diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13680 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9998846198223146,
+  "eval_steps": 500,
+  "global_step": 9750,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015384023691396484,
+      "grad_norm": 0.797955334186554,
+      "learning_rate": 4.999996755554811e-05,
+      "loss": 1.901,
+      "step": 5
+    },
+    {
+      "epoch": 0.003076804738279297,
+      "grad_norm": 1.7458454370498657,
+      "learning_rate": 4.999987022227664e-05,
+      "loss": 1.7519,
+      "step": 10
+    },
+    {
+      "epoch": 0.004615207107418946,
+      "grad_norm": 1.4588245153427124,
+      "learning_rate": 4.999970800043822e-05,
+      "loss": 1.8846,
+      "step": 15
+    },
+    {
+      "epoch": 0.006153609476558594,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 4.9999480890453916e-05,
+      "loss": 1.7849,
+      "step": 20
+    },
+    {
+      "epoch": 0.007692011845698243,
+      "grad_norm": 1.8351398706436157,
+      "learning_rate": 4.99991888929132e-05,
+      "loss": 1.5929,
+      "step": 25
+    },
+    {
+      "epoch": 0.009230414214837892,
+      "grad_norm": 1.145352840423584,
+      "learning_rate": 4.999883200857397e-05,
+      "loss": 1.4525,
+      "step": 30
+    },
+    {
+      "epoch": 0.010768816583977539,
+      "grad_norm": 2.642392873764038,
+      "learning_rate": 4.999841023836254e-05,
+      "loss": 1.676,
+      "step": 35
+    },
+    {
+      "epoch": 0.012307218953117188,
+      "grad_norm": 1.1834251880645752,
+      "learning_rate": 4.999792358337363e-05,
+      "loss": 1.4053,
+      "step": 40
+    },
+    {
+      "epoch": 0.013845621322256836,
+      "grad_norm": 1.87557053565979,
+      "learning_rate": 4.999737204487039e-05,
+      "loss": 1.395,
+      "step": 45
+    },
+    {
+      "epoch": 0.015384023691396485,
+      "grad_norm": 1.6970587968826294,
+      "learning_rate": 4.999675562428437e-05,
+      "loss": 1.5014,
+      "step": 50
+    },
+    {
+      "epoch": 0.016922426060536132,
+      "grad_norm": 1.1248029470443726,
+      "learning_rate": 4.999607432321551e-05,
+      "loss": 1.2834,
+      "step": 55
+    },
+    {
+      "epoch": 0.018460828429675783,
+      "grad_norm": 5.078829765319824,
+      "learning_rate": 4.999532814343219e-05,
+      "loss": 1.4268,
+      "step": 60
+    },
+    {
+      "epoch": 0.01999923079881543,
+      "grad_norm": 1.0455766916275024,
+      "learning_rate": 4.999451708687114e-05,
+      "loss": 1.33,
+      "step": 65
+    },
+    {
+      "epoch": 0.021537633167955077,
+      "grad_norm": 1.837597370147705,
+      "learning_rate": 4.999364115563751e-05,
+      "loss": 1.3419,
+      "step": 70
+    },
+    {
+      "epoch": 0.023076035537094728,
+      "grad_norm": 1.0537432432174683,
+      "learning_rate": 4.999270035200483e-05,
+      "loss": 1.3378,
+      "step": 75
+    },
+    {
+      "epoch": 0.024614437906234375,
+      "grad_norm": 2.875886917114258,
+      "learning_rate": 4.9991694678415e-05,
+      "loss": 1.263,
+      "step": 80
+    },
+    {
+      "epoch": 0.026152840275374026,
+      "grad_norm": 0.9554703831672668,
+      "learning_rate": 4.9990624137478314e-05,
+      "loss": 1.2777,
+      "step": 85
+    },
+    {
+      "epoch": 0.027691242644513673,
+      "grad_norm": 1.6160002946853638,
+      "learning_rate": 4.998948873197342e-05,
+      "loss": 1.3095,
+      "step": 90
+    },
+    {
+      "epoch": 0.02922964501365332,
+      "grad_norm": 0.982302188873291,
+      "learning_rate": 4.998828846484732e-05,
+      "loss": 1.3788,
+      "step": 95
+    },
+    {
+      "epoch": 0.03076804738279297,
+      "grad_norm": 1.1846606731414795,
+      "learning_rate": 4.9987023339215374e-05,
+      "loss": 1.3818,
+      "step": 100
+    },
+    {
+      "epoch": 0.03230644975193262,
+      "grad_norm": 1.7230656147003174,
+      "learning_rate": 4.9985693358361296e-05,
+      "loss": 1.2638,
+      "step": 105
+    },
+    {
+      "epoch": 0.033844852121072265,
+      "grad_norm": 1.3569287061691284,
+      "learning_rate": 4.998429852573712e-05,
+      "loss": 1.1706,
+      "step": 110
+    },
+    {
+      "epoch": 0.035383254490211916,
+      "grad_norm": 1.555022120475769,
+      "learning_rate": 4.998283884496321e-05,
+      "loss": 1.287,
+      "step": 115
+    },
+    {
+      "epoch": 0.036921656859351566,
+      "grad_norm": 1.9749863147735596,
+      "learning_rate": 4.998131431982826e-05,
+      "loss": 1.1886,
+      "step": 120
+    },
+    {
+      "epoch": 0.03846005922849121,
+      "grad_norm": 0.8346091508865356,
+      "learning_rate": 4.9979724954289244e-05,
+      "loss": 1.2522,
+      "step": 125
+    },
+    {
+      "epoch": 0.03999846159763086,
+      "grad_norm": 4.200445175170898,
+      "learning_rate": 4.997807075247146e-05,
+      "loss": 1.2879,
+      "step": 130
+    },
+    {
+      "epoch": 0.04153686396677051,
+      "grad_norm": 1.5512927770614624,
+      "learning_rate": 4.9976351718668476e-05,
+      "loss": 1.3098,
+      "step": 135
+    },
+    {
+      "epoch": 0.043075266335910155,
+      "grad_norm": 1.1664855480194092,
+      "learning_rate": 4.9974567857342155e-05,
+      "loss": 1.2395,
+      "step": 140
+    },
+    {
+      "epoch": 0.044613668705049805,
+      "grad_norm": 1.1823747158050537,
+      "learning_rate": 4.997271917312259e-05,
+      "loss": 1.2825,
+      "step": 145
+    },
+    {
+      "epoch": 0.046152071074189456,
+      "grad_norm": 1.0937031507492065,
+      "learning_rate": 4.997080567080817e-05,
+      "loss": 1.1966,
+      "step": 150
+    },
+    {
+      "epoch": 0.0476904734433291,
+      "grad_norm": 0.899167001247406,
+      "learning_rate": 4.9968827355365465e-05,
+      "loss": 1.2136,
+      "step": 155
+    },
+    {
+      "epoch": 0.04922887581246875,
+      "grad_norm": 1.9292325973510742,
+      "learning_rate": 4.996678423192933e-05,
+      "loss": 1.2735,
+      "step": 160
+    },
+    {
+      "epoch": 0.0507672781816084,
+      "grad_norm": 1.0430412292480469,
+      "learning_rate": 4.9964676305802794e-05,
+      "loss": 1.3386,
+      "step": 165
+    },
+    {
+      "epoch": 0.05230568055074805,
+      "grad_norm": 1.879791498184204,
+      "learning_rate": 4.99625035824571e-05,
+      "loss": 1.2693,
+      "step": 170
+    },
+    {
+      "epoch": 0.053844082919887695,
+      "grad_norm": 1.6582942008972168,
+      "learning_rate": 4.996026606753167e-05,
+      "loss": 1.1652,
+      "step": 175
+    },
+    {
+      "epoch": 0.055382485289027346,
+      "grad_norm": 3.5379745960235596,
+      "learning_rate": 4.99579637668341e-05,
+      "loss": 1.3028,
+      "step": 180
+    },
+    {
+      "epoch": 0.056920887658166996,
+      "grad_norm": 3.7615771293640137,
+      "learning_rate": 4.9955596686340154e-05,
+      "loss": 1.1635,
+      "step": 185
+    },
+    {
+      "epoch": 0.05845929002730664,
+      "grad_norm": 0.928205668926239,
+      "learning_rate": 4.995316483219372e-05,
+      "loss": 1.4032,
+      "step": 190
+    },
+    {
+      "epoch": 0.05999769239644629,
+      "grad_norm": 1.0329424142837524,
+      "learning_rate": 4.995066821070679e-05,
+      "loss": 1.2417,
+      "step": 195
+    },
+    {
+      "epoch": 0.06153609476558594,
+      "grad_norm": 1.8608773946762085,
+      "learning_rate": 4.994810682835951e-05,
+      "loss": 1.2683,
+      "step": 200
+    },
+    {
+      "epoch": 0.06307449713472559,
+      "grad_norm": 2.3579938411712646,
+      "learning_rate": 4.9945480691800075e-05,
+      "loss": 1.252,
+      "step": 205
+    },
+    {
+      "epoch": 0.06461289950386524,
+      "grad_norm": 0.7121773958206177,
+      "learning_rate": 4.994278980784478e-05,
+      "loss": 1.2303,
+      "step": 210
+    },
+    {
+      "epoch": 0.06615130187300489,
+      "grad_norm": 0.783092200756073,
+      "learning_rate": 4.9940034183477954e-05,
+      "loss": 1.3052,
+      "step": 215
+    },
+    {
+      "epoch": 0.06768970424214453,
+      "grad_norm": 0.9442629814147949,
+      "learning_rate": 4.993721382585199e-05,
+      "loss": 1.3012,
+      "step": 220
+    },
+    {
+      "epoch": 0.06922810661128419,
+      "grad_norm": 1.0986599922180176,
+      "learning_rate": 4.9934328742287285e-05,
+      "loss": 1.1937,
+      "step": 225
+    },
+    {
+      "epoch": 0.07076650898042383,
+      "grad_norm": 1.1720629930496216,
+      "learning_rate": 4.9931378940272214e-05,
+      "loss": 1.2892,
+      "step": 230
+    },
+    {
+      "epoch": 0.07230491134956347,
+      "grad_norm": 1.6844689846038818,
+      "learning_rate": 4.992836442746317e-05,
+      "loss": 1.1496,
+      "step": 235
+    },
+    {
+      "epoch": 0.07384331371870313,
+      "grad_norm": 1.4238617420196533,
+      "learning_rate": 4.992528521168449e-05,
+      "loss": 1.2769,
+      "step": 240
+    },
+    {
+      "epoch": 0.07538171608784278,
+      "grad_norm": 1.1840909719467163,
+      "learning_rate": 4.992214130092845e-05,
+      "loss": 1.2355,
+      "step": 245
+    },
+    {
+      "epoch": 0.07692011845698242,
+      "grad_norm": 1.3611423969268799,
+      "learning_rate": 4.9918932703355256e-05,
+      "loss": 1.3105,
+      "step": 250
+    },
+    {
+      "epoch": 0.07845852082612208,
+      "grad_norm": 1.2186192274093628,
+      "learning_rate": 4.991565942729298e-05,
+      "loss": 1.3476,
+      "step": 255
+    },
+    {
+      "epoch": 0.07999692319526172,
+      "grad_norm": 1.8337148427963257,
+      "learning_rate": 4.991232148123761e-05,
+      "loss": 1.1474,
+      "step": 260
+    },
+    {
+      "epoch": 0.08153532556440136,
+      "grad_norm": 1.6369004249572754,
+      "learning_rate": 4.990891887385297e-05,
+      "loss": 1.1932,
+      "step": 265
+    },
+    {
+      "epoch": 0.08307372793354102,
+      "grad_norm": 1.483383059501648,
+      "learning_rate": 4.9905451613970725e-05,
+      "loss": 1.2814,
+      "step": 270
+    },
+    {
+      "epoch": 0.08461213030268067,
+      "grad_norm": 0.9386169910430908,
+      "learning_rate": 4.990191971059033e-05,
+      "loss": 1.337,
+      "step": 275
+    },
+    {
+      "epoch": 0.08615053267182031,
+      "grad_norm": 0.8377572894096375,
+      "learning_rate": 4.989832317287904e-05,
+      "loss": 1.1444,
+      "step": 280
+    },
+    {
+      "epoch": 0.08768893504095997,
+      "grad_norm": 2.5681023597717285,
+      "learning_rate": 4.9894662010171874e-05,
+      "loss": 1.3671,
+      "step": 285
+    },
+    {
+      "epoch": 0.08922733741009961,
+      "grad_norm": 0.8709166049957275,
+      "learning_rate": 4.98909362319716e-05,
+      "loss": 1.1663,
+      "step": 290
+    },
+    {
+      "epoch": 0.09076573977923925,
+      "grad_norm": 0.9449151158332825,
+      "learning_rate": 4.988714584794866e-05,
+      "loss": 1.2205,
+      "step": 295
+    },
+    {
+      "epoch": 0.09230414214837891,
+      "grad_norm": 1.1699813604354858,
+      "learning_rate": 4.988329086794122e-05,
+      "loss": 1.3335,
+      "step": 300
+    },
+    {
+      "epoch": 0.09384254451751856,
+      "grad_norm": 1.5694646835327148,
+      "learning_rate": 4.98793713019551e-05,
+      "loss": 1.3269,
+      "step": 305
+    },
+    {
+      "epoch": 0.0953809468866582,
+      "grad_norm": 1.073695421218872,
+      "learning_rate": 4.9875387160163744e-05,
+      "loss": 1.2983,
+      "step": 310
+    },
+    {
+      "epoch": 0.09691934925579786,
+      "grad_norm": 0.9977647066116333,
+      "learning_rate": 4.987133845290822e-05,
+      "loss": 1.2401,
+      "step": 315
+    },
+    {
+      "epoch": 0.0984577516249375,
+      "grad_norm": 1.1257829666137695,
+      "learning_rate": 4.986722519069719e-05,
+      "loss": 1.1723,
+      "step": 320
+    },
+    {
+      "epoch": 0.09999615399407714,
+      "grad_norm": 2.4303648471832275,
+      "learning_rate": 4.9863047384206835e-05,
+      "loss": 1.2318,
+      "step": 325
+    },
+    {
+      "epoch": 0.1015345563632168,
+      "grad_norm": 2.8859667778015137,
+      "learning_rate": 4.9858805044280895e-05,
+      "loss": 1.2859,
+      "step": 330
+    },
+    {
+      "epoch": 0.10307295873235645,
+      "grad_norm": 1.5526022911071777,
+      "learning_rate": 4.985449818193061e-05,
+      "loss": 1.26,
+      "step": 335
+    },
+    {
+      "epoch": 0.1046113611014961,
+      "grad_norm": 1.387646198272705,
+      "learning_rate": 4.9850126808334665e-05,
+      "loss": 1.1915,
+      "step": 340
+    },
+    {
+      "epoch": 0.10614976347063575,
+      "grad_norm": 1.4605071544647217,
+      "learning_rate": 4.984569093483922e-05,
+      "loss": 1.2653,
+      "step": 345
+    },
+    {
+      "epoch": 0.10768816583977539,
+      "grad_norm": 0.9224326610565186,
+      "learning_rate": 4.984119057295783e-05,
+      "loss": 1.1685,
+      "step": 350
+    },
+    {
+      "epoch": 0.10922656820891505,
+      "grad_norm": 0.7252030372619629,
+      "learning_rate": 4.983662573437143e-05,
+      "loss": 1.1732,
+      "step": 355
+    },
+    {
+      "epoch": 0.11076497057805469,
+      "grad_norm": 1.599922776222229,
+      "learning_rate": 4.9831996430928326e-05,
+      "loss": 1.144,
+      "step": 360
+    },
+    {
+      "epoch": 0.11230337294719434,
+      "grad_norm": 1.3098397254943848,
+      "learning_rate": 4.9827302674644126e-05,
+      "loss": 1.1434,
+      "step": 365
+    },
+    {
+      "epoch": 0.11384177531633399,
+      "grad_norm": 1.1649527549743652,
+      "learning_rate": 4.982254447770175e-05,
+      "loss": 1.2149,
+      "step": 370
+    },
+    {
+      "epoch": 0.11538017768547364,
+      "grad_norm": 1.794103741645813,
+      "learning_rate": 4.981772185245135e-05,
+      "loss": 1.1748,
+      "step": 375
+    },
+    {
+      "epoch": 0.11691858005461328,
+      "grad_norm": 0.9207557439804077,
+      "learning_rate": 4.981283481141034e-05,
+      "loss": 1.2767,
+      "step": 380
+    },
+    {
+      "epoch": 0.11845698242375294,
+      "grad_norm": 1.2299586534500122,
+      "learning_rate": 4.980788336726328e-05,
+      "loss": 1.2229,
+      "step": 385
+    },
+    {
+      "epoch": 0.11999538479289258,
+      "grad_norm": 1.6368671655654907,
+      "learning_rate": 4.980286753286195e-05,
+      "loss": 1.1892,
+      "step": 390
+    },
+    {
+      "epoch": 0.12153378716203223,
+      "grad_norm": 0.9977037906646729,
+      "learning_rate": 4.9797787321225215e-05,
+      "loss": 1.2199,
+      "step": 395
+    },
+    {
+      "epoch": 0.12307218953117188,
+      "grad_norm": 1.6430251598358154,
+      "learning_rate": 4.979264274553905e-05,
+      "loss": 1.167,
+      "step": 400
+    },
+    {
+      "epoch": 0.12461059190031153,
+      "grad_norm": 0.9530766606330872,
+      "learning_rate": 4.97874338191565e-05,
+      "loss": 1.1379,
+      "step": 405
+    },
+    {
+      "epoch": 0.12614899426945117,
+      "grad_norm": 1.4140937328338623,
+      "learning_rate": 4.978216055559761e-05,
+      "loss": 1.2004,
+      "step": 410
+    },
+    {
+      "epoch": 0.12768739663859083,
+      "grad_norm": 2.96806001663208,
+      "learning_rate": 4.9776822968549454e-05,
+      "loss": 1.1749,
+      "step": 415
+    },
+    {
+      "epoch": 0.12922579900773049,
+      "grad_norm": 1.0619330406188965,
+      "learning_rate": 4.977142107186602e-05,
+      "loss": 1.1946,
+      "step": 420
+    },
+    {
+      "epoch": 0.13076420137687012,
+      "grad_norm": 3.656414031982422,
+      "learning_rate": 4.976595487956823e-05,
+      "loss": 1.1884,
+      "step": 425
+    },
+    {
+      "epoch": 0.13230260374600977,
+      "grad_norm": 2.3072898387908936,
+      "learning_rate": 4.976042440584392e-05,
+      "loss": 1.1723,
+      "step": 430
+    },
+    {
+      "epoch": 0.13384100611514943,
+      "grad_norm": 1.3326804637908936,
+      "learning_rate": 4.975482966504772e-05,
+      "loss": 1.2658,
+      "step": 435
+    },
+    {
+      "epoch": 0.13537940848428906,
+      "grad_norm": 1.7775696516036987,
+      "learning_rate": 4.97491706717011e-05,
+      "loss": 1.2043,
+      "step": 440
+    },
+    {
+      "epoch": 0.13691781085342872,
+      "grad_norm": 2.0168044567108154,
+      "learning_rate": 4.97434474404923e-05,
+      "loss": 1.2228,
+      "step": 445
+    },
+    {
+      "epoch": 0.13845621322256838,
+      "grad_norm": 1.757367730140686,
+      "learning_rate": 4.973765998627628e-05,
+      "loss": 1.0853,
+      "step": 450
+    },
+    {
+      "epoch": 0.139994615591708,
+      "grad_norm": 1.5851964950561523,
+      "learning_rate": 4.9731808324074717e-05,
+      "loss": 1.1649,
+      "step": 455
+    },
+    {
+      "epoch": 0.14153301796084766,
+      "grad_norm": 0.7482730746269226,
+      "learning_rate": 4.9725892469075905e-05,
+      "loss": 1.1394,
+      "step": 460
+    },
+    {
+      "epoch": 0.14307142032998732,
+      "grad_norm": 1.041838526725769,
+      "learning_rate": 4.9719912436634796e-05,
+      "loss": 1.183,
+      "step": 465
+    },
+    {
+      "epoch": 0.14460982269912695,
+      "grad_norm": 1.5557304620742798,
+      "learning_rate": 4.97138682422729e-05,
+      "loss": 1.0956,
+      "step": 470
+    },
+    {
+      "epoch": 0.1461482250682666,
+      "grad_norm": 2.4157028198242188,
+      "learning_rate": 4.970775990167826e-05,
+      "loss": 1.2898,
+      "step": 475
+    },
+    {
+      "epoch": 0.14768662743740626,
+      "grad_norm": 1.1285796165466309,
+      "learning_rate": 4.9701587430705415e-05,
+      "loss": 1.2317,
+      "step": 480
+    },
+    {
+      "epoch": 0.1492250298065459,
+      "grad_norm": 1.0138782262802124,
+      "learning_rate": 4.969535084537536e-05,
+      "loss": 1.2337,
+      "step": 485
+    },
+    {
+      "epoch": 0.15076343217568555,
+      "grad_norm": 2.229642629623413,
+      "learning_rate": 4.9689050161875506e-05,
+      "loss": 1.196,
+      "step": 490
+    },
+    {
+      "epoch": 0.1523018345448252,
+      "grad_norm": 1.7063117027282715,
+      "learning_rate": 4.9682685396559627e-05,
+      "loss": 1.1881,
+      "step": 495
+    },
+    {
+      "epoch": 0.15384023691396484,
+      "grad_norm": 1.2955132722854614,
+      "learning_rate": 4.967625656594782e-05,
+      "loss": 1.1329,
+      "step": 500
+    },
+    {
+      "epoch": 0.1553786392831045,
+      "grad_norm": 1.562524437904358,
+      "learning_rate": 4.96697636867265e-05,
+      "loss": 1.1875,
+      "step": 505
+    },
+    {
+      "epoch": 0.15691704165224415,
+      "grad_norm": 1.5448559522628784,
+      "learning_rate": 4.966320677574827e-05,
+      "loss": 1.2321,
+      "step": 510
+    },
+    {
+      "epoch": 0.15845544402138378,
+      "grad_norm": 1.5567346811294556,
+      "learning_rate": 4.9656585850031987e-05,
+      "loss": 1.3752,
+      "step": 515
+    },
+    {
+      "epoch": 0.15999384639052344,
+      "grad_norm": 2.398003339767456,
+      "learning_rate": 4.964990092676263e-05,
+      "loss": 1.308,
+      "step": 520
+    },
+    {
+      "epoch": 0.1615322487596631,
+      "grad_norm": 1.0853981971740723,
+      "learning_rate": 4.964315202329127e-05,
+      "loss": 1.1672,
+      "step": 525
+    },
+    {
+      "epoch": 0.16307065112880273,
+      "grad_norm": 1.2827069759368896,
+      "learning_rate": 4.963633915713509e-05,
+      "loss": 1.1813,
+      "step": 530
+    },
+    {
+      "epoch": 0.1646090534979424,
+      "grad_norm": 1.0297755002975464,
+      "learning_rate": 4.962946234597726e-05,
+      "loss": 1.224,
+      "step": 535
+    },
+    {
+      "epoch": 0.16614745586708204,
+      "grad_norm": 1.2967716455459595,
+      "learning_rate": 4.962252160766693e-05,
+      "loss": 1.216,
+      "step": 540
+    },
+    {
+      "epoch": 0.16768585823622167,
+      "grad_norm": 1.1088382005691528,
+      "learning_rate": 4.961551696021918e-05,
+      "loss": 1.2665,
+      "step": 545
+    },
+    {
+      "epoch": 0.16922426060536133,
+      "grad_norm": 1.604500412940979,
+      "learning_rate": 4.960844842181494e-05,
+      "loss": 1.2242,
+      "step": 550
+    },
+    {
+      "epoch": 0.170762662974501,
+      "grad_norm": 0.8669219017028809,
+      "learning_rate": 4.960131601080104e-05,
+      "loss": 1.0902,
+      "step": 555
+    },
+    {
+      "epoch": 0.17230106534364062,
+      "grad_norm": 2.331838369369507,
+      "learning_rate": 4.9594119745690014e-05,
+      "loss": 1.2584,
+      "step": 560
+    },
+    {
+      "epoch": 0.17383946771278028,
+      "grad_norm": 1.547214388847351,
+      "learning_rate": 4.95868596451602e-05,
+      "loss": 1.2759,
+      "step": 565
+    },
+    {
+      "epoch": 0.17537787008191993,
+      "grad_norm": 0.90031498670578,
+      "learning_rate": 4.957953572805558e-05,
+      "loss": 1.1871,
+      "step": 570
+    },
+    {
+      "epoch": 0.17691627245105956,
+      "grad_norm": 1.2103489637374878,
+      "learning_rate": 4.957214801338581e-05,
+      "loss": 1.2548,
+      "step": 575
+    },
+    {
+      "epoch": 0.17845467482019922,
+      "grad_norm": 1.005483865737915,
+      "learning_rate": 4.956469652032609e-05,
+      "loss": 1.1955,
+      "step": 580
+    },
+    {
+      "epoch": 0.17999307718933888,
+      "grad_norm": 0.9381604790687561,
+      "learning_rate": 4.9557181268217227e-05,
+      "loss": 1.1725,
+      "step": 585
+    },
+    {
+      "epoch": 0.1815314795584785,
+      "grad_norm": 1.0934962034225464,
+      "learning_rate": 4.9549602276565435e-05,
+      "loss": 1.1608,
+      "step": 590
+    },
+    {
+      "epoch": 0.18306988192761817,
+      "grad_norm": 1.2921926975250244,
+      "learning_rate": 4.954195956504245e-05,
+      "loss": 1.2638,
+      "step": 595
+    },
+    {
+      "epoch": 0.18460828429675782,
+      "grad_norm": 1.0068799257278442,
+      "learning_rate": 4.953425315348534e-05,
+      "loss": 1.1962,
+      "step": 600
+    },
+    {
+      "epoch": 0.18614668666589745,
+      "grad_norm": 1.205253005027771,
+      "learning_rate": 4.9526483061896534e-05,
+      "loss": 1.2266,
+      "step": 605
+    },
+    {
+      "epoch": 0.1876850890350371,
+      "grad_norm": 2.4699807167053223,
+      "learning_rate": 4.951864931044374e-05,
+      "loss": 1.2421,
+      "step": 610
+    },
+    {
+      "epoch": 0.18922349140417677,
+      "grad_norm": 1.494227409362793,
+      "learning_rate": 4.9510751919459895e-05,
+      "loss": 1.0911,
+      "step": 615
+    },
+    {
+      "epoch": 0.1907618937733164,
+      "grad_norm": 1.6315932273864746,
+      "learning_rate": 4.950279090944313e-05,
+      "loss": 1.2937,
+      "step": 620
+    },
+    {
+      "epoch": 0.19230029614245606,
+      "grad_norm": 1.1606189012527466,
+      "learning_rate": 4.949476630105669e-05,
+      "loss": 1.2341,
+      "step": 625
+    },
+    {
+      "epoch": 0.19383869851159571,
+      "grad_norm": 0.7757031321525574,
+      "learning_rate": 4.94866781151289e-05,
+      "loss": 1.1867,
+      "step": 630
+    },
+    {
+      "epoch": 0.19537710088073534,
+      "grad_norm": 1.1521573066711426,
+      "learning_rate": 4.9478526372653096e-05,
+      "loss": 1.2039,
+      "step": 635
+    },
+    {
+      "epoch": 0.196915503249875,
+      "grad_norm": 0.8454951643943787,
+      "learning_rate": 4.947031109478758e-05,
+      "loss": 1.1598,
+      "step": 640
+    },
+    {
+      "epoch": 0.19845390561901466,
+      "grad_norm": 1.3610525131225586,
+      "learning_rate": 4.9462032302855576e-05,
+      "loss": 1.2351,
+      "step": 645
+    },
+    {
+      "epoch": 0.1999923079881543,
+      "grad_norm": 1.006761908531189,
+      "learning_rate": 4.9453690018345144e-05,
+      "loss": 1.3258,
+      "step": 650
+    },
+    {
+      "epoch": 0.20153071035729395,
+      "grad_norm": 1.7177060842514038,
+      "learning_rate": 4.9445284262909156e-05,
+      "loss": 1.2972,
+      "step": 655
+    },
+    {
+      "epoch": 0.2030691127264336,
+      "grad_norm": 1.063753604888916,
+      "learning_rate": 4.943681505836523e-05,
+      "loss": 1.1798,
+      "step": 660
+    },
+    {
+      "epoch": 0.20460751509557323,
+      "grad_norm": 1.1765042543411255,
+      "learning_rate": 4.9428282426695646e-05,
+      "loss": 1.2289,
+      "step": 665
+    },
+    {
+      "epoch": 0.2061459174647129,
+      "grad_norm": 1.552025556564331,
+      "learning_rate": 4.9419686390047334e-05,
+      "loss": 1.2983,
+      "step": 670
+    },
+    {
+      "epoch": 0.20768431983385255,
+      "grad_norm": 2.3030500411987305,
+      "learning_rate": 4.9411026970731805e-05,
+      "loss": 1.1431,
+      "step": 675
+    },
+    {
+      "epoch": 0.2092227222029922,
+      "grad_norm": 1.2778570652008057,
+      "learning_rate": 4.9402304191225044e-05,
+      "loss": 1.1139,
+      "step": 680
+    },
+    {
+      "epoch": 0.21076112457213184,
+      "grad_norm": 1.5465021133422852,
+      "learning_rate": 4.9393518074167536e-05,
+      "loss": 1.1983,
+      "step": 685
+    },
+    {
+      "epoch": 0.2122995269412715,
+      "grad_norm": 1.079221248626709,
+      "learning_rate": 4.9384668642364126e-05,
+      "loss": 1.1707,
+      "step": 690
+    },
+    {
+      "epoch": 0.21383792931041115,
+      "grad_norm": 0.9178045988082886,
+      "learning_rate": 4.937575591878403e-05,
+      "loss": 1.2633,
+      "step": 695
+    },
+    {
+      "epoch": 0.21537633167955078,
+      "grad_norm": 1.1767947673797607,
+      "learning_rate": 4.93667799265607e-05,
+      "loss": 1.1424,
+      "step": 700
+    },
+    {
+      "epoch": 0.21691473404869044,
+      "grad_norm": 1.7399604320526123,
+      "learning_rate": 4.935774068899184e-05,
+      "loss": 1.2006,
+      "step": 705
+    },
+    {
+      "epoch": 0.2184531364178301,
+      "grad_norm": 1.282569169998169,
+      "learning_rate": 4.934863822953929e-05,
+      "loss": 1.31,
+      "step": 710
+    },
+    {
+      "epoch": 0.21999153878696973,
+      "grad_norm": 1.0059244632720947,
+      "learning_rate": 4.933947257182901e-05,
+      "loss": 1.2209,
+      "step": 715
+    },
+    {
+      "epoch": 0.22152994115610938,
+      "grad_norm": 1.1119279861450195,
+      "learning_rate": 4.9330243739650964e-05,
+      "loss": 1.2628,
+      "step": 720
+    },
+    {
+      "epoch": 0.22306834352524904,
+      "grad_norm": 1.2987867593765259,
+      "learning_rate": 4.932095175695911e-05,
+      "loss": 1.2771,
+      "step": 725
+    },
+    {
+      "epoch": 0.22460674589438867,
+      "grad_norm": 2.659963846206665,
+      "learning_rate": 4.9311596647871317e-05,
+      "loss": 1.2315,
+      "step": 730
+    },
+    {
+      "epoch": 0.22614514826352833,
+      "grad_norm": 1.846644639968872,
+      "learning_rate": 4.9302178436669286e-05,
+      "loss": 1.1513,
+      "step": 735
+    },
+    {
+      "epoch": 0.22768355063266799,
+      "grad_norm": 1.6089102029800415,
+      "learning_rate": 4.929269714779852e-05,
+      "loss": 1.2085,
+      "step": 740
+    },
+    {
+      "epoch": 0.22922195300180762,
+      "grad_norm": 1.953465223312378,
+      "learning_rate": 4.9283152805868235e-05,
+      "loss": 1.2173,
+      "step": 745
+    },
+    {
+      "epoch": 0.23076035537094727,
+      "grad_norm": 1.127911925315857,
+      "learning_rate": 4.92735454356513e-05,
+      "loss": 1.2868,
+      "step": 750
+    },
+    {
+      "epoch": 0.23229875774008693,
+      "grad_norm": 1.9900950193405151,
+      "learning_rate": 4.9263875062084194e-05,
+      "loss": 1.2541,
+      "step": 755
+    },
+    {
+      "epoch": 0.23383716010922656,
+      "grad_norm": 1.0306469202041626,
+      "learning_rate": 4.925414171026691e-05,
+      "loss": 1.2301,
+      "step": 760
+    },
+    {
+      "epoch": 0.23537556247836622,
+      "grad_norm": 2.1911792755126953,
+      "learning_rate": 4.9244345405462903e-05,
+      "loss": 1.2235,
