epoch": 1.7990773962070734,
+      "grad_norm": 0.4932118100724615,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.2207,
+      "step": 3510
+    },
+    {
+      "epoch": 1.7995899538698104,
+      "grad_norm": 0.4975700821855478,
+      "learning_rate": 5.222805925136298e-07,
+      "loss": 0.2144,
+      "step": 3511
+    },
+    {
+      "epoch": 1.8001025115325473,
+      "grad_norm": 0.5786813734472438,
+      "learning_rate": 5.196358484961251e-07,
+      "loss": 0.2656,
+      "step": 3512
+    },
+    {
+      "epoch": 1.8006150691952845,
+      "grad_norm": 0.5368053354652499,
+      "learning_rate": 5.169976391309228e-07,
+      "loss": 0.2534,
+      "step": 3513
+    },
+    {
+      "epoch": 1.8011276268580216,
+      "grad_norm": 0.5541357256938974,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.2614,
+      "step": 3514
+    },
+    {
+      "epoch": 1.8016401845207586,
+      "grad_norm": 0.51398313075432,
+      "learning_rate": 5.117408316268047e-07,
+      "loss": 0.2524,
+      "step": 3515
+    },
+    {
+      "epoch": 1.8021527421834955,
+      "grad_norm": 0.5362932224934296,
+      "learning_rate": 5.091222371113158e-07,
+      "loss": 0.2449,
+      "step": 3516
+    },
+    {
+      "epoch": 1.8026652998462327,
+      "grad_norm": 0.531850603349053,
+      "learning_rate": 5.065101844949793e-07,
+      "loss": 0.2412,
+      "step": 3517
+    },
+    {
+      "epoch": 1.8031778575089699,
+      "grad_norm": 0.4928405853754215,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2087,
+      "step": 3518
+    },
+    {
+      "epoch": 1.8036904151717068,
+      "grad_norm": 0.5423563392214303,
+      "learning_rate": 5.013057121570308e-07,
+      "loss": 0.2625,
+      "step": 3519
+    },
+    {
+      "epoch": 1.8042029728344438,
+      "grad_norm": 0.5432168901882702,
+      "learning_rate": 4.987132960227681e-07,
+      "loss": 0.2353,
+      "step": 3520
+    },
+    {
+      "epoch": 1.804715530497181,
+      "grad_norm": 0.47920856265018874,
+      "learning_rate": 4.961274289623641e-07,
+      "loss": 0.2084,
+      "step": 3521
+    },
+    {
+      "epoch": 1.805228088159918,
+      "grad_norm": 0.5681655717044908,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2427,
+      "step": 3522
+    },
+    {
+      "epoch": 1.805740645822655,
+      "grad_norm": 0.4998028147883861,
+      "learning_rate": 4.909753491881896e-07,
+      "loss": 0.2348,
+      "step": 3523
+    },
+    {
+      "epoch": 1.806253203485392,
+      "grad_norm": 0.5571730915772515,
+      "learning_rate": 4.884091400256574e-07,
+      "loss": 0.2427,
+      "step": 3524
+    },
+    {
+      "epoch": 1.8067657611481291,
+      "grad_norm": 0.5044315096285428,
+      "learning_rate": 4.858494870394625e-07,
+      "loss": 0.239,
+      "step": 3525
+    },
+    {
+      "epoch": 1.8072783188108663,
+      "grad_norm": 0.4908795160629024,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.2093,
+      "step": 3526
+    },
+    {
+      "epoch": 1.8077908764736033,
+      "grad_norm": 0.49661639492298937,
+      "learning_rate": 4.807498566488545e-07,
+      "loss": 0.2641,
+      "step": 3527
+    },
+    {
+      "epoch": 1.8083034341363402,
+      "grad_norm": 0.5302189123196902,
+      "learning_rate": 4.782098827595305e-07,
+      "loss": 0.2585,
+      "step": 3528
+    },
+    {
+      "epoch": 1.8088159917990774,
+      "grad_norm": 0.5026559112225567,
+      "learning_rate": 4.75676472076716e-07,
+      "loss": 0.246,
+      "step": 3529
+    },
+    {
+      "epoch": 1.8093285494618145,
+      "grad_norm": 0.5222540415344199,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.2335,
+      "step": 3530
+    },
+    {
+      "epoch": 1.8098411071245515,
+      "grad_norm": 0.5016444789354628,
+      "learning_rate": 4.706293473110313e-07,
+      "loss": 0.2403,
+      "step": 3531
+    },
+    {
+      "epoch": 1.8103536647872884,
+      "grad_norm": 0.5346792499222005,
+      "learning_rate": 4.681156367070594e-07,
+      "loss": 0.2266,
+      "step": 3532
+    },
+    {
+      "epoch": 1.8108662224500256,
+      "grad_norm": 0.5421349236821194,
+      "learning_rate": 4.656084962673857e-07,
+      "loss": 0.2637,
+      "step": 3533
+    },
+    {
+      "epoch": 1.8113787801127628,
+      "grad_norm": 0.5471576790063788,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2351,
+      "step": 3534
+    },
+    {
+      "epoch": 1.8118913377754997,
+      "grad_norm": 0.5458185359887727,
+      "learning_rate": 4.606139327889181e-07,
+      "loss": 0.242,
+      "step": 3535
+    },
+    {
+      "epoch": 1.8124038954382367,
+      "grad_norm": 0.516325069540423,
+      "learning_rate": 4.581265131927948e-07,
+      "loss": 0.2499,
+      "step": 3536
+    },
+    {
+      "epoch": 1.8129164531009738,
+      "grad_norm": 0.5605991580698305,
+      "learning_rate": 4.5564567064630085e-07,
+      "loss": 0.2572,
+      "step": 3537
+    },
+    {
+      "epoch": 1.813429010763711,
+      "grad_norm": 0.5226927066469013,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2421,
+      "step": 3538
+    },
+    {
+      "epoch": 1.813941568426448,
+      "grad_norm": 0.4896948865284463,
+      "learning_rate": 4.5070372353767543e-07,
+      "loss": 0.2397,
+      "step": 3539
+    },
+    {
+      "epoch": 1.8144541260891849,
+      "grad_norm": 0.4410497224366311,
+      "learning_rate": 4.482426223819458e-07,
+      "loss": 0.206,
+      "step": 3540
+    },
+    {
+      "epoch": 1.814966683751922,
+      "grad_norm": 0.5361273744920065,
+      "learning_rate": 4.4578810508864145e-07,
+      "loss": 0.2312,
+      "step": 3541
+    },
+    {
+      "epoch": 1.8154792414146592,
+      "grad_norm": 0.5498310840521278,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.2462,
+      "step": 3542
+    },
+    {
+      "epoch": 1.8159917990773962,
+      "grad_norm": 0.5234024205670531,
+      "learning_rate": 4.4089882885220204e-07,
+      "loss": 0.2362,
+      "step": 3543
+    },
+    {
+      "epoch": 1.816504356740133,
+      "grad_norm": 0.4796287299240725,
+      "learning_rate": 4.3846407327915784e-07,
+      "loss": 0.2534,
+      "step": 3544
+    },
+    {
+      "epoch": 1.8170169144028703,
+      "grad_norm": 0.49107048752459403,
+      "learning_rate": 4.3603590830872843e-07,
+      "loss": 0.2035,
+      "step": 3545
+    },
+    {
+      "epoch": 1.8175294720656074,
+      "grad_norm": 0.5381259852434463,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.2559,
+      "step": 3546
+    },
+    {
+      "epoch": 1.8180420297283444,
+      "grad_norm": 0.5190216264338235,
+      "learning_rate": 4.311993568659323e-07,
+      "loss": 0.2303,
+      "step": 3547
+    },
+    {
+      "epoch": 1.8185545873910813,
+      "grad_norm": 0.5369195647813269,
+      "learning_rate": 4.287909737273177e-07,
+      "loss": 0.2515,
+      "step": 3548
+    },
+    {
+      "epoch": 1.8190671450538185,
+      "grad_norm": 0.519935804634435,
+      "learning_rate": 4.263891878588178e-07,
+      "loss": 0.2424,
+      "step": 3549
+    },
+    {
+      "epoch": 1.8195797027165557,
+      "grad_norm": 0.5084684073240556,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.2421,
+      "step": 3550
+    },
+    {
+      "epoch": 1.8200922603792926,
+      "grad_norm": 0.5317416198404695,
+      "learning_rate": 4.2160541454964954e-07,
+      "loss": 0.2577,
+      "step": 3551
+    },
+    {
+      "epoch": 1.8206048180420298,
+      "grad_norm": 0.4954031659262068,
+      "learning_rate": 4.192234304063558e-07,
+      "loss": 0.2353,
+      "step": 3552
+    },
+    {
+      "epoch": 1.821117375704767,
+      "grad_norm": 0.4962147978309273,
+      "learning_rate": 4.1684805012791706e-07,
+      "loss": 0.2224,
+      "step": 3553
+    },
+    {
+      "epoch": 1.821629933367504,
+      "grad_norm": 0.49121160201167396,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.2176,
+      "step": 3554
+    },
+    {
+      "epoch": 1.8221424910302408,
+      "grad_norm": 0.5008382717527089,
+      "learning_rate": 4.1211710771030034e-07,
+      "loss": 0.2438,
+      "step": 3555
+    },
+    {
+      "epoch": 1.822655048692978,
+      "grad_norm": 0.49721239492517294,
+      "learning_rate": 4.0976154883207497e-07,
+      "loss": 0.2125,
+      "step": 3556
+    },
+    {
+      "epoch": 1.8231676063557152,
+      "grad_norm": 0.5094333708243655,
+      "learning_rate": 4.074126003406198e-07,
+      "loss": 0.2229,
+      "step": 3557
+    },
+    {
+      "epoch": 1.8236801640184521,
+      "grad_norm": 0.5154211255877035,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2202,
+      "step": 3558
+    },
+    {
+      "epoch": 1.824192721681189,
+      "grad_norm": 0.5278595678699352,
