diff --git "a/checkpoint-723/trainer_state.json" "b/checkpoint-723/trainer_state.json"
deleted file mode 100644--- "a/checkpoint-723/trainer_state.json"
+++ /dev/null
@@ -1,5126 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.0,
-  "eval_steps": 181,
-  "global_step": 723,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0013831258644536654,
-      "grad_norm": 0.5777013897895813,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.8618,
-      "step": 1
-    },
-    {
-      "epoch": 0.0013831258644536654,
-      "eval_loss": 1.4421294927597046,
-      "eval_runtime": 109.7132,
-      "eval_samples_per_second": 13.162,
-      "eval_steps_per_second": 0.829,
-      "step": 1
-    },
-    {
-      "epoch": 0.0027662517289073307,
-      "grad_norm": 0.6331229209899902,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.7491,
-      "step": 2
-    },
-    {
-      "epoch": 0.004149377593360996,
-      "grad_norm": 0.5830245018005371,
-      "learning_rate": 6e-06,
-      "loss": 1.8016,
-      "step": 3
-    },
-    {
-      "epoch": 0.005532503457814661,
-      "grad_norm": 0.5829038023948669,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 1.9693,
-      "step": 4
-    },
-    {
-      "epoch": 0.006915629322268326,
-      "grad_norm": 0.552374005317688,
-      "learning_rate": 1e-05,
-      "loss": 1.9286,
-      "step": 5
-    },
-    {
-      "epoch": 0.008298755186721992,
-      "grad_norm": 0.5776746869087219,
-      "learning_rate": 1.2e-05,
-      "loss": 1.7761,
-      "step": 6
-    },
-    {
-      "epoch": 0.009681881051175657,
-      "grad_norm": 0.5832110643386841,
-      "learning_rate": 1.4e-05,
-      "loss": 1.818,
-      "step": 7
-    },
-    {
-      "epoch": 0.011065006915629323,
-      "grad_norm": 0.5438941717147827,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.863,
-      "step": 8
-    },
-    {
-      "epoch": 0.012448132780082987,
-      "grad_norm": 0.583209216594696,
-      "learning_rate": 1.8e-05,
-      "loss": 1.9574,
-      "step": 9
-    },
-    {
-      "epoch": 0.013831258644536652,
-      "grad_norm": 0.5854378938674927,
-      "learning_rate": 2e-05,
-      "loss": 1.9059,
-      "step": 10
-    },
-    {
-      "epoch": 0.015214384508990318,
-      "grad_norm": 0.49761173129081726,
-      "learning_rate": 1.999997606900827e-05,
-      "loss": 1.8309,
-      "step": 11
-    },
-    {
-      "epoch": 0.016597510373443983,
-      "grad_norm": 0.4992309808731079,
-      "learning_rate": 1.999990427614762e-05,
-      "loss": 1.7712,
-      "step": 12
-    },
-    {
-      "epoch": 0.017980636237897647,
-      "grad_norm": 0.49240681529045105,
-      "learning_rate": 1.999978462176166e-05,
-      "loss": 1.8119,
-      "step": 13
-    },
-    {
-      "epoch": 0.019363762102351315,
-      "grad_norm": 0.469261109828949,
-      "learning_rate": 1.999961710642308e-05,
-      "loss": 1.8405,
-      "step": 14
-    },
-    {
-      "epoch": 0.02074688796680498,
-      "grad_norm": 0.4924378991127014,
-      "learning_rate": 1.999940173093365e-05,
-      "loss": 1.7209,
-      "step": 15
-    },
-    {
-      "epoch": 0.022130013831258646,
-      "grad_norm": 0.5745378732681274,
-      "learning_rate": 1.999913849632419e-05,
-      "loss": 1.8454,
-      "step": 16
-    },
-    {
-      "epoch": 0.02351313969571231,
-      "grad_norm": 0.5211701989173889,
-      "learning_rate": 1.9998827403854596e-05,
-      "loss": 1.8306,
-      "step": 17
-    },
-    {
-      "epoch": 0.024896265560165973,
-      "grad_norm": 0.5535211563110352,
-      "learning_rate": 1.9998468455013825e-05,
-      "loss": 1.8723,
-      "step": 18
-    },
-    {
-      "epoch": 0.02627939142461964,
-      "grad_norm": 0.5848811268806458,
-      "learning_rate": 1.9998061651519868e-05,
-      "loss": 1.7475,
-      "step": 19
-    },
-    {
-      "epoch": 0.027662517289073305,
-      "grad_norm": 0.6234457492828369,
-      "learning_rate": 1.999760699531977e-05,
-      "loss": 1.7606,
-      "step": 20
-    },
-    {
-      "epoch": 0.029045643153526972,
-      "grad_norm": 0.559583842754364,
-      "learning_rate": 1.9997104488589607e-05,
-      "loss": 1.7085,
-      "step": 21
-    },
-    {
-      "epoch": 0.030428769017980636,
-      "grad_norm": 0.5410158038139343,
-      "learning_rate": 1.9996554133734473e-05,
-      "loss": 1.7188,
-      "step": 22
-    },
-    {
-      "epoch": 0.0318118948824343,
-      "grad_norm": 0.49627774953842163,
-      "learning_rate": 1.999595593338848e-05,
-      "loss": 1.8304,
-      "step": 23
-    },
-    {
-      "epoch": 0.03319502074688797,
-      "grad_norm": 0.49211692810058594,
-      "learning_rate": 1.9995309890414735e-05,
-      "loss": 1.9025,
-      "step": 24
-    },
-    {
-      "epoch": 0.034578146611341634,
-      "grad_norm": 0.5097021460533142,
-      "learning_rate": 1.9994616007905318e-05,
-      "loss": 1.8533,
-      "step": 25
-    },
-    {
-      "epoch": 0.035961272475795295,
-      "grad_norm": 0.4716247320175171,
-      "learning_rate": 1.99938742891813e-05,
-      "loss": 1.7933,
-      "step": 26
-    },
-    {
-      "epoch": 0.03734439834024896,
-      "grad_norm": 0.5546444058418274,
-      "learning_rate": 1.9993084737792687e-05,
-      "loss": 1.7098,
-      "step": 27
-    },
-    {
-      "epoch": 0.03872752420470263,
-      "grad_norm": 0.6025681495666504,
-      "learning_rate": 1.9992247357518428e-05,
-      "loss": 1.8161,
-      "step": 28
-    },
-    {
-      "epoch": 0.040110650069156296,
-      "grad_norm": 0.5782088041305542,
-      "learning_rate": 1.9991362152366393e-05,
-      "loss": 1.7695,
-      "step": 29
-    },
-    {
-      "epoch": 0.04149377593360996,
-      "grad_norm": 0.5509287714958191,
-      "learning_rate": 1.9990429126573353e-05,
-      "loss": 1.8172,
-      "step": 30
-    },
-    {
-      "epoch": 0.042876901798063624,
-      "grad_norm": 0.504231333732605,
-      "learning_rate": 1.9989448284604947e-05,
-      "loss": 1.8554,
-      "step": 31
-    },
-    {
-      "epoch": 0.04426002766251729,
-      "grad_norm": 0.5206706523895264,
-      "learning_rate": 1.9988419631155686e-05,
-      "loss": 1.7808,
-      "step": 32
-    },
-    {
-      "epoch": 0.04564315352697095,
-      "grad_norm": 0.49199178814888,
-      "learning_rate": 1.9987343171148904e-05,
-      "loss": 1.7783,
-      "step": 33
-    },
-    {
-      "epoch": 0.04702627939142462,
-      "grad_norm": 0.5891382694244385,
-      "learning_rate": 1.9986218909736758e-05,
-      "loss": 1.7706,
-      "step": 34
-    },
-    {
-      "epoch": 0.048409405255878286,
-      "grad_norm": 0.5374609231948853,
-      "learning_rate": 1.9985046852300183e-05,
-      "loss": 1.7426,
-      "step": 35
-    },
-    {
-      "epoch": 0.04979253112033195,
-      "grad_norm": 0.5391809344291687,
-      "learning_rate": 1.9983827004448875e-05,
-      "loss": 1.7105,
-      "step": 36
-    },
-    {
-      "epoch": 0.051175656984785614,
-      "grad_norm": 0.5461614727973938,
-      "learning_rate": 1.9982559372021274e-05,
-      "loss": 1.694,
-      "step": 37
-    },
-    {
-      "epoch": 0.05255878284923928,
-      "grad_norm": 0.531283974647522,
-      "learning_rate": 1.9981243961084516e-05,
-      "loss": 1.7295,
-      "step": 38
-    },
-    {
-      "epoch": 0.05394190871369295,
-      "grad_norm": 0.512266218662262,
-      "learning_rate": 1.997988077793442e-05,
-      "loss": 1.6836,
-      "step": 39
-    },
-    {
-      "epoch": 0.05532503457814661,
-      "grad_norm": 0.5940636396408081,
-      "learning_rate": 1.997846982909545e-05,
-      "loss": 1.7573,
-      "step": 40
-    },
-    {
-      "epoch": 0.056708160442600276,
-      "grad_norm": 0.5649189352989197,
-      "learning_rate": 1.9977011121320687e-05,
-      "loss": 1.6611,
-      "step": 41
-    },
-    {
-      "epoch": 0.058091286307053944,
-      "grad_norm": 0.5476697087287903,
-      "learning_rate": 1.99755046615918e-05,
-      "loss": 1.6048,
-      "step": 42
-    },
-    {
-      "epoch": 0.059474412171507604,
-      "grad_norm": 0.5268926024436951,
-      "learning_rate": 1.9973950457119e-05,
-      "loss": 1.6819,
-      "step": 43
-    },
-    {
-      "epoch": 0.06085753803596127,
-      "grad_norm": 0.5020188689231873,
-      "learning_rate": 1.9972348515341018e-05,
-      "loss": 1.8042,
-      "step": 44
-    },
-    {
-      "epoch": 0.06224066390041494,
-      "grad_norm": 0.5357793569564819,
-      "learning_rate": 1.9970698843925064e-05,
-      "loss": 1.799,
-      "step": 45
-    },
-    {
-      "epoch": 0.0636237897648686,
-      "grad_norm": 0.5506536364555359,
-      "learning_rate": 1.9969001450766795e-05,
-      "loss": 1.6811,
-      "step": 46
-    },
-    {
-      "epoch": 0.06500691562932227,
-      "grad_norm": 0.5344098806381226,
-      "learning_rate": 1.9967256343990272e-05,
-      "loss": 1.6994,
-      "step": 47
-    },
-    {
-      "epoch": 0.06639004149377593,
-      "grad_norm": 0.6598169803619385,
-      "learning_rate": 1.996546353194792e-05,
-      "loss": 1.6529,
-      "step": 48
-    },
-    {
-      "epoch": 0.0677731673582296,
-      "grad_norm": 0.5558203458786011,
-      "learning_rate": 1.9963623023220493e-05,
-      "loss": 1.6161,
-      "step": 49
-    },
-    {
-      "epoch": 0.06915629322268327,
-      "grad_norm": 0.5793808102607727,
-      "learning_rate": 1.9961734826617033e-05,
-      "loss": 1.3158,
-      "step": 50
-    },
-    {
-      "epoch": 0.07053941908713693,
-      "grad_norm": 0.5689916610717773,
-      "learning_rate": 1.995979895117482e-05,
-      "loss": 1.5892,
-      "step": 51
-    },
-    {
-      "epoch": 0.07192254495159059,
-      "grad_norm": 0.5733948349952698,
-      "learning_rate": 1.9957815406159344e-05,
-      "loss": 1.6697,
-      "step": 52
-    },
-    {
-      "epoch": 0.07330567081604426,
-      "grad_norm": 0.5705085396766663,
-      "learning_rate": 1.995578420106424e-05,
-      "loss": 1.6955,
-      "step": 53
-    },
-    {
-      "epoch": 0.07468879668049792,
-      "grad_norm": 0.6449640393257141,
-      "learning_rate": 1.995370534561125e-05,
-      "loss": 1.6568,
-      "step": 54
-    },
-    {
-      "epoch": 0.07607192254495158,
-      "grad_norm": 0.7000141739845276,
-      "learning_rate": 1.995157884975021e-05,
-      "loss": 1.6543,
-      "step": 55
-    },
-    {
-      "epoch": 0.07745504840940526,
-      "grad_norm": 0.725715696811676,
-      "learning_rate": 1.994940472365893e-05,
-      "loss": 1.5375,
-      "step": 56
-    },
-    {
-      "epoch": 0.07883817427385892,
-      "grad_norm": 0.616287350654602,
-      "learning_rate": 1.994718297774322e-05,
-      "loss": 1.7027,
-      "step": 57
-    },
-    {
-      "epoch": 0.08022130013831259,
-      "grad_norm": 0.561693012714386,
-      "learning_rate": 1.9944913622636798e-05,
-      "loss": 1.7051,
-      "step": 58
-    },
-    {
-      "epoch": 0.08160442600276625,
-      "grad_norm": 0.6091241240501404,
-      "learning_rate": 1.994259666920124e-05,
-      "loss": 1.6528,
-      "step": 59
-    },
-    {
-      "epoch": 0.08298755186721991,
-      "grad_norm": 0.6698527932167053,
-      "learning_rate": 1.994023212852595e-05,
-      "loss": 1.6039,
-      "step": 60
-    },
-    {
-      "epoch": 0.08437067773167359,
-      "grad_norm": 0.6914900541305542,
-      "learning_rate": 1.9937820011928086e-05,
-      "loss": 1.4681,
-      "step": 61
-    },
-    {
-      "epoch": 0.08575380359612725,
-      "grad_norm": 0.6026729345321655,
-      "learning_rate": 1.993536033095252e-05,
-      "loss": 1.6564,
-      "step": 62
-    },
-    {
-      "epoch": 0.08713692946058091,
-      "grad_norm": 0.6261776089668274,
-      "learning_rate": 1.9932853097371765e-05,
-      "loss": 1.7278,
-      "step": 63
-    },
-    {
-      "epoch": 0.08852005532503458,
-      "grad_norm": 0.6159116625785828,
-      "learning_rate": 1.9930298323185945e-05,
-      "loss": 1.604,
-      "step": 64
-    },
-    {
-      "epoch": 0.08990318118948824,
-      "grad_norm": 0.633053719997406,
-      "learning_rate": 1.992769602062272e-05,
-      "loss": 1.7534,
-      "step": 65
-    },
-    {
-      "epoch": 0.0912863070539419,
-      "grad_norm": 0.6861281394958496,
-      "learning_rate": 1.9925046202137215e-05,
-      "loss": 1.7307,
-      "step": 66
-    },
-    {
-      "epoch": 0.09266943291839558,
-      "grad_norm": 0.6718122959136963,
-      "learning_rate": 1.9922348880411997e-05,
-      "loss": 1.6336,
-      "step": 67
-    },
-    {
-      "epoch": 0.09405255878284924,
-      "grad_norm": 0.6402599811553955,
-      "learning_rate": 1.9919604068356978e-05,
-      "loss": 1.6736,
-      "step": 68
-    },
-    {
-      "epoch": 0.0954356846473029,
-      "grad_norm": 0.6972026824951172,
-      "learning_rate": 1.9916811779109374e-05,
-      "loss": 1.6701,
-      "step": 69
-    },
-    {
-      "epoch": 0.09681881051175657,
-      "grad_norm": 0.6567057967185974,
-      "learning_rate": 1.991397202603363e-05,
-      "loss": 1.716,
-      "step": 70
-    },
-    {
-      "epoch": 0.09820193637621023,
-      "grad_norm": 0.705405592918396,
-      "learning_rate": 1.991108482272138e-05,
-      "loss": 1.6328,
-      "step": 71
-    },
-    {
-      "epoch": 0.0995850622406639,
-      "grad_norm": 0.6641591787338257,
-      "learning_rate": 1.9908150182991338e-05,
-      "loss": 1.71,
-      "step": 72
-    },
-    {
-      "epoch": 0.10096818810511757,
-      "grad_norm": 0.7807027101516724,
-      "learning_rate": 1.990516812088928e-05,
-      "loss": 1.6361,
-      "step": 73
-    },
-    {
-      "epoch": 0.10235131396957123,
-      "grad_norm": 0.7211569547653198,
-      "learning_rate": 1.9902138650687943e-05,
-      "loss": 1.5969,
-      "step": 74
-    },
-    {
-      "epoch": 0.1037344398340249,
-      "grad_norm": 0.6972200870513916,
-      "learning_rate": 1.9899061786886978e-05,
-      "loss": 1.6331,
-      "step": 75
-    },
-    {
-      "epoch": 0.10511756569847856,
-      "grad_norm": 0.722425103187561,
-      "learning_rate": 1.9895937544212856e-05,
-      "loss": 1.6524,
-      "step": 76
-    },
-    {
-      "epoch": 0.10650069156293222,
-      "grad_norm": 0.8138152360916138,
-      "learning_rate": 1.9892765937618826e-05,
-      "loss": 1.6209,
-      "step": 77
-    },
-    {
-      "epoch": 0.1078838174273859,
-      "grad_norm": 0.6803196668624878,
-      "learning_rate": 1.9889546982284833e-05,
-      "loss": 1.6756,
-      "step": 78
-    },
-    {
-      "epoch": 0.10926694329183956,
-      "grad_norm": 0.7364890575408936,
-      "learning_rate": 1.988628069361743e-05,
-      "loss": 1.5753,
-      "step": 79
-    },
-    {
-      "epoch": 0.11065006915629322,
-      "grad_norm": 0.6657965183258057,
-      "learning_rate": 1.988296708724972e-05,
-      "loss": 1.62,
-      "step": 80
-    },
-    {
-      "epoch": 0.11203319502074689,
-      "grad_norm": 0.709164559841156,
-      "learning_rate": 1.9879606179041283e-05,
-      "loss": 1.7221,
-      "step": 81
-    },
-    {
-      "epoch": 0.11341632088520055,
-      "grad_norm": 0.6902270317077637,
-      "learning_rate": 1.987619798507809e-05,
-      "loss": 1.7295,
-      "step": 82
-    },
-    {
-      "epoch": 0.11479944674965421,
-      "grad_norm": 0.6978942155838013,
-      "learning_rate": 1.987274252167244e-05,
-      "loss": 1.7619,
-      "step": 83
-    },
-    {
-      "epoch": 0.11618257261410789,
-      "grad_norm": 0.6876259446144104,
-      "learning_rate": 1.986923980536286e-05,
-      "loss": 1.6597,
-      "step": 84
-    },
-    {
-      "epoch": 0.11756569847856155,
-      "grad_norm": 0.7877823114395142,
-      "learning_rate": 1.9865689852914048e-05,
-      "loss": 1.5334,
-      "step": 85
-    },
-    {
-      "epoch": 0.11894882434301521,
-      "grad_norm": 0.748802125453949,
-      "learning_rate": 1.9862092681316774e-05,
-      "loss": 1.6225,
-      "step": 86
-    },
-    {
-      "epoch": 0.12033195020746888,
-      "grad_norm": 0.8329505920410156,
-      "learning_rate": 1.9858448307787827e-05,
-      "loss": 1.5343,
-      "step": 87
-    },
-    {
-      "epoch": 0.12171507607192254,
-      "grad_norm": 0.8424030542373657,
-      "learning_rate": 1.9854756749769893e-05,
-      "loss": 1.6712,
-      "step": 88
-    },
-    {
-      "epoch": 0.12309820193637622,
-      "grad_norm": 0.7446318864822388,
-      "learning_rate": 1.98510180249315e-05,
-      "loss": 1.7452,
-      "step": 89
-    },
-    {
-      "epoch": 0.12448132780082988,