+      "step": 765
+    },
+    {
+      "epoch": 0.23691396484750588,
+      "grad_norm": 1.6600091457366943,
+      "learning_rate": 4.923448617309905e-05,
+      "loss": 1.2173,
+      "step": 770
+    },
+    {
+      "epoch": 0.2384523672166455,
+      "grad_norm": 1.517933964729309,
+      "learning_rate": 4.922456403876552e-05,
+      "loss": 1.2161,
+      "step": 775
+    },
+    {
+      "epoch": 0.23999076958578516,
+      "grad_norm": 2.4637537002563477,
+      "learning_rate": 4.9214579028215776e-05,
+      "loss": 1.1367,
+      "step": 780
+    },
+    {
+      "epoch": 0.24152917195492482,
+      "grad_norm": 1.1146929264068604,
+      "learning_rate": 4.9204531167366485e-05,
+      "loss": 1.2063,
+      "step": 785
+    },
+    {
+      "epoch": 0.24306757432406445,
+      "grad_norm": 1.044700026512146,
+      "learning_rate": 4.919442048229743e-05,
+      "loss": 1.2129,
+      "step": 790
+    },
+    {
+      "epoch": 0.2446059766932041,
+      "grad_norm": 1.2291778326034546,
+      "learning_rate": 4.918424699925145e-05,
+      "loss": 1.1697,
+      "step": 795
+    },
+    {
+      "epoch": 0.24614437906234377,
+      "grad_norm": 2.0124595165252686,
+      "learning_rate": 4.917401074463441e-05,
+      "loss": 1.1384,
+      "step": 800
+    },
+    {
+      "epoch": 0.2476827814314834,
+      "grad_norm": 1.1294913291931152,
+      "learning_rate": 4.916371174501507e-05,
+      "loss": 1.1691,
+      "step": 805
+    },
+    {
+      "epoch": 0.24922118380062305,
+      "grad_norm": 2.2073304653167725,
+      "learning_rate": 4.9153350027125064e-05,
+      "loss": 1.1847,
+      "step": 810
+    },
+    {
+      "epoch": 0.2507595861697627,
+      "grad_norm": 0.9351872205734253,
+      "learning_rate": 4.9142925617858814e-05,
+      "loss": 1.1932,
+      "step": 815
+    },
+    {
+      "epoch": 0.25229798853890234,
+      "grad_norm": 1.3290927410125732,
+      "learning_rate": 4.913243854427346e-05,
+      "loss": 1.2509,
+      "step": 820
+    },
+    {
+      "epoch": 0.253836390908042,
+      "grad_norm": 1.7626986503601074,
+      "learning_rate": 4.9121888833588795e-05,
+      "loss": 1.2265,
+      "step": 825
+    },
+    {
+      "epoch": 0.25537479327718166,
+      "grad_norm": 1.9079293012619019,
+      "learning_rate": 4.911127651318717e-05,
+      "loss": 1.2083,
+      "step": 830
+    },
+    {
+      "epoch": 0.2569131956463213,
+      "grad_norm": 1.2099313735961914,
+      "learning_rate": 4.910060161061347e-05,
+      "loss": 1.0772,
+      "step": 835
+    },
+    {
+      "epoch": 0.25845159801546097,
+      "grad_norm": 2.2780282497406006,
+      "learning_rate": 4.9089864153575016e-05,
+      "loss": 1.1247,
+      "step": 840
+    },
+    {
+      "epoch": 0.2599900003846006,
+      "grad_norm": 1.4330651760101318,
+      "learning_rate": 4.907906416994146e-05,
+      "loss": 1.2317,
+      "step": 845
+    },
+    {
+      "epoch": 0.26152840275374023,
+      "grad_norm": 1.3475781679153442,
+      "learning_rate": 4.906820168774477e-05,
+      "loss": 1.2299,
+      "step": 850
+    },
+    {
+      "epoch": 0.2630668051228799,
+      "grad_norm": 1.4789565801620483,
+      "learning_rate": 4.905727673517914e-05,
+      "loss": 1.1717,
+      "step": 855
+    },
+    {
+      "epoch": 0.26460520749201955,
+      "grad_norm": 1.7257953882217407,
+      "learning_rate": 4.904628934060088e-05,
+      "loss": 1.1482,
+      "step": 860
+    },
+    {
+      "epoch": 0.2661436098611592,
+      "grad_norm": 1.5117738246917725,
+      "learning_rate": 4.903523953252841e-05,
+      "loss": 1.1459,
+      "step": 865
+    },
+    {
+      "epoch": 0.26768201223029886,
+      "grad_norm": 1.9107036590576172,
+      "learning_rate": 4.902412733964211e-05,
+      "loss": 1.1814,
+      "step": 870
+    },
+    {
+      "epoch": 0.26922041459943846,
+      "grad_norm": 1.9499353170394897,
+      "learning_rate": 4.901295279078431e-05,
+      "loss": 1.1091,
+      "step": 875
+    },
+    {
+      "epoch": 0.2707588169685781,
+      "grad_norm": 0.968377947807312,
+      "learning_rate": 4.900171591495918e-05,
+      "loss": 1.1259,
+      "step": 880
+    },
+    {
+      "epoch": 0.2722972193377178,
+      "grad_norm": 1.4124119281768799,
+      "learning_rate": 4.899041674133265e-05,
+      "loss": 1.227,
+      "step": 885
+    },
+    {
+      "epoch": 0.27383562170685743,
+      "grad_norm": 2.3951656818389893,
+      "learning_rate": 4.8979055299232376e-05,
+      "loss": 1.2486,
+      "step": 890
+    },
+    {
+      "epoch": 0.2753740240759971,
+      "grad_norm": 1.0433672666549683,
+      "learning_rate": 4.896763161814761e-05,
+      "loss": 1.2767,
+      "step": 895
+    },
+    {
+      "epoch": 0.27691242644513675,
+      "grad_norm": 3.254819393157959,
+      "learning_rate": 4.8956145727729156e-05,
+      "loss": 1.1728,
+      "step": 900
+    },
+    {
+      "epoch": 0.27845082881427635,
+      "grad_norm": 1.5969161987304688,
+      "learning_rate": 4.894459765778929e-05,
+      "loss": 1.0947,
+      "step": 905
+    },
+    {
+      "epoch": 0.279989231183416,
+      "grad_norm": 1.0399632453918457,
+      "learning_rate": 4.893298743830168e-05,
+      "loss": 1.1474,
+      "step": 910
+    },
+    {
+      "epoch": 0.28152763355255567,
+      "grad_norm": 0.8495728373527527,
+      "learning_rate": 4.89213150994013e-05,
+      "loss": 1.1966,
+      "step": 915
+    },
+    {
+      "epoch": 0.2830660359216953,
+      "grad_norm": 1.0868245363235474,
+      "learning_rate": 4.890958067138436e-05,
+      "loss": 1.194,
+      "step": 920
+    },
+    {
+      "epoch": 0.284604438290835,
+      "grad_norm": 1.8804415464401245,
+      "learning_rate": 4.889778418470823e-05,
+      "loss": 1.1109,
+      "step": 925
+    },
+    {
+      "epoch": 0.28614284065997464,
+      "grad_norm": 2.711575508117676,
+      "learning_rate": 4.8885925669991346e-05,
+      "loss": 1.1817,
+      "step": 930
+    },
+    {
+      "epoch": 0.28768124302911424,
+      "grad_norm": 1.0217710733413696,
+      "learning_rate": 4.887400515801315e-05,
+      "loss": 1.1885,
+      "step": 935
+    },
+    {
+      "epoch": 0.2892196453982539,
+      "grad_norm": 2.7488129138946533,
+      "learning_rate": 4.886202267971401e-05,
+      "loss": 1.1094,
+      "step": 940
+    },
+    {
+      "epoch": 0.29075804776739356,
+      "grad_norm": 1.5893644094467163,
+      "learning_rate": 4.8849978266195114e-05,
+      "loss": 1.1363,
+      "step": 945
+    },
+    {
+      "epoch": 0.2922964501365332,
+      "grad_norm": 1.7660548686981201,
+      "learning_rate": 4.883787194871841e-05,
+      "loss": 1.1993,
+      "step": 950
+    },
+    {
+      "epoch": 0.29383485250567287,
+      "grad_norm": 1.6953208446502686,
+      "learning_rate": 4.882570375870653e-05,
+      "loss": 1.2974,
+      "step": 955
+    },
+    {
+      "epoch": 0.29537325487481253,
+      "grad_norm": 1.191604733467102,
+      "learning_rate": 4.88134737277427e-05,
+      "loss": 1.0904,
+      "step": 960
+    },
+    {
+      "epoch": 0.29691165724395213,
+      "grad_norm": 1.317108392715454,
+      "learning_rate": 4.880118188757064e-05,
+      "loss": 1.1945,
+      "step": 965
+    },
+    {
+      "epoch": 0.2984500596130918,
+      "grad_norm": 1.4262393712997437,
+      "learning_rate": 4.878882827009452e-05,
+      "loss": 1.1332,
+      "step": 970
+    },
+    {
+      "epoch": 0.29998846198223145,
+      "grad_norm": 1.2089824676513672,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 1.1619,
+      "step": 975
+    },
+    {
+      "epoch": 0.3015268643513711,
+      "grad_norm": 1.008278727531433,
+      "learning_rate": 4.8763935831648374e-05,
+      "loss": 1.2385,
+      "step": 980
+    },
+    {
+      "epoch": 0.30306526672051076,
+      "grad_norm": 1.3324832916259766,
+      "learning_rate": 4.8751397075288084e-05,
+      "loss": 1.1853,
+      "step": 985
+    },
+    {
+      "epoch": 0.3046036690896504,
+      "grad_norm": 1.293948769569397,
+      "learning_rate": 4.8738796670843004e-05,
+      "loss": 1.17,
+      "step": 990
+    },
+    {
+      "epoch": 0.30614207145879,
+      "grad_norm": 1.5684713125228882,
+      "learning_rate": 4.8726134651018194e-05,
+      "loss": 1.1836,
+      "step": 995
+    },
+    {
+      "epoch": 0.3076804738279297,
+      "grad_norm": 1.8155791759490967,
+      "learning_rate": 4.8713411048678635e-05,
+      "loss": 1.0884,
+      "step": 1000
+    },
+    {
+      "epoch": 0.30921887619706934,
+      "grad_norm": 1.2308273315429688,
+      "learning_rate": 4.870062589684916e-05,
+      "loss": 1.2415,
+      "step": 1005
+    },
+    {
+      "epoch": 0.310757278566209,
+      "grad_norm": 1.091968297958374,
+      "learning_rate": 4.868777922871434e-05,
+      "loss": 1.1733,
+      "step": 1010
+    },
+    {
+      "epoch": 0.31229568093534865,
+      "grad_norm": 1.1831902265548706,
+      "learning_rate": 4.8674871077618424e-05,
+      "loss": 1.2495,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3138340833044883,
+      "grad_norm": 1.0130064487457275,
+      "learning_rate": 4.8661901477065244e-05,
+      "loss": 1.2221,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3153724856736279,
+      "grad_norm": 1.518776535987854,
+      "learning_rate": 4.864887046071813e-05,
+      "loss": 1.2491,
+      "step": 1025
+    },
+    {
+      "epoch": 0.31691088804276757,
+      "grad_norm": 2.234370708465576,
+      "learning_rate": 4.863577806239982e-05,
+      "loss": 1.2308,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3184492904119072,
+      "grad_norm": 1.2248209714889526,
+      "learning_rate": 4.862262431609235e-05,
+      "loss": 1.1884,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3199876927810469,
+      "grad_norm": 1.0601451396942139,
+      "learning_rate": 4.860940925593703e-05,
+      "loss": 1.0993,
+      "step": 1040
+    },
+    {
+      "epoch": 0.32152609515018654,
+      "grad_norm": 1.8158398866653442,
+      "learning_rate": 4.859613291623428e-05,
+      "loss": 1.267,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3230644975193262,
+      "grad_norm": 1.0854904651641846,
+      "learning_rate": 4.858279533144358e-05,
+      "loss": 1.185,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3246028998884658,
+      "grad_norm": 1.392730474472046,
+      "learning_rate": 4.856939653618339e-05,
+      "loss": 1.1568,
+      "step": 1055
+    },
+    {
+      "epoch": 0.32614130225760546,
+      "grad_norm": 1.2848827838897705,
+      "learning_rate": 4.855593656523103e-05,
+      "loss": 0.9792,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3276797046267451,
+      "grad_norm": 1.0641793012619019,
+      "learning_rate": 4.8542415453522615e-05,
+      "loss": 1.0783,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3292181069958848,
+      "grad_norm": 1.3434879779815674,
+      "learning_rate": 4.852883323615295e-05,
+      "loss": 1.1971,
+      "step": 1070
+    },
+    {
+      "epoch": 0.33075650936502443,
+      "grad_norm": 1.8678319454193115,
+      "learning_rate": 4.8515189948375434e-05,
+      "loss": 1.1928,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3322949117341641,
+      "grad_norm": 0.9058634638786316,
+      "learning_rate": 4.8501485625601996e-05,
+      "loss": 1.2081,
+      "step": 1080
+    },
+    {
+      "epoch": 0.33383331410330375,
+      "grad_norm": 1.7286523580551147,
+      "learning_rate": 4.848772030340297e-05,
+      "loss": 1.1119,
+      "step": 1085
+    },
+    {
+      "epoch": 0.33537171647244335,
+      "grad_norm": 0.9546146392822266,
+      "learning_rate": 4.847389401750701e-05,
+      "loss": 1.2201,
+      "step": 1090
+    },
+    {
+      "epoch": 0.336910118841583,
+      "grad_norm": 1.6596899032592773,
+      "learning_rate": 4.846000680380105e-05,
+      "loss": 1.2032,
+      "step": 1095
+    },
+    {
+      "epoch": 0.33844852121072266,
+      "grad_norm": 1.3017821311950684,
+      "learning_rate": 4.8446058698330115e-05,
+      "loss": 1.0737,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3399869235798623,
+      "grad_norm": 1.556097149848938,
+      "learning_rate": 4.843204973729729e-05,
+      "loss": 1.2315,
+      "step": 1105
+    },
+    {
+      "epoch": 0.341525325949002,
+      "grad_norm": 1.3806447982788086,
+      "learning_rate": 4.8417979957063624e-05,
+      "loss": 1.1993,
+      "step": 1110
+    },
+    {
+      "epoch": 0.34306372831814164,
+      "grad_norm": 1.5958844423294067,
+      "learning_rate": 4.8403849394148025e-05,
+      "loss": 1.1318,
+      "step": 1115
+    },
+    {
+      "epoch": 0.34460213068728124,
+      "grad_norm": 1.1803624629974365,
+      "learning_rate": 4.838965808522716e-05,
+      "loss": 1.1404,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3461405330564209,
+      "grad_norm": 1.1995015144348145,
+      "learning_rate": 4.837540606713538e-05,
+      "loss": 1.0803,
+      "step": 1125
+    },
+    {
+      "epoch": 0.34767893542556055,
+      "grad_norm": 1.4723166227340698,
+      "learning_rate": 4.836109337686457e-05,
+      "loss": 1.2123,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3492173377947002,
+      "grad_norm": 1.0283982753753662,
+      "learning_rate": 4.8346720051564144e-05,
+      "loss": 1.1247,
+      "step": 1135
+    },
+    {
+      "epoch": 0.35075574016383987,
+      "grad_norm": 1.0858715772628784,
+      "learning_rate": 4.833228612854087e-05,
+      "loss": 1.1716,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3522941425329795,
+      "grad_norm": 0.8430686593055725,
+      "learning_rate": 4.831779164525881e-05,
+      "loss": 1.3617,
+      "step": 1145
+    },
+    {
+      "epoch": 0.35383254490211913,
+      "grad_norm": 3.4253408908843994,
+      "learning_rate": 4.83032366393392e-05,
+      "loss": 1.1845,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3553709472712588,
+      "grad_norm": 1.2169939279556274,
+      "learning_rate": 4.828862114856038e-05,
+      "loss": 1.1945,
+      "step": 1155
+    },
+    {
+      "epoch": 0.35690934964039844,
+      "grad_norm": 0.8925756216049194,
+      "learning_rate": 4.827394521085767e-05,
+      "loss": 1.2021,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3584477520095381,
+      "grad_norm": 1.1597083806991577,
+      "learning_rate": 4.8259208864323304e-05,
+      "loss": 1.1994,
+      "step": 1165
+    },
+    {
+      "epoch": 0.35998615437867776,
+      "grad_norm": 1.1406500339508057,
+      "learning_rate": 4.8244412147206284e-05,
+      "loss": 1.1749,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3615245567478174,
+      "grad_norm": 0.8532546162605286,
+      "learning_rate": 4.822955509791233e-05,
+      "loss": 1.0968,
+      "step": 1175
+    },
+    {
+      "epoch": 0.363062959116957,
+      "grad_norm": 1.6647778749465942,
+      "learning_rate": 4.8214637755003745e-05,
+      "loss": 1.1396,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3646013614860967,
+      "grad_norm": 4.057651042938232,
+      "learning_rate": 4.819966015719933e-05,
+      "loss": 1.2997,
+      "step": 1185
+    },
+    {
+      "epoch": 0.36613976385523633,
+      "grad_norm": 0.9485501050949097,
+      "learning_rate": 4.8184622343374275e-05,
+      "loss": 1.2487,
+      "step": 1190
+    },
+    {
+      "epoch": 0.367678166224376,
+      "grad_norm": 1.2841589450836182,
+      "learning_rate": 4.8169524352560076e-05,
+      "loss": 1.1749,
+      "step": 1195
+    },
+    {
+      "epoch": 0.36921656859351565,
+      "grad_norm": 1.6274491548538208,
+      "learning_rate": 4.815436622394441e-05,
+      "loss": 1.2342,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3707549709626553,
+      "grad_norm": 1.1453361511230469,
+      "learning_rate": 4.813914799687107e-05,
+      "loss": 1.1538,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3722933733317949,
+      "grad_norm": 1.7069777250289917,
+      "learning_rate": 4.812386971083979e-05,
+      "loss": 1.0581,
+      "step": 1210
+    },
+    {
+      "epoch": 0.37383177570093457,
+      "grad_norm": 1.478285789489746,
+      "learning_rate": 4.810853140550624e-05,
+      "loss": 1.1604,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3753701780700742,
+      "grad_norm": 2.3933002948760986,
+      "learning_rate": 4.809313312068185e-05,
+      "loss": 1.2667,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3769085804392139,
+      "grad_norm": 1.7244596481323242,
+      "learning_rate": 4.8077674896333725e-05,
+      "loss": 1.2328,
+      "step": 1225
+    },
+    {
+      "epoch": 0.37844698280835354,
+      "grad_norm": 0.8237632513046265,
+      "learning_rate": 4.806215677258456e-05,
+      "loss": 1.1748,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3799853851774932,
+      "grad_norm": 1.9397246837615967,
+      "learning_rate": 4.8046578789712515e-05,
+      "loss": 1.1827,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3815237875466328,
+      "grad_norm": 1.060905933380127,
+      "learning_rate": 4.803094098815112e-05,
+      "loss": 1.1902,
+      "step": 1240
+    },
+    {
+      "epoch": 0.38306218991577246,
+      "grad_norm": 1.2361557483673096,
+      "learning_rate": 4.801524340848917e-05,
+      "loss": 1.1679,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3846005922849121,
+      "grad_norm": 1.1130070686340332,
+      "learning_rate": 4.799948609147061e-05,
+      "loss": 1.2108,
+      "step": 1250
+    },
+    {
+      "epoch": 0.38613899465405177,
+      "grad_norm": 1.0235941410064697,
+      "learning_rate": 4.798366907799444e-05,
+      "loss": 1.2093,
+      "step": 1255
+    },
+    {
+      "epoch": 0.38767739702319143,
+      "grad_norm": 1.1517040729522705,
+      "learning_rate": 4.7967792409114606e-05,
+      "loss": 1.1166,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3892157993923311,
+      "grad_norm": 1.4021333456039429,
+      "learning_rate": 4.795185612603991e-05,
+      "loss": 1.1535,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3907542017614707,
+      "grad_norm": 1.059432864189148,
+      "learning_rate": 4.7935860270133844e-05,
+      "loss": 1.266,
+      "step": 1270
+    },
+    {
+      "epoch": 0.39229260413061035,
+      "grad_norm": 1.1297060251235962,
+      "learning_rate": 4.791980488291456e-05,
+      "loss": 1.2552,
+      "step": 1275
+    },
+    {
+      "epoch": 0.39383100649975,
+      "grad_norm": 2.4244351387023926,
+      "learning_rate": 4.7903690006054726e-05,
+      "loss": 1.2457,
+      "step": 1280
+    },
+    {
+      "epoch": 0.39536940886888966,
+      "grad_norm": 1.4860352277755737,
+      "learning_rate": 4.7887515681381403e-05,
+      "loss": 1.0594,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3969078112380293,
+      "grad_norm": 1.5482337474822998,
+      "learning_rate": 4.787128195087596e-05,
+      "loss": 1.1454,
+      "step": 1290
+    },
+    {
+      "epoch": 0.398446213607169,
+      "grad_norm": 1.3812161684036255,
+      "learning_rate": 4.785498885667395e-05,
+      "loss": 1.212,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3999846159763086,
+      "grad_norm": 1.555119276046753,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 1.2743,
+      "step": 1300
+    },
+    {
+      "epoch": 0.40152301834544823,
+      "grad_norm": 0.8972929120063782,
+      "learning_rate": 4.782222474649279e-05,
+      "loss": 1.1012,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4030614207145879,
+      "grad_norm": 1.5616463422775269,
+      "learning_rate": 4.780575381555472e-05,
+      "loss": 1.0773,
+      "step": 1310
+    },
+    {
+      "epoch": 0.40459982308372755,
+      "grad_norm": 1.2159663438796997,
+      "learning_rate": 4.778922369100204e-05,
+      "loss": 1.3538,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4061382254528672,
+      "grad_norm": 1.9486379623413086,
+      "learning_rate": 4.7772634415739624e-05,
+      "loss": 1.1075,
+      "step": 1320
+    },
+    {
+      "epoch": 0.40767662782200687,
+      "grad_norm": 0.9763344526290894,
+      "learning_rate": 4.7755986032825864e-05,
+      "loss": 1.2541,
+      "step": 1325
+    },
+    {
+      "epoch": 0.40921503019114647,
+      "grad_norm": 1.8571903705596924,
+      "learning_rate": 4.7739278585472573e-05,
+      "loss": 1.1718,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4107534325602861,
+      "grad_norm": 1.4408457279205322,
+      "learning_rate": 4.7722512117044865e-05,
+      "loss": 1.3156,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4122918349294258,
+      "grad_norm": 0.9971082806587219,
+      "learning_rate": 4.7705686671061054e-05,
+      "loss": 1.2319,
+      "step": 1340
+    },
+    {
+      "epoch": 0.41383023729856544,
+      "grad_norm": 1.3525968790054321,
+      "learning_rate": 4.768880229119253e-05,
+      "loss": 1.1831,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4153686396677051,
+      "grad_norm": 1.2504761219024658,
+      "learning_rate": 4.767185902126364e-05,
+      "loss": 1.1658,
+      "step": 1350
+    },
+    {
+      "epoch": 0.41690704203684475,
+      "grad_norm": 1.413370132446289,
+      "learning_rate": 4.76548569052516e-05,
+      "loss": 1.1391,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4184454444059844,
+      "grad_norm": 1.1849104166030884,
+      "learning_rate": 4.763779598728636e-05,
+      "loss": 1.1059,
+      "step": 1360
+    },
+    {
+      "epoch": 0.419983846775124,
+      "grad_norm": 1.1771825551986694,
+      "learning_rate": 4.762067631165049e-05,
+      "loss": 1.1448,
+      "step": 1365
+    },
+    {
+      "epoch": 0.42152224914426367,
+      "grad_norm": 1.4707238674163818,
+      "learning_rate": 4.760349792277906e-05,
+      "loss": 1.1634,
+      "step": 1370
+    },
+    {
+      "epoch": 0.42306065151340333,
+      "grad_norm": 1.577568769454956,
+      "learning_rate": 4.758626086525956e-05,
+      "loss": 1.1911,
+      "step": 1375
+    },
+    {
+      "epoch": 0.424599053882543,
+      "grad_norm": NaN,
+      "learning_rate": 4.757242900787734e-05,
+      "loss": 1.1134,
+      "step": 1380
+    },
+    {
+      "epoch": 0.42613745625168264,
+      "grad_norm": 1.0331732034683228,
+      "learning_rate": 4.755508645963771e-05,
+      "loss": 1.2352,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4276758586208223,
+      "grad_norm": 1.8902751207351685,
+      "learning_rate": 4.7537685368404684e-05,
+      "loss": 1.1271,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4292142609899619,
+      "grad_norm": 1.1057363748550415,
+      "learning_rate": 4.7520225779343754e-05,
+      "loss": 1.1633,
+      "step": 1395
+    },
+    {
+      "epoch": 0.43075266335910156,
+      "grad_norm": 1.183310627937317,
+      "learning_rate": 4.7502707737772264e-05,
+      "loss": 1.0869,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4322910657282412,
+      "grad_norm": 1.2959816455841064,
+      "learning_rate": 4.7485131289159276e-05,
+      "loss": 1.3126,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4338294680973809,
+      "grad_norm": 1.5542187690734863,
+      "learning_rate": 4.746749647912546e-05,
+      "loss": 1.2092,
+      "step": 1410
+    },
+    {
+      "epoch": 0.43536787046652053,
+      "grad_norm": 0.8856958746910095,
+      "learning_rate": 4.7449803353442945e-05,
+      "loss": 1.2647,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4369062728356602,
+      "grad_norm": 1.3472665548324585,
+      "learning_rate": 4.743205195803523e-05,
+      "loss": 1.2187,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4384446752047998,
+      "grad_norm": 1.7011195421218872,
+      "learning_rate": 4.741424233897707e-05,
+      "loss": 1.2245,
+      "step": 1425
+    },
+    {
+      "epoch": 0.43998307757393945,
+      "grad_norm": 2.0465426445007324,
+      "learning_rate": 4.7396374542494314e-05,
+      "loss": 1.2666,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4415214799430791,
+      "grad_norm": 1.2540628910064697,
+      "learning_rate": 4.737844861496385e-05,
+      "loss": 1.0815,
+      "step": 1435
+    },
+    {
+      "epoch": 0.44305988231221877,
+      "grad_norm": 1.3363851308822632,
+      "learning_rate": 4.736046460291342e-05,
+      "loss": 1.2059,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4445982846813584,
+      "grad_norm": 1.0009596347808838,
+      "learning_rate": 4.734242255302154e-05,
+      "loss": 1.3611,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4461366870504981,
+      "grad_norm": 0.9863151907920837,
+      "learning_rate": 4.732432251211735e-05,
+      "loss": 1.1654,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4476750894196377,
+      "grad_norm": 0.8969554901123047,
+      "learning_rate": 4.7306164527180546e-05,
+      "loss": 1.1962,
+      "step": 1455
+    },
+    {
+      "epoch": 0.44921349178877734,
+      "grad_norm": 1.1396467685699463,
+      "learning_rate": 4.7287948645341185e-05,
+      "loss": 1.2541,
+      "step": 1460
+    },
+    {
+      "epoch": 0.450751894157917,
+      "grad_norm": 1.4073244333267212,
+      "learning_rate": 4.72696749138796e-05,
+      "loss": 1.1695,
+      "step": 1465
+    },
+    {
+      "epoch": 0.45229029652705666,
+      "grad_norm": 2.4806954860687256,
+      "learning_rate": 4.725134338022631e-05,
+      "loss": 1.2109,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4538286988961963,
+      "grad_norm": 1.7406140565872192,
+      "learning_rate": 4.723295409196183e-05,
+      "loss": 1.1345,
+      "step": 1475
+    },
+    {
+      "epoch": 0.45536710126533597,
+      "grad_norm": 0.8337318897247314,
+      "learning_rate": 4.721450709681658e-05,
+      "loss": 1.1668,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4569055036344756,
+      "grad_norm": 2.3858370780944824,
+      "learning_rate": 4.7196002442670794e-05,
+      "loss": 1.2392,
+      "step": 1485
+    },
+    {
+      "epoch": 0.45844390600361523,
+      "grad_norm": 1.21785306930542,
+      "learning_rate": 4.7177440177554324e-05,
+      "loss": 1.2615,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4599823083727549,
+      "grad_norm": 1.8882992267608643,
+      "learning_rate": 4.715882034964657e-05,
+      "loss": 1.3384,
+      "step": 1495
+    },
+    {
+      "epoch": 0.46152071074189455,
+      "grad_norm": 1.1433368921279907,
+      "learning_rate": 4.714014300727634e-05,
+      "loss": 1.2788,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4630591131110342,
+      "grad_norm": 1.1393554210662842,
+      "learning_rate": 4.7121408198921745e-05,
+      "loss": 1.1502,
+      "step": 1505
+    },
+    {
+      "epoch": 0.46459751548017386,
+      "grad_norm": 1.1349945068359375,
+      "learning_rate": 4.7102615973210004e-05,
+      "loss": 1.1128,
+      "step": 1510
+    },
+    {
+      "epoch": 0.46613591784931346,
+      "grad_norm": 1.2287825345993042,
+      "learning_rate": 4.708376637891742e-05,
+      "loss": 1.1704,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4676743202184531,
+      "grad_norm": 1.257469892501831,
+      "learning_rate": 4.706485946496916e-05,
+      "loss": 1.1092,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4692127225875928,
+      "grad_norm": 1.4764573574066162,
+      "learning_rate": 4.704589528043918e-05,
+      "loss": 1.1372,
+      "step": 1525
+    },
+    {
+      "epoch": 0.47075112495673244,
+      "grad_norm": 1.0339967012405396,
+      "learning_rate": 4.702687387455008e-05,
+      "loss": 1.0919,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4722895273258721,
+      "grad_norm": 1.5123881101608276,
+      "learning_rate": 4.7007795296673006e-05,
+      "loss": 1.2167,
+      "step": 1535