+      "learning_rate": 4.0273454098982424e-07,
+      "loss": 0.2709,
+      "step": 3559
+    },
+    {
+      "epoch": 1.8247052793439262,
+      "grad_norm": 0.5567631435374337,
+      "learning_rate": 4.0040543335498783e-07,
+      "loss": 0.2632,
+      "step": 3560
+    },
+    {
+      "epoch": 1.8252178370066634,
+      "grad_norm": 0.5302709109465966,
+      "learning_rate": 3.9808294255593293e-07,
+      "loss": 0.2412,
+      "step": 3561
+    },
+    {
+      "epoch": 1.8257303946694003,
+      "grad_norm": 0.5375345470928922,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.239,
+      "step": 3562
+    },
+    {
+      "epoch": 1.8262429523321373,
+      "grad_norm": 0.49982942156186894,
+      "learning_rate": 3.934578178640103e-07,
+      "loss": 0.2365,
+      "step": 3563
+    },
+    {
+      "epoch": 1.8267555099948745,
+      "grad_norm": 0.5267374541015465,
+      "learning_rate": 3.9115518715916345e-07,
+      "loss": 0.2162,
+      "step": 3564
+    },
+    {
+      "epoch": 1.8272680676576116,
+      "grad_norm": 0.5345868102722754,
+      "learning_rate": 3.8885917966613227e-07,
+      "loss": 0.247,
+      "step": 3565
+    },
+    {
+      "epoch": 1.8277806253203486,
+      "grad_norm": 0.49278492278394637,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.2378,
+      "step": 3566
+    },
+    {
+      "epoch": 1.8282931829830855,
+      "grad_norm": 0.5034467503990371,
+      "learning_rate": 3.842870406413457e-07,
+      "loss": 0.2583,
+      "step": 3567
+    },
+    {
+      "epoch": 1.8288057406458227,
+      "grad_norm": 0.5443312606768393,
+      "learning_rate": 3.82010912261086e-07,
+      "loss": 0.2189,
+      "step": 3568
+    },
+    {
+      "epoch": 1.8293182983085599,
+      "grad_norm": 0.5456250206129672,
+      "learning_rate": 3.7974141339563673e-07,
+      "loss": 0.2237,
+      "step": 3569
+    },
+    {
+      "epoch": 1.8298308559712968,
+      "grad_norm": 0.5614865485388217,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.2339,
+      "step": 3570
+    },
+    {
+      "epoch": 1.8303434136340337,
+      "grad_norm": 0.5876040765351455,
+      "learning_rate": 3.7522231046189237e-07,
+      "loss": 0.263,
+      "step": 3571
+    },
+    {
+      "epoch": 1.830855971296771,
+      "grad_norm": 0.5492426549852079,
+      "learning_rate": 3.7297270950854224e-07,
+      "loss": 0.2418,
+      "step": 3572
+    },
+    {
+      "epoch": 1.831368528959508,
+      "grad_norm": 0.49176069468593053,
+      "learning_rate": 3.707297442998758e-07,
+      "loss": 0.2539,
+      "step": 3573
+    },
+    {
+      "epoch": 1.831881086622245,
+      "grad_norm": 0.4682438977701545,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.2127,
+      "step": 3574
+    },
+    {
+      "epoch": 1.832393644284982,
+      "grad_norm": 0.4867980805768709,
+      "learning_rate": 3.6626372729617335e-07,
+      "loss": 0.2074,
+      "step": 3575
+    },
+    {
+      "epoch": 1.8329062019477191,
+      "grad_norm": 0.483069623192399,
+      "learning_rate": 3.640406785794892e-07,
+      "loss": 0.2231,
+      "step": 3576
+    },
+    {
+      "epoch": 1.8334187596104563,
+      "grad_norm": 0.4908743154026192,
+      "learning_rate": 3.618242717641829e-07,
+      "loss": 0.2052,
+      "step": 3577
+    },
+    {
+      "epoch": 1.8339313172731933,
+      "grad_norm": 0.4865621520755326,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2182,
+      "step": 3578
+    },
+    {
+      "epoch": 1.8344438749359302,
+      "grad_norm": 0.5079025573275948,
+      "learning_rate": 3.5741138994406346e-07,
+      "loss": 0.2212,
+      "step": 3579
+    },
+    {
+      "epoch": 1.8349564325986674,
+      "grad_norm": 0.5499565708582799,
+      "learning_rate": 3.5521491798097163e-07,
+      "loss": 0.2837,
+      "step": 3580
+    },
+    {
+      "epoch": 1.8354689902614045,
+      "grad_norm": 0.5425777059773343,
+      "learning_rate": 3.5302509400270693e-07,
+      "loss": 0.2288,
+      "step": 3581
+    },
+    {
+      "epoch": 1.8359815479241415,
+      "grad_norm": 0.4771092205061793,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.231,
+      "step": 3582
+    },
+    {
+      "epoch": 1.8364941055868784,
+      "grad_norm": 0.4890135745880503,
+      "learning_rate": 3.486653960337061e-07,
+      "loss": 0.2048,
+      "step": 3583
+    },
+    {
+      "epoch": 1.8370066632496156,
+      "grad_norm": 0.4593239678393964,
+      "learning_rate": 3.464955250480351e-07,
+      "loss": 0.199,
+      "step": 3584
+    },
+    {
+      "epoch": 1.8375192209123528,
+      "grad_norm": 0.5148072269537669,
+      "learning_rate": 3.443323080573191e-07,
+      "loss": 0.2502,
+      "step": 3585
+    },
+    {
+      "epoch": 1.8380317785750897,
+      "grad_norm": 0.47924755146609627,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.1926,
+      "step": 3586
+    },
+    {
+      "epoch": 1.8385443362378266,
+      "grad_norm": 0.5031761277226091,
+      "learning_rate": 3.40025842020435e-07,
+      "loss": 0.2278,
+      "step": 3587
+    },
+    {
+      "epoch": 1.8390568939005638,
+      "grad_norm": 0.5719309723660364,
+      "learning_rate": 3.3788259594264126e-07,
+      "loss": 0.2635,
+      "step": 3588
+    },
+    {
+      "epoch": 1.839569451563301,
+      "grad_norm": 0.4845702914128649,
+      "learning_rate": 3.357460097965448e-07,
+      "loss": 0.236,
+      "step": 3589
+    },
+    {
+      "epoch": 1.840082009226038,
+      "grad_norm": 0.5541548643476712,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.2357,
+      "step": 3590
+    },
+    {
+      "epoch": 1.8405945668887749,
+      "grad_norm": 0.5135468578311855,
+      "learning_rate": 3.314928231857084e-07,
+      "loss": 0.2103,
+      "step": 3591
+    },
+    {
+      "epoch": 1.841107124551512,
+      "grad_norm": 0.49091302610083926,
+      "learning_rate": 3.293762256526134e-07,
+      "loss": 0.2175,
+      "step": 3592
+    },
+    {
+      "epoch": 1.8416196822142492,
+      "grad_norm": 0.4989945813678562,
+      "learning_rate": 3.2726629391451436e-07,
+      "loss": 0.2209,
+      "step": 3593
+    },
+    {
+      "epoch": 1.8421322398769862,
+      "grad_norm": 0.5172635394808794,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.241,
+      "step": 3594
+    },
+    {
+      "epoch": 1.842644797539723,
+      "grad_norm": 0.5104926572720613,
+      "learning_rate": 3.2306643363605893e-07,
+      "loss": 0.232,
+      "step": 3595
+    },
+    {
+      "epoch": 1.8431573552024603,
+      "grad_norm": 0.5685790829611937,
+      "learning_rate": 3.2097650799059464e-07,
+      "loss": 0.2504,
+      "step": 3596
+    },
+    {
+      "epoch": 1.8436699128651974,
+      "grad_norm": 0.5217849986845448,
+      "learning_rate": 3.1889325392990613e-07,
+      "loss": 0.2562,
+      "step": 3597
+    },
+    {
+      "epoch": 1.8441824705279344,
+      "grad_norm": 0.532519600577765,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.2278,
+      "step": 3598
+    },
+    {
+      "epoch": 1.8446950281906713,
+      "grad_norm": 0.5589190033178515,
+      "learning_rate": 3.147467663020598e-07,
+      "loss": 0.2479,
+      "step": 3599
+    },
+    {
+      "epoch": 1.8452075858534085,
+      "grad_norm": 0.5251861522917574,
+      "learning_rate": 3.1268353559300467e-07,
+      "loss": 0.2493,
+      "step": 3600
+    },
+    {
+      "epoch": 1.8457201435161457,
+      "grad_norm": 0.500387499292152,
+      "learning_rate": 3.106269821849273e-07,
+      "loss": 0.2533,
+      "step": 3601
+    },
+    {
+      "epoch": 1.8462327011788826,
+      "grad_norm": 0.4784957877268638,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2136,
+      "step": 3602
+    },
+    {
+      "epoch": 1.8467452588416196,
+      "grad_norm": 0.5208637766300916,
+      "learning_rate": 3.065339129372935e-07,
+      "loss": 0.2395,
+      "step": 3603
+    },
+    {
+      "epoch": 1.8472578165043567,
+      "grad_norm": 0.4738100699736681,
+      "learning_rate": 3.0449739991901703e-07,
+      "loss": 0.2028,
+      "step": 3604
+    },
+    {
+      "epoch": 1.8477703741670939,
+      "grad_norm": 0.5561829869618911,
+      "learning_rate": 3.024675698442858e-07,
+      "loss": 0.244,
+      "step": 3605
+    },
+    {
+      "epoch": 1.8482829318298308,
+      "grad_norm": 0.5196435762841888,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2381,
+      "step": 3606
+    },
+    {
+      "epoch": 1.8487954894925678,
+      "grad_norm": 0.5494952650818208,
+      "learning_rate": 2.9842796411734263e-07,
+      "loss": 0.227,
+      "step": 3607
+    },
+    {
+      "epoch": 1.849308047155305,
+      "grad_norm": 0.5195937611652829,
+      "learning_rate": 2.964181912495656e-07,
+      "loss": 0.2316,
+      "step": 3608