-      "grad_norm": 0.7498570680618286,
-      "learning_rate": 1.984723215116693e-05,
-      "loss": 1.5573,
-      "step": 90
-    },
-    {
-      "epoch": 0.12586445366528354,
-      "grad_norm": 0.7254923582077026,
-      "learning_rate": 1.9843399146596125e-05,
-      "loss": 1.6581,
-      "step": 91
-    },
-    {
-      "epoch": 0.1272475795297372,
-      "grad_norm": 0.8855713605880737,
-      "learning_rate": 1.9839519029564608e-05,
-      "loss": 1.5751,
-      "step": 92
-    },
-    {
-      "epoch": 0.12863070539419086,
-      "grad_norm": 0.7330020070075989,
-      "learning_rate": 1.983559181864338e-05,
-      "loss": 1.5733,
-      "step": 93
-    },
-    {
-      "epoch": 0.13001383125864455,
-      "grad_norm": 0.7649924755096436,
-      "learning_rate": 1.983161753262886e-05,
-      "loss": 1.7353,
-      "step": 94
-    },
-    {
-      "epoch": 0.1313969571230982,
-      "grad_norm": 0.7761242985725403,
-      "learning_rate": 1.982759619054277e-05,
-      "loss": 1.6374,
-      "step": 95
-    },
-    {
-      "epoch": 0.13278008298755187,
-      "grad_norm": 0.8000247478485107,
-      "learning_rate": 1.982352781163204e-05,
-      "loss": 1.6466,
-      "step": 96
-    },
-    {
-      "epoch": 0.13416320885200553,
-      "grad_norm": 0.7133256196975708,
-      "learning_rate": 1.9819412415368753e-05,
-      "loss": 1.5993,
-      "step": 97
-    },
-    {
-      "epoch": 0.1355463347164592,
-      "grad_norm": 0.781935453414917,
-      "learning_rate": 1.9815250021449998e-05,
-      "loss": 1.5487,
-      "step": 98
-    },
-    {
-      "epoch": 0.13692946058091288,
-      "grad_norm": 0.7965666055679321,
-      "learning_rate": 1.981104064979783e-05,
-      "loss": 1.5498,
-      "step": 99
-    },
-    {
-      "epoch": 0.13831258644536654,
-      "grad_norm": 0.8179916143417358,
-      "learning_rate": 1.980678432055913e-05,
-      "loss": 1.6324,
-      "step": 100
-    },
-    {
-      "epoch": 0.1396957123098202,
-      "grad_norm": 0.8549492955207825,
-      "learning_rate": 1.9802481054105527e-05,
-      "loss": 1.6359,
-      "step": 101
-    },
-    {
-      "epoch": 0.14107883817427386,
-      "grad_norm": 0.8777151107788086,
-      "learning_rate": 1.9798130871033322e-05,
-      "loss": 1.5975,
-      "step": 102
-    },
-    {
-      "epoch": 0.14246196403872752,
-      "grad_norm": 1.004165530204773,
-      "learning_rate": 1.979373379216335e-05,
-      "loss": 1.623,
-      "step": 103
-    },
-    {
-      "epoch": 0.14384508990318118,
-      "grad_norm": 0.9654561877250671,
-      "learning_rate": 1.9789289838540897e-05,
-      "loss": 1.5919,
-      "step": 104
-    },
-    {
-      "epoch": 0.14522821576763487,
-      "grad_norm": 0.8471197485923767,
-      "learning_rate": 1.978479903143561e-05,
-      "loss": 1.6244,
-      "step": 105
-    },
-    {
-      "epoch": 0.14661134163208853,
-      "grad_norm": 0.813434362411499,
-      "learning_rate": 1.9780261392341383e-05,
-      "loss": 1.6015,
-      "step": 106
-    },
-    {
-      "epoch": 0.1479944674965422,
-      "grad_norm": 0.9421816468238831,
-      "learning_rate": 1.9775676942976253e-05,
-      "loss": 1.6022,
-      "step": 107
-    },
-    {
-      "epoch": 0.14937759336099585,
-      "grad_norm": 0.7814284563064575,
-      "learning_rate": 1.9771045705282313e-05,
-      "loss": 1.6354,
-      "step": 108
-    },
-    {
-      "epoch": 0.1507607192254495,
-      "grad_norm": 0.9225528240203857,
-      "learning_rate": 1.9766367701425575e-05,
-      "loss": 1.6794,
-      "step": 109
-    },
-    {
-      "epoch": 0.15214384508990317,
-      "grad_norm": 0.9190025329589844,
-      "learning_rate": 1.9761642953795896e-05,
-      "loss": 1.511,
-      "step": 110
-    },
-    {
-      "epoch": 0.15352697095435686,
-      "grad_norm": 0.857140302658081,
-      "learning_rate": 1.9756871485006862e-05,
-      "loss": 1.5262,
-      "step": 111
-    },
-    {
-      "epoch": 0.15491009681881052,
-      "grad_norm": 0.7790897488594055,
-      "learning_rate": 1.975205331789566e-05,
-      "loss": 1.2074,
-      "step": 112
-    },
-    {
-      "epoch": 0.15629322268326418,
-      "grad_norm": 0.870851993560791,
-      "learning_rate": 1.9747188475523e-05,
-      "loss": 1.6356,
-      "step": 113
-    },
-    {
-      "epoch": 0.15767634854771784,
-      "grad_norm": 0.9374619722366333,
-      "learning_rate": 1.9742276981172978e-05,
-      "loss": 1.5746,
-      "step": 114
-    },
-    {
-      "epoch": 0.1590594744121715,
-      "grad_norm": 0.8879979252815247,
-      "learning_rate": 1.973731885835298e-05,
-      "loss": 1.5728,
-      "step": 115
-    },
-    {
-      "epoch": 0.16044260027662519,
-      "grad_norm": 0.8657084107398987,
-      "learning_rate": 1.973231413079357e-05,
-      "loss": 1.6845,
-      "step": 116
-    },
-    {
-      "epoch": 0.16182572614107885,
-      "grad_norm": 1.0091602802276611,
-      "learning_rate": 1.9727262822448364e-05,
-      "loss": 1.5589,
-      "step": 117
-    },
-    {
-      "epoch": 0.1632088520055325,
-      "grad_norm": 0.8246741890907288,
-      "learning_rate": 1.9722164957493925e-05,
-      "loss": 1.6564,
-      "step": 118
-    },
-    {
-      "epoch": 0.16459197786998617,
-      "grad_norm": 0.8533387780189514,
-      "learning_rate": 1.9717020560329644e-05,
-      "loss": 1.5895,
-      "step": 119
-    },
-    {
-      "epoch": 0.16597510373443983,
-      "grad_norm": 0.8713412284851074,
-      "learning_rate": 1.971182965557763e-05,
-      "loss": 1.6223,
-      "step": 120
-    },
-    {
-      "epoch": 0.1673582295988935,
-      "grad_norm": 0.948832631111145,
-      "learning_rate": 1.9706592268082584e-05,
-      "loss": 1.5907,
-      "step": 121
-    },
-    {
-      "epoch": 0.16874135546334718,
-      "grad_norm": 0.9240509867668152,
-      "learning_rate": 1.9701308422911674e-05,
-      "loss": 1.6396,
-      "step": 122
-    },
-    {
-      "epoch": 0.17012448132780084,
-      "grad_norm": 0.8856998085975647,
-      "learning_rate": 1.969597814535444e-05,
-      "loss": 1.5642,
-      "step": 123
-    },
-    {
-      "epoch": 0.1715076071922545,
-      "grad_norm": 1.0024968385696411,
-      "learning_rate": 1.969060146092264e-05,
-      "loss": 1.5414,
-      "step": 124
-    },
-    {
-      "epoch": 0.17289073305670816,
-      "grad_norm": 1.0026429891586304,
-      "learning_rate": 1.9685178395350157e-05,
-      "loss": 1.4907,
-      "step": 125
-    },
-    {
-      "epoch": 0.17427385892116182,
-      "grad_norm": 1.0451340675354004,
-      "learning_rate": 1.967970897459286e-05,
-      "loss": 1.6097,
-      "step": 126
-    },
-    {
-      "epoch": 0.17565698478561548,
-      "grad_norm": 0.9511062502861023,
-      "learning_rate": 1.9674193224828477e-05,
-      "loss": 1.5482,
-      "step": 127
-    },
-    {
-      "epoch": 0.17704011065006917,
-      "grad_norm": 0.9325770139694214,
-      "learning_rate": 1.966863117245648e-05,
-      "loss": 1.6144,
-      "step": 128
-    },
-    {
-      "epoch": 0.17842323651452283,
-      "grad_norm": 0.9411192536354065,
-      "learning_rate": 1.9663022844097956e-05,
-      "loss": 1.6372,
-      "step": 129
-    },
-    {
-      "epoch": 0.1798063623789765,
-      "grad_norm": 0.9200664758682251,
-      "learning_rate": 1.9657368266595477e-05,
-      "loss": 1.6128,
-      "step": 130
-    },
-    {
-      "epoch": 0.18118948824343015,
-      "grad_norm": 1.013554334640503,
-      "learning_rate": 1.9651667467012977e-05,
-      "loss": 1.4844,
-      "step": 131
-    },
-    {
-      "epoch": 0.1825726141078838,
-      "grad_norm": 0.9959779381752014,
-      "learning_rate": 1.964592047263561e-05,
-      "loss": 1.5979,
-      "step": 132
-    },
-    {
-      "epoch": 0.1839557399723375,
-      "grad_norm": 0.8918667435646057,
-      "learning_rate": 1.9640127310969626e-05,
-      "loss": 1.6407,
-      "step": 133
-    },
-    {
-      "epoch": 0.18533886583679116,
-      "grad_norm": 0.9420216083526611,
-      "learning_rate": 1.9634288009742254e-05,
-      "loss": 1.6495,
-      "step": 134
-    },
-    {
-      "epoch": 0.18672199170124482,
-      "grad_norm": 1.0406219959259033,
-      "learning_rate": 1.9628402596901545e-05,
-      "loss": 1.6287,
-      "step": 135
-    },
-    {
-      "epoch": 0.18810511756569848,
-      "grad_norm": 0.9881539940834045,
-      "learning_rate": 1.9622471100616253e-05,
-      "loss": 1.6145,
-      "step": 136
-    },
-    {
-      "epoch": 0.18948824343015214,
-      "grad_norm": 0.9983912706375122,
-      "learning_rate": 1.961649354927569e-05,
-      "loss": 1.5708,
-      "step": 137
-    },
-    {
-      "epoch": 0.1908713692946058,
-      "grad_norm": 0.9389325380325317,
-      "learning_rate": 1.961046997148961e-05,
-      "loss": 1.654,
-      "step": 138
-    },
-    {
-      "epoch": 0.19225449515905949,
-      "grad_norm": 0.9617229104042053,
-      "learning_rate": 1.9604400396088047e-05,
-      "loss": 1.5198,
-      "step": 139
-    },
-    {
-      "epoch": 0.19363762102351315,
-      "grad_norm": 0.9675804972648621,
-      "learning_rate": 1.959828485212119e-05,
-      "loss": 1.635,
-      "step": 140
-    },
-    {
-      "epoch": 0.1950207468879668,
-      "grad_norm": 1.0724852085113525,
-      "learning_rate": 1.959212336885925e-05,
-      "loss": 1.4736,
-      "step": 141
-    },
-    {
-      "epoch": 0.19640387275242047,
-      "grad_norm": 1.0052822828292847,
-      "learning_rate": 1.958591597579231e-05,
-      "loss": 1.5661,
-      "step": 142
-    },
-    {
-      "epoch": 0.19778699861687413,
-      "grad_norm": 0.890238881111145,
-      "learning_rate": 1.957966270263018e-05,
-      "loss": 1.252,
-      "step": 143
-    },
-    {
-      "epoch": 0.1991701244813278,
-      "grad_norm": 0.9959819912910461,
-      "learning_rate": 1.957336357930227e-05,
-      "loss": 1.4612,
-      "step": 144
-    },
-    {
-      "epoch": 0.20055325034578148,
-      "grad_norm": 1.03453528881073,
-      "learning_rate": 1.9567018635957426e-05,
-      "loss": 1.4836,
-      "step": 145
-    },
-    {
-      "epoch": 0.20193637621023514,
-      "grad_norm": 1.0642746686935425,
-      "learning_rate": 1.9560627902963808e-05,
-      "loss": 1.5676,
-      "step": 146
-    },
-    {
-      "epoch": 0.2033195020746888,
-      "grad_norm": 1.0111668109893799,
-      "learning_rate": 1.955419141090874e-05,
-      "loss": 1.5925,
-      "step": 147
-    },
-    {
-      "epoch": 0.20470262793914246,
-      "grad_norm": 1.0494191646575928,
-      "learning_rate": 1.9547709190598538e-05,
-      "loss": 1.6079,
-      "step": 148
-    },
-    {
-      "epoch": 0.20608575380359612,
-      "grad_norm": 0.9530560970306396,
-      "learning_rate": 1.95411812730584e-05,
-      "loss": 1.5284,
-      "step": 149
-    },
-    {
-      "epoch": 0.2074688796680498,
-      "grad_norm": 1.0679057836532593,
-      "learning_rate": 1.9534607689532236e-05,
-      "loss": 1.515,
-      "step": 150
-    },
-    {
-      "epoch": 0.20885200553250347,
-      "grad_norm": 1.0620648860931396,
-      "learning_rate": 1.9527988471482517e-05,
-      "loss": 1.6281,
-      "step": 151
-    },
-    {
-      "epoch": 0.21023513139695713,
-      "grad_norm": 0.9668426513671875,
-      "learning_rate": 1.9521323650590135e-05,
-      "loss": 1.5903,
-      "step": 152
-    },
-    {
-      "epoch": 0.21161825726141079,
-      "grad_norm": 1.0088167190551758,
-      "learning_rate": 1.9514613258754244e-05,
-      "loss": 1.5094,
-      "step": 153
-    },
-    {
-      "epoch": 0.21300138312586445,
-      "grad_norm": 1.113911747932434,
-      "learning_rate": 1.950785732809211e-05,
-      "loss": 1.5212,
-      "step": 154
-    },
-    {
-      "epoch": 0.2143845089903181,
-      "grad_norm": 0.95964515209198,
-      "learning_rate": 1.9501055890938957e-05,
-      "loss": 1.5151,
-      "step": 155
-    },
-    {
-      "epoch": 0.2157676348547718,
-      "grad_norm": 0.978797972202301,
-      "learning_rate": 1.9494208979847814e-05,
-      "loss": 1.5552,
-      "step": 156
-    },
-    {
-      "epoch": 0.21715076071922546,
-      "grad_norm": 1.2046525478363037,
-      "learning_rate": 1.9487316627589353e-05,
-      "loss": 1.6375,
-      "step": 157
-    },
-    {
-      "epoch": 0.21853388658367912,
-      "grad_norm": 1.1309069395065308,
-      "learning_rate": 1.9480378867151746e-05,
-      "loss": 1.5061,
-      "step": 158
-    },
-    {
-      "epoch": 0.21991701244813278,
-      "grad_norm": 1.0797772407531738,
-      "learning_rate": 1.9473395731740483e-05,
-      "loss": 1.4682,
-      "step": 159
-    },
-    {
-      "epoch": 0.22130013831258644,
-      "grad_norm": 1.1034504175186157,
-      "learning_rate": 1.9466367254778234e-05,
-      "loss": 1.5552,
-      "step": 160
-    },
-    {
-      "epoch": 0.22268326417704012,
-      "grad_norm": 1.03969407081604,
-      "learning_rate": 1.945929346990469e-05,
-      "loss": 1.6028,
-      "step": 161
-    },
-    {
-      "epoch": 0.22406639004149378,
-      "grad_norm": 1.2426897287368774,
-      "learning_rate": 1.9452174410976383e-05,
-      "loss": 1.5764,
-      "step": 162
-    },
-    {
-      "epoch": 0.22544951590594745,
-      "grad_norm": 1.0647516250610352,
-      "learning_rate": 1.9445010112066543e-05,
-      "loss": 1.5122,
-      "step": 163
-    },
-    {
-      "epoch": 0.2268326417704011,
-      "grad_norm": 0.976096510887146,
-      "learning_rate": 1.943780060746493e-05,
-      "loss": 1.6876,
-      "step": 164
-    },
-    {
-      "epoch": 0.22821576763485477,
-      "grad_norm": 1.1224089860916138,
-      "learning_rate": 1.9430545931677657e-05,
-      "loss": 1.5798,
-      "step": 165
-    },
-    {
-      "epoch": 0.22959889349930843,
-      "grad_norm": 1.1245198249816895,
-      "learning_rate": 1.9423246119427044e-05,
-      "loss": 1.5093,
-      "step": 166
-    },
-    {
-      "epoch": 0.23098201936376211,
-      "grad_norm": 1.0290110111236572,
-      "learning_rate": 1.941590120565144e-05,
-      "loss": 1.5308,
-      "step": 167
-    },
-    {
-      "epoch": 0.23236514522821577,
-      "grad_norm": 1.0023036003112793,
-      "learning_rate": 1.940851122550506e-05,
-      "loss": 1.6366,
-      "step": 168
-    },
-    {
-      "epoch": 0.23374827109266944,
-      "grad_norm": 1.287907361984253,
-      "learning_rate": 1.940107621435781e-05,
-      "loss": 1.5091,
-      "step": 169
-    },
-    {
-      "epoch": 0.2351313969571231,
-      "grad_norm": 1.159561276435852,
-      "learning_rate": 1.9393596207795135e-05,
-      "loss": 1.3876,
-      "step": 170
-    },
-    {
-      "epoch": 0.23651452282157676,
-      "grad_norm": 1.150119662284851,
-      "learning_rate": 1.9386071241617827e-05,
-      "loss": 1.5695,
-      "step": 171
-    },
-    {
-      "epoch": 0.23789764868603042,
-      "grad_norm": 1.1534556150436401,
-      "learning_rate": 1.9378501351841864e-05,
-      "loss": 1.4763,
-      "step": 172
-    },
-    {
-      "epoch": 0.2392807745504841,
-      "grad_norm": 1.2043812274932861,
-      "learning_rate": 1.9370886574698244e-05,
-      "loss": 1.5643,
-      "step": 173
-    },
-    {
-      "epoch": 0.24066390041493776,
-      "grad_norm": 1.049975872039795,
-      "learning_rate": 1.93632269466328e-05,
-      "loss": 1.6063,
-      "step": 174
-    },
-    {
-      "epoch": 0.24204702627939143,
-      "grad_norm": 1.168745517730713,
-      "learning_rate": 1.9355522504306027e-05,
-      "loss": 1.5692,
-      "step": 175
-    },
-    {
-      "epoch": 0.24343015214384509,
-      "grad_norm": 1.119295597076416,
-      "learning_rate": 1.934777328459292e-05,
-      "loss": 1.4857,
-      "step": 176
-    },
-    {
-      "epoch": 0.24481327800829875,
-      "grad_norm": 1.1553229093551636,
-      "learning_rate": 1.9339979324582782e-05,
-      "loss": 1.4596,
-      "step": 177
-    },
-    {
-      "epoch": 0.24619640387275243,
-      "grad_norm": 1.2437880039215088,
-      "learning_rate": 1.933214066157904e-05,
-      "loss": 1.601,
-      "step": 178
-    },
-    {
-      "epoch": 0.2475795297372061,
-      "grad_norm": 1.1771876811981201,
-      "learning_rate": 1.9324257333099104e-05,
-      "loss": 1.5632,
-      "step": 179
-    },
-    {
-      "epoch": 0.24896265560165975,
-      "grad_norm": 1.140103816986084,
-      "learning_rate": 1.9316329376874146e-05,
-      "loss": 1.5909,
-      "step": 180
-    },
-    {
-      "epoch": 0.2503457814661134,