+    },
+    {
+      "epoch": 0.47382792969501175,
+      "grad_norm": 1.475289225578308,
+      "learning_rate": 4.6988659596327465e-05,
+      "loss": 1.2096,
+      "step": 1540
+    },
+    {
+      "epoch": 0.47536633206415135,
+      "grad_norm": 1.229851245880127,
+      "learning_rate": 4.696946682318124e-05,
+      "loss": 1.1183,
+      "step": 1545
+    },
+    {
+      "epoch": 0.476904734433291,
+      "grad_norm": 0.7345393896102905,
+      "learning_rate": 4.695021702705026e-05,
+      "loss": 1.2528,
+      "step": 1550
+    },
+    {
+      "epoch": 0.47844313680243067,
+      "grad_norm": 1.239741563796997,
+      "learning_rate": 4.693091025789845e-05,
+      "loss": 1.1355,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4799815391715703,
+      "grad_norm": 1.3871427774429321,
+      "learning_rate": 4.6911546565837605e-05,
+      "loss": 1.1942,
+      "step": 1560
+    },
+    {
+      "epoch": 0.48151994154071,
+      "grad_norm": 1.9207137823104858,
+      "learning_rate": 4.689212600112728e-05,
+      "loss": 1.1756,
+      "step": 1565
+    },
+    {
+      "epoch": 0.48305834390984964,
+      "grad_norm": 1.4990715980529785,
+      "learning_rate": 4.687264861417464e-05,
+      "loss": 1.1363,
+      "step": 1570
+    },
+    {
+      "epoch": 0.48459674627898924,
+      "grad_norm": 1.9793407917022705,
+      "learning_rate": 4.6853114455534345e-05,
+      "loss": 1.1298,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4861351486481289,
+      "grad_norm": 1.1142261028289795,
+      "learning_rate": 4.683352357590839e-05,
+      "loss": 1.0793,
+      "step": 1580
+    },
+    {
+      "epoch": 0.48767355101726856,
+      "grad_norm": 1.4216675758361816,
+      "learning_rate": 4.6813876026146007e-05,
+      "loss": 1.1616,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4892119533864082,
+      "grad_norm": 1.3200958967208862,
+      "learning_rate": 4.679417185724352e-05,
+      "loss": 1.2206,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4907503557555479,
+      "grad_norm": 2.0705955028533936,
+      "learning_rate": 4.67744111203442e-05,
+      "loss": 1.1871,
+      "step": 1595
+    },
+    {
+      "epoch": 0.49228875812468753,
+      "grad_norm": 1.3912644386291504,
+      "learning_rate": 4.675459386673815e-05,
+      "loss": 1.2779,
+      "step": 1600
+    },
+    {
+      "epoch": 0.49382716049382713,
+      "grad_norm": 1.1627599000930786,
+      "learning_rate": 4.673472014786216e-05,
+      "loss": 1.0949,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4953655628629668,
+      "grad_norm": 1.0313643217086792,
+      "learning_rate": 4.671479001529958e-05,
+      "loss": 1.2541,
+      "step": 1610
+    },
+    {
+      "epoch": 0.49690396523210645,
+      "grad_norm": 0.8410005569458008,
+      "learning_rate": 4.6694803520780204e-05,
+      "loss": 1.1915,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4984423676012461,
+      "grad_norm": 1.9390909671783447,
+      "learning_rate": 4.66747607161801e-05,
+      "loss": 1.1954,
+      "step": 1620
+    },
+    {
+      "epoch": 0.49998076997038576,
+      "grad_norm": 1.324788212776184,
+      "learning_rate": 4.665466165352147e-05,
+      "loss": 1.163,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5015191723395254,
+      "grad_norm": 1.9445561170578003,
+      "learning_rate": 4.663450638497259e-05,
+      "loss": 1.2891,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5030575747086651,
+      "grad_norm": 1.2497957944869995,
+      "learning_rate": 4.661429496284757e-05,
+      "loss": 1.2295,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5045959770778047,
+      "grad_norm": 1.6469943523406982,
+      "learning_rate": 4.65940274396063e-05,
+      "loss": 1.0742,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5061343794469444,
+      "grad_norm": 1.3002201318740845,
+      "learning_rate": 4.657370386785427e-05,
+      "loss": 1.1575,
+      "step": 1645
+    },
+    {
+      "epoch": 0.507672781816084,
+      "grad_norm": 1.3192851543426514,
+      "learning_rate": 4.6553324300342446e-05,
+      "loss": 1.2364,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5092111841852236,
+      "grad_norm": 2.825892925262451,
+      "learning_rate": 4.653288878996716e-05,
+      "loss": 1.3216,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5107495865543633,
+      "grad_norm": 1.0338250398635864,
+      "learning_rate": 4.651239738976991e-05,
+      "loss": 1.1948,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5122879889235029,
+      "grad_norm": 1.3987075090408325,
+      "learning_rate": 4.649185015293728e-05,
+      "loss": 1.1198,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5138263912926426,
+      "grad_norm": 1.920446515083313,
+      "learning_rate": 4.647124713280078e-05,
+      "loss": 1.1055,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5153647936617822,
+      "grad_norm": 2.621272325515747,
+      "learning_rate": 4.645058838283669e-05,
+      "loss": 1.1013,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5169031960309219,
+      "grad_norm": 1.3560134172439575,
+      "learning_rate": 4.642987395666598e-05,
+      "loss": 1.2337,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5184415984000615,
+      "grad_norm": 1.2603050470352173,
+      "learning_rate": 4.64091039080541e-05,
+      "loss": 1.1338,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5199800007692011,
+      "grad_norm": 1.3172502517700195,
+      "learning_rate": 4.638827829091086e-05,
+      "loss": 1.2974,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5215184031383409,
+      "grad_norm": 1.0840466022491455,
+      "learning_rate": 4.636739715929033e-05,
+      "loss": 1.0235,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5230568055074805,
+      "grad_norm": 0.9899886250495911,
+      "learning_rate": 4.634646056739066e-05,
+      "loss": 1.3254,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5245952078766202,
+      "grad_norm": 1.1562014818191528,
+      "learning_rate": 4.632546856955396e-05,
+      "loss": 1.2105,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5261336102457598,
+      "grad_norm": 0.9220596551895142,
+      "learning_rate": 4.630442122026613e-05,
+      "loss": 1.2339,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5276720126148994,
+      "grad_norm": 2.6827752590179443,
+      "learning_rate": 4.628331857415673e-05,
+      "loss": 1.1655,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5292104149840391,
+      "grad_norm": 1.4331889152526855,
+      "learning_rate": 4.626216068599889e-05,
+      "loss": 1.1398,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5307488173531787,
+      "grad_norm": 1.2920069694519043,
+      "learning_rate": 4.624094761070909e-05,
+      "loss": 1.155,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5322872197223184,
+      "grad_norm": 0.9748442769050598,
+      "learning_rate": 4.621967940334705e-05,
+      "loss": 1.162,
+      "step": 1730
+    },
+    {
+      "epoch": 0.533825622091458,
+      "grad_norm": 1.2728887796401978,
+      "learning_rate": 4.6198356119115595e-05,
+      "loss": 1.2505,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5353640244605977,
+      "grad_norm": 1.119135856628418,
+      "learning_rate": 4.617697781336052e-05,
+      "loss": 1.2112,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5369024268297373,
+      "grad_norm": 1.0900635719299316,
+      "learning_rate": 4.6155544541570406e-05,
+      "loss": 1.1836,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5384408291988769,
+      "grad_norm": 1.4258182048797607,
+      "learning_rate": 4.613405635937651e-05,
+      "loss": 1.1892,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5399792315680166,
+      "grad_norm": 1.1550543308258057,
+      "learning_rate": 4.611251332255264e-05,
+      "loss": 1.1996,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5415176339371562,
+      "grad_norm": 1.2096160650253296,
+      "learning_rate": 4.609091548701493e-05,
+      "loss": 1.1016,
+      "step": 1760
+    },
+    {
+      "epoch": 0.543056036306296,
+      "grad_norm": 0.9595054984092712,
+      "learning_rate": 4.6069262908821784e-05,
+      "loss": 1.233,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5445944386754356,
+      "grad_norm": 1.478603482246399,
+      "learning_rate": 4.604755564417369e-05,
+      "loss": 1.2789,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5461328410445752,
+      "grad_norm": 1.2694469690322876,
+      "learning_rate": 4.602579374941307e-05,
+      "loss": 1.3355,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5476712434137149,
+      "grad_norm": 1.3089395761489868,
+      "learning_rate": 4.600397728102414e-05,
+      "loss": 1.0967,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5492096457828545,
+      "grad_norm": 2.684046506881714,
+      "learning_rate": 4.5982106295632765e-05,
+      "loss": 1.0866,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5507480481519942,
+      "grad_norm": 1.414984107017517,
+      "learning_rate": 4.596018085000633e-05,
+      "loss": 1.2533,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5522864505211338,
+      "grad_norm": 1.0765364170074463,
+      "learning_rate": 4.593820100105355e-05,
+      "loss": 1.2313,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5538248528902735,
+      "grad_norm": 0.9591373801231384,
+      "learning_rate": 4.5916166805824353e-05,
+      "loss": 1.1038,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5553632552594131,
+      "grad_norm": 1.9848989248275757,
+      "learning_rate": 4.589407832150974e-05,
+      "loss": 1.1897,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5569016576285527,
+      "grad_norm": 1.1997350454330444,
+      "learning_rate": 4.5871935605441606e-05,
+      "loss": 1.1637,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5584400599976924,
+      "grad_norm": 0.9119892716407776,
+      "learning_rate": 4.5849738715092624e-05,
+      "loss": 1.3695,
+      "step": 1815
+    },
+    {
+      "epoch": 0.559978462366832,
+      "grad_norm": 1.0208933353424072,
+      "learning_rate": 4.582748770807605e-05,
+      "loss": 1.1085,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5615168647359717,
+      "grad_norm": 1.3944216966629028,
+      "learning_rate": 4.580518264214564e-05,
+      "loss": 1.2322,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5630552671051113,
+      "grad_norm": 0.9918113946914673,
+      "learning_rate": 4.5782823575195444e-05,
+      "loss": 1.0482,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5645936694742509,
+      "grad_norm": 0.9428990483283997,
+      "learning_rate": 4.576041056525966e-05,
+      "loss": 1.132,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5661320718433906,
+      "grad_norm": 1.1514095067977905,
+      "learning_rate": 4.5737943670512534e-05,
+      "loss": 1.2385,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5676704742125303,
+      "grad_norm": 2.3423311710357666,
+      "learning_rate": 4.5715422949268136e-05,
+      "loss": 1.2718,
+      "step": 1845
+    },
+    {
+      "epoch": 0.56920887658167,
+      "grad_norm": 1.2614145278930664,
+      "learning_rate": 4.5692848459980275e-05,
+      "loss": 1.1639,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5707472789508096,
+      "grad_norm": 1.355564832687378,
+      "learning_rate": 4.56702202612423e-05,
+      "loss": 1.2753,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5722856813199493,
+      "grad_norm": 1.5028793811798096,
+      "learning_rate": 4.564753841178697e-05,
+      "loss": 1.2298,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5738240836890889,
+      "grad_norm": 2.331164598464966,
+      "learning_rate": 4.5624802970486295e-05,
+      "loss": 1.1079,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5753624860582285,
+      "grad_norm": 1.438877820968628,
+      "learning_rate": 4.56020139963514e-05,
+      "loss": 1.1423,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5769008884273682,
+      "grad_norm": 1.6183146238327026,
+      "learning_rate": 4.557917154853234e-05,
+      "loss": 1.3155,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5784392907965078,
+      "grad_norm": 1.206664800643921,
+      "learning_rate": 4.555627568631798e-05,
+      "loss": 1.1703,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5799776931656475,
+      "grad_norm": 0.9220191240310669,
+      "learning_rate": 4.553332646913581e-05,
+      "loss": 1.2033,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5815160955347871,
+      "grad_norm": 1.4525690078735352,
+      "learning_rate": 4.551032395655181e-05,
+      "loss": 1.128,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5830544979039268,
+      "grad_norm": 0.9040252566337585,
+      "learning_rate": 4.5487268208270284e-05,
+      "loss": 1.1954,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5845929002730664,
+      "grad_norm": 1.0403451919555664,
+      "learning_rate": 4.546415928413373e-05,
+      "loss": 1.1627,
+      "step": 1900
+    },
+    {
+      "epoch": 0.586131302642206,
+      "grad_norm": 0.8466317057609558,
+      "learning_rate": 4.544099724412267e-05,
+      "loss": 1.2379,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5876697050113457,
+      "grad_norm": 1.0065761804580688,
+      "learning_rate": 4.5417782148355464e-05,
+      "loss": 1.2146,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5892081073804853,
+      "grad_norm": 1.1972275972366333,
+      "learning_rate": 4.5394514057088197e-05,
+      "loss": 1.1367,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5907465097496251,
+      "grad_norm": 1.0356810092926025,
+      "learning_rate": 4.5371193030714524e-05,
+      "loss": 1.123,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5922849121187647,
+      "grad_norm": 2.044755458831787,
+      "learning_rate": 4.534781912976546e-05,
+      "loss": 1.1501,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5938233144879043,
+      "grad_norm": 4.12185525894165,
+      "learning_rate": 4.532439241490928e-05,
+      "loss": 1.1789,
+      "step": 1930
+    },
+    {
+      "epoch": 0.595361716857044,
+      "grad_norm": 1.5785492658615112,
+      "learning_rate": 4.530091294695134e-05,
+      "loss": 1.2626,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5969001192261836,
+      "grad_norm": 1.0973812341690063,
+      "learning_rate": 4.527738078683391e-05,
+      "loss": 1.1505,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5984385215953233,
+      "grad_norm": 0.8266382813453674,
+      "learning_rate": 4.525379599563606e-05,
+      "loss": 1.2476,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5999769239644629,
+      "grad_norm": 1.5638591051101685,
+      "learning_rate": 4.5230158634573406e-05,
+      "loss": 1.0609,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6015153263336026,
+      "grad_norm": 1.0881614685058594,
+      "learning_rate": 4.5206468764998065e-05,
+      "loss": 1.1339,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6030537287027422,
+      "grad_norm": 1.1539703607559204,
+      "learning_rate": 4.518272644839843e-05,
+      "loss": 1.0505,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6045921310718818,
+      "grad_norm": 1.331925630569458,
+      "learning_rate": 4.515893174639899e-05,
+      "loss": 1.1124,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6061305334410215,
+      "grad_norm": 1.0339001417160034,
+      "learning_rate": 4.5135084720760254e-05,
+      "loss": 1.1829,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6076689358101611,
+      "grad_norm": 0.8567355275154114,
+      "learning_rate": 4.5111185433378514e-05,
+      "loss": 1.1963,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6092073381793008,
+      "grad_norm": 1.3100231885910034,
+      "learning_rate": 4.50872339462857e-05,
+      "loss": 1.2722,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6107457405484404,
+      "grad_norm": 1.4597526788711548,
+      "learning_rate": 4.506323032164925e-05,
+      "loss": 1.3127,
+      "step": 1985
+    },
+    {
+      "epoch": 0.61228414291758,
+      "grad_norm": 1.6040161848068237,
+      "learning_rate": 4.503917462177192e-05,
+      "loss": 1.2423,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6138225452867198,
+      "grad_norm": 1.1720243692398071,
+      "learning_rate": 4.5015066909091625e-05,
+      "loss": 1.1971,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6153609476558594,
+      "grad_norm": 0.9491767883300781,
+      "learning_rate": 4.499090724618129e-05,
+      "loss": 1.1892,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6168993500249991,
+      "grad_norm": 1.83636474609375,
+      "learning_rate": 4.4966695695748686e-05,
+      "loss": 1.0706,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6184377523941387,
+      "grad_norm": 1.8804785013198853,
+      "learning_rate": 4.494243232063623e-05,
+      "loss": 1.2081,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6199761547632784,
+      "grad_norm": 1.1638429164886475,
+      "learning_rate": 4.4918117183820894e-05,
+      "loss": 1.1153,
+      "step": 2015
+    },
+    {
+      "epoch": 0.621514557132418,
+      "grad_norm": 1.120448112487793,
+      "learning_rate": 4.489375034841397e-05,
+      "loss": 1.26,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6230529595015576,
+      "grad_norm": 1.4325777292251587,
+      "learning_rate": 4.486933187766095e-05,
+      "loss": 1.1023,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6245913618706973,
+      "grad_norm": 1.2294832468032837,
+      "learning_rate": 4.484486183494136e-05,
+      "loss": 1.1005,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6261297642398369,
+      "grad_norm": 1.6747307777404785,
+      "learning_rate": 4.4820340283768544e-05,
+      "loss": 1.2555,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6276681666089766,
+      "grad_norm": 1.192826509475708,
+      "learning_rate": 4.479576728778958e-05,
+      "loss": 1.0987,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6292065689781162,
+      "grad_norm": 1.767940878868103,
+      "learning_rate": 4.477114291078506e-05,
+      "loss": 1.0992,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6307449713472558,
+      "grad_norm": 1.5168949365615845,
+      "learning_rate": 4.474646721666893e-05,
+      "loss": 1.1127,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6322833737163955,
+      "grad_norm": 2.4314072132110596,
+      "learning_rate": 4.4721740269488355e-05,
+      "loss": 1.141,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6338217760855351,
+      "grad_norm": 1.7688580751419067,
+      "learning_rate": 4.46969621334235e-05,
+      "loss": 1.1934,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6353601784546749,
+      "grad_norm": 1.0726224184036255,
+      "learning_rate": 4.467213287278741e-05,
+      "loss": 1.129,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6368985808238145,
+      "grad_norm": 1.0851497650146484,
+      "learning_rate": 4.464725255202582e-05,
+      "loss": 1.1553,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6384369831929542,
+      "grad_norm": 1.4708757400512695,
+      "learning_rate": 4.462232123571702e-05,
+      "loss": 1.1845,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6399753855620938,
+      "grad_norm": 1.3815021514892578,
+      "learning_rate": 4.459733898857162e-05,
+      "loss": 1.164,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6415137879312334,
+      "grad_norm": 0.9713008999824524,
+      "learning_rate": 4.4572305875432465e-05,
+      "loss": 1.2635,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6430521903003731,
+      "grad_norm": 1.7279797792434692,
+      "learning_rate": 4.4547221961274385e-05,
+      "loss": 1.2015,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6445905926695127,
+      "grad_norm": 3.183756113052368,
+      "learning_rate": 4.4522087311204096e-05,
+      "loss": 1.0928,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6461289950386524,
+      "grad_norm": 2.138756513595581,
+      "learning_rate": 4.449690199046e-05,
+      "loss": 1.2496,
+      "step": 2100
+    },
+    {
+      "epoch": 0.647667397407792,
+      "grad_norm": 1.7964316606521606,
+      "learning_rate": 4.447166606441201e-05,
+      "loss": 1.123,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6492057997769316,
+      "grad_norm": 2.1838841438293457,
+      "learning_rate": 4.444637959856137e-05,
+      "loss": 1.0909,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6507442021460713,
+      "grad_norm": 1.1749359369277954,
+      "learning_rate": 4.442104265854055e-05,
+      "loss": 1.275,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6522826045152109,
+      "grad_norm": 1.2564892768859863,
+      "learning_rate": 4.439565531011299e-05,
+      "loss": 1.1814,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6538210068843506,
+      "grad_norm": 1.1078180074691772,
+      "learning_rate": 4.4370217619172964e-05,
+      "loss": 1.1717,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6553594092534902,
+      "grad_norm": 1.0620262622833252,
+      "learning_rate": 4.434472965174545e-05,
+      "loss": 1.2006,
+      "step": 2130
+    },
+    {
+      "epoch": 0.65689781162263,
+      "grad_norm": 1.668600082397461,
+      "learning_rate": 4.4319191473985884e-05,
+      "loss": 1.16,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6584362139917695,
+      "grad_norm": 0.7764678597450256,
+      "learning_rate": 4.429360315218005e-05,
+      "loss": 1.3311,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6599746163609091,
+      "grad_norm": 0.9446391463279724,
+      "learning_rate": 4.4267964752743854e-05,
+      "loss": 1.2131,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6615130187300489,
+      "grad_norm": 0.8063297271728516,
+      "learning_rate": 4.4242276342223235e-05,
+      "loss": 1.15,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6630514210991885,
+      "grad_norm": 1.4049417972564697,
+      "learning_rate": 4.421653798729387e-05,
+      "loss": 1.2054,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6645898234683282,
+      "grad_norm": 1.716774344444275,
+      "learning_rate": 4.4190749754761126e-05,
+      "loss": 1.055,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6661282258374678,
+      "grad_norm": 2.2953073978424072,
+      "learning_rate": 4.4164911711559803e-05,
+      "loss": 1.2199,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6676666282066075,
+      "grad_norm": 0.9870630502700806,
+      "learning_rate": 4.4139023924753995e-05,
+      "loss": 1.1933,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6692050305757471,
+      "grad_norm": 1.69342839717865,
+      "learning_rate": 4.41130864615369e-05,
+      "loss": 1.1602,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6707434329448867,
+      "grad_norm": 1.0127511024475098,
+      "learning_rate": 4.408709938923067e-05,
+      "loss": 1.2509,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6722818353140264,
+      "grad_norm": 1.2484458684921265,
+      "learning_rate": 4.40610627752862e-05,
+      "loss": 1.1144,
+      "step": 2185
+    },
+    {
+      "epoch": 0.673820237683166,
+      "grad_norm": 1.0753147602081299,
+      "learning_rate": 4.403497668728299e-05,
+      "loss": 1.2816,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6753586400523057,
+      "grad_norm": 1.2890567779541016,
+      "learning_rate": 4.400884119292894e-05,
+      "loss": 1.1233,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6768970424214453,
+      "grad_norm": 1.108391284942627,
+      "learning_rate": 4.39826563600602e-05,
+      "loss": 1.198,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6784354447905849,
+      "grad_norm": 0.9689030051231384,
+      "learning_rate": 4.395642225664097e-05,
+      "loss": 1.2354,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6799738471597246,
+      "grad_norm": 1.4813648462295532,
+      "learning_rate": 4.393013895076335e-05,
+      "loss": 1.1986,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6815122495288642,
+      "grad_norm": 1.4557727575302124,
+      "learning_rate": 4.3903806510647115e-05,
+      "loss": 1.2441,
+      "step": 2215
+    },
+    {
+      "epoch": 0.683050651898004,
+      "grad_norm": 5.211197853088379,
+      "learning_rate": 4.3877425004639616e-05,
+      "loss": 1.2142,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6845890542671436,
+      "grad_norm": 1.0981544256210327,
+      "learning_rate": 4.385099450121551e-05,
+      "loss": 1.1192,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6861274566362833,
+      "grad_norm": 1.4392608404159546,
+      "learning_rate": 4.3824515068976666e-05,
+      "loss": 1.0806,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6876658590054229,
+      "grad_norm": 0.9992257952690125,
+      "learning_rate": 4.3797986776651934e-05,
+      "loss": 1.2906,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6892042613745625,
+      "grad_norm": 1.0846220254898071,
+      "learning_rate": 4.3771409693096985e-05,
+      "loss": 1.145,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6907426637437022,
+      "grad_norm": 1.7244092226028442,
+      "learning_rate": 4.374478388729414e-05,
+      "loss": 1.2479,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6922810661128418,
+      "grad_norm": 1.187263011932373,
+      "learning_rate": 4.371810942835215e-05,
+      "loss": 1.2843,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6938194684819815,
+      "grad_norm": 1.1405870914459229,
+      "learning_rate": 4.369138638550611e-05,
+      "loss": 1.2483,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6953578708511211,
+      "grad_norm": 1.399223804473877,
+      "learning_rate": 4.3664614828117137e-05,
+      "loss": 1.1166,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6968962732202607,
+      "grad_norm": 0.8091042637825012,
+      "learning_rate": 4.363779482567234e-05,
+      "loss": 1.2809,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6984346755894004,
+      "grad_norm": 1.0107675790786743,
+      "learning_rate": 4.3610926447784534e-05,
+      "loss": 1.1957,
+      "step": 2270
+    },
+    {
+      "epoch": 0.69997307795854,
+      "grad_norm": 1.485467791557312,
+      "learning_rate": 4.3584009764192094e-05,
+      "loss": 1.1612,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7015114803276797,
+      "grad_norm": 1.0193709135055542,
+      "learning_rate": 4.3557044844758796e-05,
+      "loss": 1.1595,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7030498826968193,
+      "grad_norm": 1.1308696269989014,
+      "learning_rate": 4.353003175947359e-05,
+      "loss": 1.1202,
+      "step": 2285
+    },
+    {
+      "epoch": 0.704588285065959,
+      "grad_norm": 1.233419418334961,
+      "learning_rate": 4.3502970578450466e-05,
+      "loss": 1.1749,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7061266874350987,
+      "grad_norm": 0.9433870315551758,
+      "learning_rate": 4.3475861371928225e-05,
+      "loss": 1.1464,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7076650898042383,
+      "grad_norm": 1.7299931049346924,
+      "learning_rate": 4.344870421027036e-05,
+      "loss": 1.1479,
+      "step": 2300
+    },
+    {
+      "epoch": 0.709203492173378,
+      "grad_norm": 1.1724836826324463,
+      "learning_rate": 4.3421499163964784e-05,
+      "loss": 1.1548,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7107418945425176,
+      "grad_norm": 1.5910905599594116,