+    },
+    {
+      "epoch": 1.8498206048180421,
+      "grad_norm": 0.5161330087327759,
+      "learning_rate": 2.9441510689418806e-07,
+      "loss": 0.2427,
+      "step": 3609
+    },
+    {
+      "epoch": 1.850333162480779,
+      "grad_norm": 0.5143560084281018,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.2602,
+      "step": 3610
+    },
+    {
+      "epoch": 1.850845720143516,
+      "grad_norm": 0.5303469243084885,
+      "learning_rate": 2.9042900923879515e-07,
+      "loss": 0.2277,
+      "step": 3611
+    },
+    {
+      "epoch": 1.8513582778062532,
+      "grad_norm": 0.524607843981106,
+      "learning_rate": 2.884459986863264e-07,
+      "loss": 0.2335,
+      "step": 3612
+    },
+    {
+      "epoch": 1.8518708354689903,
+      "grad_norm": 0.5518500363126033,
+      "learning_rate": 2.864696821413548e-07,
+      "loss": 0.262,
+      "step": 3613
+    },
+    {
+      "epoch": 1.8523833931317273,
+      "grad_norm": 0.5453478301112678,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.275,
+      "step": 3614
+    },
+    {
+      "epoch": 1.8528959507944642,
+      "grad_norm": 0.497773415628103,
+      "learning_rate": 2.825371365182505e-07,
+      "loss": 0.2144,
+      "step": 3615
+    },
+    {
+      "epoch": 1.8534085084572014,
+      "grad_norm": 0.5349206799611417,
+      "learning_rate": 2.805809101507539e-07,
+      "loss": 0.2356,
+      "step": 3616
+    },
+    {
+      "epoch": 1.8539210661199386,
+      "grad_norm": 0.5360090512459518,
+      "learning_rate": 2.786313832120258e-07,
+      "loss": 0.2491,
+      "step": 3617
+    },
+    {
+      "epoch": 1.8544336237826755,
+      "grad_norm": 0.5681786596777532,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2478,
+      "step": 3618
+    },
+    {
+      "epoch": 1.8549461814454125,
+      "grad_norm": 0.5469182509000368,
+      "learning_rate": 2.7475243299135315e-07,
+      "loss": 0.2443,
+      "step": 3619
+    },
+    {
+      "epoch": 1.8554587391081496,
+      "grad_norm": 0.47457190636541086,
+      "learning_rate": 2.7282301238310326e-07,
+      "loss": 0.2135,
+      "step": 3620
+    },
+    {
+      "epoch": 1.8559712967708868,
+      "grad_norm": 0.4977581103568282,
+      "learning_rate": 2.7090029655100434e-07,
+      "loss": 0.2531,
+      "step": 3621
+    },
+    {
+      "epoch": 1.8564838544336237,
+      "grad_norm": 0.5017321197531781,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.2201,
+      "step": 3622
+    },
+    {
+      "epoch": 1.8569964120963607,
+      "grad_norm": 0.5255717043733541,
+      "learning_rate": 2.670749845118281e-07,
+      "loss": 0.2437,
+      "step": 3623
+    },
+    {
+      "epoch": 1.8575089697590978,
+      "grad_norm": 0.4894777130372874,
+      "learning_rate": 2.651723909414694e-07,
+      "loss": 0.2455,
+      "step": 3624
+    },
+    {
+      "epoch": 1.858021527421835,
+      "grad_norm": 0.4924245526958838,
+      "learning_rate": 2.6327650742070665e-07,
+      "loss": 0.2323,
+      "step": 3625
+    },
+    {
+      "epoch": 1.858534085084572,
+      "grad_norm": 0.5429144186396638,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.2455,
+      "step": 3626
+    },
+    {
+      "epoch": 1.859046642747309,
+      "grad_norm": 0.5243639657218363,
+      "learning_rate": 2.595048757505392e-07,
+      "loss": 0.2443,
+      "step": 3627
+    },
+    {
+      "epoch": 1.859559200410046,
+      "grad_norm": 0.5052415741364668,
+      "learning_rate": 2.576291302008571e-07,
+      "loss": 0.2282,
+      "step": 3628
+    },
+    {
+      "epoch": 1.8600717580727832,
+      "grad_norm": 0.5423937643280791,
+      "learning_rate": 2.5576009990020856e-07,
+      "loss": 0.2559,
+      "step": 3629
+    },
+    {
+      "epoch": 1.8605843157355202,
+      "grad_norm": 0.5275144895605228,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.2332,
+      "step": 3630
+    },
+    {
+      "epoch": 1.8610968733982574,
+      "grad_norm": 0.5154107423033625,
+      "learning_rate": 2.5204219019455243e-07,
+      "loss": 0.2426,
+      "step": 3631
+    },
+    {
+      "epoch": 1.8616094310609945,
+      "grad_norm": 0.5259871595386015,
+      "learning_rate": 2.5019331335223383e-07,
+      "loss": 0.2297,
+      "step": 3632
+    },
+    {
+      "epoch": 1.8621219887237315,
+      "grad_norm": 0.5091192215264454,
+      "learning_rate": 2.483511568843322e-07,
+      "loss": 0.2266,
+      "step": 3633
+    },
+    {
+      "epoch": 1.8626345463864684,
+      "grad_norm": 0.4736432986858121,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.1836,
+      "step": 3634
+    },
+    {
+      "epoch": 1.8631471040492056,
+      "grad_norm": 0.5015694482181836,
+      "learning_rate": 2.4468701014621177e-07,
+      "loss": 0.2274,
+      "step": 3635
+    },
+    {
+      "epoch": 1.8636596617119427,
+      "grad_norm": 0.5391157054190374,
+      "learning_rate": 2.42865022401626e-07,
+      "loss": 0.2313,
+      "step": 3636
+    },
+    {
+      "epoch": 1.8641722193746797,
+      "grad_norm": 0.5008010715957326,
+      "learning_rate": 2.4104976008272217e-07,
+      "loss": 0.223,
+      "step": 3637
+    },
+    {
+      "epoch": 1.8646847770374166,
+      "grad_norm": 0.516228998893934,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2467,
+      "step": 3638
+    },
+    {
+      "epoch": 1.8651973347001538,
+      "grad_norm": 0.5916604928670749,
+      "learning_rate": 2.3743941672223825e-07,
+      "loss": 0.2653,
+      "step": 3639
+    },
+    {
+      "epoch": 1.865709892362891,
+      "grad_norm": 0.5172648375632176,
+      "learning_rate": 2.3564433816920528e-07,
+      "loss": 0.2384,
+      "step": 3640
+    },
+    {
+      "epoch": 1.866222450025628,
+      "grad_norm": 0.5403172024332535,
+      "learning_rate": 2.3385599001894965e-07,
+      "loss": 0.2753,
+      "step": 3641
+    },
+    {
+      "epoch": 1.8667350076883649,
+      "grad_norm": 0.49266592912790946,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2285,
+      "step": 3642
+    },
+    {
+      "epoch": 1.867247565351102,
+      "grad_norm": 0.5252431159509895,
+      "learning_rate": 2.30299489852831e-07,
+      "loss": 0.246,
+      "step": 3643
+    },
+    {
+      "epoch": 1.8677601230138392,
+      "grad_norm": 0.5088560485751302,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.2235,
+      "step": 3644
+    },
+    {
+      "epoch": 1.8682726806765761,
+      "grad_norm": 0.5662875079594808,
+      "learning_rate": 2.2676992602962078e-07,
+      "loss": 0.2582,
+      "step": 3645
+    },
+    {
+      "epoch": 1.868785238339313,
+      "grad_norm": 0.5715373285129906,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2456,
+      "step": 3646
+    },
+    {
+      "epoch": 1.8692977960020503,
+      "grad_norm": 0.550299914425732,
+      "learning_rate": 2.2326730828078792e-07,
+      "loss": 0.253,
+      "step": 3647
+    },
+    {
+      "epoch": 1.8698103536647874,
+      "grad_norm": 0.5286079067825049,
+      "learning_rate": 2.215261072050323e-07,
+      "loss": 0.2258,
+      "step": 3648
+    },
+    {
+      "epoch": 1.8703229113275244,
+      "grad_norm": 0.5445664465612351,
+      "learning_rate": 2.1979164626350745e-07,
+      "loss": 0.2529,
+      "step": 3649
+    },
+    {
+      "epoch": 1.8708354689902613,
+      "grad_norm": 0.5382423647622107,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2652,
+      "step": 3650
+    },
+    {
+      "epoch": 1.8713480266529985,
+      "grad_norm": 0.5044606499146652,
+      "learning_rate": 2.1634294956063284e-07,
+      "loss": 0.2129,
+      "step": 3651
+    },
+    {
+      "epoch": 1.8718605843157357,
+      "grad_norm": 0.5278951667195844,
+      "learning_rate": 2.1462871617641379e-07,
+      "loss": 0.2438,
+      "step": 3652
+    },
+    {
+      "epoch": 1.8723731419784726,
+      "grad_norm": 0.49667368127591094,
+      "learning_rate": 2.129212276806769e-07,
+      "loss": 0.2237,
+      "step": 3653
+    },
+    {
+      "epoch": 1.8728856996412095,
+      "grad_norm": 0.5424607878637132,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.2549,
+      "step": 3654
+    },
+    {
+      "epoch": 1.8733982573039467,
+      "grad_norm": 0.5374925784535272,
+      "learning_rate": 2.0952649005777093e-07,
+      "loss": 0.2638,
+      "step": 3655
+    },
+    {
+      "epoch": 1.8739108149666839,
+      "grad_norm": 0.551695138455092,
+      "learning_rate": 2.0783924327053562e-07,
+      "loss": 0.2372,
+      "step": 3656
+    },
+    {
+      "epoch": 1.8744233726294208,
+      "grad_norm": 0.5362821912919683,
+      "learning_rate": 2.061587460516501e-07,
+      "loss": 0.2299,
+      "step": 3657
+    },
+    {
+      "epoch": 1.8749359302921578,