-      "grad_norm": 1.1156872510910034,
-      "learning_rate": 1.9308356830848925e-05,
-      "loss": 1.6356,
-      "step": 181
-    },
-    {
-      "epoch": 0.2503457814661134,
-      "eval_loss": 1.163191318511963,
-      "eval_runtime": 108.8957,
-      "eval_samples_per_second": 13.26,
-      "eval_steps_per_second": 0.836,
-      "step": 181
-    },
-    {
-      "epoch": 0.2517289073305671,
-      "grad_norm": 1.1111009120941162,
-      "learning_rate": 1.930033973318164e-05,
-      "loss": 1.4833,
-      "step": 182
-    },
-    {
-      "epoch": 0.25311203319502074,
-      "grad_norm": 1.1635496616363525,
-      "learning_rate": 1.9292278122243705e-05,
-      "loss": 1.5605,
-      "step": 183
-    },
-    {
-      "epoch": 0.2544951590594744,
-      "grad_norm": 1.1545699834823608,
-      "learning_rate": 1.9284172036619597e-05,
-      "loss": 1.6157,
-      "step": 184
-    },
-    {
-      "epoch": 0.25587828492392806,
-      "grad_norm": 1.1379010677337646,
-      "learning_rate": 1.9276021515106635e-05,
-      "loss": 1.4126,
-      "step": 185
-    },
-    {
-      "epoch": 0.2572614107883817,
-      "grad_norm": 1.153485894203186,
-      "learning_rate": 1.926782659671484e-05,
-      "loss": 1.5618,
-      "step": 186
-    },
-    {
-      "epoch": 0.25864453665283543,
-      "grad_norm": 1.1652241945266724,
-      "learning_rate": 1.925958732066672e-05,
-      "loss": 1.6186,
-      "step": 187
-    },
-    {
-      "epoch": 0.2600276625172891,
-      "grad_norm": 1.1322290897369385,
-      "learning_rate": 1.9251303726397076e-05,
-      "loss": 1.5684,
-      "step": 188
-    },
-    {
-      "epoch": 0.26141078838174275,
-      "grad_norm": 1.0313161611557007,
-      "learning_rate": 1.924297585355284e-05,
-      "loss": 1.5584,
-      "step": 189
-    },
-    {
-      "epoch": 0.2627939142461964,
-      "grad_norm": 1.2757902145385742,
-      "learning_rate": 1.9234603741992864e-05,
-      "loss": 1.5329,
-      "step": 190
-    },
-    {
-      "epoch": 0.2641770401106501,
-      "grad_norm": 1.0278979539871216,
-      "learning_rate": 1.9226187431787727e-05,
-      "loss": 1.5397,
-      "step": 191
-    },
-    {
-      "epoch": 0.26556016597510373,
-      "grad_norm": 1.227102518081665,
-      "learning_rate": 1.9217726963219567e-05,
-      "loss": 1.4856,
-      "step": 192
-    },
-    {
-      "epoch": 0.2669432918395574,
-      "grad_norm": 1.2148761749267578,
-      "learning_rate": 1.9209222376781864e-05,
-      "loss": 1.511,
-      "step": 193
-    },
-    {
-      "epoch": 0.26832641770401106,
-      "grad_norm": 1.1404523849487305,
-      "learning_rate": 1.9200673713179245e-05,
-      "loss": 1.6732,
-      "step": 194
-    },
-    {
-      "epoch": 0.2697095435684647,
-      "grad_norm": 1.3106427192687988,
-      "learning_rate": 1.9192081013327325e-05,
-      "loss": 1.4351,
-      "step": 195
-    },
-    {
-      "epoch": 0.2710926694329184,
-      "grad_norm": 1.1526203155517578,
-      "learning_rate": 1.9183444318352458e-05,
-      "loss": 1.5936,
-      "step": 196
-    },
-    {
-      "epoch": 0.27247579529737204,
-      "grad_norm": 1.1274484395980835,
-      "learning_rate": 1.9174763669591583e-05,
-      "loss": 1.5429,
-      "step": 197
-    },
-    {
-      "epoch": 0.27385892116182575,
-      "grad_norm": 1.1309480667114258,
-      "learning_rate": 1.9166039108592008e-05,
-      "loss": 1.5481,
-      "step": 198
-    },
-    {
-      "epoch": 0.2752420470262794,
-      "grad_norm": 1.1612128019332886,
-      "learning_rate": 1.9157270677111214e-05,
-      "loss": 1.4978,
-      "step": 199
-    },
-    {
-      "epoch": 0.2766251728907331,
-      "grad_norm": 1.1628715991973877,
-      "learning_rate": 1.9148458417116645e-05,
-      "loss": 1.5816,
-      "step": 200
-    },
-    {
-      "epoch": 0.27800829875518673,
-      "grad_norm": 1.3227707147598267,
-      "learning_rate": 1.9139602370785536e-05,
-      "loss": 1.4715,
-      "step": 201
-    },
-    {
-      "epoch": 0.2793914246196404,
-      "grad_norm": 1.2166869640350342,
-      "learning_rate": 1.9130702580504678e-05,
-      "loss": 1.6047,
-      "step": 202
-    },
-    {
-      "epoch": 0.28077455048409405,
-      "grad_norm": 1.0572668313980103,
-      "learning_rate": 1.9121759088870228e-05,
-      "loss": 1.6119,
-      "step": 203
-    },
-    {
-      "epoch": 0.2821576763485477,
-      "grad_norm": 1.2285197973251343,
-      "learning_rate": 1.911277193868751e-05,
-      "loss": 1.558,
-      "step": 204
-    },
-    {
-      "epoch": 0.2835408022130014,
-      "grad_norm": 1.1655161380767822,
-      "learning_rate": 1.9103741172970816e-05,
-      "loss": 1.5855,
-      "step": 205
-    },
-    {
-      "epoch": 0.28492392807745504,
-      "grad_norm": 1.232515573501587,
-      "learning_rate": 1.9094666834943177e-05,
-      "loss": 1.5637,
-      "step": 206
-    },
-    {
-      "epoch": 0.2863070539419087,
-      "grad_norm": 1.2622716426849365,
-      "learning_rate": 1.9085548968036174e-05,
-      "loss": 1.4948,
-      "step": 207
-    },
-    {
-      "epoch": 0.28769017980636236,
-      "grad_norm": 1.2562229633331299,
-      "learning_rate": 1.9076387615889728e-05,
-      "loss": 1.5645,
-      "step": 208
-    },
-    {
-      "epoch": 0.28907330567081607,
-      "grad_norm": 1.1468416452407837,
-      "learning_rate": 1.9067182822351884e-05,
-      "loss": 1.6337,
-      "step": 209
-    },
-    {
-      "epoch": 0.29045643153526973,
-      "grad_norm": 1.250429391860962,
-      "learning_rate": 1.9057934631478616e-05,
-      "loss": 1.5544,
-      "step": 210
-    },
-    {
-      "epoch": 0.2918395573997234,
-      "grad_norm": 1.1409355401992798,
-      "learning_rate": 1.9048643087533594e-05,
-      "loss": 1.5368,
-      "step": 211
-    },
-    {
-      "epoch": 0.29322268326417705,
-      "grad_norm": 1.1958142518997192,
-      "learning_rate": 1.903930823498799e-05,
-      "loss": 1.5344,
-      "step": 212
-    },
-    {
-      "epoch": 0.2946058091286307,
-      "grad_norm": 1.1556063890457153,
-      "learning_rate": 1.9029930118520266e-05,
-      "loss": 1.591,
-      "step": 213
-    },
-    {
-      "epoch": 0.2959889349930844,
-      "grad_norm": 1.1471951007843018,
-      "learning_rate": 1.9020508783015942e-05,
-      "loss": 1.5892,
-      "step": 214
-    },
-    {
-      "epoch": 0.29737206085753803,
-      "grad_norm": 1.114162802696228,
-      "learning_rate": 1.9011044273567405e-05,
-      "loss": 1.4876,
-      "step": 215
-    },
-    {
-      "epoch": 0.2987551867219917,
-      "grad_norm": 1.2788852453231812,
-      "learning_rate": 1.9001536635473664e-05,
-      "loss": 1.5746,
-      "step": 216
-    },
-    {
-      "epoch": 0.30013831258644535,
-      "grad_norm": 1.2140027284622192,
-      "learning_rate": 1.8991985914240166e-05,
-      "loss": 1.5842,
-      "step": 217
-    },
-    {
-      "epoch": 0.301521438450899,
-      "grad_norm": 1.3167227506637573,
-      "learning_rate": 1.898239215557856e-05,
-      "loss": 1.5455,
-      "step": 218
-    },
-    {
-      "epoch": 0.3029045643153527,
-      "grad_norm": 1.22543466091156,
-      "learning_rate": 1.8972755405406475e-05,
-      "loss": 1.5469,
-      "step": 219
-    },
-    {
-      "epoch": 0.30428769017980634,
-      "grad_norm": 1.2558698654174805,
-      "learning_rate": 1.8963075709847308e-05,
-      "loss": 1.4874,
-      "step": 220
-    },
-    {
-      "epoch": 0.30567081604426005,
-      "grad_norm": 1.1993651390075684,
-      "learning_rate": 1.8953353115230005e-05,
-      "loss": 1.5977,
-      "step": 221
-    },
-    {
-      "epoch": 0.3070539419087137,
-      "grad_norm": 1.422924280166626,
-      "learning_rate": 1.894358766808883e-05,
-      "loss": 1.371,
-      "step": 222
-    },
-    {
-      "epoch": 0.3084370677731674,
-      "grad_norm": 1.2783629894256592,
-      "learning_rate": 1.893377941516315e-05,
-      "loss": 1.536,
-      "step": 223
-    },
-    {
-      "epoch": 0.30982019363762103,
-      "grad_norm": 1.2318593263626099,
-      "learning_rate": 1.892392840339721e-05,
-      "loss": 1.6051,
-      "step": 224
-    },
-    {
-      "epoch": 0.3112033195020747,
-      "grad_norm": 1.2269023656845093,
-      "learning_rate": 1.8914034679939905e-05,
-      "loss": 1.4488,
-      "step": 225
-    },
-    {
-      "epoch": 0.31258644536652835,
-      "grad_norm": 1.2582275867462158,
-      "learning_rate": 1.8904098292144556e-05,
-      "loss": 1.4979,
-      "step": 226
-    },
-    {
-      "epoch": 0.313969571230982,
-      "grad_norm": 1.4117553234100342,
-      "learning_rate": 1.889411928756869e-05,
-      "loss": 1.5858,
-      "step": 227
-    },
-    {
-      "epoch": 0.3153526970954357,
-      "grad_norm": 1.472165584564209,
-      "learning_rate": 1.8884097713973798e-05,
-      "loss": 1.4512,
-      "step": 228
-    },
-    {
-      "epoch": 0.31673582295988933,
-      "grad_norm": 1.240386962890625,
-      "learning_rate": 1.8874033619325124e-05,
-      "loss": 1.5531,
-      "step": 229
-    },
-    {
-      "epoch": 0.318118948824343,
-      "grad_norm": 1.289332389831543,
-      "learning_rate": 1.8863927051791418e-05,
-      "loss": 1.4774,
-      "step": 230
-    },
-    {
-      "epoch": 0.31950207468879666,
-      "grad_norm": 1.5051286220550537,
-      "learning_rate": 1.8853778059744716e-05,
-      "loss": 1.2835,
-      "step": 231
-    },
-    {
-      "epoch": 0.32088520055325037,
-      "grad_norm": 1.4394385814666748,
-      "learning_rate": 1.884358669176011e-05,
-      "loss": 1.4348,
-      "step": 232
-    },
-    {
-      "epoch": 0.32226832641770403,
-      "grad_norm": 1.3271335363388062,
-      "learning_rate": 1.8833352996615507e-05,
-      "loss": 1.4749,
-      "step": 233
-    },
-    {
-      "epoch": 0.3236514522821577,
-      "grad_norm": 1.3286136388778687,
-      "learning_rate": 1.88230770232914e-05,
-      "loss": 1.5111,
-      "step": 234
-    },
-    {
-      "epoch": 0.32503457814661135,
-      "grad_norm": 1.2873828411102295,
-      "learning_rate": 1.8812758820970637e-05,
-      "loss": 1.4664,
-      "step": 235
-    },
-    {
-      "epoch": 0.326417704011065,
-      "grad_norm": 1.3673721551895142,
-      "learning_rate": 1.8802398439038175e-05,
-      "loss": 1.3966,
-      "step": 236
-    },
-    {
-      "epoch": 0.3278008298755187,
-      "grad_norm": 1.306979775428772,
-      "learning_rate": 1.879199592708087e-05,
-      "loss": 1.5416,
-      "step": 237
-    },
-    {
-      "epoch": 0.32918395573997233,
-      "grad_norm": 1.191098690032959,
-      "learning_rate": 1.8781551334887204e-05,
-      "loss": 1.4901,
-      "step": 238
-    },
-    {
-      "epoch": 0.330567081604426,
-      "grad_norm": 1.3898661136627197,
-      "learning_rate": 1.8771064712447054e-05,
-      "loss": 1.6548,
-      "step": 239
-    },
-    {
-      "epoch": 0.33195020746887965,
-      "grad_norm": 1.3041338920593262,
-      "learning_rate": 1.876053610995149e-05,
-      "loss": 1.4909,
-      "step": 240
-    },
-    {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 1.3957127332687378,
-      "learning_rate": 1.8749965577792482e-05,
-      "loss": 1.5244,
-      "step": 241
-    },
-    {
-      "epoch": 0.334716459197787,
-      "grad_norm": 1.3891314268112183,
-      "learning_rate": 1.87393531665627e-05,
-      "loss": 1.4816,
-      "step": 242
-    },
-    {
-      "epoch": 0.3360995850622407,
-      "grad_norm": 1.483026385307312,
-      "learning_rate": 1.872869892705525e-05,
-      "loss": 1.3794,
-      "step": 243
-    },
-    {
-      "epoch": 0.33748271092669435,
-      "grad_norm": 1.2564777135849,
-      "learning_rate": 1.8718002910263426e-05,
-      "loss": 1.5489,
-      "step": 244
-    },
-    {
-      "epoch": 0.338865836791148,
-      "grad_norm": 1.255433440208435,
-      "learning_rate": 1.8707265167380497e-05,
-      "loss": 1.4259,
-      "step": 245
-    },
-    {
-      "epoch": 0.34024896265560167,
-      "grad_norm": 1.4052809476852417,
-      "learning_rate": 1.869648574979942e-05,
-      "loss": 1.6745,
-      "step": 246
-    },
-    {
-      "epoch": 0.34163208852005533,
-      "grad_norm": 1.368813157081604,
-      "learning_rate": 1.8685664709112637e-05,
-      "loss": 1.5183,
-      "step": 247
-    },
-    {
-      "epoch": 0.343015214384509,
-      "grad_norm": 1.3451085090637207,
-      "learning_rate": 1.8674802097111784e-05,
-      "loss": 1.5095,
-      "step": 248
-    },
-    {
-      "epoch": 0.34439834024896265,
-      "grad_norm": 1.249712586402893,
-      "learning_rate": 1.8663897965787483e-05,
-      "loss": 1.5347,
-      "step": 249
-    },
-    {
-      "epoch": 0.3457814661134163,
-      "grad_norm": 1.2913545370101929,
-      "learning_rate": 1.865295236732907e-05,
-      "loss": 1.5025,
-      "step": 250
-    },
-    {
-      "epoch": 0.34716459197787,
-      "grad_norm": 1.3476046323776245,
-      "learning_rate": 1.8641965354124346e-05,
-      "loss": 1.5246,
-      "step": 251
-    },
-    {
-      "epoch": 0.34854771784232363,
-      "grad_norm": 1.253641963005066,
-      "learning_rate": 1.8630936978759337e-05,
-      "loss": 1.4977,
-      "step": 252
-    },
-    {
-      "epoch": 0.3499308437067773,
-      "grad_norm": 1.3167630434036255,
-      "learning_rate": 1.8619867294018035e-05,
-      "loss": 1.451,
-      "step": 253
-    },
-    {
-      "epoch": 0.35131396957123096,
-      "grad_norm": 1.3077702522277832,
-      "learning_rate": 1.8608756352882152e-05,
-      "loss": 1.5264,
-      "step": 254
-    },
-    {
-      "epoch": 0.35269709543568467,
-      "grad_norm": 1.2955269813537598,
-      "learning_rate": 1.8597604208530845e-05,
-      "loss": 1.4943,
-      "step": 255
-    },
-    {
-      "epoch": 0.35408022130013833,
-      "grad_norm": 1.418426513671875,
-      "learning_rate": 1.85864109143405e-05,
-      "loss": 1.5298,
-      "step": 256
-    },
-    {
-      "epoch": 0.355463347164592,
-      "grad_norm": 1.3700319528579712,
-      "learning_rate": 1.8575176523884432e-05,
-      "loss": 1.4887,
-      "step": 257
-    },
-    {
-      "epoch": 0.35684647302904565,
-      "grad_norm": 1.3327845335006714,
-      "learning_rate": 1.8563901090932673e-05,
-      "loss": 1.5672,
-      "step": 258
-    },
-    {
-      "epoch": 0.3582295988934993,
-      "grad_norm": 1.4686397314071655,
-      "learning_rate": 1.8552584669451675e-05,
-      "loss": 1.3965,
-      "step": 259
-    },
-    {
-      "epoch": 0.359612724757953,
-      "grad_norm": 1.267582654953003,
-      "learning_rate": 1.854122731360408e-05,
-      "loss": 1.5229,
-      "step": 260
-    },
-    {
-      "epoch": 0.36099585062240663,
-      "grad_norm": 1.347273826599121,
-      "learning_rate": 1.8529829077748442e-05,
-      "loss": 1.4324,
-      "step": 261
-    },
-    {
-      "epoch": 0.3623789764868603,
-      "grad_norm": 1.4890333414077759,
-      "learning_rate": 1.851839001643898e-05,
-      "loss": 1.4877,
-      "step": 262
-    },
-    {
-      "epoch": 0.36376210235131395,
-      "grad_norm": 1.374370813369751,
-      "learning_rate": 1.850691018442531e-05,
-      "loss": 1.5164,
-      "step": 263
-    },
-    {
-      "epoch": 0.3651452282157676,
-      "grad_norm": 1.3033865690231323,
-      "learning_rate": 1.8495389636652185e-05,
-      "loss": 1.4799,
-      "step": 264
-    },
-    {
-      "epoch": 0.3665283540802213,
-      "grad_norm": 1.321038007736206,
-      "learning_rate": 1.8483828428259235e-05,
-      "loss": 1.4816,
-      "step": 265
-    },
-    {
-      "epoch": 0.367911479944675,
-      "grad_norm": 1.4516509771347046,
-      "learning_rate": 1.847222661458069e-05,
-      "loss": 1.3904,
-      "step": 266
-    },
-    {
-      "epoch": 0.36929460580912865,
-      "grad_norm": 1.3580619096755981,
-      "learning_rate": 1.8460584251145137e-05,
-      "loss": 1.501,
-      "step": 267
-    },
-    {
-      "epoch": 0.3706777316735823,
-      "grad_norm": 1.3735978603363037,
-      "learning_rate": 1.8448901393675233e-05,
-      "loss": 1.5034,
-      "step": 268
-    },
-    {
-      "epoch": 0.37206085753803597,
-      "grad_norm": 1.522796630859375,
-      "learning_rate": 1.8437178098087452e-05,
-      "loss": 1.5287,
-      "step": 269
-    },
-    {
-      "epoch": 0.37344398340248963,
-      "grad_norm": 1.4429105520248413,
-      "learning_rate": 1.8425414420491817e-05,
-      "loss": 1.4459,
-      "step": 270
-    },
-    {
-      "epoch": 0.3748271092669433,