+      "learning_rate": 4.339424630362373e-05,
+      "loss": 1.1267,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7122802969116573,
+      "grad_norm": 1.4339747428894043,
+      "learning_rate": 4.336694569998354e-05,
+      "loss": 1.2297,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7138186992807969,
+      "grad_norm": 0.7568414211273193,
+      "learning_rate": 4.333959742390444e-05,
+      "loss": 1.228,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7153571016499365,
+      "grad_norm": 1.1691060066223145,
+      "learning_rate": 4.331220154637044e-05,
+      "loss": 1.0902,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7168955040190762,
+      "grad_norm": 1.296523094177246,
+      "learning_rate": 4.328475813848906e-05,
+      "loss": 1.2011,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7184339063882158,
+      "grad_norm": 1.0982245206832886,
+      "learning_rate": 4.325726727149122e-05,
+      "loss": 1.1082,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7199723087573555,
+      "grad_norm": 1.2875556945800781,
+      "learning_rate": 4.3229729016731005e-05,
+      "loss": 1.2202,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7215107111264951,
+      "grad_norm": 2.59907865524292,
+      "learning_rate": 4.320214344568549e-05,
+      "loss": 1.1947,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7230491134956348,
+      "grad_norm": 1.3555256128311157,
+      "learning_rate": 4.317451062995458e-05,
+      "loss": 1.1544,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7245875158647744,
+      "grad_norm": 2.351750135421753,
+      "learning_rate": 4.3146830641260815e-05,
+      "loss": 1.1858,
+      "step": 2355
+    },
+    {
+      "epoch": 0.726125918233914,
+      "grad_norm": 1.5076963901519775,
+      "learning_rate": 4.311910355144914e-05,
+      "loss": 1.2452,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7276643206030537,
+      "grad_norm": 1.020558476448059,
+      "learning_rate": 4.309132943248678e-05,
+      "loss": 1.1731,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7292027229721934,
+      "grad_norm": 1.3791615962982178,
+      "learning_rate": 4.306350835646303e-05,
+      "loss": 1.1717,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7307411253413331,
+      "grad_norm": 1.8552170991897583,
+      "learning_rate": 4.303564039558904e-05,
+      "loss": 1.1315,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7322795277104727,
+      "grad_norm": 1.5803345441818237,
+      "learning_rate": 4.3007725622197674e-05,
+      "loss": 1.1341,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7338179300796123,
+      "grad_norm": 1.0923699140548706,
+      "learning_rate": 4.2979764108743296e-05,
+      "loss": 1.0577,
+      "step": 2385
+    },
+    {
+      "epoch": 0.735356332448752,
+      "grad_norm": 0.9132150411605835,
+      "learning_rate": 4.295175592780158e-05,
+      "loss": 1.1672,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7368947348178916,
+      "grad_norm": 1.6605010032653809,
+      "learning_rate": 4.2923701152069336e-05,
+      "loss": 1.1087,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7384331371870313,
+      "grad_norm": 1.2207224369049072,
+      "learning_rate": 4.28955998543643e-05,
+      "loss": 1.1708,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7399715395561709,
+      "grad_norm": 2.0855581760406494,
+      "learning_rate": 4.286745210762499e-05,
+      "loss": 1.1594,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7415099419253106,
+      "grad_norm": 1.4070476293563843,
+      "learning_rate": 4.283925798491044e-05,
+      "loss": 1.2085,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7430483442944502,
+      "grad_norm": 1.0494474172592163,
+      "learning_rate": 4.281101755940009e-05,
+      "loss": 1.1877,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7445867466635898,
+      "grad_norm": 1.0690234899520874,
+      "learning_rate": 4.2782730904393546e-05,
+      "loss": 1.0301,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7461251490327295,
+      "grad_norm": 1.0572775602340698,
+      "learning_rate": 4.275439809331041e-05,
+      "loss": 1.1589,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7476635514018691,
+      "grad_norm": 1.0381057262420654,
+      "learning_rate": 4.2726019199690093e-05,
+      "loss": 1.1945,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7492019537710088,
+      "grad_norm": 2.4231226444244385,
+      "learning_rate": 4.269759429719159e-05,
+      "loss": 1.1851,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7507403561401484,
+      "grad_norm": 1.1113394498825073,
+      "learning_rate": 4.266912345959335e-05,
+      "loss": 1.1269,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7522787585092882,
+      "grad_norm": 1.7000584602355957,
+      "learning_rate": 4.264060676079302e-05,
+      "loss": 1.1546,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7538171608784278,
+      "grad_norm": 4.711215972900391,
+      "learning_rate": 4.2612044274807295e-05,
+      "loss": 1.1901,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7553555632475674,
+      "grad_norm": 1.1425622701644897,
+      "learning_rate": 4.2583436075771706e-05,
+      "loss": 1.2222,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7568939656167071,
+      "grad_norm": 1.1163311004638672,
+      "learning_rate": 4.255478223794045e-05,
+      "loss": 1.066,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7584323679858467,
+      "grad_norm": 0.741192102432251,
+      "learning_rate": 4.252608283568616e-05,
+      "loss": 1.3143,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7599707703549864,
+      "grad_norm": 0.9434729218482971,
+      "learning_rate": 4.249733794349976e-05,
+      "loss": 1.1981,
+      "step": 2470
+    },
+    {
+      "epoch": 0.761509172724126,
+      "grad_norm": 1.200185775756836,
+      "learning_rate": 4.246854763599022e-05,
+      "loss": 1.1042,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7630475750932656,
+      "grad_norm": 1.1456506252288818,
+      "learning_rate": 4.2439711987884406e-05,
+      "loss": 1.1917,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7645859774624053,
+      "grad_norm": 1.4507484436035156,
+      "learning_rate": 4.241083107402687e-05,
+      "loss": 1.249,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7661243798315449,
+      "grad_norm": 1.5303988456726074,
+      "learning_rate": 4.238190496937962e-05,
+      "loss": 1.2334,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7676627822006846,
+      "grad_norm": 1.067226529121399,
+      "learning_rate": 4.235293374902201e-05,
+      "loss": 1.211,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7692011845698242,
+      "grad_norm": 1.0933020114898682,
+      "learning_rate": 4.232391748815046e-05,
+      "loss": 1.1696,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7707395869389639,
+      "grad_norm": 0.9041593670845032,
+      "learning_rate": 4.2294856262078296e-05,
+      "loss": 1.0995,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7722779893081035,
+      "grad_norm": 0.886325478553772,
+      "learning_rate": 4.226575014623557e-05,
+      "loss": 1.2787,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7738163916772431,
+      "grad_norm": 0.8392224311828613,
+      "learning_rate": 4.223659921616885e-05,
+      "loss": 1.1928,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7753547940463829,
+      "grad_norm": 2.0462377071380615,
+      "learning_rate": 4.2207403547541e-05,
+      "loss": 1.2198,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7768931964155225,
+      "grad_norm": 0.8376229405403137,
+      "learning_rate": 4.2178163216131015e-05,
+      "loss": 1.1459,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7784315987846622,
+      "grad_norm": 1.8360308408737183,
+      "learning_rate": 4.214887829783383e-05,
+      "loss": 1.2045,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7799700011538018,
+      "grad_norm": 1.962247610092163,
+      "learning_rate": 4.2119548868660084e-05,
+      "loss": 1.1892,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7815084035229414,
+      "grad_norm": 1.0114338397979736,
+      "learning_rate": 4.209017500473596e-05,
+      "loss": 1.2576,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7830468058920811,
+      "grad_norm": 2.040184259414673,
+      "learning_rate": 4.206075678230297e-05,
+      "loss": 1.0698,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7845852082612207,
+      "grad_norm": 1.3583984375,
+      "learning_rate": 4.203129427771776e-05,
+      "loss": 1.1858,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7861236106303604,
+      "grad_norm": 2.056542158126831,
+      "learning_rate": 4.200178756745192e-05,
+      "loss": 1.2202,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7876620129995,
+      "grad_norm": 1.1544386148452759,
+      "learning_rate": 4.197223672809177e-05,
+      "loss": 1.1182,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7892004153686397,
+      "grad_norm": 1.046268343925476,
+      "learning_rate": 4.194264183633818e-05,
+      "loss": 1.1568,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7907388177377793,
+      "grad_norm": 1.4613105058670044,
+      "learning_rate": 4.1913002969006344e-05,
+      "loss": 1.2269,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7922772201069189,
+      "grad_norm": 0.9770887494087219,
+      "learning_rate": 4.188332020302561e-05,
+      "loss": 1.0355,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7938156224760586,
+      "grad_norm": 1.2040941715240479,
+      "learning_rate": 4.185359361543927e-05,
+      "loss": 1.2607,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7953540248451982,
+      "grad_norm": 1.4072517156600952,
+      "learning_rate": 4.182382328340434e-05,
+      "loss": 1.2368,
+      "step": 2585
+    },
+    {
+      "epoch": 0.796892427214338,
+      "grad_norm": 1.2777091264724731,
+      "learning_rate": 4.17940092841914e-05,
+      "loss": 1.3141,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7984308295834776,
+      "grad_norm": 0.8566577434539795,
+      "learning_rate": 4.176415169518434e-05,
+      "loss": 1.1347,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7999692319526172,
+      "grad_norm": 1.2150745391845703,
+      "learning_rate": 4.173425059388023e-05,
+      "loss": 0.99,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8015076343217569,
+      "grad_norm": 1.1194686889648438,
+      "learning_rate": 4.1704306057889053e-05,
+      "loss": 1.1714,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8030460366908965,
+      "grad_norm": 1.402685523033142,
+      "learning_rate": 4.167431816493352e-05,
+      "loss": 1.2142,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8045844390600362,
+      "grad_norm": 1.0395668745040894,
+      "learning_rate": 4.1644286992848916e-05,
+      "loss": 1.1401,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8061228414291758,
+      "grad_norm": 0.9008033871650696,
+      "learning_rate": 4.161421261958281e-05,
+      "loss": 1.2895,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8076612437983155,
+      "grad_norm": 1.628246545791626,
+      "learning_rate": 4.158409512319493e-05,
+      "loss": 1.0513,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8091996461674551,
+      "grad_norm": 1.2516933679580688,
+      "learning_rate": 4.1553934581856945e-05,
+      "loss": 1.1017,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8107380485365947,
+      "grad_norm": 1.1752599477767944,
+      "learning_rate": 4.1523731073852215e-05,
+      "loss": 1.2321,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8122764509057344,
+      "grad_norm": 1.0307414531707764,
+      "learning_rate": 4.149348467757566e-05,
+      "loss": 1.3027,
+      "step": 2640
+    },
+    {
+      "epoch": 0.813814853274874,
+      "grad_norm": 1.7551286220550537,
+      "learning_rate": 4.1463195471533476e-05,
+      "loss": 1.0455,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8153532556440137,
+      "grad_norm": 1.2423598766326904,
+      "learning_rate": 4.1432863534343016e-05,
+      "loss": 1.0785,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8168916580131533,
+      "grad_norm": 2.572627305984497,
+      "learning_rate": 4.140248894473253e-05,
+      "loss": 1.2205,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8184300603822929,
+      "grad_norm": 1.2887595891952515,
+      "learning_rate": 4.137207178154095e-05,
+      "loss": 1.0928,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8199684627514326,
+      "grad_norm": 1.1292917728424072,
+      "learning_rate": 4.134161212371776e-05,
+      "loss": 1.0875,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8215068651205722,
+      "grad_norm": 1.0025163888931274,
+      "learning_rate": 4.1311110050322674e-05,
+      "loss": 1.3263,
+      "step": 2670
+    },
+    {
+      "epoch": 0.823045267489712,
+      "grad_norm": 1.1535359621047974,
+      "learning_rate": 4.128056564052558e-05,
+      "loss": 1.0747,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8245836698588516,
+      "grad_norm": 1.1695550680160522,
+      "learning_rate": 4.124997897360617e-05,
+      "loss": 1.1265,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8261220722279913,
+      "grad_norm": 0.9830222725868225,
+      "learning_rate": 4.1219350128953885e-05,
+      "loss": 1.1475,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8276604745971309,
+      "grad_norm": 1.1742221117019653,
+      "learning_rate": 4.118867918606759e-05,
+      "loss": 1.2344,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8291988769662705,
+      "grad_norm": 0.9631347060203552,
+      "learning_rate": 4.115796622455544e-05,
+      "loss": 1.1101,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8307372793354102,
+      "grad_norm": 0.9511464238166809,
+      "learning_rate": 4.112721132413467e-05,
+      "loss": 1.2955,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8322756817045498,
+      "grad_norm": 1.5212576389312744,
+      "learning_rate": 4.109641456463135e-05,
+      "loss": 1.1907,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8338140840736895,
+      "grad_norm": 1.1152446269989014,
+      "learning_rate": 4.106557602598017e-05,
+      "loss": 1.143,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8353524864428291,
+      "grad_norm": 1.1330682039260864,
+      "learning_rate": 4.103469578822432e-05,
+      "loss": 1.1143,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8368908888119688,
+      "grad_norm": 1.170376181602478,
+      "learning_rate": 4.1003773931515175e-05,
+      "loss": 1.1621,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8384292911811084,
+      "grad_norm": 1.571735143661499,
+      "learning_rate": 4.097281053611215e-05,
+      "loss": 1.1911,
+      "step": 2725
+    },
+    {
+      "epoch": 0.839967693550248,
+      "grad_norm": 0.9364776611328125,
+      "learning_rate": 4.0941805682382484e-05,
+      "loss": 1.171,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8415060959193877,
+      "grad_norm": 1.2696608304977417,
+      "learning_rate": 4.091075945080101e-05,
+      "loss": 1.2642,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8430444982885273,
+      "grad_norm": 4.6916584968566895,
+      "learning_rate": 4.087967192194997e-05,
+      "loss": 1.2212,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8445829006576671,
+      "grad_norm": 0.9691728353500366,
+      "learning_rate": 4.0848543176518784e-05,
+      "loss": 1.1457,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8461213030268067,
+      "grad_norm": 1.154585599899292,
+      "learning_rate": 4.081737329530386e-05,
+      "loss": 1.2012,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8476597053959463,
+      "grad_norm": 1.3288275003433228,
+      "learning_rate": 4.0786162359208386e-05,
+      "loss": 1.2624,
+      "step": 2755
+    },
+    {
+      "epoch": 0.849198107765086,
+      "grad_norm": 4.425066947937012,
+      "learning_rate": 4.075491044924209e-05,
+      "loss": 1.0656,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8507365101342256,
+      "grad_norm": 0.917151927947998,
+      "learning_rate": 4.072361764652105e-05,
+      "loss": 1.0405,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8522749125033653,
+      "grad_norm": 0.9385243654251099,
+      "learning_rate": 4.0692284032267516e-05,
+      "loss": 1.2393,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8538133148725049,
+      "grad_norm": 1.1418960094451904,
+      "learning_rate": 4.0660909687809625e-05,
+      "loss": 1.2778,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8553517172416446,
+      "grad_norm": 0.9725258350372314,
+      "learning_rate": 4.062949469458125e-05,
+      "loss": 1.1223,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8568901196107842,
+      "grad_norm": 0.8571771383285522,
+      "learning_rate": 4.059803913412178e-05,
+      "loss": 1.1825,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8584285219799238,
+      "grad_norm": 1.6413207054138184,
+      "learning_rate": 4.056654308807588e-05,
+      "loss": 1.1504,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8599669243490635,
+      "grad_norm": 1.0783355236053467,
+      "learning_rate": 4.053500663819331e-05,
+      "loss": 1.2997,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8615053267182031,
+      "grad_norm": 1.5912343263626099,
+      "learning_rate": 4.05034298663287e-05,
+      "loss": 1.201,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8630437290873428,
+      "grad_norm": 1.5445067882537842,
+      "learning_rate": 4.047181285444133e-05,
+      "loss": 1.0921,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8645821314564824,
+      "grad_norm": 1.2195634841918945,
+      "learning_rate": 4.0440155684594915e-05,
+      "loss": 1.1598,
+      "step": 2810
+    },
+    {
+      "epoch": 0.866120533825622,
+      "grad_norm": 1.3299264907836914,
+      "learning_rate": 4.0408458438957454e-05,
+      "loss": 1.2355,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8676589361947618,
+      "grad_norm": 1.7162269353866577,
+      "learning_rate": 4.0376721199800896e-05,
+      "loss": 1.1072,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8691973385639014,
+      "grad_norm": 1.240971326828003,
+      "learning_rate": 4.034494404950103e-05,
+      "loss": 1.184,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8707357409330411,
+      "grad_norm": 1.2077282667160034,
+      "learning_rate": 4.0313127070537244e-05,
+      "loss": 1.1602,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8722741433021807,
+      "grad_norm": 0.9052446484565735,
+      "learning_rate": 4.028127034549229e-05,
+      "loss": 1.1715,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8738125456713204,
+      "grad_norm": 1.6603598594665527,
+      "learning_rate": 4.024937395705209e-05,
+      "loss": 1.161,
+      "step": 2840
+    },
+    {
+      "epoch": 0.87535094804046,
+      "grad_norm": 1.1748160123825073,
+      "learning_rate": 4.0217437988005515e-05,
+      "loss": 1.2401,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8768893504095996,
+      "grad_norm": 0.7816171646118164,
+      "learning_rate": 4.0185462521244146e-05,
+      "loss": 1.289,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8784277527787393,
+      "grad_norm": 2.378026247024536,
+      "learning_rate": 4.015344763976212e-05,
+      "loss": 1.1013,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8799661551478789,
+      "grad_norm": 0.9903531670570374,
+      "learning_rate": 4.012139342665586e-05,
+      "loss": 1.1572,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8815045575170186,
+      "grad_norm": 1.0575708150863647,
+      "learning_rate": 4.0089299965123875e-05,
+      "loss": 1.1393,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8830429598861582,
+      "grad_norm": 1.2809288501739502,
+      "learning_rate": 4.005716733846653e-05,
+      "loss": 1.1285,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8845813622552978,
+      "grad_norm": 2.3209915161132812,
+      "learning_rate": 4.0024995630085885e-05,
+      "loss": 1.1748,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8861197646244375,
+      "grad_norm": 1.7122613191604614,
+      "learning_rate": 3.999278492348539e-05,
+      "loss": 1.188,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8876581669935771,
+      "grad_norm": 1.0947169065475464,
+      "learning_rate": 3.996053530226977e-05,
+      "loss": 1.1547,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8891965693627168,
+      "grad_norm": 2.022311210632324,
+      "learning_rate": 3.992824685014471e-05,
+      "loss": 1.1289,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8907349717318565,
+      "grad_norm": 1.397825837135315,
+      "learning_rate": 3.9895919650916716e-05,
+      "loss": 1.2032,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8922733741009962,
+      "grad_norm": 0.9696956276893616,
+      "learning_rate": 3.9863553788492834e-05,
+      "loss": 1.0806,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8938117764701358,
+      "grad_norm": 1.9239165782928467,
+      "learning_rate": 3.983114934688048e-05,
+      "loss": 1.1205,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8953501788392754,
+      "grad_norm": 1.3101149797439575,
+      "learning_rate": 3.97987064101872e-05,
+      "loss": 1.1769,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8968885812084151,
+      "grad_norm": 0.9820067286491394,
+      "learning_rate": 3.976622506262047e-05,
+      "loss": 1.1518,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8984269835775547,
+      "grad_norm": 1.2778306007385254,
+      "learning_rate": 3.973370538848744e-05,
+      "loss": 1.156,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8999653859466944,
+      "grad_norm": 2.279306411743164,
+      "learning_rate": 3.970114747219475e-05,
+      "loss": 1.0314,
+      "step": 2925
+    },
+    {
+      "epoch": 0.901503788315834,
+      "grad_norm": 1.2938816547393799,
+      "learning_rate": 3.966855139824831e-05,
+      "loss": 1.1778,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9030421906849736,
+      "grad_norm": 0.8507199883460999,
+      "learning_rate": 3.963591725125305e-05,
+      "loss": 1.1217,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9045805930541133,
+      "grad_norm": 1.3915897607803345,
+      "learning_rate": 3.9603245115912736e-05,
+      "loss": 1.2169,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9061189954232529,
+      "grad_norm": 1.3529155254364014,
+      "learning_rate": 3.9570535077029724e-05,
+      "loss": 1.1911,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9076573977923926,
+      "grad_norm": 1.5553487539291382,
+      "learning_rate": 3.953778721950477e-05,
+      "loss": 1.1837,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9091958001615322,
+      "grad_norm": 1.1085933446884155,
+      "learning_rate": 3.9505001628336757e-05,
+      "loss": 1.2342,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9107342025306719,
+      "grad_norm": 1.0837507247924805,
+      "learning_rate": 3.947217838862255e-05,
+      "loss": 1.2045,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9122726048998115,
+      "grad_norm": 0.9050890207290649,
+      "learning_rate": 3.943931758555669e-05,
+      "loss": 1.2006,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9138110072689511,
+      "grad_norm": 1.1101486682891846,
+      "learning_rate": 3.940641930443125e-05,
+      "loss": 1.1462,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9153494096380909,
+      "grad_norm": 2.048758029937744,
+      "learning_rate": 3.9373483630635564e-05,
+      "loss": 1.1595,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9168878120072305,
+      "grad_norm": 1.0824599266052246,
+      "learning_rate": 3.934051064965602e-05,
+      "loss": 1.1756,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9184262143763702,
+      "grad_norm": 2.319460153579712,
+      "learning_rate": 3.9307500447075844e-05,
+      "loss": 1.1852,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9199646167455098,
+      "grad_norm": 0.8897768259048462,
+      "learning_rate": 3.927445310857487e-05,
+      "loss": 1.0803,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9215030191146495,
+      "grad_norm": 0.7820705771446228,
+      "learning_rate": 3.924136871992932e-05,
+      "loss": 1.2017,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9230414214837891,
+      "grad_norm": 1.2413580417633057,
+      "learning_rate": 3.9208247367011574e-05,
+      "loss": 1.1179,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9245798238529287,
+      "grad_norm": 2.127316951751709,
+      "learning_rate": 3.9175089135789987e-05,
+      "loss": 1.059,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9261182262220684,
+      "grad_norm": 2.3183646202087402,
+      "learning_rate": 3.914189411232858e-05,
+      "loss": 1.1931,
+      "step": 3010
+    },
+    {
+      "epoch": 0.927656628591208,
+      "grad_norm": 1.7735154628753662,
+      "learning_rate": 3.9108662382786925e-05,
+      "loss": 1.1452,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9291950309603477,
+      "grad_norm": 1.7767513990402222,
+      "learning_rate": 3.9075394033419826e-05,
+      "loss": 1.0446,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9307334333294873,
+      "grad_norm": 0.9649571776390076,
+      "learning_rate": 3.904208915057716e-05,
+      "loss": 1.1965,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9322718356986269,
+      "grad_norm": 1.123539686203003,
+      "learning_rate": 3.900874782070362e-05,
+      "loss": 1.1604,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9338102380677666,
+      "grad_norm": 1.48845374584198,
+      "learning_rate": 3.897537013033849e-05,
+      "loss": 1.1589,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9353486404369062,
+      "grad_norm": 0.8311707377433777,
+      "learning_rate": 3.8941956166115454e-05,
+      "loss": 1.2621,
+      "step": 3040
+    },
+    {
+      "epoch": 0.936887042806046,
+      "grad_norm": 1.6711184978485107,
+      "learning_rate": 3.890850601476233e-05,
+      "loss": 1.2114,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9384254451751856,
+      "grad_norm": 1.2482969760894775,
+      "learning_rate": 3.887501976310086e-05,
+      "loss": 1.1918,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9399638475443253,
+      "grad_norm": 1.0476139783859253,
+      "learning_rate": 3.884149749804648e-05,
+      "loss": 1.1405,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9415022499134649,
+      "grad_norm": 0.864454984664917,
+      "learning_rate": 3.880793930660813e-05,
+      "loss": 1.1581,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9430406522826045,
+      "grad_norm": 1.3563313484191895,
+      "learning_rate": 3.877434527588798e-05,
+      "loss": 1.1005,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9445790546517442,
+      "grad_norm": 1.5107038021087646,
+      "learning_rate": 3.8740715493081203e-05,
+      "loss": 1.0558,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9461174570208838,
+      "grad_norm": 1.4398322105407715,
+      "learning_rate": 3.87070500454758e-05,
+      "loss": 1.2505,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9476558593900235,
+      "grad_norm": 1.0815845727920532,
+      "learning_rate": 3.867334902045234e-05,
+      "loss": 1.2257,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9491942617591631,