+      "grad_norm": 0.5829639440840697,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.2468,
+      "step": 3658
+    },
+    {
+      "epoch": 1.875448487954895,
+      "grad_norm": 0.4849818660065177,
+      "learning_rate": 2.0281800494762803e-07,
+      "loss": 0.2311,
+      "step": 3659
+    },
+    {
+      "epoch": 1.875961045617632,
+      "grad_norm": 0.5050948211687983,
+      "learning_rate": 2.0115776336520622e-07,
+      "loss": 0.2007,
+      "step": 3660
+    },
+    {
+      "epoch": 1.876473603280369,
+      "grad_norm": 0.5425007028393566,
+      "learning_rate": 1.9950427595656352e-07,
+      "loss": 0.239,
+      "step": 3661
+    },
+    {
+      "epoch": 1.876986160943106,
+      "grad_norm": 0.5192212365378203,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.2394,
+      "step": 3662
+    },
+    {
+      "epoch": 1.8774987186058432,
+      "grad_norm": 0.5448015629368281,
+      "learning_rate": 1.962175682148426e-07,
+      "loss": 0.2343,
+      "step": 3663
+    },
+    {
+      "epoch": 1.8780112762685803,
+      "grad_norm": 0.4988808102734002,
+      "learning_rate": 1.9458435014723443e-07,
+      "loss": 0.2305,
+      "step": 3664
+    },
+    {
+      "epoch": 1.8785238339313173,
+      "grad_norm": 0.5187402517761801,
+      "learning_rate": 1.9295789078434658e-07,
+      "loss": 0.2219,
+      "step": 3665
+    },
+    {
+      "epoch": 1.8790363915940542,
+      "grad_norm": 0.4886846123215928,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2318,
+      "step": 3666
+    },
+    {
+      "epoch": 1.8795489492567914,
+      "grad_norm": 0.5242699183968589,
+      "learning_rate": 1.8972525265243512e-07,
+      "loss": 0.2408,
+      "step": 3667
+    },
+    {
+      "epoch": 1.8800615069195286,
+      "grad_norm": 0.5010707397825868,
+      "learning_rate": 1.881190761116125e-07,
+      "loss": 0.2391,
+      "step": 3668
+    },
+    {
+      "epoch": 1.8805740645822655,
+      "grad_norm": 0.4858185730732322,
+      "learning_rate": 1.8651966273191103e-07,
+      "loss": 0.2178,
+      "step": 3669
+    },
+    {
+      "epoch": 1.8810866222450024,
+      "grad_norm": 0.49194475903318535,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.2366,
+      "step": 3670
+    },
+    {
+      "epoch": 1.8815991799077396,
+      "grad_norm": 0.5430138410636419,
+      "learning_rate": 1.8334112986100994e-07,
+      "loss": 0.2555,
+      "step": 3671
+    },
+    {
+      "epoch": 1.8821117375704768,
+      "grad_norm": 0.5050331761540202,
+      "learning_rate": 1.8176201256071556e-07,
+      "loss": 0.2114,
+      "step": 3672
+    },
+    {
+      "epoch": 1.8826242952332137,
+      "grad_norm": 0.5270834556159357,
+      "learning_rate": 1.8018966280336058e-07,
+      "loss": 0.2439,
+      "step": 3673
+    },
+    {
+      "epoch": 1.8831368528959507,
+      "grad_norm": 0.4955691099294059,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.199,
+      "step": 3674
+    },
+    {
+      "epoch": 1.8836494105586878,
+      "grad_norm": 0.5041808082401743,
+      "learning_rate": 1.7706527024796805e-07,
+      "loss": 0.2443,
+      "step": 3675
+    },
+    {
+      "epoch": 1.884161968221425,
+      "grad_norm": 0.5256068183141756,
+      "learning_rate": 1.7551322960352334e-07,
+      "loss": 0.2443,
+      "step": 3676
+    },
+    {
+      "epoch": 1.884674525884162,
+      "grad_norm": 0.5202967211703224,
+      "learning_rate": 1.7396796080919599e-07,
+      "loss": 0.2322,
+      "step": 3677
+    },
+    {
+      "epoch": 1.885187083546899,
+      "grad_norm": 0.5483559865025415,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.2635,
+      "step": 3678
+    },
+    {
+      "epoch": 1.885699641209636,
+      "grad_norm": 0.5123270814955466,
+      "learning_rate": 1.708977430267278e-07,
+      "loss": 0.2361,
+      "step": 3679
+    },
+    {
+      "epoch": 1.8862121988723732,
+      "grad_norm": 0.5126476347599768,
+      "learning_rate": 1.6937279615483748e-07,
+      "loss": 0.2399,
+      "step": 3680
+    },
+    {
+      "epoch": 1.8867247565351102,
+      "grad_norm": 0.483302742038732,
+      "learning_rate": 1.6785462536555775e-07,
+      "loss": 0.2383,
+      "step": 3681
+    },
+    {
+      "epoch": 1.8872373141978471,
+      "grad_norm": 0.4788594644907392,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.2162,
+      "step": 3682
+    },
+    {
+      "epoch": 1.8877498718605843,
+      "grad_norm": 0.5074310321090203,
+      "learning_rate": 1.6483861621595875e-07,
+      "loss": 0.2393,
+      "step": 3683
+    },
+    {
+      "epoch": 1.8882624295233215,
+      "grad_norm": 0.5233988190649042,
+      "learning_rate": 1.6334077993452324e-07,
+      "loss": 0.2379,
+      "step": 3684
+    },
+    {
+      "epoch": 1.8887749871860584,
+      "grad_norm": 0.5341533102616569,
+      "learning_rate": 1.6184972389346355e-07,
+      "loss": 0.2236,
+      "step": 3685
+    },
+    {
+      "epoch": 1.8892875448487954,
+      "grad_norm": 0.4720445951608434,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.2087,
+      "step": 3686
+    },
+    {
+      "epoch": 1.8898001025115325,
+      "grad_norm": 0.514800832808244,
+      "learning_rate": 1.5888795663883904e-07,
+      "loss": 0.2497,
+      "step": 3687
+    },
+    {
+      "epoch": 1.8903126601742697,
+      "grad_norm": 0.5347055777010631,
+      "learning_rate": 1.57417247466769e-07,
+      "loss": 0.2557,
+      "step": 3688
+    },
+    {
+      "epoch": 1.8908252178370066,
+      "grad_norm": 0.5149378734119735,
+      "learning_rate": 1.559533226180654e-07,
+      "loss": 0.2495,
+      "step": 3689
+    },
+    {
+      "epoch": 1.8913377754997436,
+      "grad_norm": 0.5405744428373188,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2285,
+      "step": 3690
+    },
+    {
+      "epoch": 1.8918503331624807,
+      "grad_norm": 0.5110964568886214,
+      "learning_rate": 1.5304582992231255e-07,
+      "loss": 0.2276,
+      "step": 3691
+    },
+    {
+      "epoch": 1.892362890825218,
+      "grad_norm": 0.5321895319327338,
+      "learning_rate": 1.5160226407934797e-07,
+      "loss": 0.2557,
+      "step": 3692
+    },
+    {
+      "epoch": 1.8928754484879549,
+      "grad_norm": 0.48277517414548826,
+      "learning_rate": 1.5016548656791697e-07,
+      "loss": 0.2253,
+      "step": 3693
+    },
+    {
+      "epoch": 1.8933880061506918,
+      "grad_norm": 0.537549651380978,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.2391,
+      "step": 3694
+    },
+    {
+      "epoch": 1.893900563813429,
+      "grad_norm": 0.5216383210558537,
+      "learning_rate": 1.4731230049636503e-07,
+      "loss": 0.2479,
+      "step": 3695
+    },
+    {
+      "epoch": 1.8944131214761661,
+      "grad_norm": 0.49880887389152395,
+      "learning_rate": 1.4589589390289427e-07,
+      "loss": 0.2085,
+      "step": 3696
+    },
+    {
+      "epoch": 1.894925679138903,
+      "grad_norm": 0.5284696855334037,
+      "learning_rate": 1.4448627957426186e-07,
+      "loss": 0.245,
+      "step": 3697
+    },
+    {
+      "epoch": 1.89543823680164,
+      "grad_norm": 0.5291415172588587,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.2367,
+      "step": 3698
+    },
+    {
+      "epoch": 1.8959507944643772,
+      "grad_norm": 0.5322757912891635,
+      "learning_rate": 1.41687431593317e-07,
+      "loss": 0.253,
+      "step": 3699
+    },
+    {
+      "epoch": 1.8964633521271144,
+      "grad_norm": 0.5116412110993076,
+      "learning_rate": 1.4029819987020466e-07,
+      "loss": 0.245,
+      "step": 3700
+    },
+    {
+      "epoch": 1.8969759097898513,
+      "grad_norm": 0.5185371147087092,
+      "learning_rate": 1.3891576427032317e-07,
+      "loss": 0.2701,
+      "step": 3701
+    },
+    {
+      "epoch": 1.8974884674525883,
+      "grad_norm": 0.4962009339427864,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.2163,
+      "step": 3702
+    },
+    {
+      "epoch": 1.8980010251153254,
+      "grad_norm": 0.4873487599920759,
+      "learning_rate": 1.3617128524712643e-07,
+      "loss": 0.2537,
+      "step": 3703
+    },
+    {
+      "epoch": 1.8985135827780626,
+      "grad_norm": 0.5545726164844572,
+      "learning_rate": 1.348092437155346e-07,
+      "loss": 0.2562,
+      "step": 3704
+    },
+    {
+      "epoch": 1.8990261404407995,
+      "grad_norm": 0.5066834450120293,
+      "learning_rate": 1.334540020906172e-07,
+      "loss": 0.239,
+      "step": 3705
+    },
+    {
+      "epoch": 1.8995386981035365,
+      "grad_norm": 0.5461540868091582,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2355,
+      "step": 3706
+    },
+    {
+      "epoch": 1.9000512557662739,
+      "grad_norm": 0.5032584668786151,
+      "learning_rate": 1.3076392229270041e-07,
+      "loss": 0.2373,