-      "grad_norm": 1.3284740447998047,
-      "learning_rate": 1.841361041719161e-05,
-      "loss": 1.4334,
-      "step": 271
-    },
-    {
-      "epoch": 0.37621023513139695,
-      "grad_norm": 1.262669563293457,
-      "learning_rate": 1.8401766144683145e-05,
-      "loss": 1.5067,
-      "step": 272
-    },
-    {
-      "epoch": 0.3775933609958506,
-      "grad_norm": 1.5458612442016602,
-      "learning_rate": 1.8389881659655456e-05,
-      "loss": 1.5672,
-      "step": 273
-    },
-    {
-      "epoch": 0.3789764868603043,
-      "grad_norm": 1.3672326803207397,
-      "learning_rate": 1.8377957018990043e-05,
-      "loss": 1.4507,
-      "step": 274
-    },
-    {
-      "epoch": 0.38035961272475793,
-      "grad_norm": 1.5172324180603027,
-      "learning_rate": 1.83659922797606e-05,
-      "loss": 1.5094,
-      "step": 275
-    },
-    {
-      "epoch": 0.3817427385892116,
-      "grad_norm": 1.3924052715301514,
-      "learning_rate": 1.8353987499232747e-05,
-      "loss": 1.4878,
-      "step": 276
-    },
-    {
-      "epoch": 0.3831258644536653,
-      "grad_norm": 1.601944923400879,
-      "learning_rate": 1.834194273486374e-05,
-      "loss": 1.2526,
-      "step": 277
-    },
-    {
-      "epoch": 0.38450899031811897,
-      "grad_norm": 1.3992019891738892,
-      "learning_rate": 1.8329858044302212e-05,
-      "loss": 1.4725,
-      "step": 278
-    },
-    {
-      "epoch": 0.38589211618257263,
-      "grad_norm": 1.2960071563720703,
-      "learning_rate": 1.8317733485387893e-05,
-      "loss": 1.4154,
-      "step": 279
-    },
-    {
-      "epoch": 0.3872752420470263,
-      "grad_norm": 1.7117806673049927,
-      "learning_rate": 1.830556911615132e-05,
-      "loss": 1.4874,
-      "step": 280
-    },
-    {
-      "epoch": 0.38865836791147995,
-      "grad_norm": 1.374509572982788,
-      "learning_rate": 1.8293364994813584e-05,
-      "loss": 1.518,
-      "step": 281
-    },
-    {
-      "epoch": 0.3900414937759336,
-      "grad_norm": 1.2234368324279785,
-      "learning_rate": 1.8281121179786024e-05,
-      "loss": 1.1919,
-      "step": 282
-    },
-    {
-      "epoch": 0.3914246196403873,
-      "grad_norm": 1.4747345447540283,
-      "learning_rate": 1.826883772966997e-05,
-      "loss": 1.5335,
-      "step": 283
-    },
-    {
-      "epoch": 0.39280774550484093,
-      "grad_norm": 1.4171055555343628,
-      "learning_rate": 1.825651470325645e-05,
-      "loss": 1.4572,
-      "step": 284
-    },
-    {
-      "epoch": 0.3941908713692946,
-      "grad_norm": 1.4510976076126099,
-      "learning_rate": 1.8244152159525916e-05,
-      "loss": 1.4552,
-      "step": 285
-    },
-    {
-      "epoch": 0.39557399723374825,
-      "grad_norm": 1.3574010133743286,
-      "learning_rate": 1.823175015764795e-05,
-      "loss": 1.4206,
-      "step": 286
-    },
-    {
-      "epoch": 0.3969571230982019,
-      "grad_norm": 1.4223361015319824,
-      "learning_rate": 1.821930875698099e-05,
-      "loss": 1.5357,
-      "step": 287
-    },
-    {
-      "epoch": 0.3983402489626556,
-      "grad_norm": 1.4017705917358398,
-      "learning_rate": 1.8206828017072057e-05,
-      "loss": 1.4742,
-      "step": 288
-    },
-    {
-      "epoch": 0.3997233748271093,
-      "grad_norm": 1.5125157833099365,
-      "learning_rate": 1.819430799765644e-05,
-      "loss": 1.4549,
-      "step": 289
-    },
-    {
-      "epoch": 0.40110650069156295,
-      "grad_norm": 1.6485176086425781,
-      "learning_rate": 1.818174875865744e-05,
-      "loss": 1.4493,
-      "step": 290
-    },
-    {
-      "epoch": 0.4024896265560166,
-      "grad_norm": 1.5857341289520264,
-      "learning_rate": 1.8169150360186062e-05,
-      "loss": 1.5252,
-      "step": 291
-    },
-    {
-      "epoch": 0.40387275242047027,
-      "grad_norm": 1.4653831720352173,
-      "learning_rate": 1.815651286254074e-05,
-      "loss": 1.5078,
-      "step": 292
-    },
-    {
-      "epoch": 0.40525587828492393,
-      "grad_norm": 1.3680520057678223,
-      "learning_rate": 1.8143836326207048e-05,
-      "loss": 1.5462,
-      "step": 293
-    },
-    {
-      "epoch": 0.4066390041493776,
-      "grad_norm": 1.434511661529541,
-      "learning_rate": 1.8131120811857398e-05,
-      "loss": 1.5447,
-      "step": 294
-    },
-    {
-      "epoch": 0.40802213001383125,
-      "grad_norm": 1.527446985244751,
-      "learning_rate": 1.8118366380350773e-05,
-      "loss": 1.4645,
-      "step": 295
-    },
-    {
-      "epoch": 0.4094052558782849,
-      "grad_norm": 1.2728767395019531,
-      "learning_rate": 1.81055730927324e-05,
-      "loss": 1.1936,
-      "step": 296
-    },
-    {
-      "epoch": 0.4107883817427386,
-      "grad_norm": 1.5309098958969116,
-      "learning_rate": 1.8092741010233496e-05,
-      "loss": 1.3846,
-      "step": 297
-    },
-    {
-      "epoch": 0.41217150760719223,
-      "grad_norm": 1.6225330829620361,
-      "learning_rate": 1.8079870194270958e-05,
-      "loss": 1.455,
-      "step": 298
-    },
-    {
-      "epoch": 0.4135546334716459,
-      "grad_norm": 1.4223120212554932,
-      "learning_rate": 1.806696070644706e-05,
-      "loss": 1.4813,
-      "step": 299
-    },
-    {
-      "epoch": 0.4149377593360996,
-      "grad_norm": 1.55156672000885,
-      "learning_rate": 1.8054012608549167e-05,
-      "loss": 1.5128,
-      "step": 300
-    },
-    {
-      "epoch": 0.41632088520055327,
-      "grad_norm": 1.8663321733474731,
-      "learning_rate": 1.804102596254945e-05,
-      "loss": 1.3645,
-      "step": 301
-    },
-    {
-      "epoch": 0.41770401106500693,
-      "grad_norm": 1.597420334815979,
-      "learning_rate": 1.802800083060457e-05,
-      "loss": 1.5162,
-      "step": 302
-    },
-    {
-      "epoch": 0.4190871369294606,
-      "grad_norm": 1.5862244367599487,
-      "learning_rate": 1.8014937275055393e-05,
-      "loss": 1.4336,
-      "step": 303
-    },
-    {
-      "epoch": 0.42047026279391425,
-      "grad_norm": 1.530361533164978,
-      "learning_rate": 1.8001835358426688e-05,
-      "loss": 1.5836,
-      "step": 304
-    },
-    {
-      "epoch": 0.4218533886583679,
-      "grad_norm": 1.4132399559020996,
-      "learning_rate": 1.798869514342682e-05,
-      "loss": 1.5451,
-      "step": 305
-    },
-    {
-      "epoch": 0.42323651452282157,
-      "grad_norm": 1.5127052068710327,
-      "learning_rate": 1.7975516692947478e-05,
-      "loss": 1.3983,
-      "step": 306
-    },
-    {
-      "epoch": 0.42461964038727523,
-      "grad_norm": 1.491930365562439,
-      "learning_rate": 1.7962300070063325e-05,
-      "loss": 1.5258,
-      "step": 307
-    },
-    {
-      "epoch": 0.4260027662517289,
-      "grad_norm": 1.5085482597351074,
-      "learning_rate": 1.7949045338031744e-05,
-      "loss": 1.417,
-      "step": 308
-    },
-    {
-      "epoch": 0.42738589211618255,
-      "grad_norm": 1.5667784214019775,
-      "learning_rate": 1.793575256029252e-05,
-      "loss": 1.3179,
-      "step": 309
-    },
-    {
-      "epoch": 0.4287690179806362,
-      "grad_norm": 1.6200470924377441,
-      "learning_rate": 1.7922421800467515e-05,
-      "loss": 1.4283,
-      "step": 310
-    },
-    {
-      "epoch": 0.43015214384508993,
-      "grad_norm": 1.7292104959487915,
-      "learning_rate": 1.790905312236039e-05,
-      "loss": 1.4993,
-      "step": 311
-    },
-    {
-      "epoch": 0.4315352697095436,
-      "grad_norm": 1.44505774974823,
-      "learning_rate": 1.7895646589956294e-05,
-      "loss": 1.5516,
-      "step": 312
-    },
-    {
-      "epoch": 0.43291839557399725,
-      "grad_norm": 1.5757036209106445,
-      "learning_rate": 1.7882202267421544e-05,
-      "loss": 1.4155,
-      "step": 313
-    },
-    {
-      "epoch": 0.4343015214384509,
-      "grad_norm": 1.7449678182601929,
-      "learning_rate": 1.7868720219103343e-05,
-      "loss": 1.4392,
-      "step": 314
-    },
-    {
-      "epoch": 0.43568464730290457,
-      "grad_norm": 1.5077069997787476,
-      "learning_rate": 1.7855200509529442e-05,
-      "loss": 1.4805,
-      "step": 315
-    },
-    {
-      "epoch": 0.43706777316735823,
-      "grad_norm": 1.488542079925537,
-      "learning_rate": 1.7841643203407854e-05,
-      "loss": 1.4102,
-      "step": 316
-    },
-    {
-      "epoch": 0.4384508990318119,
-      "grad_norm": 1.5566229820251465,
-      "learning_rate": 1.7828048365626536e-05,
-      "loss": 1.4242,
-      "step": 317
-    },
-    {
-      "epoch": 0.43983402489626555,
-      "grad_norm": 1.5665713548660278,
-      "learning_rate": 1.7814416061253076e-05,
-      "loss": 1.4923,
-      "step": 318
-    },
-    {
-      "epoch": 0.4412171507607192,
-      "grad_norm": 1.4140158891677856,
-      "learning_rate": 1.780074635553439e-05,
-      "loss": 1.5374,
-      "step": 319
-    },
-    {
-      "epoch": 0.4426002766251729,
-      "grad_norm": 1.3860362768173218,
-      "learning_rate": 1.77870393138964e-05,
-      "loss": 1.564,
-      "step": 320
-    },
-    {
-      "epoch": 0.44398340248962653,
-      "grad_norm": 1.4771426916122437,
-      "learning_rate": 1.7773295001943725e-05,
-      "loss": 1.4382,
-      "step": 321
-    },
-    {
-      "epoch": 0.44536652835408025,
-      "grad_norm": 1.4840291738510132,
-      "learning_rate": 1.7759513485459367e-05,
-      "loss": 1.4851,
-      "step": 322
-    },
-    {
-      "epoch": 0.4467496542185339,
-      "grad_norm": 1.6433287858963013,
-      "learning_rate": 1.77456948304044e-05,
-      "loss": 1.4411,
-      "step": 323
-    },
-    {
-      "epoch": 0.44813278008298757,
-      "grad_norm": 1.5386046171188354,
-      "learning_rate": 1.7731839102917646e-05,
-      "loss": 1.5255,
-      "step": 324
-    },
-    {
-      "epoch": 0.44951590594744123,
-      "grad_norm": 1.466110110282898,
-      "learning_rate": 1.7717946369315365e-05,
-      "loss": 1.4595,
-      "step": 325
-    },
-    {
-      "epoch": 0.4508990318118949,
-      "grad_norm": 1.6857801675796509,
-      "learning_rate": 1.7704016696090936e-05,
-      "loss": 1.5345,
-      "step": 326
-    },
-    {
-      "epoch": 0.45228215767634855,
-      "grad_norm": 1.8097199201583862,
-      "learning_rate": 1.769005014991454e-05,
-      "loss": 1.3434,
-      "step": 327
-    },
-    {
-      "epoch": 0.4536652835408022,
-      "grad_norm": 1.843514084815979,
-      "learning_rate": 1.7676046797632834e-05,
-      "loss": 1.3947,
-      "step": 328
-    },
-    {
-      "epoch": 0.45504840940525587,
-      "grad_norm": 1.476454496383667,
-      "learning_rate": 1.7662006706268642e-05,
-      "loss": 1.5038,
-      "step": 329
-    },
-    {
-      "epoch": 0.45643153526970953,
-      "grad_norm": 1.6218907833099365,
-      "learning_rate": 1.7647929943020625e-05,
-      "loss": 1.4362,
-      "step": 330
-    },
-    {
-      "epoch": 0.4578146611341632,
-      "grad_norm": 1.481621503829956,
-      "learning_rate": 1.7633816575262966e-05,
-      "loss": 1.4619,
-      "step": 331
-    },
-    {
-      "epoch": 0.45919778699861685,
-      "grad_norm": 1.5661985874176025,
-      "learning_rate": 1.7619666670545034e-05,
-      "loss": 1.4317,
-      "step": 332
-    },
-    {
-      "epoch": 0.4605809128630705,
-      "grad_norm": 1.580833911895752,
-      "learning_rate": 1.7605480296591092e-05,
-      "loss": 1.3814,
-      "step": 333
-    },
-    {
-      "epoch": 0.46196403872752423,
-      "grad_norm": 1.473464012145996,
-      "learning_rate": 1.759125752129993e-05,
-      "loss": 1.4584,
-      "step": 334
-    },
-    {
-      "epoch": 0.4633471645919779,
-      "grad_norm": 1.6730810403823853,
-      "learning_rate": 1.757699841274458e-05,
-      "loss": 1.3358,
-      "step": 335
-    },
-    {
-      "epoch": 0.46473029045643155,
-      "grad_norm": 1.4953693151474,
-      "learning_rate": 1.7562703039171955e-05,
-      "loss": 1.5387,
-      "step": 336
-    },
-    {
-      "epoch": 0.4661134163208852,
-      "grad_norm": 1.5851799249649048,
-      "learning_rate": 1.7548371469002553e-05,
-      "loss": 1.5365,
-      "step": 337
-    },
-    {
-      "epoch": 0.46749654218533887,
-      "grad_norm": 1.5487456321716309,
-      "learning_rate": 1.753400377083011e-05,
-      "loss": 1.4721,
-      "step": 338
-    },
-    {
-      "epoch": 0.46887966804979253,
-      "grad_norm": 1.6268632411956787,
-      "learning_rate": 1.7519600013421282e-05,
-      "loss": 1.3934,
-      "step": 339
-    },
-    {
-      "epoch": 0.4702627939142462,
-      "grad_norm": 1.68012535572052,
-      "learning_rate": 1.7505160265715303e-05,
-      "loss": 1.5813,
-      "step": 340
-    },
-    {
-      "epoch": 0.47164591977869985,
-      "grad_norm": 1.6492486000061035,
-      "learning_rate": 1.7490684596823678e-05,
-      "loss": 1.4777,
-      "step": 341
-    },
-    {
-      "epoch": 0.4730290456431535,
-      "grad_norm": 1.950250267982483,
-      "learning_rate": 1.747617307602982e-05,
-      "loss": 1.5284,
-      "step": 342
-    },
-    {
-      "epoch": 0.47441217150760717,
-      "grad_norm": 1.640457272529602,
-      "learning_rate": 1.7461625772788755e-05,
-      "loss": 1.4236,
-      "step": 343
-    },
-    {
-      "epoch": 0.47579529737206083,
-      "grad_norm": 1.4680976867675781,
-      "learning_rate": 1.7447042756726756e-05,
-      "loss": 1.5607,
-      "step": 344
-    },
-    {
-      "epoch": 0.47717842323651455,
-      "grad_norm": 1.6982401609420776,
-      "learning_rate": 1.743242409764103e-05,
-      "loss": 1.511,
-      "step": 345
-    },
-    {
-      "epoch": 0.4785615491009682,
-      "grad_norm": 1.6441048383712769,
-      "learning_rate": 1.741776986549938e-05,
-      "loss": 1.4762,
-      "step": 346
-    },
-    {
-      "epoch": 0.47994467496542187,
-      "grad_norm": 1.6681658029556274,
-      "learning_rate": 1.7403080130439874e-05,
-      "loss": 1.5238,
-      "step": 347
-    },
-    {
-      "epoch": 0.48132780082987553,
-      "grad_norm": 1.6259121894836426,
-      "learning_rate": 1.7388354962770488e-05,
-      "loss": 1.459,
-      "step": 348
-    },
-    {
-      "epoch": 0.4827109266943292,
-      "grad_norm": 1.6403274536132812,
-      "learning_rate": 1.7373594432968798e-05,
-      "loss": 1.4974,
-      "step": 349
-    },
-    {
-      "epoch": 0.48409405255878285,
-      "grad_norm": 1.585744857788086,
-      "learning_rate": 1.735879861168163e-05,
-      "loss": 1.5599,
-      "step": 350
-    },
-    {
-      "epoch": 0.4854771784232365,
-      "grad_norm": 1.6174416542053223,
-      "learning_rate": 1.7343967569724716e-05,
-      "loss": 1.475,
-      "step": 351
-    },
-    {
-      "epoch": 0.48686030428769017,
-      "grad_norm": 1.6598519086837769,
-      "learning_rate": 1.7329101378082374e-05,
-      "loss": 1.1199,
-      "step": 352
-    },
-    {
-      "epoch": 0.48824343015214383,
-      "grad_norm": 1.6443450450897217,
-      "learning_rate": 1.731420010790713e-05,
-      "loss": 1.527,
-      "step": 353
-    },
-    {
-      "epoch": 0.4896265560165975,
-      "grad_norm": 1.7666903734207153,
-      "learning_rate": 1.729926383051943e-05,
-      "loss": 1.4852,
-      "step": 354
-    },
-    {
-      "epoch": 0.49100968188105115,
-      "grad_norm": 1.7390762567520142,
-      "learning_rate": 1.7284292617407253e-05,
-      "loss": 1.369,
-      "step": 355
-    },
-    {
-      "epoch": 0.49239280774550487,
-      "grad_norm": 1.7170841693878174,
-      "learning_rate": 1.7269286540225805e-05,
-      "loss": 1.4483,
-      "step": 356
-    },
-    {
-      "epoch": 0.49377593360995853,
-      "grad_norm": 1.6285370588302612,
-      "learning_rate": 1.725424567079714e-05,
-      "loss": 1.4945,
-      "step": 357
-    },
-    {
-      "epoch": 0.4951590594744122,
-      "grad_norm": 1.6297553777694702,
-      "learning_rate": 1.723917008110984e-05,
-      "loss": 1.4291,
-      "step": 358
-    },
-    {
-      "epoch": 0.49654218533886585,
-      "grad_norm": 1.5169857740402222,
-      "learning_rate": 1.722405984331867e-05,
-      "loss": 1.4683,
-      "step": 359
-    },
-    {
-      "epoch": 0.4979253112033195,
-      "grad_norm": 1.729242205619812,
-      "learning_rate": 1.720891502974423e-05,
-      "loss": 1.3923,
-      "step": 360
-    },
-    {
-      "epoch": 0.49930843706777317,
-      "grad_norm": 1.7176477909088135,
-      "learning_rate": 1.7193735712872598e-05,
-      "loss": 1.4261,
-      "step": 361
-    },
-    {
-      "epoch": 0.5006915629322268,
-      "grad_norm": 1.7728772163391113,