+      "grad_norm": 1.0675805807113647,
+      "learning_rate": 3.863961250548371e-05,
+      "loss": 1.0496,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9507326641283027,
+      "grad_norm": 1.3433740139007568,
+      "learning_rate": 3.860584058813495e-05,
+      "loss": 1.2241,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9522710664974424,
+      "grad_norm": 0.8536397218704224,
+      "learning_rate": 3.8572033356062943e-05,
+      "loss": 1.2073,
+      "step": 3095
+    },
+    {
+      "epoch": 0.953809468866582,
+      "grad_norm": 1.1579813957214355,
+      "learning_rate": 3.853819089701627e-05,
+      "loss": 1.2149,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9553478712357217,
+      "grad_norm": 1.0714601278305054,
+      "learning_rate": 3.850431329883493e-05,
+      "loss": 1.2619,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9568862736048613,
+      "grad_norm": 1.3584688901901245,
+      "learning_rate": 3.847040064945014e-05,
+      "loss": 1.1022,
+      "step": 3110
+    },
+    {
+      "epoch": 0.958424675974001,
+      "grad_norm": 3.527039051055908,
+      "learning_rate": 3.843645303688408e-05,
+      "loss": 1.3198,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9599630783431407,
+      "grad_norm": 0.9427330493927002,
+      "learning_rate": 3.840247054924968e-05,
+      "loss": 1.1009,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9615014807122803,
+      "grad_norm": 1.2566494941711426,
+      "learning_rate": 3.836845327475041e-05,
+      "loss": 1.1355,
+      "step": 3125
+    },
+    {
+      "epoch": 0.96303988308142,
+      "grad_norm": 1.2846300601959229,
+      "learning_rate": 3.833440130167999e-05,
+      "loss": 1.1365,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9645782854505596,
+      "grad_norm": 0.7137424945831299,
+      "learning_rate": 3.830031471842226e-05,
+      "loss": 1.0481,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9661166878196993,
+      "grad_norm": 2.03141188621521,
+      "learning_rate": 3.826619361345084e-05,
+      "loss": 1.32,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9676550901888389,
+      "grad_norm": 2.1782736778259277,
+      "learning_rate": 3.823203807532898e-05,
+      "loss": 1.2055,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9691934925579785,
+      "grad_norm": 0.9288005232810974,
+      "learning_rate": 3.8197848192709286e-05,
+      "loss": 1.1541,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9707318949271182,
+      "grad_norm": 1.3783565759658813,
+      "learning_rate": 3.816362405433353e-05,
+      "loss": 1.213,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9722702972962578,
+      "grad_norm": 2.103428363800049,
+      "learning_rate": 3.81293657490324e-05,
+      "loss": 1.2212,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9738086996653975,
+      "grad_norm": 0.9342910051345825,
+      "learning_rate": 3.809507336572522e-05,
+      "loss": 1.2327,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9753471020345371,
+      "grad_norm": 1.7723548412322998,
+      "learning_rate": 3.80607469934198e-05,
+      "loss": 1.1732,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9768855044036768,
+      "grad_norm": 1.1316125392913818,
+      "learning_rate": 3.80263867212122e-05,
+      "loss": 1.1486,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9784239067728164,
+      "grad_norm": 0.8731313347816467,
+      "learning_rate": 3.79919926382864e-05,
+      "loss": 1.2067,
+      "step": 3180
+    },
+    {
+      "epoch": 0.979962309141956,
+      "grad_norm": 0.8731054067611694,
+      "learning_rate": 3.795756483391419e-05,
+      "loss": 1.1672,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9815007115110957,
+      "grad_norm": 1.165216326713562,
+      "learning_rate": 3.792310339745486e-05,
+      "loss": 1.1711,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9830391138802353,
+      "grad_norm": 1.8140064477920532,
+      "learning_rate": 3.788860841835502e-05,
+      "loss": 1.1883,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9845775162493751,
+      "grad_norm": 1.9973480701446533,
+      "learning_rate": 3.785407998614831e-05,
+      "loss": 1.0442,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9861159186185147,
+      "grad_norm": 1.0596879720687866,
+      "learning_rate": 3.781951819045521e-05,
+      "loss": 1.2081,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9876543209876543,
+      "grad_norm": 0.9382377862930298,
+      "learning_rate": 3.778492312098283e-05,
+      "loss": 1.2106,
+      "step": 3210
+    },
+    {
+      "epoch": 0.989192723356794,
+      "grad_norm": 1.565450668334961,
+      "learning_rate": 3.7750294867524585e-05,
+      "loss": 1.0911,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9907311257259336,
+      "grad_norm": 1.8136720657348633,
+      "learning_rate": 3.771563351996006e-05,
+      "loss": 1.1857,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9922695280950733,
+      "grad_norm": 1.7879712581634521,
+      "learning_rate": 3.7680939168254733e-05,
+      "loss": 1.2005,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9938079304642129,
+      "grad_norm": 0.8127261996269226,
+      "learning_rate": 3.7646211902459736e-05,
+      "loss": 1.1781,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9953463328333526,
+      "grad_norm": 0.8571304082870483,
+      "learning_rate": 3.761145181271164e-05,
+      "loss": 1.2937,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9968847352024922,
+      "grad_norm": 0.9852257966995239,
+      "learning_rate": 3.757665898923223e-05,
+      "loss": 1.0958,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9984231375716318,
+      "grad_norm": 1.0459402799606323,
+      "learning_rate": 3.75418335223282e-05,
+      "loss": 1.353,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9999615399407715,
+      "grad_norm": 0.9802791476249695,
+      "learning_rate": 3.750697550239102e-05,
+      "loss": 1.1591,
+      "step": 3250
+    },
+    {
+      "epoch": 1.0014999423099111,
+      "grad_norm": 1.1630258560180664,
+      "learning_rate": 3.747208501989663e-05,
+      "loss": 1.1936,
+      "step": 3255
+    },
+    {
+      "epoch": 1.0030383446790507,
+      "grad_norm": 1.5038493871688843,
+      "learning_rate": 3.743716216540526e-05,
+      "loss": 1.1801,
+      "step": 3260
+    },
+    {
+      "epoch": 1.0045767470481906,
+      "grad_norm": 3.152625799179077,
+      "learning_rate": 3.7402207029561116e-05,
+      "loss": 1.0709,
+      "step": 3265
+    },
+    {
+      "epoch": 1.0061151494173302,
+      "grad_norm": 0.881455659866333,
+      "learning_rate": 3.736721970309222e-05,
+      "loss": 1.2295,
+      "step": 3270
+    },
+    {
+      "epoch": 1.0076535517864698,
+      "grad_norm": 1.6096093654632568,
+      "learning_rate": 3.7332200276810145e-05,
+      "loss": 1.1981,
+      "step": 3275
+    },
+    {
+      "epoch": 1.0091919541556094,
+      "grad_norm": 0.9019604325294495,
+      "learning_rate": 3.7297148841609785e-05,
+      "loss": 1.1362,
+      "step": 3280
+    },
+    {
+      "epoch": 1.010730356524749,
+      "grad_norm": 1.8702175617218018,
+      "learning_rate": 3.72620654884691e-05,
+      "loss": 1.0372,
+      "step": 3285
+    },
+    {
+      "epoch": 1.0122687588938888,
+      "grad_norm": 0.9774813652038574,
+      "learning_rate": 3.722695030844891e-05,
+      "loss": 1.0715,
+      "step": 3290
+    },
+    {
+      "epoch": 1.0138071612630284,
+      "grad_norm": 2.0321578979492188,
+      "learning_rate": 3.7191803392692626e-05,
+      "loss": 1.2701,
+      "step": 3295
+    },
+    {
+      "epoch": 1.015345563632168,
+      "grad_norm": 0.8944913744926453,
+      "learning_rate": 3.715662483242605e-05,
+      "loss": 1.1084,
+      "step": 3300
+    },
+    {
+      "epoch": 1.0168839660013076,
+      "grad_norm": 0.9244990944862366,
+      "learning_rate": 3.712141471895711e-05,
+      "loss": 1.1145,
+      "step": 3305
+    },
+    {
+      "epoch": 1.0184223683704472,
+      "grad_norm": 1.136596441268921,
+      "learning_rate": 3.708617314367562e-05,
+      "loss": 1.0873,
+      "step": 3310
+    },
+    {
+      "epoch": 1.019960770739587,
+      "grad_norm": 0.8394299745559692,
+      "learning_rate": 3.7050900198053096e-05,
+      "loss": 1.1808,
+      "step": 3315
+    },
+    {
+      "epoch": 1.0214991731087266,
+      "grad_norm": 1.818331003189087,
+      "learning_rate": 3.701559597364242e-05,
+      "loss": 1.1542,
+      "step": 3320
+    },
+    {
+      "epoch": 1.0230375754778662,
+      "grad_norm": 0.9814029932022095,
+      "learning_rate": 3.6980260562077694e-05,
+      "loss": 1.1148,
+      "step": 3325
+    },
+    {
+      "epoch": 1.0245759778470058,
+      "grad_norm": 0.9357233643531799,
+      "learning_rate": 3.6944894055073984e-05,
+      "loss": 1.1327,
+      "step": 3330
+    },
+    {
+      "epoch": 1.0261143802161454,
+      "grad_norm": 1.1413395404815674,
+      "learning_rate": 3.690949654442701e-05,
+      "loss": 1.053,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0276527825852853,
+      "grad_norm": 0.9855743050575256,
+      "learning_rate": 3.6874068122013035e-05,
+      "loss": 1.0974,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0291911849544249,
+      "grad_norm": 1.7641159296035767,
+      "learning_rate": 3.6838608879788496e-05,
+      "loss": 1.2385,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0307295873235645,
+      "grad_norm": 1.2974658012390137,
+      "learning_rate": 3.680311890978985e-05,
+      "loss": 1.2269,
+      "step": 3350
+    },
+    {
+      "epoch": 1.032267989692704,
+      "grad_norm": 1.1772881746292114,
+      "learning_rate": 3.6767598304133324e-05,
+      "loss": 1.2326,
+      "step": 3355
+    },
+    {
+      "epoch": 1.0338063920618439,
+      "grad_norm": 1.706217646598816,
+      "learning_rate": 3.673204715501461e-05,
+      "loss": 1.0957,
+      "step": 3360
+    },
+    {
+      "epoch": 1.0353447944309835,
+      "grad_norm": 1.6567659378051758,
+      "learning_rate": 3.669646555470874e-05,
+      "loss": 1.1362,
+      "step": 3365
+    },
+    {
+      "epoch": 1.036883196800123,
+      "grad_norm": 2.460679769515991,
+      "learning_rate": 3.6660853595569754e-05,
+      "loss": 1.1386,
+      "step": 3370
+    },
+    {
+      "epoch": 1.0384215991692627,
+      "grad_norm": 1.0992755889892578,
+      "learning_rate": 3.662521137003048e-05,
+      "loss": 1.0517,
+      "step": 3375
+    },
+    {
+      "epoch": 1.0399600015384023,
+      "grad_norm": 1.0326275825500488,
+      "learning_rate": 3.6589538970602325e-05,
+      "loss": 1.1088,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0414984039075421,
+      "grad_norm": 0.9226017594337463,
+      "learning_rate": 3.6553836489875e-05,
+      "loss": 1.1799,
+      "step": 3385
+    },
+    {
+      "epoch": 1.0430368062766817,
+      "grad_norm": 2.348419189453125,
+      "learning_rate": 3.65181040205163e-05,
+      "loss": 1.1934,
+      "step": 3390
+    },
+    {
+      "epoch": 1.0445752086458213,
+      "grad_norm": 0.982110857963562,
+      "learning_rate": 3.648234165527185e-05,
+      "loss": 1.1904,
+      "step": 3395
+    },
+    {
+      "epoch": 1.046113611014961,
+      "grad_norm": 1.326486587524414,
+      "learning_rate": 3.6446549486964884e-05,
+      "loss": 1.1015,
+      "step": 3400
+    },
+    {
+      "epoch": 1.0476520133841005,
+      "grad_norm": 1.5188394784927368,
+      "learning_rate": 3.641072760849599e-05,
+      "loss": 1.1911,
+      "step": 3405
+    },
+    {
+      "epoch": 1.0491904157532403,
+      "grad_norm": 0.9235185384750366,
+      "learning_rate": 3.637487611284285e-05,
+      "loss": 1.1657,
+      "step": 3410
+    },
+    {
+      "epoch": 1.05072881812238,
+      "grad_norm": 1.6619833707809448,
+      "learning_rate": 3.633899509306004e-05,
+      "loss": 1.108,
+      "step": 3415
+    },
+    {
+      "epoch": 1.0522672204915196,
+      "grad_norm": 1.4106361865997314,
+      "learning_rate": 3.630308464227877e-05,
+      "loss": 1.2603,
+      "step": 3420
+    },
+    {
+      "epoch": 1.0538056228606592,
+      "grad_norm": 1.5941157341003418,
+      "learning_rate": 3.626714485370662e-05,
+      "loss": 1.2641,
+      "step": 3425
+    },
+    {
+      "epoch": 1.0553440252297988,
+      "grad_norm": 1.66989004611969,
+      "learning_rate": 3.6231175820627344e-05,
+      "loss": 1.2124,
+      "step": 3430
+    },
+    {
+      "epoch": 1.0568824275989386,
+      "grad_norm": 1.6584402322769165,
+      "learning_rate": 3.6195177636400565e-05,
+      "loss": 1.1966,
+      "step": 3435
+    },
+    {
+      "epoch": 1.0584208299680782,
+      "grad_norm": 1.228973150253296,
+      "learning_rate": 3.615915039446162e-05,
+      "loss": 0.9945,
+      "step": 3440
+    },
+    {
+      "epoch": 1.0599592323372178,
+      "grad_norm": 2.353611946105957,
+      "learning_rate": 3.6123094188321205e-05,
+      "loss": 1.1743,
+      "step": 3445
+    },
+    {
+      "epoch": 1.0614976347063574,
+      "grad_norm": 1.3476722240447998,
+      "learning_rate": 3.608700911156525e-05,
+      "loss": 1.0994,
+      "step": 3450
+    },
+    {
+      "epoch": 1.0630360370754972,
+      "grad_norm": 2.2004075050354004,
+      "learning_rate": 3.60508952578546e-05,
+      "loss": 1.0864,
+      "step": 3455
+    },
+    {
+      "epoch": 1.0645744394446368,
+      "grad_norm": 1.1636754274368286,
+      "learning_rate": 3.601475272092478e-05,
+      "loss": 1.1245,
+      "step": 3460
+    },
+    {
+      "epoch": 1.0661128418137764,
+      "grad_norm": 1.2166703939437866,
+      "learning_rate": 3.597858159458578e-05,
+      "loss": 1.1763,
+      "step": 3465
+    },
+    {
+      "epoch": 1.067651244182916,
+      "grad_norm": 1.0816514492034912,
+      "learning_rate": 3.594238197272177e-05,
+      "loss": 1.2429,
+      "step": 3470
+    },
+    {
+      "epoch": 1.0691896465520556,
+      "grad_norm": 1.839756965637207,
+      "learning_rate": 3.5906153949290935e-05,
+      "loss": 1.1356,
+      "step": 3475
+    },
+    {
+      "epoch": 1.0707280489211954,
+      "grad_norm": 1.1021186113357544,
+      "learning_rate": 3.5869897618325126e-05,
+      "loss": 1.2321,
+      "step": 3480
+    },
+    {
+      "epoch": 1.072266451290335,
+      "grad_norm": 1.5904773473739624,
+      "learning_rate": 3.5833613073929684e-05,
+      "loss": 1.1777,
+      "step": 3485
+    },
+    {
+      "epoch": 1.0738048536594746,
+      "grad_norm": 1.026505947113037,
+      "learning_rate": 3.579730041028317e-05,
+      "loss": 1.1767,
+      "step": 3490
+    },
+    {
+      "epoch": 1.0753432560286142,
+      "grad_norm": 1.023456335067749,
+      "learning_rate": 3.576095972163718e-05,
+      "loss": 1.1832,
+      "step": 3495
+    },
+    {
+      "epoch": 1.0768816583977538,
+      "grad_norm": 0.9071354269981384,
+      "learning_rate": 3.572459110231598e-05,
+      "loss": 1.0945,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0784200607668937,
+      "grad_norm": 1.8704088926315308,
+      "learning_rate": 3.568819464671637e-05,
+      "loss": 1.1804,
+      "step": 3505
+    },
+    {
+      "epoch": 1.0799584631360333,
+      "grad_norm": 2.5177040100097656,
+      "learning_rate": 3.565177044930739e-05,
+      "loss": 1.1204,
+      "step": 3510
+    },
+    {
+      "epoch": 1.0814968655051729,
+      "grad_norm": 2.31343674659729,
+      "learning_rate": 3.56153186046301e-05,
+      "loss": 1.0483,
+      "step": 3515
+    },
+    {
+      "epoch": 1.0830352678743125,
+      "grad_norm": 1.1547645330429077,
+      "learning_rate": 3.5578839207297306e-05,
+      "loss": 1.1178,
+      "step": 3520
+    },
+    {
+      "epoch": 1.084573670243452,
+      "grad_norm": 1.569669246673584,
+      "learning_rate": 3.5542332351993324e-05,
+      "loss": 1.1746,
+      "step": 3525
+    },
+    {
+      "epoch": 1.086112072612592,
+      "grad_norm": 1.2639979124069214,
+      "learning_rate": 3.550579813347376e-05,
+      "loss": 1.0747,
+      "step": 3530
+    },
+    {
+      "epoch": 1.0876504749817315,
+      "grad_norm": 0.9912775754928589,
+      "learning_rate": 3.546923664656523e-05,
+      "loss": 1.1877,
+      "step": 3535
+    },
+    {
+      "epoch": 1.089188877350871,
+      "grad_norm": 1.9763455390930176,
+      "learning_rate": 3.54326479861651e-05,
+      "loss": 1.2445,
+      "step": 3540
+    },
+    {
+      "epoch": 1.0907272797200107,
+      "grad_norm": 0.9376320838928223,
+      "learning_rate": 3.539603224724133e-05,
+      "loss": 1.2454,
+      "step": 3545
+    },
+    {
+      "epoch": 1.0922656820891503,
+      "grad_norm": 1.8034776449203491,
+      "learning_rate": 3.535938952483211e-05,
+      "loss": 1.2255,
+      "step": 3550
+    },
+    {
+      "epoch": 1.0938040844582901,
+      "grad_norm": 1.3023470640182495,
+      "learning_rate": 3.532271991404568e-05,
+      "loss": 1.1488,
+      "step": 3555
+    },
+    {
+      "epoch": 1.0953424868274297,
+      "grad_norm": 1.3908381462097168,
+      "learning_rate": 3.528602351006006e-05,
+      "loss": 1.1498,
+      "step": 3560
+    },
+    {
+      "epoch": 1.0968808891965693,
+      "grad_norm": 2.6319515705108643,
+      "learning_rate": 3.524930040812286e-05,
+      "loss": 1.2605,
+      "step": 3565
+    },
+    {
+      "epoch": 1.098419291565709,
+      "grad_norm": 1.7958158254623413,
+      "learning_rate": 3.521255070355093e-05,
+      "loss": 1.1793,
+      "step": 3570
+    },
+    {
+      "epoch": 1.0999576939348485,
+      "grad_norm": 1.8693649768829346,
+      "learning_rate": 3.5175774491730204e-05,
+      "loss": 1.2756,
+      "step": 3575
+    },
+    {
+      "epoch": 1.1014960963039884,
+      "grad_norm": 4.634489059448242,
+      "learning_rate": 3.513897186811539e-05,
+      "loss": 1.1377,
+      "step": 3580
+    },
+    {
+      "epoch": 1.103034498673128,
+      "grad_norm": 1.2640043497085571,
+      "learning_rate": 3.510214292822978e-05,
+      "loss": 1.0172,
+      "step": 3585
+    },
+    {
+      "epoch": 1.1045729010422676,
+      "grad_norm": 0.9812289476394653,
+      "learning_rate": 3.506528776766495e-05,
+      "loss": 1.2052,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1061113034114072,
+      "grad_norm": 0.8409488797187805,
+      "learning_rate": 3.5028406482080536e-05,
+      "loss": 1.2435,
+      "step": 3595
+    },
+    {
+      "epoch": 1.107649705780547,
+      "grad_norm": 0.9872679710388184,
+      "learning_rate": 3.499149916720398e-05,
+      "loss": 1.06,
+      "step": 3600
+    },
+    {
+      "epoch": 1.1091881081496866,
+      "grad_norm": 2.1323914527893066,
+      "learning_rate": 3.495456591883031e-05,
+      "loss": 1.0506,
+      "step": 3605
+    },
+    {
+      "epoch": 1.1107265105188262,
+      "grad_norm": 1.2901802062988281,
+      "learning_rate": 3.4917606832821824e-05,
+      "loss": 1.099,
+      "step": 3610
+    },
+    {
+      "epoch": 1.1122649128879658,
+      "grad_norm": 1.414628505706787,
+      "learning_rate": 3.488062200510791e-05,
+      "loss": 1.1987,
+      "step": 3615
+    },
+    {
+      "epoch": 1.1138033152571054,
+      "grad_norm": 1.1899245977401733,
+      "learning_rate": 3.4843611531684786e-05,
+      "loss": 1.1429,
+      "step": 3620
+    },
+    {
+      "epoch": 1.1153417176262452,
+      "grad_norm": 0.9524982571601868,
+      "learning_rate": 3.480657550861518e-05,
+      "loss": 1.1281,
+      "step": 3625
+    },
+    {
+      "epoch": 1.1168801199953848,
+      "grad_norm": 1.219228744506836,
+      "learning_rate": 3.47695140320282e-05,
+      "loss": 1.1641,
+      "step": 3630
+    },
+    {
+      "epoch": 1.1184185223645244,
+      "grad_norm": 2.172743320465088,
+      "learning_rate": 3.473242719811897e-05,
+      "loss": 1.2067,
+      "step": 3635
+    },
+    {
+      "epoch": 1.119956924733664,
+      "grad_norm": 1.288135051727295,
+      "learning_rate": 3.4695315103148454e-05,
+      "loss": 1.1588,
+      "step": 3640
+    },
+    {
+      "epoch": 1.1214953271028036,
+      "grad_norm": 1.3018666505813599,
+      "learning_rate": 3.465817784344318e-05,
+      "loss": 1.1281,
+      "step": 3645
+    },
+    {
+      "epoch": 1.1230337294719435,
+      "grad_norm": 1.5916048288345337,
+      "learning_rate": 3.462101551539499e-05,
+      "loss": 1.2701,
+      "step": 3650
+    },
+    {
+      "epoch": 1.124572131841083,
+      "grad_norm": 1.3709497451782227,
+      "learning_rate": 3.45838282154608e-05,
+      "loss": 1.223,
+      "step": 3655
+    },
+    {
+      "epoch": 1.1261105342102227,
+      "grad_norm": 1.505096673965454,
+      "learning_rate": 3.4546616040162334e-05,
+      "loss": 1.1068,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1276489365793623,
+      "grad_norm": 1.1039677858352661,
+      "learning_rate": 3.450937908608587e-05,
+      "loss": 1.216,
+      "step": 3665
+    },
+    {
+      "epoch": 1.129187338948502,
+      "grad_norm": 1.4067137241363525,
+      "learning_rate": 3.4472117449882025e-05,
+      "loss": 1.1304,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1307257413176417,
+      "grad_norm": 0.9000976085662842,
+      "learning_rate": 3.443483122826547e-05,
+      "loss": 1.1709,
+      "step": 3675
+    },
+    {
+      "epoch": 1.1322641436867813,
+      "grad_norm": 1.4269496202468872,
+      "learning_rate": 3.439752051801467e-05,
+      "loss": 1.2391,
+      "step": 3680
+    },
+    {
+      "epoch": 1.133802546055921,
+      "grad_norm": 0.959568977355957,
+      "learning_rate": 3.436018541597169e-05,
+      "loss": 1.1445,
+      "step": 3685
+    },
+    {
+      "epoch": 1.1353409484250605,
+      "grad_norm": 0.9494476318359375,
+      "learning_rate": 3.4322826019041864e-05,
+      "loss": 1.1943,
+      "step": 3690
+    },
+    {
+      "epoch": 1.1368793507942003,
+      "grad_norm": 0.8512333631515503,
+      "learning_rate": 3.428544242419362e-05,
+      "loss": 1.2019,
+      "step": 3695
+    },
+    {
+      "epoch": 1.13841775316334,
+      "grad_norm": 0.9752541780471802,
+      "learning_rate": 3.4248034728458175e-05,
+      "loss": 1.1219,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1399561555324795,
+      "grad_norm": 1.4414304494857788,
+      "learning_rate": 3.4210603028929295e-05,
+      "loss": 1.2591,
+      "step": 3705
+    },
+    {
+      "epoch": 1.1414945579016191,
+      "grad_norm": 1.0669443607330322,
+      "learning_rate": 3.417314742276308e-05,
+      "loss": 1.1153,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1430329602707587,
+      "grad_norm": 1.064900279045105,
+      "learning_rate": 3.413566800717762e-05,
+      "loss": 1.2842,
+      "step": 3715
+    },
+    {
+      "epoch": 1.1445713626398986,
+      "grad_norm": 1.052623987197876,
+      "learning_rate": 3.409816487945286e-05,
+      "loss": 1.1178,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1461097650090382,
+      "grad_norm": 1.8392632007598877,
+      "learning_rate": 3.4060638136930304e-05,
+      "loss": 1.1444,
+      "step": 3725
+    },
+    {
+      "epoch": 1.1476481673781778,
+      "grad_norm": 1.3130204677581787,
+      "learning_rate": 3.402308787701268e-05,
+      "loss": 1.1948,
+      "step": 3730
+    },
+    {
+      "epoch": 1.1491865697473174,
+      "grad_norm": 1.6042733192443848,
+      "learning_rate": 3.398551419716382e-05,
+      "loss": 1.096,
+      "step": 3735
+    },
+    {
+      "epoch": 1.150724972116457,
+      "grad_norm": 1.7846676111221313,
+      "learning_rate": 3.3947917194908306e-05,
+      "loss": 1.2706,
+      "step": 3740
+    },
+    {
+      "epoch": 1.1522633744855968,
+      "grad_norm": 1.7049797773361206,
+      "learning_rate": 3.3910296967831266e-05,
+      "loss": 1.1966,
+      "step": 3745
+    },
+    {
+      "epoch": 1.1538017768547364,
+      "grad_norm": 0.9584757685661316,
+      "learning_rate": 3.3872653613578134e-05,
+      "loss": 1.1628,
+      "step": 3750
+    },
+    {
+      "epoch": 1.155340179223876,
+      "grad_norm": 1.128525972366333,
+      "learning_rate": 3.383498722985432e-05,
+      "loss": 1.163,
+      "step": 3755
+    },
+    {
+      "epoch": 1.1568785815930156,
+      "grad_norm": 1.1508246660232544,
+      "learning_rate": 3.379729791442506e-05,
+      "loss": 1.1373,
+      "step": 3760
+    },
+    {
+      "epoch": 1.1584169839621552,
+      "grad_norm": 0.9009411334991455,
+      "learning_rate": 3.375958576511508e-05,
+      "loss": 1.1696,
+      "step": 3765
+    },
+    {
+      "epoch": 1.159955386331295,
+      "grad_norm": 1.28462553024292,
+      "learning_rate": 3.372185087980838e-05,
+      "loss": 1.0845,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1614937887004346,
+      "grad_norm": 1.2223782539367676,
+      "learning_rate": 3.368409335644798e-05,
+      "loss": 1.1714,
+      "step": 3775
+    },
+    {
+      "epoch": 1.1630321910695742,
+      "grad_norm": 1.155945062637329,
+      "learning_rate": 3.364631329303564e-05,
+      "loss": 1.1125,
+      "step": 3780
+    },
+    {
+      "epoch": 1.1645705934387138,
+      "grad_norm": 2.37724232673645,
+      "learning_rate": 3.3608510787631654e-05,
+      "loss": 1.2847,
+      "step": 3785
+    },
+    {
+      "epoch": 1.1661089958078534,
+      "grad_norm": 1.2505033016204834,
+      "learning_rate": 3.357068593835453e-05,
+      "loss": 1.1486,
+      "step": 3790
+    },
+    {
+      "epoch": 1.1676473981769933,
+      "grad_norm": 1.1727055311203003,
+      "learning_rate": 3.35328388433808e-05,
+      "loss": 1.1909,
+      "step": 3795
+    },
+    {
+      "epoch": 1.1691858005461329,
+      "grad_norm": 2.013213872909546,
+      "learning_rate": 3.3494969600944715e-05,
+      "loss": 1.1553,
+      "step": 3800
+    },
+    {
+      "epoch": 1.1707242029152725,
+      "grad_norm": 1.5769529342651367,
+      "learning_rate": 3.345707830933803e-05,
+      "loss": 1.2083,
+      "step": 3805
+    },
+    {
+      "epoch": 1.172262605284412,
+      "grad_norm": 1.1164283752441406,
+      "learning_rate": 3.3419165066909705e-05,
+      "loss": 1.1371,
+      "step": 3810
+    },
+    {
+      "epoch": 1.1738010076535517,
+      "grad_norm": 1.1431792974472046,
+      "learning_rate": 3.338122997206571e-05,
+      "loss": 1.1377,
+      "step": 3815
+    },
+    {
+      "epoch": 1.1753394100226915,
+      "grad_norm": 1.0517457723617554,
+      "learning_rate": 3.33432731232687e-05,
+      "loss": 1.1152,
+      "step": 3820
+    },
+    {
+      "epoch": 1.176877812391831,
+      "grad_norm": 1.159220814704895,
+      "learning_rate": 3.3305294619037805e-05,
+      "loss": 1.1877,
+      "step": 3825
+    },
+    {
+      "epoch": 1.1784162147609707,
+      "grad_norm": 1.6318359375,
+      "learning_rate": 3.326729455794838e-05,
+      "loss": 1.1014,
+      "step": 3830
+    },
+    {
+      "epoch": 1.1799546171301103,
+      "grad_norm": 1.4239236116409302,
+      "learning_rate": 3.322927303863171e-05,
+      "loss": 1.2185,
+      "step": 3835
+    },
+    {
+      "epoch": 1.1814930194992501,
+      "grad_norm": 2.194650650024414,
+      "learning_rate": 3.319123015977478e-05,
+      "loss": 1.1972,
+      "step": 3840
+    },
+    {
+      "epoch": 1.1830314218683897,
+      "grad_norm": 0.9801701307296753,
+      "learning_rate": 3.315316602012001e-05,
+      "loss": 1.1984,
+      "step": 3845
+    },
+    {
+      "epoch": 1.1845698242375293,
+      "grad_norm": 1.7991083860397339,
+      "learning_rate": 3.311508071846504e-05,
+      "loss": 1.1742,
+      "step": 3850
+    },
+    {
+      "epoch": 1.186108226606669,
+      "grad_norm": 2.0683629512786865,