+      "step": 3707
+    },
+    {
+      "epoch": 1.9005638134290108,
+      "grad_norm": 0.5183952466629872,
+      "learning_rate": 1.2942908597392555e-07,
+      "loss": 0.2394,
+      "step": 3708
+    },
+    {
+      "epoch": 1.9010763710917478,
+      "grad_norm": 0.5381771253526807,
+      "learning_rate": 1.2810105327027754e-07,
+      "loss": 0.2622,
+      "step": 3709
+    },
+    {
+      "epoch": 1.901588928754485,
+      "grad_norm": 0.5446890637891398,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2385,
+      "step": 3710
+    },
+    {
+      "epoch": 1.902101486417222,
+      "grad_norm": 0.5280243114664113,
+      "learning_rate": 1.2546540236522575e-07,
+      "loss": 0.2336,
+      "step": 3711
+    },
+    {
+      "epoch": 1.902614044079959,
+      "grad_norm": 0.561626210124161,
+      "learning_rate": 1.2415778598053207e-07,
+      "loss": 0.2598,
+      "step": 3712
+    },
+    {
+      "epoch": 1.903126601742696,
+      "grad_norm": 0.5068071528662733,
+      "learning_rate": 1.2285697684438104e-07,
+      "loss": 0.2078,
+      "step": 3713
+    },
+    {
+      "epoch": 1.9036391594054332,
+      "grad_norm": 0.4982388156568949,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.2352,
+      "step": 3714
+    },
+    {
+      "epoch": 1.9041517170681703,
+      "grad_norm": 0.5002246813907953,
+      "learning_rate": 1.20275783899515e-07,
+      "loss": 0.2457,
+      "step": 3715
+    },
+    {
+      "epoch": 1.9046642747309073,
+      "grad_norm": 0.5776363412426483,
+      "learning_rate": 1.1899540186997127e-07,
+      "loss": 0.255,
+      "step": 3716
+    },
+    {
+      "epoch": 1.9051768323936442,
+      "grad_norm": 0.5391985276924478,
+      "learning_rate": 1.1772183064731268e-07,
+      "loss": 0.2471,
+      "step": 3717
+    },
+    {
+      "epoch": 1.9056893900563814,
+      "grad_norm": 0.5188436904033902,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2265,
+      "step": 3718
+    },
+    {
+      "epoch": 1.9062019477191186,
+      "grad_norm": 0.467234473016996,
+      "learning_rate": 1.1519512412935807e-07,
+      "loss": 0.225,
+      "step": 3719
+    },
+    {
+      "epoch": 1.9067145053818555,
+      "grad_norm": 0.5188781303884137,
+      "learning_rate": 1.1394199057567667e-07,
+      "loss": 0.2214,
+      "step": 3720
+    },
+    {
+      "epoch": 1.9072270630445924,
+      "grad_norm": 0.514112349462239,
+      "learning_rate": 1.1269567131211079e-07,
+      "loss": 0.2322,
+      "step": 3721
+    },
+    {
+      "epoch": 1.9077396207073296,
+      "grad_norm": 0.5123248499235125,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.2252,
+      "step": 3722
+    },
+    {
+      "epoch": 1.9082521783700668,
+      "grad_norm": 0.5111898952284268,
+      "learning_rate": 1.1022347908689057e-07,
+      "loss": 0.2489,
+      "step": 3723
+    },
+    {
+      "epoch": 1.9087647360328037,
+      "grad_norm": 0.49416293034879144,
+      "learning_rate": 1.0899760782927649e-07,
+      "loss": 0.2288,
+      "step": 3724
+    },
+    {
+      "epoch": 1.9092772936955407,
+      "grad_norm": 0.5285411901693654,
+      "learning_rate": 1.0777855426985395e-07,
+      "loss": 0.2372,
+      "step": 3725
+    },
+    {
+      "epoch": 1.9097898513582778,
+      "grad_norm": 0.49719997005865424,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2253,
+      "step": 3726
+    },
+    {
+      "epoch": 1.910302409021015,
+      "grad_norm": 0.485012987820869,
+      "learning_rate": 1.0536090360198092e-07,
+      "loss": 0.226,
+      "step": 3727
+    },
+    {
+      "epoch": 1.910814966683752,
+      "grad_norm": 0.47234017699186925,
+      "learning_rate": 1.0416230815997297e-07,
+      "loss": 0.2089,
+      "step": 3728
+    },
+    {
+      "epoch": 1.911327524346489,
+      "grad_norm": 0.5338996840352941,
+      "learning_rate": 1.029705337490472e-07,
+      "loss": 0.2587,
+      "step": 3729
+    },
+    {
+      "epoch": 1.911840082009226,
+      "grad_norm": 0.4794154797738766,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2198,
+      "step": 3730
+    },
+    {
+      "epoch": 1.9123526396719632,
+      "grad_norm": 0.5442020561921698,
+      "learning_rate": 1.0060745130161753e-07,
+      "loss": 0.2283,
+      "step": 3731
+    },
+    {
+      "epoch": 1.9128651973347002,
+      "grad_norm": 0.4931391319598964,
+      "learning_rate": 9.943614489394626e-08,
+      "loss": 0.2261,
+      "step": 3732
+    },
+    {
+      "epoch": 1.9133777549974371,
+      "grad_norm": 0.5459244919888582,
+      "learning_rate": 9.827166277501909e-08,
+      "loss": 0.27,
+      "step": 3733
+    },
+    {
+      "epoch": 1.9138903126601743,
+      "grad_norm": 0.5265788049446588,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.2461,
+      "step": 3734
+    },
+    {
+      "epoch": 1.9144028703229115,
+      "grad_norm": 0.5059790820174124,
+      "learning_rate": 9.596317460932703e-08,
+      "loss": 0.2254,
+      "step": 3735
+    },
+    {
+      "epoch": 1.9149154279856484,
+      "grad_norm": 0.4394651109757752,
+      "learning_rate": 9.481917015376262e-08,
+      "loss": 0.2025,
+      "step": 3736
+    },
+    {
+      "epoch": 1.9154279856483853,
+      "grad_norm": 0.5501124592078497,
+      "learning_rate": 9.368199316934446e-08,
+      "loss": 0.2214,
+      "step": 3737
+    },
+    {
+      "epoch": 1.9159405433111225,
+      "grad_norm": 0.5902592463659867,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.2559,
+      "step": 3738
+    },
+    {
+      "epoch": 1.9164531009738597,
+      "grad_norm": 0.49360302855384686,
+      "learning_rate": 9.142812474458806e-08,
+      "loss": 0.2255,
+      "step": 3739
+    },
+    {
+      "epoch": 1.9169656586365966,
+      "grad_norm": 0.5112248134242948,
+      "learning_rate": 9.031143485780159e-08,
+      "loss": 0.2711,
+      "step": 3740
+    },
+    {
+      "epoch": 1.9174782162993336,
+      "grad_norm": 0.5116828871174915,
+      "learning_rate": 8.92015755492659e-08,
+      "loss": 0.2212,
+      "step": 3741
+    },
+    {
+      "epoch": 1.9179907739620707,
+      "grad_norm": 0.564077059443026,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.2062,
+      "step": 3742
+    },
+    {
+      "epoch": 1.918503331624808,
+      "grad_norm": 0.5764444718436824,
+      "learning_rate": 8.700235172226845e-08,
+      "loss": 0.2557,
+      "step": 3743
+    },
+    {
+      "epoch": 1.9190158892875449,
+      "grad_norm": 0.539661083417502,
+      "learning_rate": 8.591298871969411e-08,
+      "loss": 0.2389,
+      "step": 3744
+    },
+    {
+      "epoch": 1.9195284469502818,
+      "grad_norm": 0.5341295907400672,
+      "learning_rate": 8.483045932714318e-08,
+      "loss": 0.2582,
+      "step": 3745
+    },
+    {
+      "epoch": 1.920041004613019,
+      "grad_norm": 0.47054245273204826,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.1905,
+      "step": 3746
+    },
+    {
+      "epoch": 1.9205535622757561,
+      "grad_norm": 0.4901806000263513,
+      "learning_rate": 8.268590435208002e-08,
+      "loss": 0.2397,
+      "step": 3747
+    },
+    {
+      "epoch": 1.921066119938493,
+      "grad_norm": 0.5936564075089582,
+      "learning_rate": 8.162388024777202e-08,
+      "loss": 0.284,
+      "step": 3748
+    },
+    {
+      "epoch": 1.92157867760123,
+      "grad_norm": 0.5332070652554551,
+      "learning_rate": 8.05686927098981e-08,
+      "loss": 0.268,
+      "step": 3749
+    },
+    {
+      "epoch": 1.9220912352639672,
+      "grad_norm": 0.5023933063439314,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2339,
+      "step": 3750
+    },
+    {
+      "epoch": 1.9226037929267044,
+      "grad_norm": 0.5687989356934711,
+      "learning_rate": 7.847883023802793e-08,
+      "loss": 0.2736,
+      "step": 3751
+    },
+    {
+      "epoch": 1.9231163505894413,
+      "grad_norm": 0.5284049234496726,
+      "learning_rate": 7.744415674453721e-08,
+      "loss": 0.216,
+      "step": 3752
+    },
+    {
+      "epoch": 1.9236289082521782,
+      "grad_norm": 0.5192400950214053,
+      "learning_rate": 7.641632269849263e-08,
+      "loss": 0.2494,
+      "step": 3753
+    },
+    {
+      "epoch": 1.9241414659149154,
+      "grad_norm": 0.4592205424192899,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.2082,
+      "step": 3754
+    },
+    {
+      "epoch": 1.9246540235776526,
+      "grad_norm": 0.9523850430055044,
+      "learning_rate": 7.438117577789561e-08,
+      "loss": 0.2182,
+      "step": 3755
+    },
+    {
+      "epoch": 1.9251665812403895,
+      "grad_norm": 0.5081871353866111,
+      "learning_rate": 7.33738643061377e-08,
+      "loss": 0.2169,
+      "step": 3756
+    },
+    {
+      "epoch": 1.9256791389031265,