-      "learning_rate": 1.7178521965354992e-05,
-      "loss": 1.4334,
-      "step": 362
-    },
-    {
-      "epoch": 0.5006915629322268,
-      "eval_loss": 1.147789478302002,
-      "eval_runtime": 108.7922,
-      "eval_samples_per_second": 13.273,
-      "eval_steps_per_second": 0.836,
-      "step": 362
-    },
-    {
-      "epoch": 0.5020746887966805,
-      "grad_norm": 1.6068779230117798,
-      "learning_rate": 1.7163273860007434e-05,
-      "loss": 1.4989,
-      "step": 363
-    },
-    {
-      "epoch": 0.5034578146611342,
-      "grad_norm": 1.751523494720459,
-      "learning_rate": 1.714799146981037e-05,
-      "loss": 1.4278,
-      "step": 364
-    },
-    {
-      "epoch": 0.5048409405255878,
-      "grad_norm": 1.6334980726242065,
-      "learning_rate": 1.7132674867908354e-05,
-      "loss": 1.4772,
-      "step": 365
-    },
-    {
-      "epoch": 0.5062240663900415,
-      "grad_norm": 1.5768158435821533,
-      "learning_rate": 1.7117324127609686e-05,
-      "loss": 1.4354,
-      "step": 366
-    },
-    {
-      "epoch": 0.5076071922544951,
-      "grad_norm": 1.7999041080474854,
-      "learning_rate": 1.710193932238605e-05,
-      "loss": 1.3682,
-      "step": 367
-    },
-    {
-      "epoch": 0.5089903181189488,
-      "grad_norm": 1.8190585374832153,
-      "learning_rate": 1.7086520525872173e-05,
-      "loss": 1.4741,
-      "step": 368
-    },
-    {
-      "epoch": 0.5103734439834025,
-      "grad_norm": 1.825280785560608,
-      "learning_rate": 1.7071067811865477e-05,
-      "loss": 1.3532,
-      "step": 369
-    },
-    {
-      "epoch": 0.5117565698478561,
-      "grad_norm": 1.862087607383728,
-      "learning_rate": 1.7055581254325716e-05,
-      "loss": 1.4157,
-      "step": 370
-    },
-    {
-      "epoch": 0.5131396957123098,
-      "grad_norm": 1.8708950281143188,
-      "learning_rate": 1.7040060927374626e-05,
-      "loss": 1.4732,
-      "step": 371
-    },
-    {
-      "epoch": 0.5145228215767634,
-      "grad_norm": 1.5391322374343872,
-      "learning_rate": 1.7024506905295566e-05,
-      "loss": 1.4484,
-      "step": 372
-    },
-    {
-      "epoch": 0.5159059474412172,
-      "grad_norm": 1.7950142621994019,
-      "learning_rate": 1.7008919262533174e-05,
-      "loss": 1.4488,
-      "step": 373
-    },
-    {
-      "epoch": 0.5172890733056709,
-      "grad_norm": 1.7594027519226074,
-      "learning_rate": 1.6993298073693005e-05,
-      "loss": 1.4757,
-      "step": 374
-    },
-    {
-      "epoch": 0.5186721991701245,
-      "grad_norm": 1.7801642417907715,
-      "learning_rate": 1.6977643413541156e-05,
-      "loss": 1.4611,
-      "step": 375
-    },
-    {
-      "epoch": 0.5200553250345782,
-      "grad_norm": 1.7385730743408203,
-      "learning_rate": 1.6961955357003948e-05,
-      "loss": 1.5696,
-      "step": 376
-    },
-    {
-      "epoch": 0.5214384508990318,
-      "grad_norm": 2.0058438777923584,
-      "learning_rate": 1.6946233979167516e-05,
-      "loss": 1.3498,
-      "step": 377
-    },
-    {
-      "epoch": 0.5228215767634855,
-      "grad_norm": 1.7878988981246948,
-      "learning_rate": 1.693047935527751e-05,
-      "loss": 1.4722,
-      "step": 378
-    },
-    {
-      "epoch": 0.5242047026279392,
-      "grad_norm": 1.900752067565918,
-      "learning_rate": 1.6914691560738675e-05,
-      "loss": 1.418,
-      "step": 379
-    },
-    {
-      "epoch": 0.5255878284923928,
-      "grad_norm": 1.7536234855651855,
-      "learning_rate": 1.6898870671114527e-05,
-      "loss": 1.4366,
-      "step": 380
-    },
-    {
-      "epoch": 0.5269709543568465,
-      "grad_norm": 1.7499502897262573,
-      "learning_rate": 1.6883016762126986e-05,
-      "loss": 1.3753,
-      "step": 381
-    },
-    {
-      "epoch": 0.5283540802213001,
-      "grad_norm": 1.7654677629470825,
-      "learning_rate": 1.6867129909656e-05,
-      "loss": 1.3705,
-      "step": 382
-    },
-    {
-      "epoch": 0.5297372060857538,
-      "grad_norm": 1.679612159729004,
-      "learning_rate": 1.6851210189739195e-05,
-      "loss": 1.5566,
-      "step": 383
-    },
-    {
-      "epoch": 0.5311203319502075,
-      "grad_norm": 1.8122912645339966,
-      "learning_rate": 1.6835257678571515e-05,
-      "loss": 1.4993,
-      "step": 384
-    },
-    {
-      "epoch": 0.5325034578146611,
-      "grad_norm": 1.8270299434661865,
-      "learning_rate": 1.681927245250484e-05,
-      "loss": 1.4321,
-      "step": 385
-    },
-    {
-      "epoch": 0.5338865836791148,
-      "grad_norm": 1.7611494064331055,
-      "learning_rate": 1.680325458804763e-05,
-      "loss": 1.4337,
-      "step": 386
-    },
-    {
-      "epoch": 0.5352697095435685,
-      "grad_norm": 1.650876760482788,
-      "learning_rate": 1.6787204161864562e-05,
-      "loss": 1.4233,
-      "step": 387
-    },
-    {
-      "epoch": 0.5366528354080221,
-      "grad_norm": 1.6672831773757935,
-      "learning_rate": 1.6771121250776163e-05,
-      "loss": 1.3721,
-      "step": 388
-    },
-    {
-      "epoch": 0.5380359612724758,
-      "grad_norm": 1.7278107404708862,
-      "learning_rate": 1.675500593175843e-05,
-      "loss": 1.51,
-      "step": 389
-    },
-    {
-      "epoch": 0.5394190871369294,
-      "grad_norm": 1.7551602125167847,
-      "learning_rate": 1.6738858281942477e-05,
-      "loss": 1.5643,
-      "step": 390
-    },
-    {
-      "epoch": 0.5408022130013831,
-      "grad_norm": 1.679298996925354,
-      "learning_rate": 1.6722678378614164e-05,
-      "loss": 1.466,
-      "step": 391
-    },
-    {
-      "epoch": 0.5421853388658368,
-      "grad_norm": 1.9446730613708496,
-      "learning_rate": 1.6706466299213718e-05,
-      "loss": 1.4915,
-      "step": 392
-    },
-    {
-      "epoch": 0.5435684647302904,
-      "grad_norm": 1.822323203086853,
-      "learning_rate": 1.6690222121335357e-05,
-      "loss": 1.4275,
-      "step": 393
-    },
-    {
-      "epoch": 0.5449515905947441,
-      "grad_norm": 2.0493321418762207,
-      "learning_rate": 1.6673945922726945e-05,
-      "loss": 1.3294,
-      "step": 394
-    },
-    {
-      "epoch": 0.5463347164591977,
-      "grad_norm": 1.8270610570907593,
-      "learning_rate": 1.6657637781289596e-05,
-      "loss": 1.4471,
-      "step": 395
-    },
-    {
-      "epoch": 0.5477178423236515,
-      "grad_norm": 1.9434516429901123,
-      "learning_rate": 1.6641297775077313e-05,
-      "loss": 1.4606,
-      "step": 396
-    },
-    {
-      "epoch": 0.5491009681881052,
-      "grad_norm": 1.714198350906372,
-      "learning_rate": 1.66249259822966e-05,
-      "loss": 1.4398,
-      "step": 397
-    },
-    {
-      "epoch": 0.5504840940525588,
-      "grad_norm": 1.787815809249878,
-      "learning_rate": 1.660852248130611e-05,
-      "loss": 1.3972,
-      "step": 398
-    },
-    {
-      "epoch": 0.5518672199170125,
-      "grad_norm": 1.9069771766662598,
-      "learning_rate": 1.6592087350616245e-05,
-      "loss": 1.4049,
-      "step": 399
-    },
-    {
-      "epoch": 0.5532503457814661,
-      "grad_norm": 1.8910343647003174,
-      "learning_rate": 1.6575620668888812e-05,
-      "loss": 1.3682,
-      "step": 400
-    },
-    {
-      "epoch": 0.5546334716459198,
-      "grad_norm": 1.7473318576812744,
-      "learning_rate": 1.6559122514936606e-05,
-      "loss": 1.3818,
-      "step": 401
-    },
-    {
-      "epoch": 0.5560165975103735,
-      "grad_norm": 1.792608380317688,
-      "learning_rate": 1.6542592967723065e-05,
-      "loss": 1.343,
-      "step": 402
-    },
-    {
-      "epoch": 0.5573997233748271,
-      "grad_norm": 1.8886102437973022,
-      "learning_rate": 1.6526032106361888e-05,
-      "loss": 1.3608,
-      "step": 403
-    },
-    {
-      "epoch": 0.5587828492392808,
-      "grad_norm": 1.8249069452285767,
-      "learning_rate": 1.6509440010116634e-05,
-      "loss": 1.2995,
-      "step": 404
-    },
-    {
-      "epoch": 0.5601659751037344,
-      "grad_norm": 1.9114660024642944,
-      "learning_rate": 1.649281675840037e-05,
-      "loss": 1.4432,
-      "step": 405
-    },
-    {
-      "epoch": 0.5615491009681881,
-      "grad_norm": 1.9942965507507324,
-      "learning_rate": 1.6476162430775278e-05,
-      "loss": 1.4114,
-      "step": 406
-    },
-    {
-      "epoch": 0.5629322268326418,
-      "grad_norm": 1.7948976755142212,
-      "learning_rate": 1.645947710695227e-05,
-      "loss": 1.3939,
-      "step": 407
-    },
-    {
-      "epoch": 0.5643153526970954,
-      "grad_norm": 1.7326204776763916,
-      "learning_rate": 1.6442760866790616e-05,
-      "loss": 1.4526,
-      "step": 408
-    },
-    {
-      "epoch": 0.5656984785615491,
-      "grad_norm": 1.7523624897003174,
-      "learning_rate": 1.6426013790297556e-05,
-      "loss": 1.5081,
-      "step": 409
-    },
-    {
-      "epoch": 0.5670816044260027,
-      "grad_norm": 2.0211379528045654,
-      "learning_rate": 1.6409235957627926e-05,
-      "loss": 1.281,
-      "step": 410
-    },
-    {
-      "epoch": 0.5684647302904564,
-      "grad_norm": 2.0155961513519287,
-      "learning_rate": 1.639242744908375e-05,
-      "loss": 1.327,
-      "step": 411
-    },
-    {
-      "epoch": 0.5698478561549101,
-      "grad_norm": 1.7048110961914062,
-      "learning_rate": 1.6375588345113895e-05,
-      "loss": 1.5404,
-      "step": 412
-    },
-    {
-      "epoch": 0.5712309820193637,
-      "grad_norm": 1.8413456678390503,
-      "learning_rate": 1.6358718726313645e-05,
-      "loss": 1.5062,
-      "step": 413
-    },
-    {
-      "epoch": 0.5726141078838174,
-      "grad_norm": 1.6750757694244385,
-      "learning_rate": 1.6341818673424342e-05,
-      "loss": 0.9839,
-      "step": 414
-    },
-    {
-      "epoch": 0.573997233748271,
-      "grad_norm": 1.7165724039077759,
-      "learning_rate": 1.6324888267332998e-05,
-      "loss": 1.4027,
-      "step": 415
-    },
-    {
-      "epoch": 0.5753803596127247,
-      "grad_norm": 1.8710963726043701,
-      "learning_rate": 1.630792758907189e-05,
-      "loss": 1.3251,
-      "step": 416
-    },
-    {
-      "epoch": 0.5767634854771784,
-      "grad_norm": 1.9592095613479614,
-      "learning_rate": 1.6290936719818182e-05,
-      "loss": 1.4502,
-      "step": 417
-    },
-    {
-      "epoch": 0.5781466113416321,
-      "grad_norm": 1.7474467754364014,
-      "learning_rate": 1.6273915740893557e-05,
-      "loss": 1.3636,
-      "step": 418
-    },
-    {
-      "epoch": 0.5795297372060858,
-      "grad_norm": 1.893127202987671,
-      "learning_rate": 1.6256864733763787e-05,
-      "loss": 1.4927,
-      "step": 419
-    },
-    {
-      "epoch": 0.5809128630705395,
-      "grad_norm": 1.8203260898590088,
-      "learning_rate": 1.6239783780038374e-05,
-      "loss": 1.493,
-      "step": 420
-    },
-    {
-      "epoch": 0.5822959889349931,
-      "grad_norm": 2.0269949436187744,
-      "learning_rate": 1.6222672961470158e-05,
-      "loss": 1.3748,
-      "step": 421
-    },
-    {
-      "epoch": 0.5836791147994468,
-      "grad_norm": 1.7836514711380005,
-      "learning_rate": 1.6205532359954905e-05,
-      "loss": 1.3786,
-      "step": 422
-    },
-    {
-      "epoch": 0.5850622406639004,
-      "grad_norm": 1.7503560781478882,
-      "learning_rate": 1.618836205753093e-05,
-      "loss": 1.3905,
-      "step": 423
-    },
-    {
-      "epoch": 0.5864453665283541,
-      "grad_norm": 1.9079328775405884,
-      "learning_rate": 1.6171162136378716e-05,
-      "loss": 1.51,
-      "step": 424
-    },
-    {
-      "epoch": 0.5878284923928078,
-      "grad_norm": 2.0547492504119873,
-      "learning_rate": 1.6153932678820487e-05,
-      "loss": 1.5062,
-      "step": 425
-    },
-    {
-      "epoch": 0.5892116182572614,
-      "grad_norm": 1.9388115406036377,
-      "learning_rate": 1.6136673767319853e-05,
-      "loss": 1.5496,
-      "step": 426
-    },
-    {
-      "epoch": 0.5905947441217151,
-      "grad_norm": 1.8805044889450073,
-      "learning_rate": 1.611938548448138e-05,
-      "loss": 1.4072,
-      "step": 427
-    },
-    {
-      "epoch": 0.5919778699861687,
-      "grad_norm": 1.9267683029174805,
-      "learning_rate": 1.6102067913050227e-05,
-      "loss": 1.4047,
-      "step": 428
-    },
-    {
-      "epoch": 0.5933609958506224,
-      "grad_norm": 1.8546456098556519,
-      "learning_rate": 1.6084721135911715e-05,
-      "loss": 1.4631,
-      "step": 429
-    },
-    {
-      "epoch": 0.5947441217150761,
-      "grad_norm": 1.7463114261627197,
-      "learning_rate": 1.606734523609097e-05,
-      "loss": 1.4703,
-      "step": 430
-    },
-    {
-      "epoch": 0.5961272475795297,
-      "grad_norm": 2.014925003051758,
-      "learning_rate": 1.604994029675249e-05,
-      "loss": 1.294,
-      "step": 431
-    },
-    {
-      "epoch": 0.5975103734439834,
-      "grad_norm": 2.1670687198638916,
-      "learning_rate": 1.603250640119977e-05,
-      "loss": 1.2759,
-      "step": 432
-    },
-    {
-      "epoch": 0.598893499308437,
-      "grad_norm": 1.9688105583190918,
-      "learning_rate": 1.601504363287489e-05,
-      "loss": 1.4053,
-      "step": 433
-    },
-    {
-      "epoch": 0.6002766251728907,
-      "grad_norm": 2.0988128185272217,
-      "learning_rate": 1.5997552075358122e-05,
-      "loss": 1.3473,
-      "step": 434
-    },
-    {
-      "epoch": 0.6016597510373444,
-      "grad_norm": 2.176347255706787,
-      "learning_rate": 1.598003181236753e-05,
-      "loss": 1.3099,
-      "step": 435
-    },
-    {
-      "epoch": 0.603042876901798,
-      "grad_norm": 2.1589407920837402,
-      "learning_rate": 1.5962482927758568e-05,
-      "loss": 1.3499,
-      "step": 436
-    },
-    {
-      "epoch": 0.6044260027662517,
-      "grad_norm": 1.9357820749282837,
-      "learning_rate": 1.5944905505523677e-05,
-      "loss": 1.3215,
-      "step": 437
-    },
-    {
-      "epoch": 0.6058091286307054,
-      "grad_norm": 1.7777889966964722,
-      "learning_rate": 1.592729962979189e-05,
-      "loss": 1.4082,
-      "step": 438
-    },
-    {
-      "epoch": 0.607192254495159,
-      "grad_norm": 2.0036303997039795,
-      "learning_rate": 1.590966538482842e-05,
-      "loss": 1.4111,
-      "step": 439
-    },
-    {
-      "epoch": 0.6085753803596127,
-      "grad_norm": 1.9790940284729004,
-      "learning_rate": 1.589200285503426e-05,
-      "loss": 1.3613,
-      "step": 440
-    },
-    {
-      "epoch": 0.6099585062240664,
-      "grad_norm": 1.9340951442718506,
-      "learning_rate": 1.5874312124945773e-05,
-      "loss": 1.3719,
-      "step": 441
-    },
-    {
-      "epoch": 0.6113416320885201,
-      "grad_norm": 2.064161777496338,
-      "learning_rate": 1.585659327923432e-05,
-      "loss": 1.414,
-      "step": 442
-    },
-    {
-      "epoch": 0.6127247579529738,
-      "grad_norm": 1.955349326133728,
-      "learning_rate": 1.5838846402705793e-05,
-      "loss": 1.4853,
-      "step": 443
-    },
-    {
-      "epoch": 0.6141078838174274,
-      "grad_norm": 2.2459092140197754,
-      "learning_rate": 1.582107158030027e-05,
-      "loss": 1.2444,
-      "step": 444
-    },
-    {
-      "epoch": 0.6154910096818811,
-      "grad_norm": 1.9782187938690186,
-      "learning_rate": 1.5803268897091582e-05,
-      "loss": 1.3512,
-      "step": 445
-    },
-    {
-      "epoch": 0.6168741355463347,
-      "grad_norm": 1.6818398237228394,
-      "learning_rate": 1.5785438438286892e-05,
-      "loss": 1.0878,
-      "step": 446
-    },
-    {
-      "epoch": 0.6182572614107884,
-      "grad_norm": 2.271744728088379,
-      "learning_rate": 1.576758028922632e-05,
-      "loss": 1.5163,
-      "step": 447
-    },
-    {
-      "epoch": 0.6196403872752421,
-      "grad_norm": 2.1444921493530273,
-      "learning_rate": 1.574969453538251e-05,
-      "loss": 1.4333,
-      "step": 448
-    },
-    {
-      "epoch": 0.6210235131396957,
-      "grad_norm": 2.0727901458740234,
-      "learning_rate": 1.573178126236022e-05,
-      "loss": 1.5082,
-      "step": 449
-    },
-    {
-      "epoch": 0.6224066390041494,
-      "grad_norm": 1.9815788269042969,
-      "learning_rate": 1.5713840555895937e-05,
-      "loss": 1.3223,
-      "step": 450
-    },
-    {
-      "epoch": 0.623789764868603,
-      "grad_norm": 2.123006820678711,
-      "learning_rate": 1.569587250185743e-05,
-      "loss": 1.3614,
-      "step": 451
-    },
-    {
-      "epoch": 0.6251728907330567,
-      "grad_norm": 2.089308738708496,
-      "learning_rate": 1.567787718624338e-05,