+      "learning_rate": 3.307697435366237e-05,
+      "loss": 1.1949,
+      "step": 3855
+    },
+    {
+      "epoch": 1.1876466289758085,
+      "grad_norm": 1.1672157049179077,
+      "learning_rate": 3.303884702461924e-05,
+      "loss": 1.1302,
+      "step": 3860
+    },
+    {
+      "epoch": 1.1891850313449484,
+      "grad_norm": 2.2780561447143555,
+      "learning_rate": 3.300069883029727e-05,
+      "loss": 1.1606,
+      "step": 3865
+    },
+    {
+      "epoch": 1.190723433714088,
+      "grad_norm": 0.9155642986297607,
+      "learning_rate": 3.296252986971222e-05,
+      "loss": 1.2142,
+      "step": 3870
+    },
+    {
+      "epoch": 1.1922618360832276,
+      "grad_norm": 0.7124369740486145,
+      "learning_rate": 3.29243402419338e-05,
+      "loss": 1.2005,
+      "step": 3875
+    },
+    {
+      "epoch": 1.1938002384523672,
+      "grad_norm": 0.9560266137123108,
+      "learning_rate": 3.2886130046085306e-05,
+      "loss": 1.1488,
+      "step": 3880
+    },
+    {
+      "epoch": 1.195338640821507,
+      "grad_norm": 2.2523956298828125,
+      "learning_rate": 3.284789938134346e-05,
+      "loss": 1.1306,
+      "step": 3885
+    },
+    {
+      "epoch": 1.1968770431906466,
+      "grad_norm": 1.153905987739563,
+      "learning_rate": 3.2809648346938105e-05,
+      "loss": 1.2155,
+      "step": 3890
+    },
+    {
+      "epoch": 1.1984154455597862,
+      "grad_norm": 1.6172610521316528,
+      "learning_rate": 3.2771377042151944e-05,
+      "loss": 1.1306,
+      "step": 3895
+    },
+    {
+      "epoch": 1.1999538479289258,
+      "grad_norm": 1.023131012916565,
+      "learning_rate": 3.2733085566320285e-05,
+      "loss": 1.1581,
+      "step": 3900
+    },
+    {
+      "epoch": 1.2014922502980654,
+      "grad_norm": 1.280826449394226,
+      "learning_rate": 3.2694774018830816e-05,
+      "loss": 1.238,
+      "step": 3905
+    },
+    {
+      "epoch": 1.2030306526672052,
+      "grad_norm": 0.9869760870933533,
+      "learning_rate": 3.265644249912331e-05,
+      "loss": 1.2139,
+      "step": 3910
+    },
+    {
+      "epoch": 1.2045690550363448,
+      "grad_norm": 2.917651414871216,
+      "learning_rate": 3.261809110668937e-05,
+      "loss": 0.9819,
+      "step": 3915
+    },
+    {
+      "epoch": 1.2061074574054844,
+      "grad_norm": 0.8974500298500061,
+      "learning_rate": 3.25797199410722e-05,
+      "loss": 1.144,
+      "step": 3920
+    },
+    {
+      "epoch": 1.207645859774624,
+      "grad_norm": 1.1364940404891968,
+      "learning_rate": 3.254132910186631e-05,
+      "loss": 1.2498,
+      "step": 3925
+    },
+    {
+      "epoch": 1.2091842621437636,
+      "grad_norm": 1.5066908597946167,
+      "learning_rate": 3.2502918688717285e-05,
+      "loss": 1.1648,
+      "step": 3930
+    },
+    {
+      "epoch": 1.2107226645129034,
+      "grad_norm": 1.5828046798706055,
+      "learning_rate": 3.2464488801321494e-05,
+      "loss": 1.2591,
+      "step": 3935
+    },
+    {
+      "epoch": 1.212261066882043,
+      "grad_norm": 1.869602918624878,
+      "learning_rate": 3.2426039539425876e-05,
+      "loss": 1.2592,
+      "step": 3940
+    },
+    {
+      "epoch": 1.2137994692511826,
+      "grad_norm": 0.9975070357322693,
+      "learning_rate": 3.2387571002827656e-05,
+      "loss": 1.2349,
+      "step": 3945
+    },
+    {
+      "epoch": 1.2153378716203223,
+      "grad_norm": 2.0380489826202393,
+      "learning_rate": 3.234908329137406e-05,
+      "loss": 1.1056,
+      "step": 3950
+    },
+    {
+      "epoch": 1.2168762739894619,
+      "grad_norm": 1.3352590799331665,
+      "learning_rate": 3.231057650496214e-05,
+      "loss": 1.1998,
+      "step": 3955
+    },
+    {
+      "epoch": 1.2184146763586017,
+      "grad_norm": 1.1479742527008057,
+      "learning_rate": 3.2272050743538385e-05,
+      "loss": 1.1915,
+      "step": 3960
+    },
+    {
+      "epoch": 1.2199530787277413,
+      "grad_norm": 1.3033512830734253,
+      "learning_rate": 3.22335061070986e-05,
+      "loss": 1.1046,
+      "step": 3965
+    },
+    {
+      "epoch": 1.2214914810968809,
+      "grad_norm": 1.171144723892212,
+      "learning_rate": 3.219494269568753e-05,
+      "loss": 1.1206,
+      "step": 3970
+    },
+    {
+      "epoch": 1.2230298834660205,
+      "grad_norm": 1.440745234489441,
+      "learning_rate": 3.215636060939869e-05,
+      "loss": 1.112,
+      "step": 3975
+    },
+    {
+      "epoch": 1.22456828583516,
+      "grad_norm": 1.118881344795227,
+      "learning_rate": 3.211775994837405e-05,
+      "loss": 1.166,
+      "step": 3980
+    },
+    {
+      "epoch": 1.2261066882043,
+      "grad_norm": 1.8601933717727661,
+      "learning_rate": 3.207914081280379e-05,
+      "loss": 1.2695,
+      "step": 3985
+    },
+    {
+      "epoch": 1.2276450905734395,
+      "grad_norm": 1.5979291200637817,
+      "learning_rate": 3.204050330292604e-05,
+      "loss": 1.1158,
+      "step": 3990
+    },
+    {
+      "epoch": 1.2291834929425791,
+      "grad_norm": 1.368026614189148,
+      "learning_rate": 3.2001847519026646e-05,
+      "loss": 1.1164,
+      "step": 3995
+    },
+    {
+      "epoch": 1.2307218953117187,
+      "grad_norm": 1.9506460428237915,
+      "learning_rate": 3.196317356143884e-05,
+      "loss": 1.1607,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2322602976808583,
+      "grad_norm": 1.1299954652786255,
+      "learning_rate": 3.192448153054306e-05,
+      "loss": 1.1518,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2337987000499981,
+      "grad_norm": 1.389312982559204,
+      "learning_rate": 3.188577152676666e-05,
+      "loss": 1.0802,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2353371024191377,
+      "grad_norm": 1.1280195713043213,
+      "learning_rate": 3.1847043650583604e-05,
+      "loss": 1.1637,
+      "step": 4015
+    },
+    {
+      "epoch": 1.2368755047882773,
+      "grad_norm": 0.959011971950531,
+      "learning_rate": 3.180829800251428e-05,
+      "loss": 1.1652,
+      "step": 4020
+    },
+    {
+      "epoch": 1.238413907157417,
+      "grad_norm": 1.3471107482910156,
+      "learning_rate": 3.1769534683125195e-05,
+      "loss": 1.1713,
+      "step": 4025
+    },
+    {
+      "epoch": 1.2399523095265566,
+      "grad_norm": 1.7551147937774658,
+      "learning_rate": 3.1730753793028724e-05,
+      "loss": 1.1362,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2414907118956964,
+      "grad_norm": 1.71892249584198,
+      "learning_rate": 3.169195543288283e-05,
+      "loss": 1.1525,
+      "step": 4035
+    },
+    {
+      "epoch": 1.243029114264836,
+      "grad_norm": 1.1600077152252197,
+      "learning_rate": 3.165313970339087e-05,
+      "loss": 1.1085,
+      "step": 4040
+    },
+    {
+      "epoch": 1.2445675166339756,
+      "grad_norm": 0.9502371549606323,
+      "learning_rate": 3.1614306705301204e-05,
+      "loss": 1.2047,
+      "step": 4045
+    },
+    {
+      "epoch": 1.2461059190031152,
+      "grad_norm": 1.5982012748718262,
+      "learning_rate": 3.15754565394071e-05,
+      "loss": 1.167,
+      "step": 4050
+    },
+    {
+      "epoch": 1.247644321372255,
+      "grad_norm": 1.2877519130706787,
+      "learning_rate": 3.153658930654631e-05,
+      "loss": 1.0551,
+      "step": 4055
+    },
+    {
+      "epoch": 1.2491827237413946,
+      "grad_norm": 0.9973001480102539,
+      "learning_rate": 3.149770510760093e-05,
+      "loss": 1.0985,
+      "step": 4060
+    },
+    {
+      "epoch": 1.2507211261105342,
+      "grad_norm": 2.4550881385803223,
+      "learning_rate": 3.14588040434971e-05,
+      "loss": 1.2302,
+      "step": 4065
+    },
+    {
+      "epoch": 1.2522595284796738,
+      "grad_norm": 1.004104733467102,
+      "learning_rate": 3.1419886215204694e-05,
+      "loss": 1.1658,
+      "step": 4070
+    },
+    {
+      "epoch": 1.2537979308488136,
+      "grad_norm": 2.3353803157806396,
+      "learning_rate": 3.138095172373714e-05,
+      "loss": 1.1095,
+      "step": 4075
+    },
+    {
+      "epoch": 1.2553363332179532,
+      "grad_norm": 3.044813394546509,
+      "learning_rate": 3.134200067015108e-05,
+      "loss": 1.2342,
+      "step": 4080
+    },
+    {
+      "epoch": 1.2568747355870928,
+      "grad_norm": 1.2953006029129028,
+      "learning_rate": 3.130303315554617e-05,
+      "loss": 1.0888,
+      "step": 4085
+    },
+    {
+      "epoch": 1.2584131379562324,
+      "grad_norm": 1.468589425086975,
+      "learning_rate": 3.1264049281064775e-05,
+      "loss": 1.0256,
+      "step": 4090
+    },
+    {
+      "epoch": 1.259951540325372,
+      "grad_norm": 1.058424949645996,
+      "learning_rate": 3.1225049147891737e-05,
+      "loss": 1.1953,
+      "step": 4095
+    },
+    {
+      "epoch": 1.2614899426945119,
+      "grad_norm": 1.2426944971084595,
+      "learning_rate": 3.118603285725409e-05,
+      "loss": 1.2162,
+      "step": 4100
+    },
+    {
+      "epoch": 1.2630283450636515,
+      "grad_norm": 1.318311095237732,
+      "learning_rate": 3.114700051042081e-05,
+      "loss": 1.0836,
+      "step": 4105
+    },
+    {
+      "epoch": 1.264566747432791,
+      "grad_norm": 1.1298370361328125,
+      "learning_rate": 3.110795220870253e-05,
+      "loss": 1.1453,
+      "step": 4110
+    },
+    {
+      "epoch": 1.2661051498019307,
+      "grad_norm": 1.2516565322875977,
+      "learning_rate": 3.1068888053451334e-05,
+      "loss": 1.1972,
+      "step": 4115
+    },
+    {
+      "epoch": 1.2676435521710703,
+      "grad_norm": 1.364283800125122,
+      "learning_rate": 3.102980814606042e-05,
+      "loss": 1.0966,
+      "step": 4120
+    },
+    {
+      "epoch": 1.26918195454021,
+      "grad_norm": 1.593657374382019,
+      "learning_rate": 3.099071258796387e-05,
+      "loss": 1.1734,
+      "step": 4125
+    },
+    {
+      "epoch": 1.2707203569093497,
+      "grad_norm": 1.190921425819397,
+      "learning_rate": 3.0951601480636404e-05,
+      "loss": 1.2381,
+      "step": 4130
+    },
+    {
+      "epoch": 1.2722587592784893,
+      "grad_norm": 1.0939090251922607,
+      "learning_rate": 3.091247492559312e-05,
+      "loss": 1.0485,
+      "step": 4135
+    },
+    {
+      "epoch": 1.273797161647629,
+      "grad_norm": 1.5152145624160767,
+      "learning_rate": 3.087333302438916e-05,
+      "loss": 1.0804,
+      "step": 4140
+    },
+    {
+      "epoch": 1.2753355640167685,
+      "grad_norm": 1.4711779356002808,
+      "learning_rate": 3.0834175878619546e-05,
+      "loss": 1.0904,
+      "step": 4145
+    },
+    {
+      "epoch": 1.2768739663859083,
+      "grad_norm": 1.8078420162200928,
+      "learning_rate": 3.0795003589918834e-05,
+      "loss": 1.1033,
+      "step": 4150
+    },
+    {
+      "epoch": 1.278412368755048,
+      "grad_norm": 1.389979600906372,
+      "learning_rate": 3.0755816259960915e-05,
+      "loss": 1.0441,
+      "step": 4155
+    },
+    {
+      "epoch": 1.2799507711241875,
+      "grad_norm": 1.3937149047851562,
+      "learning_rate": 3.071661399045869e-05,
+      "loss": 1.235,
+      "step": 4160
+    },
+    {
+      "epoch": 1.2814891734933271,
+      "grad_norm": 1.5922768115997314,
+      "learning_rate": 3.067739688316385e-05,
+      "loss": 1.1928,
+      "step": 4165
+    },
+    {
+      "epoch": 1.2830275758624667,
+      "grad_norm": 1.0673027038574219,
+      "learning_rate": 3.0638165039866614e-05,
+      "loss": 1.0306,
+      "step": 4170
+    },
+    {
+      "epoch": 1.2845659782316066,
+      "grad_norm": 1.3722416162490845,
+      "learning_rate": 3.0598918562395425e-05,
+      "loss": 1.0907,
+      "step": 4175
+    },
+    {
+      "epoch": 1.2861043806007462,
+      "grad_norm": 0.9146732687950134,
+      "learning_rate": 3.0559657552616716e-05,
+      "loss": 1.1372,
+      "step": 4180
+    },
+    {
+      "epoch": 1.2876427829698858,
+      "grad_norm": 1.4379206895828247,
+      "learning_rate": 3.0520382112434636e-05,
+      "loss": 1.1397,
+      "step": 4185
+    },
+    {
+      "epoch": 1.2891811853390254,
+      "grad_norm": 1.0497307777404785,
+      "learning_rate": 3.0481092343790822e-05,
+      "loss": 1.1,
+      "step": 4190
+    },
+    {
+      "epoch": 1.290719587708165,
+      "grad_norm": 0.9752093553543091,
+      "learning_rate": 3.044178834866405e-05,
+      "loss": 1.1479,
+      "step": 4195
+    },
+    {
+      "epoch": 1.2922579900773048,
+      "grad_norm": 1.132630467414856,
+      "learning_rate": 3.0402470229070056e-05,
+      "loss": 1.1148,
+      "step": 4200
+    },
+    {
+      "epoch": 1.2937963924464444,
+      "grad_norm": 1.2013700008392334,
+      "learning_rate": 3.0363138087061222e-05,
+      "loss": 1.0205,
+      "step": 4205
+    },
+    {
+      "epoch": 1.295334794815584,
+      "grad_norm": 1.7345274686813354,
+      "learning_rate": 3.032379202472634e-05,
+      "loss": 1.1862,
+      "step": 4210
+    },
+    {
+      "epoch": 1.2968731971847236,
+      "grad_norm": 1.4804950952529907,
+      "learning_rate": 3.0284432144190315e-05,
+      "loss": 1.0884,
+      "step": 4215
+    },
+    {
+      "epoch": 1.2984115995538632,
+      "grad_norm": 1.7368816137313843,
+      "learning_rate": 3.0245058547613924e-05,
+      "loss": 1.2023,
+      "step": 4220
+    },
+    {
+      "epoch": 1.299950001923003,
+      "grad_norm": 1.1152502298355103,
+      "learning_rate": 3.0205671337193566e-05,
+      "loss": 1.1609,
+      "step": 4225
+    },
+    {
+      "epoch": 1.3014884042921426,
+      "grad_norm": 0.8946184515953064,
+      "learning_rate": 3.016627061516094e-05,
+      "loss": 1.0451,
+      "step": 4230
+    },
+    {
+      "epoch": 1.3030268066612822,
+      "grad_norm": 0.9706270098686218,
+      "learning_rate": 3.0126856483782838e-05,
+      "loss": 1.1886,
+      "step": 4235
+    },
+    {
+      "epoch": 1.3045652090304218,
+      "grad_norm": 2.113147020339966,
+      "learning_rate": 3.008742904536085e-05,
+      "loss": 1.1681,
+      "step": 4240
+    },
+    {
+      "epoch": 1.3061036113995614,
+      "grad_norm": 1.2672151327133179,
+      "learning_rate": 3.0047988402231113e-05,
+      "loss": 1.1769,
+      "step": 4245
+    },
+    {
+      "epoch": 1.3076420137687013,
+      "grad_norm": 0.8190154433250427,
+      "learning_rate": 3.000853465676402e-05,
+      "loss": 1.0251,
+      "step": 4250
+    },
+    {
+      "epoch": 1.3091804161378409,
+      "grad_norm": 1.9547206163406372,
+      "learning_rate": 2.9969067911363992e-05,
+      "loss": 1.2783,
+      "step": 4255
+    },
+    {
+      "epoch": 1.3107188185069805,
+      "grad_norm": 1.1879409551620483,
+      "learning_rate": 2.992958826846918e-05,
+      "loss": 1.0925,
+      "step": 4260
+    },
+    {
+      "epoch": 1.31225722087612,
+      "grad_norm": 1.0043494701385498,
+      "learning_rate": 2.9890095830551207e-05,
+      "loss": 1.135,
+      "step": 4265
+    },
+    {
+      "epoch": 1.3137956232452597,
+      "grad_norm": 1.3702094554901123,
+      "learning_rate": 2.985059070011492e-05,
+      "loss": 1.1819,
+      "step": 4270
+    },
+    {
+      "epoch": 1.3153340256143995,
+      "grad_norm": 1.28737473487854,
+      "learning_rate": 2.9811072979698107e-05,
+      "loss": 1.146,
+      "step": 4275
+    },
+    {
+      "epoch": 1.316872427983539,
+      "grad_norm": 1.6883684396743774,
+      "learning_rate": 2.9771542771871224e-05,
+      "loss": 1.0705,
+      "step": 4280
+    },
+    {
+      "epoch": 1.3184108303526787,
+      "grad_norm": 0.9764033555984497,
+      "learning_rate": 2.973200017923715e-05,
+      "loss": 1.1656,
+      "step": 4285
+    },
+    {
+      "epoch": 1.3199492327218185,
+      "grad_norm": 1.5175962448120117,
+      "learning_rate": 2.96924453044309e-05,
+      "loss": 1.1809,
+      "step": 4290
+    },
+    {
+      "epoch": 1.321487635090958,
+      "grad_norm": 1.7922492027282715,
+      "learning_rate": 2.9652878250119375e-05,
+      "loss": 1.0016,
+      "step": 4295
+    },
+    {
+      "epoch": 1.3230260374600977,
+      "grad_norm": 1.114981770515442,
+      "learning_rate": 2.9613299119001082e-05,
+      "loss": 1.2292,
+      "step": 4300
+    },
+    {
+      "epoch": 1.3245644398292373,
+      "grad_norm": 1.1757748126983643,
+      "learning_rate": 2.9573708013805885e-05,
+      "loss": 1.1243,
+      "step": 4305
+    },
+    {
+      "epoch": 1.326102842198377,
+      "grad_norm": 1.034613847732544,
+      "learning_rate": 2.953410503729471e-05,
+      "loss": 1.2089,
+      "step": 4310
+    },
+    {
+      "epoch": 1.3276412445675168,
+      "grad_norm": 1.4121217727661133,
+      "learning_rate": 2.9494490292259326e-05,
+      "loss": 1.0332,
+      "step": 4315
+    },
+    {
+      "epoch": 1.3291796469366564,
+      "grad_norm": 1.6308557987213135,
+      "learning_rate": 2.945486388152201e-05,
+      "loss": 1.1536,
+      "step": 4320
+    },
+    {
+      "epoch": 1.330718049305796,
+      "grad_norm": 1.5427446365356445,
+      "learning_rate": 2.941522590793534e-05,
+      "loss": 1.2825,
+      "step": 4325
+    },
+    {
+      "epoch": 1.3322564516749356,
+      "grad_norm": 2.0228891372680664,
+      "learning_rate": 2.9375576474381905e-05,
+      "loss": 1.0036,
+      "step": 4330
+    },
+    {
+      "epoch": 1.3337948540440752,
+      "grad_norm": 1.04116952419281,
+      "learning_rate": 2.9335915683774034e-05,
+      "loss": 1.0644,
+      "step": 4335
+    },
+    {
+      "epoch": 1.335333256413215,
+      "grad_norm": 2.5048625469207764,
+      "learning_rate": 2.9296243639053545e-05,
+      "loss": 1.0763,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3368716587823546,
+      "grad_norm": 1.2487713098526,
+      "learning_rate": 2.9256560443191434e-05,
+      "loss": 1.2327,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3384100611514942,
+      "grad_norm": 1.1335622072219849,
+      "learning_rate": 2.9216866199187697e-05,
+      "loss": 1.1097,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3399484635206338,
+      "grad_norm": 1.3442351818084717,
+      "learning_rate": 2.9177161010070946e-05,
+      "loss": 1.1278,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3414868658897734,
+      "grad_norm": 1.2338272333145142,
+      "learning_rate": 2.9137444978898244e-05,
+      "loss": 1.0971,
+      "step": 4360
+    },
+    {
+      "epoch": 1.3430252682589132,
+      "grad_norm": 1.0198732614517212,
+      "learning_rate": 2.9097718208754777e-05,
+      "loss": 1.1223,
+      "step": 4365
+    },
+    {
+      "epoch": 1.3445636706280528,
+      "grad_norm": 1.195879578590393,
+      "learning_rate": 2.90579808027536e-05,
+      "loss": 1.1469,
+      "step": 4370
+    },
+    {
+      "epoch": 1.3461020729971924,
+      "grad_norm": 1.2800102233886719,
+      "learning_rate": 2.901823286403539e-05,
+      "loss": 1.1358,
+      "step": 4375
+    },
+    {
+      "epoch": 1.347640475366332,
+      "grad_norm": 1.07575261592865,
+      "learning_rate": 2.897847449576815e-05,
+      "loss": 1.1122,
+      "step": 4380
+    },
+    {
+      "epoch": 1.3491788777354716,
+      "grad_norm": 0.9419439435005188,
+      "learning_rate": 2.8938705801146958e-05,
+      "loss": 1.1214,
+      "step": 4385
+    },
+    {
+      "epoch": 1.3507172801046115,
+      "grad_norm": 1.0005466938018799,
+      "learning_rate": 2.88989268833937e-05,
+      "loss": 1.2561,
+      "step": 4390
+    },
+    {
+      "epoch": 1.352255682473751,
+      "grad_norm": 0.9994781613349915,
+      "learning_rate": 2.8859137845756784e-05,
+      "loss": 1.125,
+      "step": 4395
+    },
+    {
+      "epoch": 1.3537940848428907,
+      "grad_norm": 1.0525974035263062,
+      "learning_rate": 2.8819338791510887e-05,
+      "loss": 1.1561,
+      "step": 4400
+    },
+    {
+      "epoch": 1.3553324872120303,
+      "grad_norm": 1.1305088996887207,
+      "learning_rate": 2.8779529823956704e-05,
+      "loss": 1.129,
+      "step": 4405
+    },
+    {
+      "epoch": 1.3568708895811699,
+      "grad_norm": 1.073445439338684,
+      "learning_rate": 2.8739711046420626e-05,
+      "loss": 1.1403,
+      "step": 4410
+    },
+    {
+      "epoch": 1.3584092919503097,
+      "grad_norm": 1.792883038520813,
+      "learning_rate": 2.8699882562254538e-05,
+      "loss": 1.0386,
+      "step": 4415
+    },
+    {
+      "epoch": 1.3599476943194493,
+      "grad_norm": 1.0318684577941895,
+      "learning_rate": 2.8660044474835514e-05,
+      "loss": 1.2134,
+      "step": 4420
+    },
+    {
+      "epoch": 1.3614860966885889,
+      "grad_norm": 0.9824214577674866,
+      "learning_rate": 2.862019688756553e-05,
+      "loss": 1.1874,
+      "step": 4425
+    },
+    {
+      "epoch": 1.3630244990577285,
+      "grad_norm": 3.1488852500915527,
+      "learning_rate": 2.858033990387125e-05,
+      "loss": 1.1967,
+      "step": 4430
+    },
+    {
+      "epoch": 1.364562901426868,
+      "grad_norm": 2.4233949184417725,
+      "learning_rate": 2.8540473627203708e-05,
+      "loss": 1.0966,
+      "step": 4435
+    },
+    {
+      "epoch": 1.366101303796008,
+      "grad_norm": 1.065972924232483,
+      "learning_rate": 2.8500598161038057e-05,
+      "loss": 1.0942,
+      "step": 4440
+    },
+    {
+      "epoch": 1.3676397061651475,
+      "grad_norm": 1.105650782585144,
+      "learning_rate": 2.8460713608873323e-05,
+      "loss": 1.2002,
+      "step": 4445
+    },
+    {
+      "epoch": 1.3691781085342871,
+      "grad_norm": 2.50470232963562,
+      "learning_rate": 2.8420820074232086e-05,
+      "loss": 1.0063,
+      "step": 4450
+    },
+    {
+      "epoch": 1.3707165109034267,
+      "grad_norm": 1.2763255834579468,
+      "learning_rate": 2.8380917660660262e-05,
+      "loss": 1.1806,
+      "step": 4455
+    },
+    {
+      "epoch": 1.3722549132725663,
+      "grad_norm": 1.4548883438110352,
+      "learning_rate": 2.8341006471726816e-05,
+      "loss": 1.091,
+      "step": 4460
+    },
+    {
+      "epoch": 1.3737933156417061,
+      "grad_norm": 4.325429916381836,
+      "learning_rate": 2.830108661102346e-05,
+      "loss": 1.1485,
+      "step": 4465
+    },
+    {
+      "epoch": 1.3753317180108457,
+      "grad_norm": 1.4059125185012817,
+      "learning_rate": 2.826115818216444e-05,
+      "loss": 1.109,
+      "step": 4470
+    },
+    {
+      "epoch": 1.3768701203799854,
+      "grad_norm": 1.088011384010315,
+      "learning_rate": 2.822122128878625e-05,
+      "loss": 1.1762,
+      "step": 4475
+    },
+    {
+      "epoch": 1.378408522749125,
+      "grad_norm": 0.8821271061897278,
+      "learning_rate": 2.818127603454732e-05,
+      "loss": 1.0974,
+      "step": 4480
+    },
+    {
+      "epoch": 1.3799469251182646,
+      "grad_norm": 1.2587283849716187,
+      "learning_rate": 2.8141322523127817e-05,
+      "loss": 1.0408,
+      "step": 4485
+    },
+    {
+      "epoch": 1.3814853274874044,
+      "grad_norm": 1.244844675064087,
+      "learning_rate": 2.810136085822931e-05,
+      "loss": 1.1508,
+      "step": 4490
+    },
+    {
+      "epoch": 1.383023729856544,
+      "grad_norm": 1.0243417024612427,
+      "learning_rate": 2.8061391143574545e-05,
+      "loss": 1.14,
+      "step": 4495
+    },
+    {
+      "epoch": 1.3845621322256836,
+      "grad_norm": 1.2699311971664429,
+      "learning_rate": 2.8021413482907176e-05,
+      "loss": 1.264,
+      "step": 4500
+    },
+    {
+      "epoch": 1.3861005345948234,
+      "grad_norm": 1.102189064025879,
+      "learning_rate": 2.798142797999144e-05,
+      "loss": 1.0759,
+      "step": 4505
+    },
+    {
+      "epoch": 1.3876389369639628,
+      "grad_norm": 1.9496086835861206,
+      "learning_rate": 2.794143473861198e-05,
+      "loss": 1.0819,
+      "step": 4510
+    },
+    {
+      "epoch": 1.3891773393331026,
+      "grad_norm": 0.8148252367973328,
+      "learning_rate": 2.7901433862573495e-05,
+      "loss": 1.1767,
+      "step": 4515
+    },
+    {
+      "epoch": 1.3907157417022422,
+      "grad_norm": 1.303236961364746,
+      "learning_rate": 2.786142545570049e-05,
+      "loss": 1.2153,
+      "step": 4520
+    },
+    {
+      "epoch": 1.3922541440713818,
+      "grad_norm": 2.4543964862823486,
+      "learning_rate": 2.782140962183704e-05,
+      "loss": 0.9841,
+      "step": 4525
+    },
+    {
+      "epoch": 1.3937925464405216,
+      "grad_norm": 1.1783264875411987,
+      "learning_rate": 2.7781386464846497e-05,
+      "loss": 1.1419,
+      "step": 4530
+    },
+    {
+      "epoch": 1.3953309488096612,
+      "grad_norm": 1.238843321800232,
+      "learning_rate": 2.7741356088611205e-05,
+      "loss": 1.0705,
+      "step": 4535
+    },
+    {
+      "epoch": 1.3968693511788008,
+      "grad_norm": 1.183280348777771,
+      "learning_rate": 2.7701318597032248e-05,
+      "loss": 1.2047,
+      "step": 4540
+    },
+    {
+      "epoch": 1.3984077535479404,
+      "grad_norm": 1.5718815326690674,
+      "learning_rate": 2.7661274094029193e-05,
+      "loss": 1.2228,
+      "step": 4545
+    },
+    {
+      "epoch": 1.39994615591708,
+      "grad_norm": 1.5735619068145752,
+      "learning_rate": 2.7621222683539792e-05,
+      "loss": 1.2255,
+      "step": 4550
+    },
+    {
+      "epoch": 1.4014845582862199,
+      "grad_norm": 1.1554558277130127,
+      "learning_rate": 2.7581164469519732e-05,
+      "loss": 1.1794,
+      "step": 4555
+    },
+    {
+      "epoch": 1.4030229606553595,
+      "grad_norm": 1.498023509979248,
+      "learning_rate": 2.754109955594235e-05,
+      "loss": 1.1865,
+      "step": 4560
+    },
+    {
+      "epoch": 1.404561363024499,
+      "grad_norm": 1.2338696718215942,
+      "learning_rate": 2.7501028046798387e-05,
+      "loss": 1.0968,
+      "step": 4565
+    },
+    {
+      "epoch": 1.4060997653936387,
+      "grad_norm": 1.8019814491271973,
+      "learning_rate": 2.7460950046095696e-05,
+      "loss": 1.1228,
+      "step": 4570
+    },
+    {
+      "epoch": 1.4076381677627783,
+      "grad_norm": 1.3556607961654663,
+      "learning_rate": 2.742086565785896e-05,
+      "loss": 1.1827,
+      "step": 4575
+    },
+    {
+      "epoch": 1.409176570131918,
+      "grad_norm": 1.5109410285949707,
+      "learning_rate": 2.738077498612949e-05,
+      "loss": 1.1437,
+      "step": 4580
+    },
+    {
+      "epoch": 1.4107149725010577,
+      "grad_norm": 1.0716326236724854,
+      "learning_rate": 2.7340678134964855e-05,
+      "loss": 1.2372,
+      "step": 4585
+    },
+    {
+      "epoch": 1.4122533748701973,
+      "grad_norm": 1.2416647672653198,
+      "learning_rate": 2.7300575208438683e-05,
+      "loss": 1.1694,
+      "step": 4590
+    },
+    {
+      "epoch": 1.413791777239337,
+      "grad_norm": 1.2061866521835327,
+      "learning_rate": 2.7260466310640377e-05,
+      "loss": 1.2479,
+      "step": 4595
+    },
+    {
+      "epoch": 1.4153301796084765,
+      "grad_norm": 1.7378590106964111,
+      "learning_rate": 2.7220351545674834e-05,
+      "loss": 1.1552,
+      "step": 4600
+    },
+    {
+      "epoch": 1.4168685819776163,
+      "grad_norm": 1.3554580211639404,
+      "learning_rate": 2.7180231017662178e-05,
+      "loss": 1.1731,
+      "step": 4605
+    },
+    {
+      "epoch": 1.418406984346756,
+      "grad_norm": 1.5967758893966675,
+      "learning_rate": 2.7140104830737496e-05,
+      "loss": 1.2213,
+      "step": 4610
+    },
+    {
+      "epoch": 1.4199453867158955,
+      "grad_norm": 1.3568000793457031,
+      "learning_rate": 2.709997308905055e-05,
+      "loss": 1.1097,
+      "step": 4615
+    },
+    {
+      "epoch": 1.4214837890850351,
+      "grad_norm": 1.3305199146270752,
+      "learning_rate": 2.705983589676554e-05,
+      "loss": 1.0566,
+      "step": 4620
+    },
+    {
+      "epoch": 1.4230221914541747,