+      "grad_norm": 0.5085280057997124,
+      "learning_rate": 7.237339508740615e-08,
+      "loss": 0.2228,
+      "step": 3757
+    },
+    {
+      "epoch": 1.9261916965658636,
+      "grad_norm": 0.507204775383841,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.2285,
+      "step": 3758
+    },
+    {
+      "epoch": 1.9267042542286008,
+      "grad_norm": 0.49055181253838404,
+      "learning_rate": 7.039298616273393e-08,
+      "loss": 0.2207,
+      "step": 3759
+    },
+    {
+      "epoch": 1.9272168118913378,
+      "grad_norm": 0.5229614737256711,
+      "learning_rate": 6.941304782185577e-08,
+      "loss": 0.2643,
+      "step": 3760
+    },
+    {
+      "epoch": 1.9277293695540747,
+      "grad_norm": 0.5615849304342084,
+      "learning_rate": 6.843995446412788e-08,
+      "loss": 0.24,
+      "step": 3761
+    },
+    {
+      "epoch": 1.9282419272168119,
+      "grad_norm": 0.5533112668168508,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.2466,
+      "step": 3762
+    },
+    {
+      "epoch": 1.928754484879549,
+      "grad_norm": 0.5207827042515006,
+      "learning_rate": 6.651430537635284e-08,
+      "loss": 0.2253,
+      "step": 3763
+    },
+    {
+      "epoch": 1.929267042542286,
+      "grad_norm": 0.5295237254988437,
+      "learning_rate": 6.556175097362061e-08,
+      "loss": 0.2495,
+      "step": 3764
+    },
+    {
+      "epoch": 1.929779600205023,
+      "grad_norm": 0.4775929150055672,
+      "learning_rate": 6.461604420867296e-08,
+      "loss": 0.1933,
+      "step": 3765
+    },
+    {
+      "epoch": 1.93029215786776,
+      "grad_norm": 0.522905681323147,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.2345,
+      "step": 3766
+    },
+    {
+      "epoch": 1.9308047155304973,
+      "grad_norm": 0.5080758806282974,
+      "learning_rate": 6.274517619484499e-08,
+      "loss": 0.2635,
+      "step": 3767
+    },
+    {
+      "epoch": 1.9313172731932342,
+      "grad_norm": 0.4910280126604571,
+      "learning_rate": 6.182001623552202e-08,
+      "loss": 0.2187,
+      "step": 3768
+    },
+    {
+      "epoch": 1.9318298308559712,
+      "grad_norm": 0.48997007698919276,
+      "learning_rate": 6.090170649309502e-08,
+      "loss": 0.2356,
+      "step": 3769
+    },
+    {
+      "epoch": 1.9323423885187083,
+      "grad_norm": 0.49984382042900266,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.241,
+      "step": 3770
+    },
+    {
+      "epoch": 1.9328549461814455,
+      "grad_norm": 0.5256545773886598,
+      "learning_rate": 5.9085640186111735e-08,
+      "loss": 0.2515,
+      "step": 3771
+    },
+    {
+      "epoch": 1.9333675038441824,
+      "grad_norm": 0.5363632929159456,
+      "learning_rate": 5.8187884873338595e-08,
+      "loss": 0.2249,
+      "step": 3772
+    },
+    {
+      "epoch": 1.9338800615069194,
+      "grad_norm": 0.546155081704222,
+      "learning_rate": 5.7296982281026534e-08,
+      "loss": 0.2417,
+      "step": 3773
+    },
+    {
+      "epoch": 1.9343926191696565,
+      "grad_norm": 0.505638325064641,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.2202,
+      "step": 3774
+    },
+    {
+      "epoch": 1.9349051768323937,
+      "grad_norm": 0.5500437204609934,
+      "learning_rate": 5.5535737709404566e-08,
+      "loss": 0.254,
+      "step": 3775
+    },
+    {
+      "epoch": 1.9354177344951307,
+      "grad_norm": 0.5398106644079539,
+      "learning_rate": 5.466539694408912e-08,
+      "loss": 0.2376,
+      "step": 3776
+    },
+    {
+      "epoch": 1.9359302921578676,
+      "grad_norm": 0.49461967003963536,
+      "learning_rate": 5.380191132722701e-08,
+      "loss": 0.2347,
+      "step": 3777
+    },
+    {
+      "epoch": 1.9364428498206048,
+      "grad_norm": 0.520194763850452,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.2336,
+      "step": 3778
+    },
+    {
+      "epoch": 1.936955407483342,
+      "grad_norm": 0.46445820452495995,
+      "learning_rate": 5.209550791487883e-08,
+      "loss": 0.1949,
+      "step": 3779
+    },
+    {
+      "epoch": 1.9374679651460789,
+      "grad_norm": 0.5185305814109444,
+      "learning_rate": 5.125259129558857e-08,
+      "loss": 0.2242,
+      "step": 3780
+    },
+    {
+      "epoch": 1.9379805228088158,
+      "grad_norm": 0.5473330353579338,
+      "learning_rate": 5.041653217713993e-08,
+      "loss": 0.2524,
+      "step": 3781
+    },
+    {
+      "epoch": 1.938493080471553,
+      "grad_norm": 0.5132361954233096,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.2267,
+      "step": 3782
+    },
+    {
+      "epoch": 1.9390056381342902,
+      "grad_norm": 0.5657336744468553,
+      "learning_rate": 4.876498874316515e-08,
+      "loss": 0.2761,
+      "step": 3783
+    },
+    {
+      "epoch": 1.9395181957970271,
+      "grad_norm": 0.5340092801543871,
+      "learning_rate": 4.7949505566018405e-08,
+      "loss": 0.2504,
+      "step": 3784
+    },
+    {
+      "epoch": 1.9400307534597643,
+      "grad_norm": 0.5027878735840453,
+      "learning_rate": 4.714088216647428e-08,
+      "loss": 0.2398,
+      "step": 3785
+    },
+    {
+      "epoch": 1.9405433111225014,
+      "grad_norm": 0.5142810007234186,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.2543,
+      "step": 3786
+    },
+    {
+      "epoch": 1.9410558687852384,
+      "grad_norm": 0.4989511301751721,
+      "learning_rate": 4.5544216924944265e-08,
+      "loss": 0.2475,
+      "step": 3787
+    },
+    {
+      "epoch": 1.9415684264479753,
+      "grad_norm": 0.5517001490216655,
+      "learning_rate": 4.475617618351136e-08,
+      "loss": 0.2361,
+      "step": 3788
+    },
+    {
+      "epoch": 1.9420809841107125,
+      "grad_norm": 0.5051106772953177,
+      "learning_rate": 4.3974997420789257e-08,
+      "loss": 0.2129,
+      "step": 3789
+    },
+    {
+      "epoch": 1.9425935417734497,
+      "grad_norm": 0.5239142532090092,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.2299,
+      "step": 3790
+    },
+    {
+      "epoch": 1.9431060994361866,
+      "grad_norm": 0.5033647108795073,
+      "learning_rate": 4.243322798055172e-08,
+      "loss": 0.2246,
+      "step": 3791
+    },
+    {
+      "epoch": 1.9436186570989236,
+      "grad_norm": 0.5276452281271756,
+      "learning_rate": 4.167263836575286e-08,
+      "loss": 0.2467,
+      "step": 3792
+    },
+    {
+      "epoch": 1.9441312147616607,
+      "grad_norm": 0.5774410025676607,
+      "learning_rate": 4.091891285509353e-08,
+      "loss": 0.2745,
+      "step": 3793
+    },
+    {
+      "epoch": 1.944643772424398,
+      "grad_norm": 0.5133462624283267,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.2541,
+      "step": 3794
+    },
+    {
+      "epoch": 1.9451563300871348,
+      "grad_norm": 0.5109933792037963,
+      "learning_rate": 3.943205621957713e-08,
+      "loss": 0.2318,
+      "step": 3795
+    },
+    {
+      "epoch": 1.9456688877498718,
+      "grad_norm": 0.5399553106285906,
+      "learning_rate": 3.869892611958692e-08,
+      "loss": 0.2657,
+      "step": 3796
+    },
+    {
+      "epoch": 1.946181445412609,
+      "grad_norm": 0.49378636961551536,
+      "learning_rate": 3.797266217346329e-08,
+      "loss": 0.2019,
+      "step": 3797
+    },
+    {
+      "epoch": 1.9466940030753461,
+      "grad_norm": 0.5307940043185995,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.2155,
+      "step": 3798
+    },
+    {
+      "epoch": 1.947206560738083,
+      "grad_norm": 0.4870172558295633,
+      "learning_rate": 3.654073474049225e-08,
+      "loss": 0.2408,
+      "step": 3799
+    },
+    {
+      "epoch": 1.94771911840082,
+      "grad_norm": 0.5116754332836094,
+      "learning_rate": 3.583507224064642e-08,
+      "loss": 0.2379,
+      "step": 3800
+    },
+    {
+      "epoch": 1.9482316760635572,
+      "grad_norm": 0.5232988364882032,
+      "learning_rate": 3.5136277868674795e-08,
+      "loss": 0.2253,
+      "step": 3801
+    },
+    {
+      "epoch": 1.9487442337262944,
+      "grad_norm": 0.4642136613239574,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2088,
+      "step": 3802
+    },
+    {
+      "epoch": 1.9492567913890313,
+      "grad_norm": 0.5335879083255208,
+      "learning_rate": 3.3759295430281226e-08,
+      "loss": 0.2759,
+      "step": 3803
+    },
+    {
+      "epoch": 1.9497693490517682,
+      "grad_norm": 0.5062588187418063,
+      "learning_rate": 3.308110831299116e-08,
+      "loss": 0.2266,
+      "step": 3804
+    },
+    {
+      "epoch": 1.9502819067145054,
+      "grad_norm": 0.5114952306570131,
+      "learning_rate": 3.240979122183352e-08,
+      "loss": 0.2239,
+      "step": 3805
+    },
+    {
+      "epoch": 1.9507944643772426,
+      "grad_norm": 0.5258013489452003,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.2391,