-      "loss": 1.3747,
-      "step": 452
-    },
-    {
-      "epoch": 0.6265560165975104,
-      "grad_norm": 2.087164878845215,
-      "learning_rate": 1.5659854695182928e-05,
-      "loss": 1.4676,
-      "step": 453
-    },
-    {
-      "epoch": 0.627939142461964,
-      "grad_norm": 1.9156312942504883,
-      "learning_rate": 1.5641805114935297e-05,
-      "loss": 1.4936,
-      "step": 454
-    },
-    {
-      "epoch": 0.6293222683264177,
-      "grad_norm": 2.051100492477417,
-      "learning_rate": 1.5623728531889356e-05,
-      "loss": 1.3817,
-      "step": 455
-    },
-    {
-      "epoch": 0.6307053941908713,
-      "grad_norm": 1.8827084302902222,
-      "learning_rate": 1.560562503256322e-05,
-      "loss": 1.4932,
-      "step": 456
-    },
-    {
-      "epoch": 0.632088520055325,
-      "grad_norm": 1.9822756052017212,
-      "learning_rate": 1.558749470360382e-05,
-      "loss": 1.3448,
-      "step": 457
-    },
-    {
-      "epoch": 0.6334716459197787,
-      "grad_norm": 2.2374722957611084,
-      "learning_rate": 1.556933763178651e-05,
-      "loss": 1.4071,
-      "step": 458
-    },
-    {
-      "epoch": 0.6348547717842323,
-      "grad_norm": 2.2126710414886475,
-      "learning_rate": 1.5551153904014645e-05,
-      "loss": 1.3177,
-      "step": 459
-    },
-    {
-      "epoch": 0.636237897648686,
-      "grad_norm": 1.9302269220352173,
-      "learning_rate": 1.5532943607319143e-05,
-      "loss": 1.3762,
-      "step": 460
-    },
-    {
-      "epoch": 0.6376210235131397,
-      "grad_norm": 2.0682942867279053,
-      "learning_rate": 1.5514706828858096e-05,
-      "loss": 1.4728,
-      "step": 461
-    },
-    {
-      "epoch": 0.6390041493775933,
-      "grad_norm": 1.912061333656311,
-      "learning_rate": 1.5496443655916348e-05,
-      "loss": 1.3751,
-      "step": 462
-    },
-    {
-      "epoch": 0.640387275242047,
-      "grad_norm": 2.072456121444702,
-      "learning_rate": 1.547815417590506e-05,
-      "loss": 1.3737,
-      "step": 463
-    },
-    {
-      "epoch": 0.6417704011065007,
-      "grad_norm": 2.1260902881622314,
-      "learning_rate": 1.5459838476361326e-05,
-      "loss": 1.4867,
-      "step": 464
-    },
-    {
-      "epoch": 0.6431535269709544,
-      "grad_norm": 2.0210676193237305,
-      "learning_rate": 1.54414966449477e-05,
-      "loss": 1.4505,
-      "step": 465
-    },
-    {
-      "epoch": 0.6445366528354081,
-      "grad_norm": 1.9175024032592773,
-      "learning_rate": 1.5423128769451832e-05,
-      "loss": 1.4385,
-      "step": 466
-    },
-    {
-      "epoch": 0.6459197786998617,
-      "grad_norm": 1.7762582302093506,
-      "learning_rate": 1.5404734937786017e-05,
-      "loss": 1.0755,
-      "step": 467
-    },
-    {
-      "epoch": 0.6473029045643154,
-      "grad_norm": 2.193601131439209,
-      "learning_rate": 1.5386315237986785e-05,
-      "loss": 1.4005,
-      "step": 468
-    },
-    {
-      "epoch": 0.648686030428769,
-      "grad_norm": 2.159917116165161,
-      "learning_rate": 1.5367869758214466e-05,
-      "loss": 1.4358,
-      "step": 469
-    },
-    {
-      "epoch": 0.6500691562932227,
-      "grad_norm": 2.10957932472229,
-      "learning_rate": 1.5349398586752794e-05,
-      "loss": 1.3415,
-      "step": 470
-    },
-    {
-      "epoch": 0.6514522821576764,
-      "grad_norm": 2.0722029209136963,
-      "learning_rate": 1.533090181200845e-05,
-      "loss": 1.5206,
-      "step": 471
-    },
-    {
-      "epoch": 0.65283540802213,
-      "grad_norm": 2.0709426403045654,
-      "learning_rate": 1.5312379522510666e-05,
-      "loss": 1.5009,
-      "step": 472
-    },
-    {
-      "epoch": 0.6542185338865837,
-      "grad_norm": 2.1032183170318604,
-      "learning_rate": 1.5293831806910803e-05,
-      "loss": 1.4927,
-      "step": 473
-    },
-    {
-      "epoch": 0.6556016597510373,
-      "grad_norm": 2.0395522117614746,
-      "learning_rate": 1.52752587539819e-05,
-      "loss": 1.4599,
-      "step": 474
-    },
-    {
-      "epoch": 0.656984785615491,
-      "grad_norm": 1.9160783290863037,
-      "learning_rate": 1.5256660452618276e-05,
-      "loss": 1.389,
-      "step": 475
-    },
-    {
-      "epoch": 0.6583679114799447,
-      "grad_norm": 2.017306089401245,
-      "learning_rate": 1.5238036991835085e-05,
-      "loss": 1.3694,
-      "step": 476
-    },
-    {
-      "epoch": 0.6597510373443983,
-      "grad_norm": 2.1166417598724365,
-      "learning_rate": 1.521938846076791e-05,
-      "loss": 1.2891,
-      "step": 477
-    },
-    {
-      "epoch": 0.661134163208852,
-      "grad_norm": 2.1959710121154785,
-      "learning_rate": 1.5200714948672313e-05,
-      "loss": 1.4218,
-      "step": 478
-    },
-    {
-      "epoch": 0.6625172890733056,
-      "grad_norm": 2.0393929481506348,
-      "learning_rate": 1.5182016544923432e-05,
-      "loss": 1.3561,
-      "step": 479
-    },
-    {
-      "epoch": 0.6639004149377593,
-      "grad_norm": 2.0378098487854004,
-      "learning_rate": 1.5163293339015535e-05,
-      "loss": 1.4816,
-      "step": 480
-    },
-    {
-      "epoch": 0.665283540802213,
-      "grad_norm": 2.1572370529174805,
-      "learning_rate": 1.5144545420561598e-05,
-      "loss": 1.4697,
-      "step": 481
-    },
-    {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 1.9824397563934326,
-      "learning_rate": 1.512577287929288e-05,
-      "loss": 1.4023,
-      "step": 482
-    },
-    {
-      "epoch": 0.6680497925311203,
-      "grad_norm": 2.1600515842437744,
-      "learning_rate": 1.5106975805058483e-05,
-      "loss": 1.3469,
-      "step": 483
-    },
-    {
-      "epoch": 0.669432918395574,
-      "grad_norm": 2.0499649047851562,
-      "learning_rate": 1.5088154287824934e-05,
-      "loss": 1.4444,
-      "step": 484
-    },
-    {
-      "epoch": 0.6708160442600276,
-      "grad_norm": 2.1112446784973145,
-      "learning_rate": 1.506930841767575e-05,
-      "loss": 1.3261,
-      "step": 485
-    },
-    {
-      "epoch": 0.6721991701244814,
-      "grad_norm": 2.0517220497131348,
-      "learning_rate": 1.5050438284811001e-05,
-      "loss": 1.398,
-      "step": 486
-    },
-    {
-      "epoch": 0.673582295988935,
-      "grad_norm": 2.2738828659057617,
-      "learning_rate": 1.5031543979546887e-05,
-      "loss": 1.3997,
-      "step": 487
-    },
-    {
-      "epoch": 0.6749654218533887,
-      "grad_norm": 2.024562358856201,
-      "learning_rate": 1.5012625592315298e-05,
-      "loss": 1.3629,
-      "step": 488
-    },
-    {
-      "epoch": 0.6763485477178424,
-      "grad_norm": 2.437912940979004,
-      "learning_rate": 1.499368321366339e-05,
-      "loss": 1.3036,
-      "step": 489
-    },
-    {
-      "epoch": 0.677731673582296,
-      "grad_norm": 1.9140431880950928,
-      "learning_rate": 1.4974716934253146e-05,
-      "loss": 1.446,
-      "step": 490
-    },
-    {
-      "epoch": 0.6791147994467497,
-      "grad_norm": 2.439692974090576,
-      "learning_rate": 1.4955726844860939e-05,
-      "loss": 1.453,
-      "step": 491
-    },
-    {
-      "epoch": 0.6804979253112033,
-      "grad_norm": 2.154038906097412,
-      "learning_rate": 1.4936713036377102e-05,
-      "loss": 1.3837,
-      "step": 492
-    },
-    {
-      "epoch": 0.681881051175657,
-      "grad_norm": 2.1363091468811035,
-      "learning_rate": 1.4917675599805497e-05,
-      "loss": 1.3759,
-      "step": 493
-    },
-    {
-      "epoch": 0.6832641770401107,
-      "grad_norm": 2.051652431488037,
-      "learning_rate": 1.4898614626263066e-05,
-      "loss": 1.3556,
-      "step": 494
-    },
-    {
-      "epoch": 0.6846473029045643,
-      "grad_norm": 2.1247611045837402,
-      "learning_rate": 1.4879530206979418e-05,
-      "loss": 1.3635,
-      "step": 495
-    },
-    {
-      "epoch": 0.686030428769018,
-      "grad_norm": 2.284594774246216,
-      "learning_rate": 1.4860422433296363e-05,
-      "loss": 1.4161,
-      "step": 496
-    },
-    {
-      "epoch": 0.6874135546334716,
-      "grad_norm": 2.339827060699463,
-      "learning_rate": 1.4841291396667494e-05,
-      "loss": 1.4127,
-      "step": 497
-    },
-    {
-      "epoch": 0.6887966804979253,
-      "grad_norm": 2.558943033218384,
-      "learning_rate": 1.4822137188657752e-05,
-      "loss": 1.384,
-      "step": 498
-    },
-    {
-      "epoch": 0.690179806362379,
-      "grad_norm": 2.1640422344207764,
-      "learning_rate": 1.4802959900942967e-05,
-      "loss": 1.3584,
-      "step": 499
-    },
-    {
-      "epoch": 0.6915629322268326,
-      "grad_norm": 2.1310927867889404,
-      "learning_rate": 1.4783759625309454e-05,
-      "loss": 1.375,
-      "step": 500
-    },
-    {
-      "epoch": 0.6929460580912863,
-      "grad_norm": 2.0568363666534424,
-      "learning_rate": 1.4764536453653536e-05,
-      "loss": 1.383,
-      "step": 501
-    },
-    {
-      "epoch": 0.69432918395574,
-      "grad_norm": 2.259617567062378,
-      "learning_rate": 1.474529047798112e-05,
-      "loss": 1.3432,
-      "step": 502
-    },
-    {
-      "epoch": 0.6957123098201936,
-      "grad_norm": 2.0728511810302734,
-      "learning_rate": 1.4726021790407268e-05,
-      "loss": 1.4592,
-      "step": 503
-    },
-    {
-      "epoch": 0.6970954356846473,
-      "grad_norm": 2.205367088317871,
-      "learning_rate": 1.4706730483155738e-05,
-      "loss": 1.4189,
-      "step": 504
-    },
-    {
-      "epoch": 0.6984785615491009,
-      "grad_norm": 2.147625207901001,
-      "learning_rate": 1.4687416648558555e-05,
-      "loss": 1.4086,
-      "step": 505
-    },
-    {
-      "epoch": 0.6998616874135546,
-      "grad_norm": 2.062960624694824,
-      "learning_rate": 1.4668080379055563e-05,
-      "loss": 1.4586,
-      "step": 506
-    },
-    {
-      "epoch": 0.7012448132780082,
-      "grad_norm": 2.2476038932800293,
-      "learning_rate": 1.4648721767193981e-05,
-      "loss": 1.4667,
-      "step": 507
-    },
-    {
-      "epoch": 0.7026279391424619,
-      "grad_norm": 2.2841742038726807,
-      "learning_rate": 1.4629340905627964e-05,
-      "loss": 1.3349,
-      "step": 508
-    },
-    {
-      "epoch": 0.7040110650069157,
-      "grad_norm": 2.1537253856658936,
-      "learning_rate": 1.4609937887118165e-05,
-      "loss": 1.4284,
-      "step": 509
-    },
-    {
-      "epoch": 0.7053941908713693,
-      "grad_norm": 2.1182830333709717,
-      "learning_rate": 1.4590512804531272e-05,
-      "loss": 1.4696,
-      "step": 510
-    },
-    {
-      "epoch": 0.706777316735823,
-      "grad_norm": 1.9890772104263306,
-      "learning_rate": 1.4571065750839586e-05,
-      "loss": 1.4727,
-      "step": 511
-    },
-    {
-      "epoch": 0.7081604426002767,
-      "grad_norm": 2.303697347640991,
-      "learning_rate": 1.4551596819120564e-05,
-      "loss": 1.3592,
-      "step": 512
-    },
-    {
-      "epoch": 0.7095435684647303,
-      "grad_norm": 2.0708582401275635,
-      "learning_rate": 1.4532106102556377e-05,
-      "loss": 1.3987,
-      "step": 513
-    },
-    {
-      "epoch": 0.710926694329184,
-      "grad_norm": 2.2212769985198975,
-      "learning_rate": 1.4512593694433455e-05,
-      "loss": 1.3956,
-      "step": 514
-    },
-    {
-      "epoch": 0.7123098201936376,
-      "grad_norm": 2.004349946975708,
-      "learning_rate": 1.4493059688142055e-05,
-      "loss": 1.3919,
-      "step": 515
-    },
-    {
-      "epoch": 0.7136929460580913,
-      "grad_norm": 2.043989658355713,
-      "learning_rate": 1.447350417717581e-05,
-      "loss": 1.5023,
-      "step": 516
-    },
-    {
-      "epoch": 0.715076071922545,
-      "grad_norm": 2.0797524452209473,
-      "learning_rate": 1.445392725513127e-05,
-      "loss": 1.4577,
-      "step": 517
-    },
-    {
-      "epoch": 0.7164591977869986,
-      "grad_norm": 2.4839534759521484,
-      "learning_rate": 1.4434329015707468e-05,
-      "loss": 1.4462,
-      "step": 518
-    },
-    {
-      "epoch": 0.7178423236514523,
-      "grad_norm": 2.393336057662964,
-      "learning_rate": 1.4414709552705465e-05,
-      "loss": 1.3544,
-      "step": 519
-    },
-    {
-      "epoch": 0.719225449515906,
-      "grad_norm": 2.1947097778320312,
-      "learning_rate": 1.4395068960027903e-05,
-      "loss": 1.3591,
-      "step": 520
-    },
-    {
-      "epoch": 0.7206085753803596,
-      "grad_norm": 2.231341600418091,
-      "learning_rate": 1.4375407331678553e-05,
-      "loss": 1.4389,
-      "step": 521
-    },
-    {
-      "epoch": 0.7219917012448133,
-      "grad_norm": 2.1521496772766113,
-      "learning_rate": 1.435572476176187e-05,
-      "loss": 1.3715,
-      "step": 522
-    },
-    {
-      "epoch": 0.7233748271092669,
-      "grad_norm": 2.3095715045928955,
-      "learning_rate": 1.4336021344482539e-05,
-      "loss": 1.4191,
-      "step": 523
-    },
-    {
-      "epoch": 0.7247579529737206,
-      "grad_norm": 2.364912986755371,
-      "learning_rate": 1.4316297174145018e-05,
-      "loss": 1.3801,
-      "step": 524
-    },
-    {
-      "epoch": 0.7261410788381742,
-      "grad_norm": 2.283719301223755,
-      "learning_rate": 1.4296552345153099e-05,
-      "loss": 1.3237,
-      "step": 525
-    },
-    {
-      "epoch": 0.7275242047026279,
-      "grad_norm": 2.2587459087371826,
-      "learning_rate": 1.427678695200945e-05,
-      "loss": 1.3574,
-      "step": 526
-    },
-    {
-      "epoch": 0.7289073305670816,
-      "grad_norm": 2.1912143230438232,
-      "learning_rate": 1.4257001089315173e-05,
-      "loss": 1.3736,
-      "step": 527
-    },
-    {
-      "epoch": 0.7302904564315352,
-      "grad_norm": 2.349207639694214,
-      "learning_rate": 1.4237194851769318e-05,
-      "loss": 1.4744,
-      "step": 528
-    },
-    {
-      "epoch": 0.7316735822959889,
-      "grad_norm": 2.1632063388824463,
-      "learning_rate": 1.4217368334168472e-05,
-      "loss": 1.4335,
-      "step": 529
-    },
-    {
-      "epoch": 0.7330567081604425,
-      "grad_norm": 2.3758816719055176,
-      "learning_rate": 1.4197521631406279e-05,
-      "loss": 1.3985,
-      "step": 530
-    },
-    {
-      "epoch": 0.7344398340248963,
-      "grad_norm": 2.2720682621002197,
-      "learning_rate": 1.4177654838472996e-05,
-      "loss": 1.2187,
-      "step": 531
-    },
-    {
-      "epoch": 0.73582295988935,
-      "grad_norm": 2.222930908203125,
-      "learning_rate": 1.4157768050455038e-05,
-      "loss": 1.3715,
-      "step": 532
-    },
-    {
-      "epoch": 0.7372060857538036,
-      "grad_norm": 2.493760824203491,
-      "learning_rate": 1.4137861362534513e-05,
-      "loss": 1.3032,
-      "step": 533
-    },
-    {
-      "epoch": 0.7385892116182573,
-      "grad_norm": 2.4699270725250244,
-      "learning_rate": 1.4117934869988776e-05,
-      "loss": 1.2739,
-      "step": 534
-    },
-    {
-      "epoch": 0.739972337482711,
-      "grad_norm": 2.443169593811035,
-      "learning_rate": 1.4097988668189977e-05,
-      "loss": 1.4135,
-      "step": 535
-    },
-    {
-      "epoch": 0.7413554633471646,
-      "grad_norm": 2.4985241889953613,
-      "learning_rate": 1.4078022852604591e-05,
-      "loss": 1.331,
-      "step": 536
-    },
-    {
-      "epoch": 0.7427385892116183,
-      "grad_norm": 2.4425034523010254,
-      "learning_rate": 1.4058037518792975e-05,
-      "loss": 1.4912,
-      "step": 537
-    },
-    {
-      "epoch": 0.7441217150760719,
-      "grad_norm": 2.4378395080566406,
-      "learning_rate": 1.4038032762408897e-05,
-      "loss": 1.4527,
-      "step": 538
-    },
-    {
-      "epoch": 0.7455048409405256,
-      "grad_norm": 2.427049160003662,
-      "learning_rate": 1.4018008679199092e-05,
-      "loss": 1.269,
-      "step": 539
-    },
-    {
-      "epoch": 0.7468879668049793,
-      "grad_norm": 2.3256077766418457,
-      "learning_rate": 1.3997965365002789e-05,
-      "loss": 1.3976,
-      "step": 540
-    },
-    {
-      "epoch": 0.7482710926694329,
-      "grad_norm": 2.26210880279541,
-      "learning_rate": 1.3977902915751268e-05,
-      "loss": 1.3096,
-      "step": 541
-    },
-    {
-      "epoch": 0.7496542185338866,
-      "grad_norm": 2.7226202487945557,
-      "learning_rate": 1.3957821427467392e-05,
-      "loss": 1.4281,
-      "step": 542
-    },
-    {
-      "epoch": 0.7510373443983402,
-      "grad_norm": 2.3421599864959717,
-      "learning_rate": 1.3937720996265147e-05,
-      "loss": 1.3555,
-      "step": 543
-    },
-    {
-      "epoch": 0.7510373443983402,