+      "grad_norm": 1.3310613632202148,
+      "learning_rate": 2.7019693358060792e-05,
+      "loss": 1.1266,
+      "step": 4625
+    },
+    {
+      "epoch": 1.4245605938233146,
+      "grad_norm": 1.6512633562088013,
+      "learning_rate": 2.6979545577128522e-05,
+      "loss": 1.1205,
+      "step": 4630
+    },
+    {
+      "epoch": 1.4260989961924542,
+      "grad_norm": 0.985725462436676,
+      "learning_rate": 2.6939392658174568e-05,
+      "loss": 1.1683,
+      "step": 4635
+    },
+    {
+      "epoch": 1.4276373985615938,
+      "grad_norm": 1.284145712852478,
+      "learning_rate": 2.6899234705418052e-05,
+      "loss": 1.1016,
+      "step": 4640
+    },
+    {
+      "epoch": 1.4291758009307334,
+      "grad_norm": 1.1945017576217651,
+      "learning_rate": 2.685907182309122e-05,
+      "loss": 1.1832,
+      "step": 4645
+    },
+    {
+      "epoch": 1.430714203299873,
+      "grad_norm": 1.2088850736618042,
+      "learning_rate": 2.681890411543908e-05,
+      "loss": 1.1577,
+      "step": 4650
+    },
+    {
+      "epoch": 1.4322526056690128,
+      "grad_norm": 2.208871841430664,
+      "learning_rate": 2.6778731686719178e-05,
+      "loss": 1.2002,
+      "step": 4655
+    },
+    {
+      "epoch": 1.4337910080381524,
+      "grad_norm": 0.9800326824188232,
+      "learning_rate": 2.6738554641201298e-05,
+      "loss": 1.1888,
+      "step": 4660
+    },
+    {
+      "epoch": 1.435329410407292,
+      "grad_norm": 5.183529853820801,
+      "learning_rate": 2.669837308316723e-05,
+      "loss": 1.0284,
+      "step": 4665
+    },
+    {
+      "epoch": 1.4368678127764316,
+      "grad_norm": 1.1594517230987549,
+      "learning_rate": 2.6658187116910455e-05,
+      "loss": 1.1301,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4384062151455712,
+      "grad_norm": 1.1288197040557861,
+      "learning_rate": 2.6617996846735904e-05,
+      "loss": 1.1265,
+      "step": 4675
+    },
+    {
+      "epoch": 1.439944617514711,
+      "grad_norm": 2.477808952331543,
+      "learning_rate": 2.6577802376959698e-05,
+      "loss": 1.1718,
+      "step": 4680
+    },
+    {
+      "epoch": 1.4414830198838506,
+      "grad_norm": 1.1020774841308594,
+      "learning_rate": 2.653760381190881e-05,
+      "loss": 1.1919,
+      "step": 4685
+    },
+    {
+      "epoch": 1.4430214222529902,
+      "grad_norm": 1.5409555435180664,
+      "learning_rate": 2.64974012559209e-05,
+      "loss": 1.1858,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4445598246221298,
+      "grad_norm": 1.6623104810714722,
+      "learning_rate": 2.6457194813343948e-05,
+      "loss": 1.2185,
+      "step": 4695
+    },
+    {
+      "epoch": 1.4460982269912694,
+      "grad_norm": 1.2908927202224731,
+      "learning_rate": 2.641698458853603e-05,
+      "loss": 1.0986,
+      "step": 4700
+    },
+    {
+      "epoch": 1.4476366293604093,
+      "grad_norm": 1.120782732963562,
+      "learning_rate": 2.637677068586505e-05,
+      "loss": 1.2856,
+      "step": 4705
+    },
+    {
+      "epoch": 1.4491750317295489,
+      "grad_norm": 0.8865787386894226,
+      "learning_rate": 2.6336553209708447e-05,
+      "loss": 1.2691,
+      "step": 4710
+    },
+    {
+      "epoch": 1.4507134340986885,
+      "grad_norm": 1.8774168491363525,
+      "learning_rate": 2.6296332264452934e-05,
+      "loss": 1.1015,
+      "step": 4715
+    },
+    {
+      "epoch": 1.4522518364678283,
+      "grad_norm": 1.0902308225631714,
+      "learning_rate": 2.6256107954494242e-05,
+      "loss": 1.1832,
+      "step": 4720
+    },
+    {
+      "epoch": 1.4537902388369677,
+      "grad_norm": 1.2977639436721802,
+      "learning_rate": 2.6215880384236818e-05,
+      "loss": 1.2232,
+      "step": 4725
+    },
+    {
+      "epoch": 1.4553286412061075,
+      "grad_norm": 1.2228929996490479,
+      "learning_rate": 2.6175649658093586e-05,
+      "loss": 1.0803,
+      "step": 4730
+    },
+    {
+      "epoch": 1.456867043575247,
+      "grad_norm": 1.099314570426941,
+      "learning_rate": 2.6135415880485654e-05,
+      "loss": 1.1727,
+      "step": 4735
+    },
+    {
+      "epoch": 1.4584054459443867,
+      "grad_norm": 1.762929081916809,
+      "learning_rate": 2.609517915584204e-05,
+      "loss": 1.2126,
+      "step": 4740
+    },
+    {
+      "epoch": 1.4599438483135265,
+      "grad_norm": 1.0587412118911743,
+      "learning_rate": 2.6054939588599448e-05,
+      "loss": 1.1546,
+      "step": 4745
+    },
+    {
+      "epoch": 1.4614822506826661,
+      "grad_norm": 1.0290122032165527,
+      "learning_rate": 2.6014697283201907e-05,
+      "loss": 1.0129,
+      "step": 4750
+    },
+    {
+      "epoch": 1.4630206530518057,
+      "grad_norm": 1.801530361175537,
+      "learning_rate": 2.597445234410058e-05,
+      "loss": 1.2372,
+      "step": 4755
+    },
+    {
+      "epoch": 1.4645590554209453,
+      "grad_norm": 2.0911014080047607,
+      "learning_rate": 2.5934204875753494e-05,
+      "loss": 1.0657,
+      "step": 4760
+    },
+    {
+      "epoch": 1.466097457790085,
+      "grad_norm": 1.0208872556686401,
+      "learning_rate": 2.589395498262519e-05,
+      "loss": 1.2039,
+      "step": 4765
+    },
+    {
+      "epoch": 1.4676358601592248,
+      "grad_norm": 1.2312899827957153,
+      "learning_rate": 2.5853702769186528e-05,
+      "loss": 1.1466,
+      "step": 4770
+    },
+    {
+      "epoch": 1.4691742625283644,
+      "grad_norm": 1.3418748378753662,
+      "learning_rate": 2.5813448339914393e-05,
+      "loss": 1.1439,
+      "step": 4775
+    },
+    {
+      "epoch": 1.470712664897504,
+      "grad_norm": 1.0246156454086304,
+      "learning_rate": 2.5773191799291417e-05,
+      "loss": 1.1856,
+      "step": 4780
+    },
+    {
+      "epoch": 1.4722510672666436,
+      "grad_norm": 1.1938049793243408,
+      "learning_rate": 2.5732933251805713e-05,
+      "loss": 1.1144,
+      "step": 4785
+    },
+    {
+      "epoch": 1.4737894696357832,
+      "grad_norm": 2.593672037124634,
+      "learning_rate": 2.569267280195059e-05,
+      "loss": 1.1321,
+      "step": 4790
+    },
+    {
+      "epoch": 1.475327872004923,
+      "grad_norm": 1.1651784181594849,
+      "learning_rate": 2.5652410554224322e-05,
+      "loss": 1.1729,
+      "step": 4795
+    },
+    {
+      "epoch": 1.4768662743740626,
+      "grad_norm": 1.3578174114227295,
+      "learning_rate": 2.5612146613129828e-05,
+      "loss": 1.0952,
+      "step": 4800
+    },
+    {
+      "epoch": 1.4784046767432022,
+      "grad_norm": 2.145015001296997,
+      "learning_rate": 2.5571881083174427e-05,
+      "loss": 1.0512,
+      "step": 4805
+    },
+    {
+      "epoch": 1.4799430791123418,
+      "grad_norm": 1.4912582635879517,
+      "learning_rate": 2.553161406886955e-05,
+      "loss": 1.0594,
+      "step": 4810
+    },
+    {
+      "epoch": 1.4814814814814814,
+      "grad_norm": 1.3633592128753662,
+      "learning_rate": 2.5491345674730522e-05,
+      "loss": 1.2463,
+      "step": 4815
+    },
+    {
+      "epoch": 1.4830198838506212,
+      "grad_norm": 1.1500188112258911,
+      "learning_rate": 2.5451076005276197e-05,
+      "loss": 1.1274,
+      "step": 4820
+    },
+    {
+      "epoch": 1.4845582862197608,
+      "grad_norm": 1.1924505233764648,
+      "learning_rate": 2.5410805165028772e-05,
+      "loss": 1.0678,
+      "step": 4825
+    },
+    {
+      "epoch": 1.4860966885889004,
+      "grad_norm": 2.115426540374756,
+      "learning_rate": 2.537053325851348e-05,
+      "loss": 1.2476,
+      "step": 4830
+    },
+    {
+      "epoch": 1.48763509095804,
+      "grad_norm": 1.191912055015564,
+      "learning_rate": 2.5330260390258302e-05,
+      "loss": 1.1183,
+      "step": 4835
+    },
+    {
+      "epoch": 1.4891734933271796,
+      "grad_norm": 0.8928827047348022,
+      "learning_rate": 2.5289986664793743e-05,
+      "loss": 1.1317,
+      "step": 4840
+    },
+    {
+      "epoch": 1.4907118956963195,
+      "grad_norm": 1.3792165517807007,
+      "learning_rate": 2.52497121866525e-05,
+      "loss": 1.2063,
+      "step": 4845
+    },
+    {
+      "epoch": 1.492250298065459,
+      "grad_norm": 1.0137156248092651,
+      "learning_rate": 2.520943706036927e-05,
+      "loss": 1.1008,
+      "step": 4850
+    },
+    {
+      "epoch": 1.4937887004345987,
+      "grad_norm": 1.1347684860229492,
+      "learning_rate": 2.5169161390480382e-05,
+      "loss": 1.1237,
+      "step": 4855
+    },
+    {
+      "epoch": 1.4953271028037383,
+      "grad_norm": 1.1808063983917236,
+      "learning_rate": 2.5128885281523606e-05,
+      "loss": 1.242,
+      "step": 4860
+    },
+    {
+      "epoch": 1.4968655051728779,
+      "grad_norm": 1.2332245111465454,
+      "learning_rate": 2.508860883803784e-05,
+      "loss": 1.1619,
+      "step": 4865
+    },
+    {
+      "epoch": 1.4984039075420177,
+      "grad_norm": 1.0898313522338867,
+      "learning_rate": 2.5048332164562872e-05,
+      "loss": 1.2496,
+      "step": 4870
+    },
+    {
+      "epoch": 1.4999423099111573,
+      "grad_norm": 1.0363882780075073,
+      "learning_rate": 2.500805536563905e-05,
+      "loss": 1.1321,
+      "step": 4875
+    },
+    {
+      "epoch": 1.501480712280297,
+      "grad_norm": 1.2297104597091675,
+      "learning_rate": 2.4967778545807074e-05,
+      "loss": 1.1335,
+      "step": 4880
+    },
+    {
+      "epoch": 1.5030191146494367,
+      "grad_norm": 1.6471554040908813,
+      "learning_rate": 2.4927501809607692e-05,
+      "loss": 1.1601,
+      "step": 4885
+    },
+    {
+      "epoch": 1.504557517018576,
+      "grad_norm": 1.9324811697006226,
+      "learning_rate": 2.4887225261581436e-05,
+      "loss": 1.1363,
+      "step": 4890
+    },
+    {
+      "epoch": 1.506095919387716,
+      "grad_norm": 1.3468040227890015,
+      "learning_rate": 2.4846949006268344e-05,
+      "loss": 1.2231,
+      "step": 4895
+    },
+    {
+      "epoch": 1.5076343217568555,
+      "grad_norm": 1.3519035577774048,
+      "learning_rate": 2.4806673148207693e-05,
+      "loss": 1.2626,
+      "step": 4900
+    },
+    {
+      "epoch": 1.5091727241259951,
+      "grad_norm": 0.9620422720909119,
+      "learning_rate": 2.476639779193776e-05,
+      "loss": 1.2636,
+      "step": 4905
+    },
+    {
+      "epoch": 1.510711126495135,
+      "grad_norm": 1.1294174194335938,
+      "learning_rate": 2.4726123041995463e-05,
+      "loss": 1.234,
+      "step": 4910
+    },
+    {
+      "epoch": 1.5122495288642743,
+      "grad_norm": 1.0007851123809814,
+      "learning_rate": 2.4685849002916183e-05,
+      "loss": 1.1924,
+      "step": 4915
+    },
+    {
+      "epoch": 1.5137879312334142,
+      "grad_norm": 1.8807976245880127,
+      "learning_rate": 2.4645575779233464e-05,
+      "loss": 1.2864,
+      "step": 4920
+    },
+    {
+      "epoch": 1.5153263336025538,
+      "grad_norm": 2.8548858165740967,
+      "learning_rate": 2.460530347547871e-05,
+      "loss": 1.1517,
+      "step": 4925
+    },
+    {
+      "epoch": 1.5168647359716934,
+      "grad_norm": 1.5825186967849731,
+      "learning_rate": 2.4565032196180952e-05,
+      "loss": 1.2202,
+      "step": 4930
+    },
+    {
+      "epoch": 1.5184031383408332,
+      "grad_norm": 1.0249121189117432,
+      "learning_rate": 2.4524762045866555e-05,
+      "loss": 1.2323,
+      "step": 4935
+    },
+    {
+      "epoch": 1.5199415407099726,
+      "grad_norm": 0.9302557706832886,
+      "learning_rate": 2.4484493129058944e-05,
+      "loss": 1.2025,
+      "step": 4940
+    },
+    {
+      "epoch": 1.5214799430791124,
+      "grad_norm": 2.6338999271392822,
+      "learning_rate": 2.444422555027837e-05,
+      "loss": 1.1265,
+      "step": 4945
+    },
+    {
+      "epoch": 1.523018345448252,
+      "grad_norm": 1.4891122579574585,
+      "learning_rate": 2.4403959414041583e-05,
+      "loss": 1.0759,
+      "step": 4950
+    },
+    {
+      "epoch": 1.5245567478173916,
+      "grad_norm": 1.1123028993606567,
+      "learning_rate": 2.4363694824861615e-05,
+      "loss": 1.0696,
+      "step": 4955
+    },
+    {
+      "epoch": 1.5260951501865314,
+      "grad_norm": 2.104362964630127,
+      "learning_rate": 2.4323431887247446e-05,
+      "loss": 1.0459,
+      "step": 4960
+    },
+    {
+      "epoch": 1.5276335525556708,
+      "grad_norm": 0.9691029787063599,
+      "learning_rate": 2.4283170705703812e-05,
+      "loss": 1.2534,
+      "step": 4965
+    },
+    {
+      "epoch": 1.5291719549248106,
+      "grad_norm": 1.148000717163086,
+      "learning_rate": 2.424291138473085e-05,
+      "loss": 1.2542,
+      "step": 4970
+    },
+    {
+      "epoch": 1.5307103572939502,
+      "grad_norm": 0.9757530093193054,
+      "learning_rate": 2.4202654028823913e-05,
+      "loss": 1.087,
+      "step": 4975
+    },
+    {
+      "epoch": 1.5322487596630898,
+      "grad_norm": 1.5252931118011475,
+      "learning_rate": 2.4162398742473214e-05,
+      "loss": 1.2709,
+      "step": 4980
+    },
+    {
+      "epoch": 1.5337871620322296,
+      "grad_norm": 1.3991789817810059,
+      "learning_rate": 2.4122145630163616e-05,
+      "loss": 1.0422,
+      "step": 4985
+    },
+    {
+      "epoch": 1.535325564401369,
+      "grad_norm": 1.2912970781326294,
+      "learning_rate": 2.408189479637432e-05,
+      "loss": 1.1019,
+      "step": 4990
+    },
+    {
+      "epoch": 1.5368639667705088,
+      "grad_norm": 1.0709604024887085,
+      "learning_rate": 2.4041646345578637e-05,
+      "loss": 1.1101,
+      "step": 4995
+    },
+    {
+      "epoch": 1.5384023691396485,
+      "grad_norm": 1.062154769897461,
+      "learning_rate": 2.4001400382243675e-05,
+      "loss": 1.1262,
+      "step": 5000
+    },
+    {
+      "epoch": 1.539940771508788,
+      "grad_norm": 1.4915518760681152,
+      "learning_rate": 2.3961157010830095e-05,
+      "loss": 1.2232,
+      "step": 5005
+    },
+    {
+      "epoch": 1.5414791738779279,
+      "grad_norm": 1.2963896989822388,
+      "learning_rate": 2.3920916335791833e-05,
+      "loss": 1.1251,
+      "step": 5010
+    },
+    {
+      "epoch": 1.5430175762470673,
+      "grad_norm": 1.115599274635315,
+      "learning_rate": 2.3880678461575805e-05,
+      "loss": 1.1748,
+      "step": 5015
+    },
+    {
+      "epoch": 1.544555978616207,
+      "grad_norm": 1.3244112730026245,
+      "learning_rate": 2.3840443492621674e-05,
+      "loss": 1.1706,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5460943809853467,
+      "grad_norm": 1.903247356414795,
+      "learning_rate": 2.380021153336158e-05,
+      "loss": 1.1706,
+      "step": 5025
+    },
+    {
+      "epoch": 1.5476327833544863,
+      "grad_norm": 2.5160722732543945,
+      "learning_rate": 2.375998268821982e-05,
+      "loss": 1.0949,
+      "step": 5030
+    },
+    {
+      "epoch": 1.549171185723626,
+      "grad_norm": 1.065032958984375,
+      "learning_rate": 2.371975706161262e-05,
+      "loss": 1.1651,
+      "step": 5035
+    },
+    {
+      "epoch": 1.5507095880927657,
+      "grad_norm": 1.1798903942108154,
+      "learning_rate": 2.3679534757947862e-05,
+      "loss": 1.0752,
+      "step": 5040
+    },
+    {
+      "epoch": 1.5522479904619053,
+      "grad_norm": 1.417855143547058,
+      "learning_rate": 2.3639315881624777e-05,
+      "loss": 1.204,
+      "step": 5045
+    },
+    {
+      "epoch": 1.553786392831045,
+      "grad_norm": 3.4984400272369385,
+      "learning_rate": 2.3599100537033728e-05,
+      "loss": 1.2201,
+      "step": 5050
+    },
+    {
+      "epoch": 1.5553247952001845,
+      "grad_norm": 1.7004238367080688,
+      "learning_rate": 2.35588888285559e-05,
+      "loss": 1.2381,
+      "step": 5055
+    },
+    {
+      "epoch": 1.5568631975693243,
+      "grad_norm": 1.1954479217529297,
+      "learning_rate": 2.3518680860563026e-05,
+      "loss": 1.291,
+      "step": 5060
+    },
+    {
+      "epoch": 1.558401599938464,
+      "grad_norm": 1.0503448247909546,
+      "learning_rate": 2.3478476737417177e-05,
+      "loss": 1.114,
+      "step": 5065
+    },
+    {
+      "epoch": 1.5599400023076035,
+      "grad_norm": 1.9155570268630981,
+      "learning_rate": 2.3438276563470382e-05,
+      "loss": 1.2755,
+      "step": 5070
+    },
+    {
+      "epoch": 1.5614784046767431,
+      "grad_norm": 1.2648513317108154,
+      "learning_rate": 2.3398080443064453e-05,
+      "loss": 1.1355,
+      "step": 5075
+    },
+    {
+      "epoch": 1.5630168070458828,
+      "grad_norm": 1.9838799238204956,
+      "learning_rate": 2.335788848053069e-05,
+      "loss": 1.1396,
+      "step": 5080
+    },
+    {
+      "epoch": 1.5645552094150226,
+      "grad_norm": 0.9827559590339661,
+      "learning_rate": 2.331770078018958e-05,
+      "loss": 1.1287,
+      "step": 5085
+    },
+    {
+      "epoch": 1.5660936117841622,
+      "grad_norm": 1.3725968599319458,
+      "learning_rate": 2.3277517446350566e-05,
+      "loss": 1.1653,
+      "step": 5090
+    },
+    {
+      "epoch": 1.5676320141533018,
+      "grad_norm": 1.1363013982772827,
+      "learning_rate": 2.3237338583311742e-05,
+      "loss": 1.1403,
+      "step": 5095
+    },
+    {
+      "epoch": 1.5691704165224416,
+      "grad_norm": 1.514285922050476,
+      "learning_rate": 2.3197164295359593e-05,
+      "loss": 1.1604,
+      "step": 5100
+    },
+    {
+      "epoch": 1.570708818891581,
+      "grad_norm": 1.5780757665634155,
+      "learning_rate": 2.3156994686768753e-05,
+      "loss": 1.1634,
+      "step": 5105
+    },
+    {
+      "epoch": 1.5722472212607208,
+      "grad_norm": 2.1901051998138428,
+      "learning_rate": 2.3116829861801686e-05,
+      "loss": 1.0262,
+      "step": 5110
+    },
+    {
+      "epoch": 1.5737856236298604,
+      "grad_norm": 0.9369887113571167,
+      "learning_rate": 2.307666992470845e-05,
+      "loss": 1.1561,
+      "step": 5115
+    },
+    {
+      "epoch": 1.575324025999,
+      "grad_norm": 1.2366549968719482,
+      "learning_rate": 2.3036514979726442e-05,
+      "loss": 1.1065,
+      "step": 5120
+    },
+    {
+      "epoch": 1.5768624283681398,
+      "grad_norm": 1.257912516593933,
+      "learning_rate": 2.2996365131080046e-05,
+      "loss": 1.1071,
+      "step": 5125
+    },
+    {
+      "epoch": 1.5784008307372792,
+      "grad_norm": 1.1271820068359375,
+      "learning_rate": 2.295622048298045e-05,
+      "loss": 1.1964,
+      "step": 5130
+    },
+    {
+      "epoch": 1.579939233106419,
+      "grad_norm": 1.9496513605117798,
+      "learning_rate": 2.2916081139625362e-05,
+      "loss": 1.1061,
+      "step": 5135
+    },
+    {
+      "epoch": 1.5814776354755586,
+      "grad_norm": 1.8635252714157104,
+      "learning_rate": 2.287594720519869e-05,
+      "loss": 1.0035,
+      "step": 5140
+    },
+    {
+      "epoch": 1.5830160378446982,
+      "grad_norm": 0.9544889330863953,
+      "learning_rate": 2.2835818783870312e-05,
+      "loss": 1.2609,
+      "step": 5145
+    },
+    {
+      "epoch": 1.584554440213838,
+      "grad_norm": 1.032304048538208,
+      "learning_rate": 2.2795695979795813e-05,
+      "loss": 1.2609,
+      "step": 5150
+    },
+    {
+      "epoch": 1.5860928425829774,
+      "grad_norm": 1.1428989171981812,
+      "learning_rate": 2.275557889711617e-05,
+      "loss": 1.1503,
+      "step": 5155
+    },
+    {
+      "epoch": 1.5876312449521173,
+      "grad_norm": 1.0252608060836792,
+      "learning_rate": 2.271546763995752e-05,
+      "loss": 1.0164,
+      "step": 5160
+    },
+    {
+      "epoch": 1.5891696473212569,
+      "grad_norm": 0.9309441447257996,
+      "learning_rate": 2.2675362312430894e-05,
+      "loss": 1.118,
+      "step": 5165
+    },
+    {
+      "epoch": 1.5907080496903965,
+      "grad_norm": 1.2354007959365845,
+      "learning_rate": 2.2635263018631915e-05,
+      "loss": 1.1657,
+      "step": 5170
+    },
+    {
+      "epoch": 1.5922464520595363,
+      "grad_norm": 1.468483805656433,
+      "learning_rate": 2.2595169862640568e-05,
+      "loss": 1.1834,
+      "step": 5175
+    },
+    {
+      "epoch": 1.5937848544286757,
+      "grad_norm": 1.2876496315002441,
+      "learning_rate": 2.255508294852086e-05,
+      "loss": 1.152,
+      "step": 5180
+    },
+    {
+      "epoch": 1.5953232567978155,
+      "grad_norm": 1.2103363275527954,
+      "learning_rate": 2.2515002380320655e-05,
+      "loss": 1.3024,
+      "step": 5185
+    },
+    {
+      "epoch": 1.596861659166955,
+      "grad_norm": 1.977842926979065,
+      "learning_rate": 2.2474928262071307e-05,
+      "loss": 1.1682,
+      "step": 5190
+    },
+    {
+      "epoch": 1.5984000615360947,
+      "grad_norm": 2.3454501628875732,
+      "learning_rate": 2.243486069778744e-05,
+      "loss": 1.1026,
+      "step": 5195
+    },
+    {
+      "epoch": 1.5999384639052345,
+      "grad_norm": 1.8084458112716675,
+      "learning_rate": 2.239479979146667e-05,
+      "loss": 1.1451,
+      "step": 5200
+    },
+    {
+      "epoch": 1.601476866274374,
+      "grad_norm": 1.0935570001602173,
+      "learning_rate": 2.235474564708933e-05,
+      "loss": 0.9931,
+      "step": 5205
+    },
+    {
+      "epoch": 1.6030152686435137,
+      "grad_norm": 1.21275794506073,
+      "learning_rate": 2.2314698368618198e-05,
+      "loss": 1.2305,
+      "step": 5210
+    },
+    {
+      "epoch": 1.6045536710126533,
+      "grad_norm": 1.6089550256729126,
+      "learning_rate": 2.227465805999823e-05,
+      "loss": 1.1902,
+      "step": 5215
+    },
+    {
+      "epoch": 1.606092073381793,
+      "grad_norm": 1.717024803161621,
+      "learning_rate": 2.2234624825156293e-05,
+      "loss": 1.204,
+      "step": 5220
+    },
+    {
+      "epoch": 1.6076304757509328,
+      "grad_norm": 1.2060118913650513,
+      "learning_rate": 2.219459876800091e-05,
+      "loss": 1.1242,
+      "step": 5225
+    },
+    {
+      "epoch": 1.6091688781200721,
+      "grad_norm": 0.8598418831825256,
+      "learning_rate": 2.2154579992421964e-05,
+      "loss": 1.0447,
+      "step": 5230
+    },
+    {
+      "epoch": 1.610707280489212,
+      "grad_norm": 0.9737907648086548,
+      "learning_rate": 2.2114568602290406e-05,
+      "loss": 1.1645,
+      "step": 5235
+    },
+    {
+      "epoch": 1.6122456828583516,
+      "grad_norm": 1.1507982015609741,
+      "learning_rate": 2.2074564701458065e-05,
+      "loss": 1.1268,
+      "step": 5240
+    },
+    {
+      "epoch": 1.6137840852274912,
+      "grad_norm": 1.7262670993804932,
+      "learning_rate": 2.2034568393757313e-05,
+      "loss": 1.1256,
+      "step": 5245
+    },
+    {
+      "epoch": 1.615322487596631,
+      "grad_norm": 1.7809712886810303,
+      "learning_rate": 2.1994579783000804e-05,
+      "loss": 1.2046,
+      "step": 5250
+    },
+    {
+      "epoch": 1.6168608899657706,
+      "grad_norm": 2.4736218452453613,
+      "learning_rate": 2.1954598972981237e-05,
+      "loss": 1.1503,
+      "step": 5255
+    },
+    {
+      "epoch": 1.6183992923349102,
+      "grad_norm": 1.1869739294052124,
+      "learning_rate": 2.1914626067471032e-05,
+      "loss": 1.0835,
+      "step": 5260
+    },
+    {
+      "epoch": 1.6199376947040498,
+      "grad_norm": 1.188988447189331,
+      "learning_rate": 2.187466117022212e-05,
+      "loss": 1.0705,
+      "step": 5265
+    },
+    {
+      "epoch": 1.6214760970731894,
+      "grad_norm": 1.4258966445922852,
+      "learning_rate": 2.183470438496563e-05,
+      "loss": 1.1984,
+      "step": 5270
+    },
+    {
+      "epoch": 1.6230144994423292,
+      "grad_norm": 0.839988112449646,
+      "learning_rate": 2.1794755815411642e-05,
+      "loss": 1.1258,
+      "step": 5275
+    },
+    {
+      "epoch": 1.6245529018114688,
+      "grad_norm": 1.4149938821792603,
+      "learning_rate": 2.175481556524892e-05,
+      "loss": 1.1313,
+      "step": 5280
+    },
+    {
+      "epoch": 1.6260913041806084,
+      "grad_norm": 1.8420780897140503,
+      "learning_rate": 2.1714883738144627e-05,
+      "loss": 1.1325,
+      "step": 5285
+    },
+    {
+      "epoch": 1.627629706549748,
+      "grad_norm": 1.2860522270202637,
+      "learning_rate": 2.1674960437744044e-05,
+      "loss": 1.2157,
+      "step": 5290
+    },
+    {
+      "epoch": 1.6291681089188876,
+      "grad_norm": 1.9698175191879272,
+      "learning_rate": 2.1635045767670356e-05,
+      "loss": 1.1402,
+      "step": 5295
+    },
+    {
+      "epoch": 1.6307065112880275,
+      "grad_norm": 1.716629147529602,
+      "learning_rate": 2.1595139831524326e-05,
+      "loss": 1.1031,
+      "step": 5300
+    },
+    {
+      "epoch": 1.632244913657167,
+      "grad_norm": 1.4469795227050781,
+      "learning_rate": 2.155524273288405e-05,
+      "loss": 1.1826,
+      "step": 5305
+    },
+    {
+      "epoch": 1.6337833160263067,
+      "grad_norm": 1.2318928241729736,
+      "learning_rate": 2.1515354575304695e-05,
+      "loss": 1.1256,
+      "step": 5310
+    },
+    {
+      "epoch": 1.6353217183954465,
+      "grad_norm": 1.3652349710464478,
+      "learning_rate": 2.1475475462318202e-05,
+      "loss": 1.2348,
+      "step": 5315
+    },
+    {
+      "epoch": 1.6368601207645859,
+      "grad_norm": 1.40614914894104,
+      "learning_rate": 2.1435605497433057e-05,
+      "loss": 1.1109,
+      "step": 5320
+    },
+    {
+      "epoch": 1.6383985231337257,
+      "grad_norm": 1.7324262857437134,
+      "learning_rate": 2.139574478413398e-05,
+      "loss": 1.0686,
+      "step": 5325
+    },
+    {
+      "epoch": 1.6399369255028653,
+      "grad_norm": 1.711517333984375,
+      "learning_rate": 2.135589342588171e-05,
+      "loss": 0.9758,
+      "step": 5330
+    },
+    {
+      "epoch": 1.641475327872005,
+      "grad_norm": 1.4670315980911255,
+      "learning_rate": 2.1316051526112672e-05,
+      "loss": 1.0558,
+      "step": 5335
+    },
+    {
+      "epoch": 1.6430137302411447,
+      "grad_norm": 1.625055193901062,
+      "learning_rate": 2.1276219188238768e-05,
+      "loss": 1.1067,
+      "step": 5340
+    },
+    {
+      "epoch": 1.644552132610284,
+      "grad_norm": 2.5813450813293457,
+      "learning_rate": 2.1236396515647046e-05,
+      "loss": 1.0579,
+      "step": 5345
+    },
+    {
+      "epoch": 1.646090534979424,
+      "grad_norm": 1.4112730026245117,
+      "learning_rate": 2.1196583611699503e-05,
+      "loss": 1.2685,
+      "step": 5350
+    },
+    {
+      "epoch": 1.6476289373485635,
+      "grad_norm": 0.8786386251449585,
+      "learning_rate": 2.1156780579732764e-05,
+      "loss": 1.1172,
+      "step": 5355
+    },
+    {
+      "epoch": 1.6491673397177031,
+      "grad_norm": 0.9437426328659058,
+      "learning_rate": 2.111698752305783e-05,
+      "loss": 1.0846,
+      "step": 5360
+    },
+    {
+      "epoch": 1.650705742086843,
+      "grad_norm": 1.353835940361023,
+      "learning_rate": 2.1077204544959825e-05,
+      "loss": 1.1903,
+      "step": 5365
+    },
+    {
+      "epoch": 1.6522441444559823,
+      "grad_norm": 1.4217884540557861,
+      "learning_rate": 2.1037431748697688e-05,
+      "loss": 1.0995,
+      "step": 5370
+    },
+    {
+      "epoch": 1.6537825468251222,
+      "grad_norm": 1.296500325202942,
+      "learning_rate": 2.099766923750395e-05,
+      "loss": 1.1283,
+      "step": 5375
+    },
+    {
+      "epoch": 1.6553209491942618,
+      "grad_norm": 1.42235267162323,
+      "learning_rate": 2.095791711458444e-05,
+      "loss": 1.1127,
+      "step": 5380
+    },
+    {
+      "epoch": 1.6568593515634014,
+      "grad_norm": 1.1460708379745483,
+      "learning_rate": 2.0918175483118036e-05,
+      "loss": 1.0967,
+      "step": 5385
+    },
+    {
+      "epoch": 1.6583977539325412,
+      "grad_norm": 1.1948250532150269,
+      "learning_rate": 2.0878444446256364e-05,
+      "loss": 1.2248,
+      "step": 5390