+      "step": 3806
+    },
+    {
+      "epoch": 1.9513070220399795,
+      "grad_norm": 0.6131504437517615,
+      "learning_rate": 3.1087768964092046e-08,
+      "loss": 0.2416,
+      "step": 3807
+    },
+    {
+      "epoch": 1.9518195797027165,
+      "grad_norm": 0.5005779401661012,
+      "learning_rate": 3.043706470875374e-08,
+      "loss": 0.2303,
+      "step": 3808
+    },
+    {
+      "epoch": 1.9523321373654536,
+      "grad_norm": 0.5208767614404453,
+      "learning_rate": 2.9793232302043342e-08,
+      "loss": 0.2263,
+      "step": 3809
+    },
+    {
+      "epoch": 1.9528446950281908,
+      "grad_norm": 0.5351961906696017,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.2377,
+      "step": 3810
+    },
+    {
+      "epoch": 1.9533572526909277,
+      "grad_norm": 0.49468438035853846,
+      "learning_rate": 2.852618480489566e-08,
+      "loss": 0.2462,
+      "step": 3811
+    },
+    {
+      "epoch": 1.9538698103536647,
+      "grad_norm": 0.5582684897874637,
+      "learning_rate": 2.7902970587813104e-08,
+      "loss": 0.2412,
+      "step": 3812
+    },
+    {
+      "epoch": 1.9543823680164019,
+      "grad_norm": 0.5520320491369907,
+      "learning_rate": 2.728662996606457e-08,
+      "loss": 0.244,
+      "step": 3813
+    },
+    {
+      "epoch": 1.954894925679139,
+      "grad_norm": 0.5156388962466082,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.2065,
+      "step": 3814
+    },
+    {
+      "epoch": 1.955407483341876,
+      "grad_norm": 0.5043470682033674,
+      "learning_rate": 2.6074571203165143e-08,
+      "loss": 0.2183,
+      "step": 3815
+    },
+    {
+      "epoch": 1.955920041004613,
+      "grad_norm": 0.5835712848325693,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.2697,
+      "step": 3816
+    },
+    {
+      "epoch": 1.95643259866735,
+      "grad_norm": 0.5608084982378412,
+      "learning_rate": 2.4890011858000886e-08,
+      "loss": 0.2495,
+      "step": 3817
+    },
+    {
+      "epoch": 1.9569451563300873,
+      "grad_norm": 0.501349761983295,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.2017,
+      "step": 3818
+    },
+    {
+      "epoch": 1.9574577139928242,
+      "grad_norm": 0.5013401613131485,
+      "learning_rate": 2.373295519655927e-08,
+      "loss": 0.2355,
+      "step": 3819
+    },
+    {
+      "epoch": 1.9579702716555611,
+      "grad_norm": 0.5486481086120026,
+      "learning_rate": 2.3164741372119215e-08,
+      "loss": 0.2435,
+      "step": 3820
+    },
+    {
+      "epoch": 1.9584828293182983,
+      "grad_norm": 0.5353604086754027,
+      "learning_rate": 2.2603404408994e-08,
+      "loss": 0.2442,
+      "step": 3821
+    },
+    {
+      "epoch": 1.9589953869810355,
+      "grad_norm": 0.5178150715123552,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.2544,
+      "step": 3822
+    },
+    {
+      "epoch": 1.9595079446437724,
+      "grad_norm": 0.5357461472859986,
+      "learning_rate": 2.150136260962721e-08,
+      "loss": 0.2497,
+      "step": 3823
+    },
+    {
+      "epoch": 1.9600205023065094,
+      "grad_norm": 0.47108457489152106,
+      "learning_rate": 2.0960658533003554e-08,
+      "loss": 0.2032,
+      "step": 3824
+    },
+    {
+      "epoch": 1.9605330599692465,
+      "grad_norm": 0.508176852729797,
+      "learning_rate": 2.0426832836930588e-08,
+      "loss": 0.2426,
+      "step": 3825
+    },
+    {
+      "epoch": 1.9610456176319837,
+      "grad_norm": 0.5226423777909627,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.2359,
+      "step": 3826
+    },
+    {
+      "epoch": 1.9615581752947207,
+      "grad_norm": 0.511426199083,
+      "learning_rate": 1.9379818053523182e-08,
+      "loss": 0.2361,
+      "step": 3827
+    },
+    {
+      "epoch": 1.9620707329574576,
+      "grad_norm": 0.5282912575983394,
+      "learning_rate": 1.8866629687879222e-08,
+      "loss": 0.2557,
+      "step": 3828
+    },
+    {
+      "epoch": 1.9625832906201948,
+      "grad_norm": 0.5678627783583752,
+      "learning_rate": 1.8360321146163595e-08,
+      "loss": 0.2644,
+      "step": 3829
+    },
+    {
+      "epoch": 1.963095848282932,
+      "grad_norm": 0.5556441595369315,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.2645,
+      "step": 3830
+    },
+    {
+      "epoch": 1.9636084059456689,
+      "grad_norm": 0.5361552661400072,
+      "learning_rate": 1.7368344925736692e-08,
+      "loss": 0.2472,
+      "step": 3831
+    },
+    {
+      "epoch": 1.9641209636084058,
+      "grad_norm": 0.5019920377906681,
+      "learning_rate": 1.6882677930777357e-08,
+      "loss": 0.2353,
+      "step": 3832
+    },
+    {
+      "epoch": 1.964633521271143,
+      "grad_norm": 0.5007653528366266,
+      "learning_rate": 1.6403892127251352e-08,
+      "loss": 0.241,
+      "step": 3833
+    },
+    {
+      "epoch": 1.9651460789338802,
+      "grad_norm": 0.565507081552663,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.1997,
+      "step": 3834
+    },
+    {
+      "epoch": 1.965658636596617,
+      "grad_norm": 0.5010549908155537,
+      "learning_rate": 1.546696540982939e-08,
+      "loss": 0.2045,
+      "step": 3835
+    },
+    {
+      "epoch": 1.966171194259354,
+      "grad_norm": 0.4865847531014788,
+      "learning_rate": 1.50088251417424e-08,
+      "loss": 0.1984,
+      "step": 3836
+    },
+    {
+      "epoch": 1.9666837519220912,
+      "grad_norm": 0.5372173105129358,
+      "learning_rate": 1.4557567356702218e-08,
+      "loss": 0.2611,
+      "step": 3837
+    },
+    {
+      "epoch": 1.9671963095848284,
+      "grad_norm": 0.5645956019363307,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2595,
+      "step": 3838
+    },
+    {
+      "epoch": 1.9677088672475653,
+      "grad_norm": 0.46771684676675235,
+      "learning_rate": 1.3675700475196395e-08,
+      "loss": 0.1962,
+      "step": 3839
+    },
+    {
+      "epoch": 1.9682214249103023,
+      "grad_norm": 0.5486979897303109,
+      "learning_rate": 1.3245091986587855e-08,
+      "loss": 0.2253,
+      "step": 3840
+    },
+    {
+      "epoch": 1.9687339825730394,
+      "grad_norm": 0.5338118877771281,
+      "learning_rate": 1.2821367196736988e-08,
+      "loss": 0.2376,
+      "step": 3841
+    },
+    {
+      "epoch": 1.9692465402357766,
+      "grad_norm": 0.5457392819626573,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.2321,
+      "step": 3842
+    },
+    {
+      "epoch": 1.9697590978985136,
+      "grad_norm": 0.49842192729603885,
+      "learning_rate": 1.1994569876830897e-08,
+      "loss": 0.2338,
+      "step": 3843
+    },
+    {
+      "epoch": 1.9702716555612505,
+      "grad_norm": 0.5337132392914319,
+      "learning_rate": 1.1591497916673133e-08,
+      "loss": 0.2515,
+      "step": 3844
+    },
+    {
+      "epoch": 1.9707842132239877,
+      "grad_norm": 0.49521668360449583,
+      "learning_rate": 1.119531079506797e-08,
+      "loss": 0.234,
+      "step": 3845
+    },
+    {
+      "epoch": 1.9712967708867248,
+      "grad_norm": 0.5273401190452901,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2309,
+      "step": 3846
+    },
+    {
+      "epoch": 1.9718093285494618,
+      "grad_norm": 0.4722848404013851,
+      "learning_rate": 1.0423592155108798e-08,
+      "loss": 0.2059,
+      "step": 3847
+    },
+    {
+      "epoch": 1.9723218862121987,
+      "grad_norm": 0.47543825791541255,
+      "learning_rate": 1.0048061168688173e-08,
+      "loss": 0.2296,
+      "step": 3848
+    },
+    {
+      "epoch": 1.972834443874936,
+      "grad_norm": 0.48858093025518895,
+      "learning_rate": 9.67941608468359e-09,
+      "loss": 0.2231,
+      "step": 3849
+    },
+    {
+      "epoch": 1.973347001537673,
+      "grad_norm": 0.525534702136645,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.2461,
+      "step": 3850
+    },
+    {
+      "epoch": 1.97385955920041,
+      "grad_norm": 0.5568057694505545,
+      "learning_rate": 8.96278463558331e-09,
+      "loss": 0.2605,
+      "step": 3851
+    },
+    {
+      "epoch": 1.974372116863147,
+      "grad_norm": 0.5178548639342521,
+      "learning_rate": 8.61479876444804e-09,
+      "loss": 0.2347,
+      "step": 3852
+    },
+    {
+      "epoch": 1.9748846745258841,
+      "grad_norm": 0.5166285949263214,
+      "learning_rate": 8.273699783652999e-09,
+      "loss": 0.2377,
+      "step": 3853
+    },
+    {
+      "epoch": 1.9753972321886213,
+      "grad_norm": 0.46896384018069226,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.209,
+      "step": 3854
+    },
+    {
+      "epoch": 1.9759097898513582,
+      "grad_norm": 0.5394883085319654,
+      "learning_rate": 7.612163428791786e-09,
+      "loss": 0.2473,
+      "step": 3855
+    },
+    {
+      "epoch": 1.9764223475140952,