-      "eval_loss": 1.1585720777511597,
-      "eval_runtime": 108.9727,
-      "eval_samples_per_second": 13.251,
-      "eval_steps_per_second": 0.835,
-      "step": 543
-    },
-    {
-      "epoch": 0.7524204702627939,
-      "grad_norm": 2.397235870361328,
-      "learning_rate": 1.3917601718349183e-05,
-      "loss": 1.362,
-      "step": 544
-    },
-    {
-      "epoch": 0.7538035961272476,
-      "grad_norm": 2.4912915229797363,
-      "learning_rate": 1.3897463690014353e-05,
-      "loss": 1.2447,
-      "step": 545
-    },
-    {
-      "epoch": 0.7551867219917012,
-      "grad_norm": 2.4246633052825928,
-      "learning_rate": 1.3877307007645256e-05,
-      "loss": 1.4679,
-      "step": 546
-    },
-    {
-      "epoch": 0.7565698478561549,
-      "grad_norm": 2.182666778564453,
-      "learning_rate": 1.385713176771577e-05,
-      "loss": 1.1187,
-      "step": 547
-    },
-    {
-      "epoch": 0.7579529737206085,
-      "grad_norm": 2.2946391105651855,
-      "learning_rate": 1.3836938066788599e-05,
-      "loss": 1.3262,
-      "step": 548
-    },
-    {
-      "epoch": 0.7593360995850622,
-      "grad_norm": 2.28464412689209,
-      "learning_rate": 1.3816726001514802e-05,
-      "loss": 1.4728,
-      "step": 549
-    },
-    {
-      "epoch": 0.7607192254495159,
-      "grad_norm": 2.2874138355255127,
-      "learning_rate": 1.3796495668633325e-05,
-      "loss": 1.3737,
-      "step": 550
-    },
-    {
-      "epoch": 0.7621023513139695,
-      "grad_norm": 2.259789228439331,
-      "learning_rate": 1.377624716497056e-05,
-      "loss": 1.3688,
-      "step": 551
-    },
-    {
-      "epoch": 0.7634854771784232,
-      "grad_norm": 2.3231420516967773,
-      "learning_rate": 1.3755980587439857e-05,
-      "loss": 1.4047,
-      "step": 552
-    },
-    {
-      "epoch": 0.7648686030428768,
-      "grad_norm": 2.470014810562134,
-      "learning_rate": 1.3735696033041079e-05,
-      "loss": 1.4894,
-      "step": 553
-    },
-    {
-      "epoch": 0.7662517289073306,
-      "grad_norm": 2.389801263809204,
-      "learning_rate": 1.3715393598860129e-05,
-      "loss": 1.3389,
-      "step": 554
-    },
-    {
-      "epoch": 0.7676348547717843,
-      "grad_norm": 2.190786123275757,
-      "learning_rate": 1.369507338206848e-05,
-      "loss": 1.3945,
-      "step": 555
-    },
-    {
-      "epoch": 0.7690179806362379,
-      "grad_norm": 2.5276217460632324,
-      "learning_rate": 1.367473547992272e-05,
-      "loss": 1.3594,
-      "step": 556
-    },
-    {
-      "epoch": 0.7704011065006916,
-      "grad_norm": 2.6756093502044678,
-      "learning_rate": 1.3654379989764084e-05,
-      "loss": 1.3529,
-      "step": 557
-    },
-    {
-      "epoch": 0.7717842323651453,
-      "grad_norm": 2.585706949234009,
-      "learning_rate": 1.3634007009017986e-05,
-      "loss": 1.3762,
-      "step": 558
-    },
-    {
-      "epoch": 0.7731673582295989,
-      "grad_norm": 2.3161914348602295,
-      "learning_rate": 1.3613616635193551e-05,
-      "loss": 1.4073,
-      "step": 559
-    },
-    {
-      "epoch": 0.7745504840940526,
-      "grad_norm": 2.3607048988342285,
-      "learning_rate": 1.3593208965883156e-05,
-      "loss": 1.4318,
-      "step": 560
-    },
-    {
-      "epoch": 0.7759336099585062,
-      "grad_norm": 2.048332691192627,
-      "learning_rate": 1.357278409876195e-05,
-      "loss": 1.1218,
-      "step": 561
-    },
-    {
-      "epoch": 0.7773167358229599,
-      "grad_norm": 2.6801364421844482,
-      "learning_rate": 1.3552342131587399e-05,
-      "loss": 1.4298,
-      "step": 562
-    },
-    {
-      "epoch": 0.7786998616874136,
-      "grad_norm": 2.579087972640991,
-      "learning_rate": 1.3531883162198815e-05,
-      "loss": 1.3948,
-      "step": 563
-    },
-    {
-      "epoch": 0.7800829875518672,
-      "grad_norm": 2.5193817615509033,
-      "learning_rate": 1.351140728851688e-05,
-      "loss": 1.347,
-      "step": 564
-    },
-    {
-      "epoch": 0.7814661134163209,
-      "grad_norm": 2.2663426399230957,
-      "learning_rate": 1.3490914608543189e-05,
-      "loss": 1.4367,
-      "step": 565
-    },
-    {
-      "epoch": 0.7828492392807745,
-      "grad_norm": 2.56097412109375,
-      "learning_rate": 1.3470405220359773e-05,
-      "loss": 1.256,
-      "step": 566
-    },
-    {
-      "epoch": 0.7842323651452282,
-      "grad_norm": 2.6170148849487305,
-      "learning_rate": 1.3449879222128628e-05,
-      "loss": 1.3289,
-      "step": 567
-    },
-    {
-      "epoch": 0.7856154910096819,
-      "grad_norm": 2.7682836055755615,
-      "learning_rate": 1.3429336712091258e-05,
-      "loss": 1.3766,
-      "step": 568
-    },
-    {
-      "epoch": 0.7869986168741355,
-      "grad_norm": 2.343817949295044,
-      "learning_rate": 1.340877778856819e-05,
-      "loss": 1.4599,
-      "step": 569
-    },
-    {
-      "epoch": 0.7883817427385892,
-      "grad_norm": 2.5406904220581055,
-      "learning_rate": 1.3388202549958507e-05,
-      "loss": 1.46,
-      "step": 570
-    },
-    {
-      "epoch": 0.7897648686030428,
-      "grad_norm": 2.5566861629486084,
-      "learning_rate": 1.3367611094739384e-05,
-      "loss": 1.38,
-      "step": 571
-    },
-    {
-      "epoch": 0.7911479944674965,
-      "grad_norm": 2.6734111309051514,
-      "learning_rate": 1.334700352146561e-05,
-      "loss": 1.3167,
-      "step": 572
-    },
-    {
-      "epoch": 0.7925311203319502,
-      "grad_norm": 2.242243766784668,
-      "learning_rate": 1.3326379928769114e-05,
-      "loss": 1.3883,
-      "step": 573
-    },
-    {
-      "epoch": 0.7939142461964038,
-      "grad_norm": 2.456390142440796,
-      "learning_rate": 1.3305740415358506e-05,
-      "loss": 1.4072,
-      "step": 574
-    },
-    {
-      "epoch": 0.7952973720608575,
-      "grad_norm": 2.449395179748535,
-      "learning_rate": 1.3285085080018589e-05,
-      "loss": 1.38,
-      "step": 575
-    },
-    {
-      "epoch": 0.7966804979253111,
-      "grad_norm": 2.6714186668395996,
-      "learning_rate": 1.3264414021609899e-05,
-      "loss": 1.3204,
-      "step": 576
-    },
-    {
-      "epoch": 0.7980636237897649,
-      "grad_norm": 2.479647397994995,
-      "learning_rate": 1.3243727339068216e-05,
-      "loss": 1.3234,
-      "step": 577
-    },
-    {
-      "epoch": 0.7994467496542186,
-      "grad_norm": 2.5289547443389893,
-      "learning_rate": 1.3223025131404106e-05,
-      "loss": 1.3533,
-      "step": 578
-    },
-    {
-      "epoch": 0.8008298755186722,
-      "grad_norm": 2.6476147174835205,
-      "learning_rate": 1.3202307497702443e-05,
-      "loss": 1.3419,
-      "step": 579
-    },
-    {
-      "epoch": 0.8022130013831259,
-      "grad_norm": 2.7068285942077637,
-      "learning_rate": 1.3181574537121933e-05,
-      "loss": 1.3636,
-      "step": 580
-    },
-    {
-      "epoch": 0.8035961272475796,
-      "grad_norm": 2.6454317569732666,
-      "learning_rate": 1.3160826348894635e-05,
-      "loss": 1.352,
-      "step": 581
-    },
-    {
-      "epoch": 0.8049792531120332,
-      "grad_norm": 2.513296127319336,
-      "learning_rate": 1.3140063032325491e-05,
-      "loss": 1.3446,
-      "step": 582
-    },
-    {
-      "epoch": 0.8063623789764869,
-      "grad_norm": 2.563950538635254,
-      "learning_rate": 1.3119284686791859e-05,
-      "loss": 1.4223,
-      "step": 583
-    },
-    {
-      "epoch": 0.8077455048409405,
-      "grad_norm": 2.553441047668457,
-      "learning_rate": 1.3098491411743014e-05,
-      "loss": 1.4375,
-      "step": 584
-    },
-    {
-      "epoch": 0.8091286307053942,
-      "grad_norm": 2.6420164108276367,
-      "learning_rate": 1.3077683306699702e-05,
-      "loss": 1.2831,
-      "step": 585
-    },
-    {
-      "epoch": 0.8105117565698479,
-      "grad_norm": 2.67268443107605,
-      "learning_rate": 1.3056860471253639e-05,
-      "loss": 1.4366,
-      "step": 586
-    },
-    {
-      "epoch": 0.8118948824343015,
-      "grad_norm": 2.650383472442627,
-      "learning_rate": 1.3036023005067042e-05,
-      "loss": 1.2001,
-      "step": 587
-    },
-    {
-      "epoch": 0.8132780082987552,
-      "grad_norm": 2.6096444129943848,
-      "learning_rate": 1.3015171007872161e-05,
-      "loss": 1.4101,
-      "step": 588
-    },
-    {
-      "epoch": 0.8146611341632088,
-      "grad_norm": 2.8249430656433105,
-      "learning_rate": 1.2994304579470787e-05,
-      "loss": 1.3862,
-      "step": 589
-    },
-    {
-      "epoch": 0.8160442600276625,
-      "grad_norm": 2.864114999771118,
-      "learning_rate": 1.297342381973379e-05,
-      "loss": 1.4244,
-      "step": 590
-    },
-    {
-      "epoch": 0.8174273858921162,
-      "grad_norm": 2.7902991771698,
-      "learning_rate": 1.2952528828600623e-05,
-      "loss": 1.3664,
-      "step": 591
-    },
-    {
-      "epoch": 0.8188105117565698,
-      "grad_norm": 2.8960015773773193,
-      "learning_rate": 1.2931619706078862e-05,
-      "loss": 1.3758,
-      "step": 592
-    },
-    {
-      "epoch": 0.8201936376210235,
-      "grad_norm": 2.313051223754883,
-      "learning_rate": 1.2910696552243708e-05,
-      "loss": 1.438,
-      "step": 593
-    },
-    {
-      "epoch": 0.8215767634854771,
-      "grad_norm": 2.627824544906616,
-      "learning_rate": 1.2889759467237532e-05,
-      "loss": 1.3203,
-      "step": 594
-    },
-    {
-      "epoch": 0.8229598893499308,
-      "grad_norm": 2.7432913780212402,
-      "learning_rate": 1.2868808551269374e-05,
-      "loss": 1.3742,
-      "step": 595
-    },
-    {
-      "epoch": 0.8243430152143845,
-      "grad_norm": 2.6407742500305176,
-      "learning_rate": 1.2847843904614474e-05,
-      "loss": 1.3906,
-      "step": 596
-    },
-    {
-      "epoch": 0.8257261410788381,
-      "grad_norm": 2.8636410236358643,
-      "learning_rate": 1.2826865627613785e-05,
-      "loss": 1.3633,
-      "step": 597
-    },
-    {
-      "epoch": 0.8271092669432918,
-      "grad_norm": 2.397425413131714,
-      "learning_rate": 1.2805873820673509e-05,
-      "loss": 1.4116,
-      "step": 598
-    },
-    {
-      "epoch": 0.8284923928077456,
-      "grad_norm": 2.8172311782836914,
-      "learning_rate": 1.2784868584264587e-05,
-      "loss": 1.3482,
-      "step": 599
-    },
-    {
-      "epoch": 0.8298755186721992,
-      "grad_norm": 2.766284227371216,
-      "learning_rate": 1.2763850018922257e-05,
-      "loss": 1.3548,
-      "step": 600
-    },
-    {
-      "epoch": 0.8312586445366529,
-      "grad_norm": 2.8403494358062744,
-      "learning_rate": 1.2742818225245538e-05,
-      "loss": 1.4169,
-      "step": 601
-    },
-    {
-      "epoch": 0.8326417704011065,
-      "grad_norm": 2.8391079902648926,
-      "learning_rate": 1.2721773303896765e-05,
-      "loss": 1.3907,
-      "step": 602
-    },
-    {
-      "epoch": 0.8340248962655602,
-      "grad_norm": 3.021026372909546,
-      "learning_rate": 1.2700715355601107e-05,
-      "loss": 1.2413,
-      "step": 603
-    },
-    {
-      "epoch": 0.8354080221300139,
-      "grad_norm": 2.439363956451416,
-      "learning_rate": 1.2679644481146081e-05,
-      "loss": 1.3966,
-      "step": 604
-    },
-    {
-      "epoch": 0.8367911479944675,
-      "grad_norm": 2.6423158645629883,
-      "learning_rate": 1.265856078138107e-05,
-      "loss": 1.3411,
-      "step": 605
-    },
-    {
-      "epoch": 0.8381742738589212,
-      "grad_norm": 2.848080635070801,
-      "learning_rate": 1.2637464357216847e-05,
-      "loss": 1.3249,
-      "step": 606
-    },
-    {
-      "epoch": 0.8395573997233748,
-      "grad_norm": 2.5756824016571045,
-      "learning_rate": 1.2616355309625076e-05,
-      "loss": 1.4163,
-      "step": 607
-    },
-    {
-      "epoch": 0.8409405255878285,
-      "grad_norm": 2.7762558460235596,
-      "learning_rate": 1.2595233739637851e-05,
-      "loss": 1.3655,
-      "step": 608
-    },
-    {
-      "epoch": 0.8423236514522822,
-      "grad_norm": 2.599550485610962,
-      "learning_rate": 1.2574099748347195e-05,
-      "loss": 1.3372,
-      "step": 609
-    },
-    {
-      "epoch": 0.8437067773167358,
-      "grad_norm": 2.7273144721984863,
-      "learning_rate": 1.2552953436904578e-05,
-      "loss": 1.2719,
-      "step": 610
-    },
-    {
-      "epoch": 0.8450899031811895,
-      "grad_norm": 2.3901562690734863,
-      "learning_rate": 1.2531794906520447e-05,
-      "loss": 1.3207,
-      "step": 611
-    },
-    {
-      "epoch": 0.8464730290456431,
-      "grad_norm": 2.7380173206329346,
-      "learning_rate": 1.2510624258463719e-05,
-      "loss": 1.3461,
-      "step": 612
-    },
-    {
-      "epoch": 0.8478561549100968,
-      "grad_norm": 2.6389694213867188,
-      "learning_rate": 1.248944159406132e-05,
-      "loss": 1.3443,
-      "step": 613
-    },
-    {
-      "epoch": 0.8492392807745505,
-      "grad_norm": 2.623025417327881,
-      "learning_rate": 1.246824701469768e-05,
-      "loss": 1.3893,
-      "step": 614
-    },
-    {
-      "epoch": 0.8506224066390041,
-      "grad_norm": 2.646023988723755,
-      "learning_rate": 1.2447040621814262e-05,
-      "loss": 1.372,
-      "step": 615
-    },
-    {
-      "epoch": 0.8520055325034578,
-      "grad_norm": 2.4752604961395264,
-      "learning_rate": 1.2425822516909065e-05,
-      "loss": 1.413,
-      "step": 616
-    },
-    {
-      "epoch": 0.8533886583679114,
-      "grad_norm": 2.928985834121704,
-      "learning_rate": 1.2404592801536151e-05,
-      "loss": 1.3141,
-      "step": 617
-    },
-    {
-      "epoch": 0.8547717842323651,
-      "grad_norm": 2.6695446968078613,
-      "learning_rate": 1.2383351577305148e-05,
-      "loss": 1.3596,
-      "step": 618
-    },
-    {
-      "epoch": 0.8561549100968188,
-      "grad_norm": 2.6358587741851807,
-      "learning_rate": 1.2362098945880765e-05,
-      "loss": 1.2506,
-      "step": 619
-    },
-    {
-      "epoch": 0.8575380359612724,
-      "grad_norm": 2.5874736309051514,
-      "learning_rate": 1.2340835008982315e-05,
-      "loss": 1.3993,
-      "step": 620
-    },
-    {
-      "epoch": 0.8589211618257261,
-      "grad_norm": 2.909454822540283,
-      "learning_rate": 1.2319559868383215e-05,
-      "loss": 1.3344,
-      "step": 621
-    },
-    {
-      "epoch": 0.8603042876901799,
-      "grad_norm": 2.6511762142181396,
-      "learning_rate": 1.2298273625910512e-05,
-      "loss": 1.3711,
-      "step": 622
-    },
-    {
-      "epoch": 0.8616874135546335,
-      "grad_norm": 2.508411169052124,
-      "learning_rate": 1.2276976383444384e-05,
-      "loss": 1.3315,
-      "step": 623
-    },
-    {
-      "epoch": 0.8630705394190872,
-      "grad_norm": 2.6246912479400635,
-      "learning_rate": 1.2255668242917651e-05,
-      "loss": 1.3493,
-      "step": 624
-    },
-    {
-      "epoch": 0.8644536652835408,
-      "grad_norm": 2.5388360023498535,
-      "learning_rate": 1.2234349306315308e-05,
-      "loss": 1.394,
-      "step": 625
-    },
-    {
-      "epoch": 0.8658367911479945,
-      "grad_norm": 2.443962812423706,
-      "learning_rate": 1.2213019675674008e-05,
-      "loss": 1.3458,
-      "step": 626
-    },
-    {
-      "epoch": 0.8672199170124482,
-      "grad_norm": 2.2994024753570557,
-      "learning_rate": 1.2191679453081598e-05,
-      "loss": 1.1168,
-      "step": 627
-    },
-    {
-      "epoch": 0.8686030428769018,
-      "grad_norm": 2.854074239730835,
-      "learning_rate": 1.2170328740676613e-05,
-      "loss": 1.4029,
-      "step": 628
-    },
-    {
-      "epoch": 0.8699861687413555,
-      "grad_norm": 2.6301426887512207,
-      "learning_rate": 1.2148967640647801e-05,
-      "loss": 1.2258,
-      "step": 629
-    },
-    {
-      "epoch": 0.8713692946058091,
-      "grad_norm": 2.680180311203003,
-      "learning_rate": 1.2127596255233622e-05,
-      "loss": 1.324,
-      "step": 630
-    },
-    {
-      "epoch": 0.8727524204702628,
-      "grad_norm": 2.532710075378418,
-      "learning_rate": 1.2106214686721763e-05,
-      "loss": 1.3227,
-      "step": 631
-    },
-    {
-      "epoch": 0.8741355463347165,
-      "grad_norm": 2.6774704456329346,
-      "learning_rate": 1.2084823037448654e-05,
-      "loss": 1.3242,
-      "step": 632
-    },
-    {
-      "epoch": 0.8755186721991701,
-      "grad_norm": 2.6629583835601807,
-      "learning_rate": 1.2063421409798974e-05,
-      "loss": 1.3086,
-      "step": 633
-    },
-    {
-      "epoch": 0.8769017980636238,
-      "grad_norm": 2.9211409091949463,
-      "learning_rate": 1.2042009906205152e-05,
-      "loss": 1.2639,
-      "step": 634
-    },
-    {
-      "epoch": 0.8782849239280774,
-      "grad_norm": 2.9951515197753906,
-      "learning_rate": 1.2020588629146897e-05,
-      "loss": 1.3449,
-      "step": 635
-    },
-    {
-      "epoch": 0.8796680497925311,
-      "grad_norm": 3.1883134841918945,
-      "learning_rate": 1.1999157681150683e-05,
-      "loss": 1.3168,
-      "step": 636
-    },
-    {
-      "epoch": 0.8810511756569848,
-      "grad_norm": 2.857069253921509,
-      "learning_rate": 1.1977717164789286e-05,
-      "loss": 1.3151,
-      "step": 637
-    },
-    {
-      "epoch": 0.8824343015214384,
-      "grad_norm": 2.866753339767456,
-      "learning_rate": 1.1956267182681265e-05,
-      "loss": 1.4015,
-      "step": 638
-    },
-    {
-      "epoch": 0.8838174273858921,
-      "grad_norm": 2.751404047012329,
-      "learning_rate": 1.193480783749049e-05,
-      "loss": 1.417,
-      "step": 639
-    },
-    {
-      "epoch": 0.8852005532503457,
-      "grad_norm": 2.753148078918457,
-      "learning_rate": 1.1913339231925642e-05,
-      "loss": 1.2989,
-      "step": 640
-    },
-    {
-      "epoch": 0.8865836791147994,
-      "grad_norm": 2.9508845806121826,
-      "learning_rate": 1.1891861468739729e-05,
-      "loss": 1.3506,
-      "step": 641
-    },
-    {
-      "epoch": 0.8879668049792531,
-      "grad_norm": 2.8244197368621826,
-      "learning_rate": 1.1870374650729582e-05,
-      "loss": 1.3815,
-      "step": 642
-    },
-    {
-      "epoch": 0.8893499308437067,
-      "grad_norm": 2.7276086807250977,
-      "learning_rate": 1.1848878880735374e-05,
-      "loss": 1.4006,
-      "step": 643
-    },
-    {
-      "epoch": 0.8907330567081605,
-      "grad_norm": 2.7729735374450684,
-      "learning_rate": 1.1827374261640128e-05,
-      "loss": 1.3746,
-      "step": 644
-    },
-    {
-      "epoch": 0.8921161825726142,
-      "grad_norm": 3.0188214778900146,
-      "learning_rate": 1.1805860896369212e-05,
-      "loss": 1.2675,
-      "step": 645
-    },
-    {
-      "epoch": 0.8934993084370678,
-      "grad_norm": 2.7433032989501953,
-      "learning_rate": 1.1784338887889858e-05,
-      "loss": 1.406,
-      "step": 646
-    },
-    {
-      "epoch": 0.8948824343015215,
-      "grad_norm": 3.0397491455078125,
-      "learning_rate": 1.1762808339210672e-05,
-      "loss": 1.2682,
-      "step": 647
-    },
-    {
-      "epoch": 0.8962655601659751,
-      "grad_norm": 2.876121997833252,
-      "learning_rate": 1.1741269353381128e-05,
-      "loss": 1.4072,
-      "step": 648
-    },
-    {
-      "epoch": 0.8976486860304288,
-      "grad_norm": 2.9146461486816406,
-      "learning_rate": 1.1719722033491086e-05,
-      "loss": 1.3016,
-      "step": 649
-    },
-    {
-      "epoch": 0.8990318118948825,
-      "grad_norm": 2.6939985752105713,
-      "learning_rate": 1.1698166482670293e-05,
-      "loss": 1.4551,
-      "step": 650
-    },
-    {
-      "epoch": 0.9004149377593361,
-      "grad_norm": 2.525695323944092,
-      "learning_rate": 1.1676602804087887e-05,
-      "loss": 1.3592,
-      "step": 651
-    },
-    {
-      "epoch": 0.9017980636237898,
-      "grad_norm": 2.9363038539886475,
-      "learning_rate": 1.165503110095191e-05,
-      "loss": 1.3402,
-      "step": 652
-    },
-    {
-      "epoch": 0.9031811894882434,
-      "grad_norm": 2.9058566093444824,
-      "learning_rate": 1.1633451476508819e-05,
-      "loss": 1.3014,
-      "step": 653
-    },
-    {
-      "epoch": 0.9045643153526971,
-      "grad_norm": 2.6156883239746094,
-      "learning_rate": 1.1611864034042972e-05,
-      "loss": 1.4221,
-      "step": 654
-    },
-    {
-      "epoch": 0.9059474412171508,
-      "grad_norm": 2.717287302017212,
-      "learning_rate": 1.1590268876876151e-05,
-      "loss": 1.2827,
-      "step": 655
-    },
-    {
-      "epoch": 0.9073305670816044,
-      "grad_norm": 2.8226826190948486,
-      "learning_rate": 1.1568666108367066e-05,
-      "loss": 1.2393,
-      "step": 656
-    },
-    {
-      "epoch": 0.9087136929460581,
-      "grad_norm": 3.023271322250366,
-      "learning_rate": 1.1547055831910841e-05,
-      "loss": 1.2443,
-      "step": 657
-    },
-    {
-      "epoch": 0.9100968188105117,
-      "grad_norm": 3.041276693344116,
-      "learning_rate": 1.1525438150938554e-05,
-      "loss": 1.3386,
-      "step": 658
-    },
-    {
-      "epoch": 0.9114799446749654,
-      "grad_norm": 3.0393760204315186,
-      "learning_rate": 1.1503813168916715e-05,
-      "loss": 1.4778,
-      "step": 659
-    },
-    {
-      "epoch": 0.9128630705394191,
-      "grad_norm": 2.7392125129699707,
-      "learning_rate": 1.1482180989346771e-05,
-      "loss": 1.3789,
-      "step": 660
-    },
-    {
-      "epoch": 0.9142461964038727,
-      "grad_norm": 2.8566339015960693,
-      "learning_rate": 1.1460541715764628e-05,
-      "loss": 1.4119,
-      "step": 661
-    },
-    {
-      "epoch": 0.9156293222683264,
-      "grad_norm": 2.5865392684936523,
-      "learning_rate": 1.1438895451740141e-05,
-      "loss": 1.4265,
-      "step": 662
-    },
-    {
-      "epoch": 0.91701244813278,
-      "grad_norm": 2.861591100692749,
-      "learning_rate": 1.1417242300876621e-05,
-      "loss": 1.3768,
-      "step": 663
-    },
-    {
-      "epoch": 0.9183955739972337,
-      "grad_norm": 2.6707632541656494,
-      "learning_rate": 1.1395582366810348e-05,
-      "loss": 1.3279,
-      "step": 664
-    },
-    {
-      "epoch": 0.9197786998616874,
-      "grad_norm": 3.7327070236206055,
-      "learning_rate": 1.1373915753210056e-05,
-      "loss": 1.34,
-      "step": 665
-    },
-    {
-      "epoch": 0.921161825726141,
-      "grad_norm": 2.8060691356658936,
-      "learning_rate": 1.135224256377646e-05,
-      "loss": 1.3169,
-      "step": 666
-    },
-    {
-      "epoch": 0.9225449515905948,
-      "grad_norm": 2.822150707244873,
-      "learning_rate": 1.1330562902241742e-05,
-      "loss": 1.4016,
-      "step": 667
-    },
-    {
-      "epoch": 0.9239280774550485,
-      "grad_norm": 3.038985252380371,
-      "learning_rate": 1.1308876872369062e-05,
-      "loss": 1.2763,
-      "step": 668
-    },
-    {
-      "epoch": 0.9253112033195021,
-      "grad_norm": 2.8417439460754395,
-      "learning_rate": 1.128718457795206e-05,
-      "loss": 1.3405,
-      "step": 669
-    },
-    {
-      "epoch": 0.9266943291839558,
-      "grad_norm": 2.795173406600952,
-      "learning_rate": 1.1265486122814359e-05,
-      "loss": 1.3664,
-      "step": 670
-    },
-    {
-      "epoch": 0.9280774550484094,
-      "grad_norm": 2.7899844646453857,
-      "learning_rate": 1.124378161080907e-05,
-      "loss": 1.3715,
-      "step": 671
-    },
-    {
-      "epoch": 0.9294605809128631,
-      "grad_norm": 2.6709322929382324,
-      "learning_rate": 1.1222071145818293e-05,
-      "loss": 1.3386,
-      "step": 672
-    },
-    {
-      "epoch": 0.9308437067773168,
-      "grad_norm": 2.7561426162719727,
-      "learning_rate": 1.120035483175262e-05,
-      "loss": 1.3753,
-      "step": 673
-    },
-    {
-      "epoch": 0.9322268326417704,
-      "grad_norm": 2.7683603763580322,
-      "learning_rate": 1.1178632772550636e-05,
-      "loss": 1.2831,
-      "step": 674
-    },
-    {
-      "epoch": 0.9336099585062241,
-      "grad_norm": 3.2102935314178467,
-      "learning_rate": 1.1156905072178425e-05,
-      "loss": 1.2388,
-      "step": 675
-    },
-    {
-      "epoch": 0.9349930843706777,
-      "grad_norm": 2.959686040878296,
-      "learning_rate": 1.113517183462907e-05,
-      "loss": 1.3251,
-      "step": 676
-    },
-    {
-      "epoch": 0.9363762102351314,
-      "grad_norm": 2.9151217937469482,
-      "learning_rate": 1.1113433163922161e-05,
-      "loss": 1.3598,
-      "step": 677
-    },
-    {
-      "epoch": 0.9377593360995851,
-      "grad_norm": 2.778136730194092,
-      "learning_rate": 1.1091689164103281e-05,
-      "loss": 1.3768,
-      "step": 678
-    },
-    {
-      "epoch": 0.9391424619640387,
-      "grad_norm": 2.6228294372558594,
-      "learning_rate": 1.1069939939243531e-05,
-      "loss": 1.2864,
-      "step": 679
-    },
-    {
-      "epoch": 0.9405255878284924,
-      "grad_norm": 2.5928773880004883,
-      "learning_rate": 1.1048185593439014e-05,
-      "loss": 1.3992,
-      "step": 680
-    },
-    {
-      "epoch": 0.941908713692946,
-      "grad_norm": 3.3055930137634277,
-      "learning_rate": 1.1026426230810342e-05,
-      "loss": 1.2491,
-      "step": 681
-    },
-    {
-      "epoch": 0.9432918395573997,
-      "grad_norm": 3.021859884262085,
-      "learning_rate": 1.1004661955502143e-05,
-      "loss": 1.3086,
-      "step": 682
-    },
-    {
-      "epoch": 0.9446749654218534,
-      "grad_norm": 3.3848326206207275,
-      "learning_rate": 1.0982892871682556e-05,
-      "loss": 1.363,
-      "step": 683
-    },
-    {
-      "epoch": 0.946058091286307,
-      "grad_norm": 3.09657883644104,
-      "learning_rate": 1.0961119083542727e-05,
-      "loss": 1.2815,
-      "step": 684
-    },
-    {
-      "epoch": 0.9474412171507607,
-      "grad_norm": 2.734449625015259,
-      "learning_rate": 1.0939340695296332e-05,
-      "loss": 1.326,
-      "step": 685
-    },
-    {
-      "epoch": 0.9488243430152143,
-      "grad_norm": 3.091350555419922,
-      "learning_rate": 1.0917557811179057e-05,
-      "loss": 1.3682,
-      "step": 686
-    },
-    {
-      "epoch": 0.950207468879668,
-      "grad_norm": 2.9399776458740234,
-      "learning_rate": 1.0895770535448103e-05,
-      "loss": 1.1755,
-      "step": 687
-    },
-    {
-      "epoch": 0.9515905947441217,
-      "grad_norm": 3.312972068786621,
-      "learning_rate": 1.0873978972381692e-05,
-      "loss": 1.4613,
-      "step": 688
-    },
-    {
-      "epoch": 0.9529737206085753,
-      "grad_norm": 3.2397265434265137,
-      "learning_rate": 1.0852183226278568e-05,
-      "loss": 1.2265,
-      "step": 689
-    },
-    {
-      "epoch": 0.9543568464730291,
-      "grad_norm": 3.401766300201416,
-      "learning_rate": 1.0830383401457499e-05,
-      "loss": 1.1583,
-      "step": 690
-    },
-    {
-      "epoch": 0.9557399723374828,
-      "grad_norm": 3.0945234298706055,
-      "learning_rate": 1.0808579602256766e-05,
-      "loss": 1.2549,
-      "step": 691
-    },
-    {
-      "epoch": 0.9571230982019364,
-      "grad_norm": 2.892479658126831,
-      "learning_rate": 1.0786771933033677e-05,
-      "loss": 1.2499,
-      "step": 692
-    },
-    {
-      "epoch": 0.9585062240663901,
-      "grad_norm": 3.107635736465454,
-      "learning_rate": 1.0764960498164066e-05,
-      "loss": 1.3087,
-      "step": 693
-    },
-    {
-      "epoch": 0.9598893499308437,
-      "grad_norm": 3.117335319519043,
-      "learning_rate": 1.0743145402041781e-05,
-      "loss": 1.2303,
-      "step": 694
-    },
-    {
-      "epoch": 0.9612724757952974,
-      "grad_norm": 3.034398078918457,
-      "learning_rate": 1.0721326749078205e-05,
-      "loss": 1.315,
-      "step": 695
-    },
-    {
-      "epoch": 0.9626556016597511,
-      "grad_norm": 3.142618417739868,
-      "learning_rate": 1.0699504643701732e-05,
-      "loss": 1.3106,
-      "step": 696
-    },
-    {
-      "epoch": 0.9640387275242047,
-      "grad_norm": 2.986525535583496,
-      "learning_rate": 1.0677679190357292e-05,
-      "loss": 1.1909,
-      "step": 697
-    },
-    {
-      "epoch": 0.9654218533886584,
-      "grad_norm": 3.2569093704223633,
-      "learning_rate": 1.0655850493505834e-05,
-      "loss": 1.2402,
-      "step": 698
-    },
-    {
-      "epoch": 0.966804979253112,
-      "grad_norm": 3.035935163497925,
-      "learning_rate": 1.0634018657623827e-05,
-      "loss": 1.3137,
-      "step": 699
-    },
-    {
-      "epoch": 0.9681881051175657,
-      "grad_norm": 2.9152543544769287,
-      "learning_rate": 1.0612183787202768e-05,
-      "loss": 1.299,
-      "step": 700
-    },
-    {
-      "epoch": 0.9695712309820194,
-      "grad_norm": 3.0205094814300537,
-      "learning_rate": 1.059034598674868e-05,
-      "loss": 1.2816,
-      "step": 701
-    },
-    {
-      "epoch": 0.970954356846473,
-      "grad_norm": 3.073659658432007,
-      "learning_rate": 1.0568505360781606e-05,
-      "loss": 1.2782,
-      "step": 702
-    },
-    {
-      "epoch": 0.9723374827109267,
-      "grad_norm": 3.1953279972076416,
-      "learning_rate": 1.0546662013835119e-05,
-      "loss": 1.3486,
-      "step": 703
-    },
-    {
-      "epoch": 0.9737206085753803,
-      "grad_norm": 3.613943576812744,
-      "learning_rate": 1.0524816050455801e-05,
-      "loss": 1.1737,
-      "step": 704
-    },
-    {
-      "epoch": 0.975103734439834,
-      "grad_norm": 3.117497682571411,
-      "learning_rate": 1.0502967575202769e-05,
-      "loss": 1.3392,
-      "step": 705
-    },
-    {
-      "epoch": 0.9764868603042877,
-      "grad_norm": 3.1448299884796143,
-      "learning_rate": 1.0481116692647165e-05,
-      "loss": 1.2706,
-      "step": 706
-    },
-    {
-      "epoch": 0.9778699861687413,
-      "grad_norm": 2.8863043785095215,
-      "learning_rate": 1.045926350737164e-05,
-      "loss": 1.3248,
-      "step": 707
-    },
-    {
-      "epoch": 0.979253112033195,
-      "grad_norm": 3.1689250469207764,
-      "learning_rate": 1.0437408123969877e-05,
-      "loss": 1.2143,
-      "step": 708
-    },
-    {
-      "epoch": 0.9806362378976486,
-      "grad_norm": 2.828787088394165,
-      "learning_rate": 1.0415550647046074e-05,
-      "loss": 1.3921,
-      "step": 709
-    },
-    {
-      "epoch": 0.9820193637621023,
-      "grad_norm": 2.9206900596618652,
-      "learning_rate": 1.039369118121445e-05,
-      "loss": 1.4469,
-      "step": 710
-    },
-    {
-      "epoch": 0.983402489626556,
-      "grad_norm": 2.994562864303589,
-      "learning_rate": 1.0371829831098747e-05,
-      "loss": 1.3058,
-      "step": 711
-    },
-    {
-      "epoch": 0.9847856154910097,
-      "grad_norm": 2.9757678508758545,
-      "learning_rate": 1.0349966701331721e-05,
-      "loss": 1.3384,
-      "step": 712
-    },
-    {
-      "epoch": 0.9861687413554634,
-      "grad_norm": 3.144249200820923,
-      "learning_rate": 1.0328101896554647e-05,
-      "loss": 1.3066,
-      "step": 713
-    },
-    {
-      "epoch": 0.9875518672199171,
-      "grad_norm": 3.0815372467041016,
-      "learning_rate": 1.0306235521416822e-05,
-      "loss": 1.3301,
-      "step": 714
-    },
-    {
-      "epoch": 0.9889349930843707,
-      "grad_norm": 2.8530967235565186,
-      "learning_rate": 1.0284367680575045e-05,
-      "loss": 1.3116,
-      "step": 715
-    },
-    {
-      "epoch": 0.9903181189488244,
-      "grad_norm": 2.911203145980835,
-      "learning_rate": 1.0262498478693148e-05,
-      "loss": 1.3735,
-      "step": 716
-    },
-    {
-      "epoch": 0.991701244813278,
-      "grad_norm": 2.647660493850708,
-      "learning_rate": 1.0240628020441468e-05,
-      "loss": 1.3103,
-      "step": 717
-    },
-    {
-      "epoch": 0.9930843706777317,
-      "grad_norm": 3.142230749130249,
-      "learning_rate": 1.0218756410496353e-05,
-      "loss": 1.2823,
-      "step": 718
-    },
-    {
-      "epoch": 0.9944674965421854,
-      "grad_norm": 2.798553705215454,
-      "learning_rate": 1.019688375353967e-05,
-      "loss": 1.3124,
-      "step": 719
-    },
-    {
-      "epoch": 0.995850622406639,
-      "grad_norm": 3.3929383754730225,
-      "learning_rate": 1.0175010154258288e-05,
-      "loss": 1.2478,
-      "step": 720
-    },
-    {
-      "epoch": 0.9972337482710927,
-      "grad_norm": 3.2481846809387207,
-      "learning_rate": 1.0153135717343599e-05,
-      "loss": 1.2861,
-      "step": 721
-    },
-    {
-      "epoch": 0.9986168741355463,
-      "grad_norm": 2.964385747909546,
-      "learning_rate": 1.013126054749099e-05,
-      "loss": 1.3865,
-      "step": 722
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 3.1757357120513916,
-      "learning_rate": 1.0109384749399369e-05,
-      "loss": 1.3613,
-      "step": 723
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 1446,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
-  "save_steps": 723,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": false
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 1.4040967111099023e+18,
-  "train_batch_size": 4,
-  "trial_name": null,
-  "trial_params": null
-}