+      "grad_norm": 0.5710988748505464,
+      "learning_rate": 7.291726510710861e-09,
+      "loss": 0.2639,
+      "step": 3856
+    },
+    {
+      "epoch": 1.9769349051768323,
+      "grad_norm": 0.564114870349268,
+      "learning_rate": 6.978177394940666e-09,
+      "loss": 0.2485,
+      "step": 3857
+    },
+    {
+      "epoch": 1.9774474628395695,
+      "grad_norm": 0.5728754661720231,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.2619,
+      "step": 3858
+    },
+    {
+      "epoch": 1.9779600205023065,
+      "grad_norm": 0.5265974187237238,
+      "learning_rate": 6.371743430082511e-09,
+      "loss": 0.2431,
+      "step": 3859
+    },
+    {
+      "epoch": 1.9784725781650434,
+      "grad_norm": 0.5203635758253577,
+      "learning_rate": 6.078858998999071e-09,
+      "loss": 0.2513,
+      "step": 3860
+    },
+    {
+      "epoch": 1.9789851358277806,
+      "grad_norm": 0.5015807451068299,
+      "learning_rate": 5.79286320623651e-09,
+      "loss": 0.2409,
+      "step": 3861
+    },
+    {
+      "epoch": 1.9794976934905177,
+      "grad_norm": 0.5257974508309339,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.2476,
+      "step": 3862
+    },
+    {
+      "epoch": 1.9800102511532547,
+      "grad_norm": 0.5864343616907348,
+      "learning_rate": 5.241538319448181e-09,
+      "loss": 0.2735,
+      "step": 3863
+    },
+    {
+      "epoch": 1.9805228088159919,
+      "grad_norm": 0.5335335510487438,
+      "learning_rate": 4.9762096054428724e-09,
+      "loss": 0.2778,
+      "step": 3864
+    },
+    {
+      "epoch": 1.981035366478729,
+      "grad_norm": 0.526042473992512,
+      "learning_rate": 4.717770289794921e-09,
+      "loss": 0.2261,
+      "step": 3865
+    },
+    {
+      "epoch": 1.981547924141466,
+      "grad_norm": 0.53426076551551,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.2311,
+      "step": 3866
+    },
+    {
+      "epoch": 1.982060481804203,
+      "grad_norm": 0.5481477383768364,
+      "learning_rate": 4.221560561372684e-09,
+      "loss": 0.2686,
+      "step": 3867
+    },
+    {
+      "epoch": 1.98257303946694,
+      "grad_norm": 0.5351147272751389,
+      "learning_rate": 3.983790490628137e-09,
+      "loss": 0.2536,
+      "step": 3868
+    },
+    {
+      "epoch": 1.9830855971296772,
+      "grad_norm": 0.5295327645371228,
+      "learning_rate": 3.7529105022970915e-09,
+      "loss": 0.2438,
+      "step": 3869
+    },
+    {
+      "epoch": 1.9835981547924142,
+      "grad_norm": 0.5646987075674598,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.2606,
+      "step": 3870
+    },
+    {
+      "epoch": 1.9841107124551511,
+      "grad_norm": 0.5233111610367157,
+      "learning_rate": 3.311821404697879e-09,
+      "loss": 0.212,
+      "step": 3871
+    },
+    {
+      "epoch": 1.9846232701178883,
+      "grad_norm": 0.5650531040026108,
+      "learning_rate": 3.1016125994631776e-09,
+      "loss": 0.2227,
+      "step": 3872
+    },
+    {
+      "epoch": 1.9851358277806255,
+      "grad_norm": 0.49285624596649336,
+      "learning_rate": 2.898294484714459e-09,
+      "loss": 0.2071,
+      "step": 3873
+    },
+    {
+      "epoch": 1.9856483854433624,
+      "grad_norm": 0.4855997713466467,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2138,
+      "step": 3874
+    },
+    {
+      "epoch": 1.9861609431060994,
+      "grad_norm": 0.4977562975860101,
+      "learning_rate": 2.512330882494807e-09,
+      "loss": 0.2117,
+      "step": 3875
+    },
+    {
+      "epoch": 1.9866735007688365,
+      "grad_norm": 0.4940865340913999,
+      "learning_rate": 2.329685661063286e-09,
+      "loss": 0.2342,
+      "step": 3876
+    },
+    {
+      "epoch": 1.9871860584315737,
+      "grad_norm": 0.49216246569846817,
+      "learning_rate": 2.1539316621921326e-09,
+      "loss": 0.2254,
+      "step": 3877
+    },
+    {
+      "epoch": 1.9876986160943106,
+      "grad_norm": 0.5197301174489108,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.2376,
+      "step": 3878
+    },
+    {
+      "epoch": 1.9882111737570476,
+      "grad_norm": 0.5444696670324548,
+      "learning_rate": 1.823097811959329e-09,
+      "loss": 0.2661,
+      "step": 3879
+    },
+    {
+      "epoch": 1.9887237314197848,
+      "grad_norm": 0.4648902709838693,
+      "learning_rate": 1.6680181886352676e-09,
+      "loss": 0.2004,
+      "step": 3880
+    },
+    {
+      "epoch": 1.989236289082522,
+      "grad_norm": 0.5278671620730218,
+      "learning_rate": 1.519830243948972e-09,
+      "loss": 0.2236,
+      "step": 3881
+    },
+    {
+      "epoch": 1.9897488467452589,
+      "grad_norm": 0.48368822344884654,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2088,
+      "step": 3882
+    },
+    {
+      "epoch": 1.9902614044079958,
+      "grad_norm": 0.5551484073004841,
+      "learning_rate": 1.244129794307769e-09,
+      "loss": 0.2395,
+      "step": 3883
+    },
+    {
+      "epoch": 1.990773962070733,
+      "grad_norm": 0.5535463997172264,
+      "learning_rate": 1.1166174793908468e-09,
+      "loss": 0.2589,
+      "step": 3884
+    },
+    {
+      "epoch": 1.9912865197334702,
+      "grad_norm": 0.5606017185108414,
+      "learning_rate": 9.959972231798898e-10,
+      "loss": 0.2605,
+      "step": 3885
+    },
+    {
+      "epoch": 1.991799077396207,
+      "grad_norm": 0.47028143904352643,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.2215,
+      "step": 3886
+    },
+    {
+      "epoch": 1.992311635058944,
+      "grad_norm": 0.5121708919949861,
+      "learning_rate": 7.75433214699195e-10,
+      "loss": 0.2338,
+      "step": 3887
+    },
+    {
+      "epoch": 1.9928241927216812,
+      "grad_norm": 0.5076911334453652,
+      "learning_rate": 6.754896144589573e-10,
+      "loss": 0.2378,
+      "step": 3888
+    },
+    {
+      "epoch": 1.9933367503844184,
+      "grad_norm": 0.5051573895066178,
+      "learning_rate": 5.824383769892361e-10,
+      "loss": 0.2137,
+      "step": 3889
+    },
+    {
+      "epoch": 1.9938493080471553,
+      "grad_norm": 0.5268445770776763,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.2238,
+      "step": 3890
+    },
+    {
+      "epoch": 1.9943618657098923,
+      "grad_norm": 0.5453230915401894,
+      "learning_rate": 4.170132421610351e-10,
+      "loss": 0.2499,
+      "step": 3891
+    },
+    {
+      "epoch": 1.9948744233726294,
+      "grad_norm": 0.5714417908545991,
+      "learning_rate": 3.446394588280111e-10,
+      "loss": 0.2579,
+      "step": 3892
+    },
+    {
+      "epoch": 1.9953869810353666,
+      "grad_norm": 0.5150244256912673,
+      "learning_rate": 2.791582663141945e-10,
+      "loss": 0.2135,
+      "step": 3893
+    },
+    {
+      "epoch": 1.9958995386981035,
+      "grad_norm": 0.5260831412025676,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.2165,
+      "step": 3894
+    },
+    {
+      "epoch": 1.9964120963608405,
+      "grad_norm": 0.5023202959449917,
+      "learning_rate": 1.688738295335668e-10,
+      "loss": 0.204,
+      "step": 3895
+    },
+    {
+      "epoch": 1.9969246540235777,
+      "grad_norm": 0.5055406060398261,
+      "learning_rate": 1.2407066128372613e-10,
+      "loss": 0.2441,
+      "step": 3896
+    },
+    {
+      "epoch": 1.9974372116863148,
+      "grad_norm": 0.5363921083784502,
+      "learning_rate": 8.61602358859237e-11,
+      "loss": 0.2445,
+      "step": 3897
+    },
+    {
+      "epoch": 1.9979497693490518,
+      "grad_norm": 0.4991605528885618,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.2476,
+      "step": 3898
+    },
+    {
+      "epoch": 1.9984623270117887,
+      "grad_norm": 0.5453711022715952,
+      "learning_rate": 3.101771342550741e-11,
+      "loss": 0.2359,
+      "step": 3899
+    },
+    {
+      "epoch": 1.9989748846745259,
+      "grad_norm": 0.49074061659798995,
+      "learning_rate": 1.3785654370268575e-11,
+      "loss": 0.2209,
+      "step": 3900
+    },
+    {
+      "epoch": 1.999487442337263,
+      "grad_norm": 0.47395603763525723,
+      "learning_rate": 3.4464141862589064e-12,
+      "loss": 0.2102,
+      "step": 3901
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5203750108228126,
+      "learning_rate": 0.0,
+      "loss": 0.2422,
+      "step": 3902
+    },
+    {
+      "epoch": 2.0,
+      "step": 3902,
+      "total_flos": 1.6038479007842304e+16,
+      "train_loss": 0.3359356340286305,
+      "train_runtime": 61896.2014,
+      "train_samples_per_second": 8.069,
+      "train_steps_per_second": 0.063
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3902,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 24000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6038479007842304e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}