+    },
+    {
+      "epoch": 1.6599361563016806,
+      "grad_norm": 0.8646104335784912,
+      "learning_rate": 2.083872410712357e-05,
+      "loss": 1.1827,
+      "step": 5395
+    },
+    {
+      "epoch": 1.6614745586708204,
+      "grad_norm": 1.0145443677902222,
+      "learning_rate": 2.079901456881601e-05,
+      "loss": 1.0875,
+      "step": 5400
+    },
+    {
+      "epoch": 1.66301296103996,
+      "grad_norm": 1.4740755558013916,
+      "learning_rate": 2.075931593440203e-05,
+      "loss": 1.1525,
+      "step": 5405
+    },
+    {
+      "epoch": 1.6645513634090996,
+      "grad_norm": 1.4376249313354492,
+      "learning_rate": 2.0719628306921664e-05,
+      "loss": 1.1863,
+      "step": 5410
+    },
+    {
+      "epoch": 1.6660897657782394,
+      "grad_norm": 1.4571655988693237,
+      "learning_rate": 2.067995178938638e-05,
+      "loss": 0.8931,
+      "step": 5415
+    },
+    {
+      "epoch": 1.6676281681473788,
+      "grad_norm": 1.8079279661178589,
+      "learning_rate": 2.0640286484778804e-05,
+      "loss": 0.9931,
+      "step": 5420
+    },
+    {
+      "epoch": 1.6691665705165186,
+      "grad_norm": 1.1777446269989014,
+      "learning_rate": 2.0600632496052457e-05,
+      "loss": 1.2278,
+      "step": 5425
+    },
+    {
+      "epoch": 1.6707049728856582,
+      "grad_norm": 1.0672292709350586,
+      "learning_rate": 2.05609899261315e-05,
+      "loss": 1.2236,
+      "step": 5430
+    },
+    {
+      "epoch": 1.6722433752547978,
+      "grad_norm": 1.5936694145202637,
+      "learning_rate": 2.0521358877910444e-05,
+      "loss": 1.2029,
+      "step": 5435
+    },
+    {
+      "epoch": 1.6737817776239376,
+      "grad_norm": 1.6731094121932983,
+      "learning_rate": 2.0481739454253904e-05,
+      "loss": 1.1809,
+      "step": 5440
+    },
+    {
+      "epoch": 1.675320179993077,
+      "grad_norm": 2.1021690368652344,
+      "learning_rate": 2.044213175799632e-05,
+      "loss": 1.1373,
+      "step": 5445
+    },
+    {
+      "epoch": 1.6768585823622169,
+      "grad_norm": 2.2051169872283936,
+      "learning_rate": 2.0402535891941695e-05,
+      "loss": 1.1215,
+      "step": 5450
+    },
+    {
+      "epoch": 1.6783969847313565,
+      "grad_norm": 2.0235435962677,
+      "learning_rate": 2.0362951958863306e-05,
+      "loss": 1.0841,
+      "step": 5455
+    },
+    {
+      "epoch": 1.679935387100496,
+      "grad_norm": 1.1775131225585938,
+      "learning_rate": 2.0323380061503494e-05,
+      "loss": 1.0182,
+      "step": 5460
+    },
+    {
+      "epoch": 1.6814737894696359,
+      "grad_norm": 1.5589635372161865,
+      "learning_rate": 2.0283820302573327e-05,
+      "loss": 1.1266,
+      "step": 5465
+    },
+    {
+      "epoch": 1.6830121918387755,
+      "grad_norm": 1.2760752439498901,
+      "learning_rate": 2.024427278475239e-05,
+      "loss": 1.0757,
+      "step": 5470
+    },
+    {
+      "epoch": 1.684550594207915,
+      "grad_norm": 1.4246011972427368,
+      "learning_rate": 2.0204737610688482e-05,
+      "loss": 1.0888,
+      "step": 5475
+    },
+    {
+      "epoch": 1.6860889965770547,
+      "grad_norm": 1.7687841653823853,
+      "learning_rate": 2.0165214882997363e-05,
+      "loss": 1.1195,
+      "step": 5480
+    },
+    {
+      "epoch": 1.6876273989461943,
+      "grad_norm": 1.873183012008667,
+      "learning_rate": 2.012570470426249e-05,
+      "loss": 1.2103,
+      "step": 5485
+    },
+    {
+      "epoch": 1.6891658013153341,
+      "grad_norm": 1.3162018060684204,
+      "learning_rate": 2.0086207177034765e-05,
+      "loss": 1.1606,
+      "step": 5490
+    },
+    {
+      "epoch": 1.6907042036844737,
+      "grad_norm": 1.9198535680770874,
+      "learning_rate": 2.0046722403832227e-05,
+      "loss": 1.1806,
+      "step": 5495
+    },
+    {
+      "epoch": 1.6922426060536133,
+      "grad_norm": 1.908752202987671,
+      "learning_rate": 2.000725048713983e-05,
+      "loss": 1.2071,
+      "step": 5500
+    },
+    {
+      "epoch": 1.693781008422753,
+      "grad_norm": 1.1114752292633057,
+      "learning_rate": 1.996779152940914e-05,
+      "loss": 1.1524,
+      "step": 5505
+    },
+    {
+      "epoch": 1.6953194107918925,
+      "grad_norm": 1.055211067199707,
+      "learning_rate": 1.99283456330581e-05,
+      "loss": 1.1581,
+      "step": 5510
+    },
+    {
+      "epoch": 1.6968578131610323,
+      "grad_norm": 3.748779773712158,
+      "learning_rate": 1.988891290047075e-05,
+      "loss": 1.1191,
+      "step": 5515
+    },
+    {
+      "epoch": 1.698396215530172,
+      "grad_norm": 1.2784082889556885,
+      "learning_rate": 1.9849493433996963e-05,
+      "loss": 1.0007,
+      "step": 5520
+    },
+    {
+      "epoch": 1.6999346178993116,
+      "grad_norm": 1.2205626964569092,
+      "learning_rate": 1.9810087335952172e-05,
+      "loss": 1.1451,
+      "step": 5525
+    },
+    {
+      "epoch": 1.7014730202684514,
+      "grad_norm": 0.999777615070343,
+      "learning_rate": 1.977069470861714e-05,
+      "loss": 1.1493,
+      "step": 5530
+    },
+    {
+      "epoch": 1.7030114226375908,
+      "grad_norm": 1.3446388244628906,
+      "learning_rate": 1.9731315654237613e-05,
+      "loss": 1.1549,
+      "step": 5535
+    },
+    {
+      "epoch": 1.7045498250067306,
+      "grad_norm": 1.1934943199157715,
+      "learning_rate": 1.9691950275024144e-05,
+      "loss": 1.2842,
+      "step": 5540
+    },
+    {
+      "epoch": 1.7060882273758702,
+      "grad_norm": 1.0282137393951416,
+      "learning_rate": 1.9652598673151798e-05,
+      "loss": 0.9822,
+      "step": 5545
+    },
+    {
+      "epoch": 1.7076266297450098,
+      "grad_norm": 1.6977639198303223,
+      "learning_rate": 1.961326095075986e-05,
+      "loss": 1.1677,
+      "step": 5550
+    },
+    {
+      "epoch": 1.7091650321141496,
+      "grad_norm": 1.1296579837799072,
+      "learning_rate": 1.9573937209951604e-05,
+      "loss": 1.1473,
+      "step": 5555
+    },
+    {
+      "epoch": 1.710703434483289,
+      "grad_norm": 1.5597577095031738,
+      "learning_rate": 1.9534627552793998e-05,
+      "loss": 1.1344,
+      "step": 5560
+    },
+    {
+      "epoch": 1.7122418368524288,
+      "grad_norm": 1.7576704025268555,
+      "learning_rate": 1.9495332081317464e-05,
+      "loss": 1.1523,
+      "step": 5565
+    },
+    {
+      "epoch": 1.7137802392215684,
+      "grad_norm": 0.7898119688034058,
+      "learning_rate": 1.945605089751561e-05,
+      "loss": 1.2694,
+      "step": 5570
+    },
+    {
+      "epoch": 1.715318641590708,
+      "grad_norm": 1.6534613370895386,
+      "learning_rate": 1.9416784103344958e-05,
+      "loss": 1.2023,
+      "step": 5575
+    },
+    {
+      "epoch": 1.7168570439598478,
+      "grad_norm": 1.4977833032608032,
+      "learning_rate": 1.937753180072466e-05,
+      "loss": 1.134,
+      "step": 5580
+    },
+    {
+      "epoch": 1.7183954463289872,
+      "grad_norm": 1.0111984014511108,
+      "learning_rate": 1.93382940915363e-05,
+      "loss": 1.1758,
+      "step": 5585
+    },
+    {
+      "epoch": 1.719933848698127,
+      "grad_norm": 1.2132163047790527,
+      "learning_rate": 1.9299071077623536e-05,
+      "loss": 1.0628,
+      "step": 5590
+    },
+    {
+      "epoch": 1.7214722510672666,
+      "grad_norm": 1.3343091011047363,
+      "learning_rate": 1.9259862860791894e-05,
+      "loss": 1.0998,
+      "step": 5595
+    },
+    {
+      "epoch": 1.7230106534364062,
+      "grad_norm": 1.314704179763794,
+      "learning_rate": 1.922066954280852e-05,
+      "loss": 1.1529,
+      "step": 5600
+    },
+    {
+      "epoch": 1.724549055805546,
+      "grad_norm": 1.7054861783981323,
+      "learning_rate": 1.918149122540187e-05,
+      "loss": 1.1795,
+      "step": 5605
+    },
+    {
+      "epoch": 1.7260874581746855,
+      "grad_norm": 1.477265477180481,
+      "learning_rate": 1.9142328010261463e-05,
+      "loss": 1.1547,
+      "step": 5610
+    },
+    {
+      "epoch": 1.7276258605438253,
+      "grad_norm": 3.371182918548584,
+      "learning_rate": 1.910317999903762e-05,
+      "loss": 1.13,
+      "step": 5615
+    },
+    {
+      "epoch": 1.7291642629129649,
+      "grad_norm": 1.1603204011917114,
+      "learning_rate": 1.9064047293341205e-05,
+      "loss": 1.2422,
+      "step": 5620
+    },
+    {
+      "epoch": 1.7307026652821045,
+      "grad_norm": 1.587916612625122,
+      "learning_rate": 1.9024929994743354e-05,
+      "loss": 1.2718,
+      "step": 5625
+    },
+    {
+      "epoch": 1.7322410676512443,
+      "grad_norm": 1.9051862955093384,
+      "learning_rate": 1.8985828204775206e-05,
+      "loss": 1.0266,
+      "step": 5630
+    },
+    {
+      "epoch": 1.7337794700203837,
+      "grad_norm": 0.9715479016304016,
+      "learning_rate": 1.8946742024927662e-05,
+      "loss": 1.0988,
+      "step": 5635
+    },
+    {
+      "epoch": 1.7353178723895235,
+      "grad_norm": 1.471206545829773,
+      "learning_rate": 1.8907671556651102e-05,
+      "loss": 1.2839,
+      "step": 5640
+    },
+    {
+      "epoch": 1.736856274758663,
+      "grad_norm": 1.127846598625183,
+      "learning_rate": 1.8868616901355096e-05,
+      "loss": 1.084,
+      "step": 5645
+    },
+    {
+      "epoch": 1.7383946771278027,
+      "grad_norm": 1.6066924333572388,
+      "learning_rate": 1.8829578160408216e-05,
+      "loss": 1.1079,
+      "step": 5650
+    },
+    {
+      "epoch": 1.7399330794969425,
+      "grad_norm": 1.393292784690857,
+      "learning_rate": 1.8790555435137697e-05,
+      "loss": 1.0559,
+      "step": 5655
+    },
+    {
+      "epoch": 1.741471481866082,
+      "grad_norm": 1.9487007856369019,
+      "learning_rate": 1.875154882682922e-05,
+      "loss": 1.1889,
+      "step": 5660
+    },
+    {
+      "epoch": 1.7430098842352217,
+      "grad_norm": 0.9660601019859314,
+      "learning_rate": 1.8712558436726623e-05,
+      "loss": 1.1461,
+      "step": 5665
+    },
+    {
+      "epoch": 1.7445482866043613,
+      "grad_norm": 0.9336276054382324,
+      "learning_rate": 1.8673584366031647e-05,
+      "loss": 1.1877,
+      "step": 5670
+    },
+    {
+      "epoch": 1.746086688973501,
+      "grad_norm": 1.3800424337387085,
+      "learning_rate": 1.8634626715903693e-05,
+      "loss": 1.2214,
+      "step": 5675
+    },
+    {
+      "epoch": 1.7476250913426408,
+      "grad_norm": 0.946338415145874,
+      "learning_rate": 1.8595685587459522e-05,
+      "loss": 1.2095,
+      "step": 5680
+    },
+    {
+      "epoch": 1.7491634937117804,
+      "grad_norm": 2.1087701320648193,
+      "learning_rate": 1.8556761081773013e-05,
+      "loss": 1.0836,
+      "step": 5685
+    },
+    {
+      "epoch": 1.75070189608092,
+      "grad_norm": 0.8675004243850708,
+      "learning_rate": 1.851785329987492e-05,
+      "loss": 1.1866,
+      "step": 5690
+    },
+    {
+      "epoch": 1.7522402984500596,
+      "grad_norm": 1.0243099927902222,
+      "learning_rate": 1.8478962342752583e-05,
+      "loss": 1.2131,
+      "step": 5695
+    },
+    {
+      "epoch": 1.7537787008191992,
+      "grad_norm": 1.2281181812286377,
+      "learning_rate": 1.8440088311349634e-05,
+      "loss": 1.2623,
+      "step": 5700
+    },
+    {
+      "epoch": 1.755317103188339,
+      "grad_norm": 1.4797272682189941,
+      "learning_rate": 1.840123130656583e-05,
+      "loss": 1.2044,
+      "step": 5705
+    },
+    {
+      "epoch": 1.7568555055574786,
+      "grad_norm": 1.8646970987319946,
+      "learning_rate": 1.8362391429256698e-05,
+      "loss": 1.152,
+      "step": 5710
+    },
+    {
+      "epoch": 1.7583939079266182,
+      "grad_norm": 1.331153154373169,
+      "learning_rate": 1.8323568780233325e-05,
+      "loss": 1.2001,
+      "step": 5715
+    },
+    {
+      "epoch": 1.7599323102957578,
+      "grad_norm": 1.3193144798278809,
+      "learning_rate": 1.8284763460262085e-05,
+      "loss": 1.2019,
+      "step": 5720
+    },
+    {
+      "epoch": 1.7614707126648974,
+      "grad_norm": 2.1348347663879395,
+      "learning_rate": 1.824597557006434e-05,
+      "loss": 1.1155,
+      "step": 5725
+    },
+    {
+      "epoch": 1.7630091150340372,
+      "grad_norm": 0.8556899428367615,
+      "learning_rate": 1.820720521031626e-05,
+      "loss": 1.208,
+      "step": 5730
+    },
+    {
+      "epoch": 1.7645475174031768,
+      "grad_norm": 1.1282209157943726,
+      "learning_rate": 1.8168452481648476e-05,
+      "loss": 1.2334,
+      "step": 5735
+    },
+    {
+      "epoch": 1.7660859197723164,
+      "grad_norm": 1.082572102546692,
+      "learning_rate": 1.8129717484645876e-05,
+      "loss": 1.0867,
+      "step": 5740
+    },
+    {
+      "epoch": 1.7676243221414563,
+      "grad_norm": 1.2311683893203735,
+      "learning_rate": 1.809100031984734e-05,
+      "loss": 1.2453,
+      "step": 5745
+    },
+    {
+      "epoch": 1.7691627245105956,
+      "grad_norm": 1.6507987976074219,
+      "learning_rate": 1.805230108774541e-05,
+      "loss": 1.1704,
+      "step": 5750
+    },
+    {
+      "epoch": 1.7707011268797355,
+      "grad_norm": 1.152384877204895,
+      "learning_rate": 1.8013619888786127e-05,
+      "loss": 1.2105,
+      "step": 5755
+    },
+    {
+      "epoch": 1.772239529248875,
+      "grad_norm": 1.057258129119873,
+      "learning_rate": 1.7974956823368727e-05,
+      "loss": 1.2076,
+      "step": 5760
+    },
+    {
+      "epoch": 1.7737779316180147,
+      "grad_norm": 0.8284866809844971,
+      "learning_rate": 1.7936311991845355e-05,
+      "loss": 1.1031,
+      "step": 5765
+    },
+    {
+      "epoch": 1.7753163339871545,
+      "grad_norm": 0.9494788646697998,
+      "learning_rate": 1.789768549452085e-05,
+      "loss": 1.1328,
+      "step": 5770
+    },
+    {
+      "epoch": 1.7768547363562939,
+      "grad_norm": 0.9356427192687988,
+      "learning_rate": 1.785907743165245e-05,
+      "loss": 1.1613,
+      "step": 5775
+    },
+    {
+      "epoch": 1.7783931387254337,
+      "grad_norm": 1.4174987077713013,
+      "learning_rate": 1.7820487903449544e-05,
+      "loss": 1.1835,
+      "step": 5780
+    },
+    {
+      "epoch": 1.7799315410945733,
+      "grad_norm": 0.9856932163238525,
+      "learning_rate": 1.778191701007343e-05,
+      "loss": 1.1883,
+      "step": 5785
+    },
+    {
+      "epoch": 1.781469943463713,
+      "grad_norm": 1.5140010118484497,
+      "learning_rate": 1.7743364851637017e-05,
+      "loss": 1.0919,
+      "step": 5790
+    },
+    {
+      "epoch": 1.7830083458328527,
+      "grad_norm": 1.0665122270584106,
+      "learning_rate": 1.7704831528204608e-05,
+      "loss": 1.2349,
+      "step": 5795
+    },
+    {
+      "epoch": 1.784546748201992,
+      "grad_norm": 1.7453125715255737,
+      "learning_rate": 1.7666317139791618e-05,
+      "loss": 1.0932,
+      "step": 5800
+    },
+    {
+      "epoch": 1.786085150571132,
+      "grad_norm": 1.5194388628005981,
+      "learning_rate": 1.7627821786364265e-05,
+      "loss": 1.1284,
+      "step": 5805
+    },
+    {
+      "epoch": 1.7876235529402715,
+      "grad_norm": 1.0937379598617554,
+      "learning_rate": 1.7589345567839433e-05,
+      "loss": 1.1492,
+      "step": 5810
+    },
+    {
+      "epoch": 1.7891619553094111,
+      "grad_norm": 1.1306276321411133,
+      "learning_rate": 1.75508885840843e-05,
+      "loss": 1.1348,
+      "step": 5815
+    },
+    {
+      "epoch": 1.790700357678551,
+      "grad_norm": 1.172468900680542,
+      "learning_rate": 1.7512450934916128e-05,
+      "loss": 1.1805,
+      "step": 5820
+    },
+    {
+      "epoch": 1.7922387600476903,
+      "grad_norm": 1.2849513292312622,
+      "learning_rate": 1.747403272010199e-05,
+      "loss": 1.2076,
+      "step": 5825
+    },
+    {
+      "epoch": 1.7937771624168302,
+      "grad_norm": 1.0420297384262085,
+      "learning_rate": 1.7435634039358527e-05,
+      "loss": 1.1641,
+      "step": 5830
+    },
+    {
+      "epoch": 1.7953155647859698,
+      "grad_norm": 0.999232828617096,
+      "learning_rate": 1.7397254992351662e-05,
+      "loss": 1.0551,
+      "step": 5835
+    },
+    {
+      "epoch": 1.7968539671551094,
+      "grad_norm": 1.499186635017395,
+      "learning_rate": 1.7358895678696368e-05,
+      "loss": 1.2383,
+      "step": 5840
+    },
+    {
+      "epoch": 1.7983923695242492,
+      "grad_norm": 1.5105886459350586,
+      "learning_rate": 1.73205561979564e-05,
+      "loss": 1.1292,
+      "step": 5845
+    },
+    {
+      "epoch": 1.7999307718933886,
+      "grad_norm": 1.0507991313934326,
+      "learning_rate": 1.7282236649644035e-05,
+      "loss": 1.1165,
+      "step": 5850
+    },
+    {
+      "epoch": 1.8014691742625284,
+      "grad_norm": 1.8862839937210083,
+      "learning_rate": 1.7243937133219818e-05,
+      "loss": 1.1018,
+      "step": 5855
+    },
+    {
+      "epoch": 1.803007576631668,
+      "grad_norm": 1.0384774208068848,
+      "learning_rate": 1.7205657748092275e-05,
+      "loss": 1.2132,
+      "step": 5860
+    },
+    {
+      "epoch": 1.8045459790008076,
+      "grad_norm": 1.6291218996047974,
+      "learning_rate": 1.716739859361771e-05,
+      "loss": 1.1987,
+      "step": 5865
+    },
+    {
+      "epoch": 1.8060843813699474,
+      "grad_norm": 0.9344973564147949,
+      "learning_rate": 1.712915976909992e-05,
+      "loss": 1.1942,
+      "step": 5870
+    },
+    {
+      "epoch": 1.8076227837390868,
+      "grad_norm": 1.1848701238632202,
+      "learning_rate": 1.7090941373789898e-05,
+      "loss": 1.2246,
+      "step": 5875
+    },
+    {
+      "epoch": 1.8091611861082266,
+      "grad_norm": 2.381784439086914,
+      "learning_rate": 1.7052743506885652e-05,
+      "loss": 1.1942,
+      "step": 5880
+    },
+    {
+      "epoch": 1.8106995884773662,
+      "grad_norm": 2.1836729049682617,
+      "learning_rate": 1.701456626753189e-05,
+      "loss": 1.1698,
+      "step": 5885
+    },
+    {
+      "epoch": 1.8122379908465058,
+      "grad_norm": 1.0437754392623901,
+      "learning_rate": 1.6976409754819767e-05,
+      "loss": 1.195,
+      "step": 5890
+    },
+    {
+      "epoch": 1.8137763932156457,
+      "grad_norm": 1.6375466585159302,
+      "learning_rate": 1.6938274067786663e-05,
+      "loss": 1.2909,
+      "step": 5895
+    },
+    {
+      "epoch": 1.815314795584785,
+      "grad_norm": 1.204934000968933,
+      "learning_rate": 1.6900159305415892e-05,
+      "loss": 1.1238,
+      "step": 5900
+    },
+    {
+      "epoch": 1.8168531979539249,
+      "grad_norm": 1.3872859477996826,
+      "learning_rate": 1.6862065566636466e-05,
+      "loss": 1.2359,
+      "step": 5905
+    },
+    {
+      "epoch": 1.8183916003230645,
+      "grad_norm": 1.733851432800293,
+      "learning_rate": 1.682399295032283e-05,
+      "loss": 1.1205,
+      "step": 5910
+    },
+    {
+      "epoch": 1.819930002692204,
+      "grad_norm": 1.224089503288269,
+      "learning_rate": 1.6785941555294573e-05,
+      "loss": 1.1741,
+      "step": 5915
+    },
+    {
+      "epoch": 1.8214684050613439,
+      "grad_norm": 1.3531755208969116,
+      "learning_rate": 1.675551578496907e-05,
+      "loss": 1.1854,
+      "step": 5920
+    },
+    {
+      "epoch": 1.8230068074304835,
+      "grad_norm": 1.2245914936065674,
+      "learning_rate": 1.6717502837103975e-05,
+      "loss": 1.1473,
+      "step": 5925
+    },
+    {
+      "epoch": 1.824545209799623,
+      "grad_norm": 1.7382968664169312,
+      "learning_rate": 1.6679511386925337e-05,
+      "loss": 1.1192,
+      "step": 5930
+    },
+    {
+      "epoch": 1.8260836121687627,
+      "grad_norm": 2.019582748413086,
+      "learning_rate": 1.6641541533042098e-05,
+      "loss": 1.1244,
+      "step": 5935
+    },
+    {
+      "epoch": 1.8276220145379023,
+      "grad_norm": 1.7222234010696411,
+      "learning_rate": 1.6603593374007153e-05,
+      "loss": 1.0861,
+      "step": 5940
+    },
+    {
+      "epoch": 1.8291604169070421,
+      "grad_norm": 1.1440834999084473,
+      "learning_rate": 1.656566700831708e-05,
+      "loss": 1.1462,
+      "step": 5945
+    },
+    {
+      "epoch": 1.8306988192761817,
+      "grad_norm": 2.6508567333221436,
+      "learning_rate": 1.6527762534411888e-05,
+      "loss": 1.0422,
+      "step": 5950
+    },
+    {
+      "epoch": 1.8322372216453213,
+      "grad_norm": 1.4697010517120361,
+      "learning_rate": 1.6489880050674767e-05,
+      "loss": 1.184,
+      "step": 5955
+    },
+    {
+      "epoch": 1.8337756240144611,
+      "grad_norm": 1.2767016887664795,
+      "learning_rate": 1.6452019655431828e-05,
+      "loss": 1.1303,
+      "step": 5960
+    },
+    {
+      "epoch": 1.8353140263836005,
+      "grad_norm": 1.345672607421875,
+      "learning_rate": 1.641418144695185e-05,
+      "loss": 1.0967,
+      "step": 5965
+    },
+    {
+      "epoch": 1.8368524287527404,
+      "grad_norm": 0.7754538059234619,
+      "learning_rate": 1.637636552344604e-05,
+      "loss": 1.1405,
+      "step": 5970
+    },
+    {
+      "epoch": 1.83839083112188,
+      "grad_norm": 1.5453096628189087,
+      "learning_rate": 1.6338571983067754e-05,
+      "loss": 1.0661,
+      "step": 5975
+    },
+    {
+      "epoch": 1.8399292334910196,
+      "grad_norm": 1.0012327432632446,
+      "learning_rate": 1.6300800923912224e-05,
+      "loss": 1.1432,
+      "step": 5980
+    },
+    {
+      "epoch": 1.8414676358601594,
+      "grad_norm": 0.9194159507751465,
+      "learning_rate": 1.6263052444016374e-05,
+      "loss": 1.1303,
+      "step": 5985
+    },
+    {
+      "epoch": 1.8430060382292988,
+      "grad_norm": 0.9644936323165894,
+      "learning_rate": 1.62253266413585e-05,
+      "loss": 1.177,
+      "step": 5990
+    },
+    {
+      "epoch": 1.8445444405984386,
+      "grad_norm": 1.5765095949172974,
+      "learning_rate": 1.6187623613858038e-05,
+      "loss": 1.1667,
+      "step": 5995
+    },
+    {
+      "epoch": 1.8460828429675782,
+      "grad_norm": 0.9766362309455872,
+      "learning_rate": 1.6149943459375312e-05,
+      "loss": 1.0883,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8476212453367178,
+      "grad_norm": 1.214890480041504,
+      "learning_rate": 1.6112286275711298e-05,
+      "loss": 1.1184,
+      "step": 6005
+    },
+    {
+      "epoch": 1.8491596477058576,
+      "grad_norm": 1.128957986831665,
+      "learning_rate": 1.6074652160607302e-05,
+      "loss": 1.2097,
+      "step": 6010
+    },
+    {
+      "epoch": 1.850698050074997,
+      "grad_norm": 0.8823169469833374,
+      "learning_rate": 1.603704121174479e-05,
+      "loss": 1.1738,
+      "step": 6015
+    },
+    {
+      "epoch": 1.8522364524441368,
+      "grad_norm": 1.1140327453613281,
+      "learning_rate": 1.5999453526745104e-05,
+      "loss": 1.116,
+      "step": 6020
+    },
+    {
+      "epoch": 1.8537748548132764,
+      "grad_norm": 1.464568853378296,
+      "learning_rate": 1.5961889203169184e-05,
+      "loss": 1.1432,
+      "step": 6025
+    },
+    {
+      "epoch": 1.855313257182416,
+      "grad_norm": 1.134320616722107,
+      "learning_rate": 1.592434833851734e-05,
+      "loss": 1.1429,
+      "step": 6030
+    },
+    {
+      "epoch": 1.8568516595515558,
+      "grad_norm": 1.3063602447509766,
+      "learning_rate": 1.5886831030229e-05,
+      "loss": 1.1284,
+      "step": 6035
+    },
+    {
+      "epoch": 1.8583900619206952,
+      "grad_norm": 2.385805130004883,
+      "learning_rate": 1.5849337375682435e-05,
+      "loss": 1.1124,
+      "step": 6040
+    },
+    {
+      "epoch": 1.859928464289835,
+      "grad_norm": 0.9437244534492493,
+      "learning_rate": 1.5811867472194535e-05,
+      "loss": 1.173,
+      "step": 6045
+    },
+    {
+      "epoch": 1.8614668666589747,
+      "grad_norm": 1.2181042432785034,
+      "learning_rate": 1.577442141702054e-05,
+      "loss": 1.3157,
+      "step": 6050
+    },
+    {
+      "epoch": 1.8630052690281143,
+      "grad_norm": 1.382906198501587,
+      "learning_rate": 1.5736999307353785e-05,
+      "loss": 1.0986,
+      "step": 6055
+    },
+    {
+      "epoch": 1.864543671397254,
+      "grad_norm": 1.4896674156188965,
+      "learning_rate": 1.5699601240325474e-05,
+      "loss": 1.2179,
+      "step": 6060
+    },
+    {
+      "epoch": 1.8660820737663935,
+      "grad_norm": 1.4793058633804321,
+      "learning_rate": 1.5662227313004364e-05,
+      "loss": 1.0728,
+      "step": 6065
+    },
+    {
+      "epoch": 1.8676204761355333,
+      "grad_norm": 1.6949820518493652,
+      "learning_rate": 1.5624877622396588e-05,
+      "loss": 1.1914,
+      "step": 6070
+    },
+    {
+      "epoch": 1.8691588785046729,
+      "grad_norm": 1.629065990447998,
+      "learning_rate": 1.5587552265445375e-05,
+      "loss": 1.2429,
+      "step": 6075
+    },
+    {
+      "epoch": 1.8706972808738125,
+      "grad_norm": 1.2364530563354492,
+      "learning_rate": 1.5550251339030783e-05,
+      "loss": 1.0742,
+      "step": 6080
+    },
+    {
+      "epoch": 1.8722356832429523,
+      "grad_norm": 0.8806670904159546,
+      "learning_rate": 1.5512974939969464e-05,
+      "loss": 1.1999,
+      "step": 6085
+    },
+    {
+      "epoch": 1.8737740856120917,
+      "grad_norm": 1.9689141511917114,
+      "learning_rate": 1.5475723165014393e-05,
+      "loss": 1.1117,
+      "step": 6090
+    },
+    {
+      "epoch": 1.8753124879812315,
+      "grad_norm": 0.9261854887008667,
+      "learning_rate": 1.543849611085465e-05,
+      "loss": 1.0828,
+      "step": 6095
+    },
+    {
+      "epoch": 1.8768508903503711,
+      "grad_norm": 1.5326859951019287,
+      "learning_rate": 1.5401293874115147e-05,
+      "loss": 1.0189,
+      "step": 6100
+    },
+    {
+      "epoch": 1.8783892927195107,
+      "grad_norm": 1.155781865119934,
+      "learning_rate": 1.5364116551356376e-05,
+      "loss": 1.0945,
+      "step": 6105
+    },
+    {
+      "epoch": 1.8799276950886505,
+      "grad_norm": 1.5384670495986938,
+      "learning_rate": 1.532696423907416e-05,
+      "loss": 1.205,
+      "step": 6110
+    },
+    {
+      "epoch": 1.88146609745779,
+      "grad_norm": 1.558389663696289,
+      "learning_rate": 1.528983703369943e-05,
+      "loss": 1.1512,
+      "step": 6115
+    },
+    {
+      "epoch": 1.8830044998269297,
+      "grad_norm": 1.9130889177322388,
+      "learning_rate": 1.5252735031597915e-05,
+      "loss": 1.1126,
+      "step": 6120
+    },
+    {
+      "epoch": 1.8845429021960693,
+      "grad_norm": 1.6255749464035034,
+      "learning_rate": 1.521565832906994e-05,
+      "loss": 1.1858,
+      "step": 6125
+    },
+    {
+      "epoch": 1.886081304565209,
+      "grad_norm": 1.8116388320922852,
+      "learning_rate": 1.5178607022350186e-05,
+      "loss": 1.0117,
+      "step": 6130
+    },
+    {
+      "epoch": 1.8876197069343488,
+      "grad_norm": 1.0656421184539795,
+      "learning_rate": 1.5141581207607391e-05,
+      "loss": 1.1256,
+      "step": 6135
+    },
+    {
+      "epoch": 1.8891581093034884,
+      "grad_norm": 1.6898198127746582,
+      "learning_rate": 1.5104580980944141e-05,
+      "loss": 1.0737,
+      "step": 6140
+    },
+    {
+      "epoch": 1.890696511672628,
+      "grad_norm": 1.3169077634811401,
+      "learning_rate": 1.5067606438396595e-05,
+      "loss": 1.1341,
+      "step": 6145
+    },
+    {
+      "epoch": 1.8922349140417676,
+      "grad_norm": 0.8919572234153748,
+      "learning_rate": 1.5030657675934256e-05,
+      "loss": 1.0732,
+      "step": 6150
+    },
+    {
+      "epoch": 1.8937733164109072,
+      "grad_norm": 2.686988353729248,
+      "learning_rate": 1.4993734789459718e-05,
+      "loss": 1.229,
+      "step": 6155
+    },
+    {
+      "epoch": 1.895311718780047,
+      "grad_norm": 1.9947311878204346,
+      "learning_rate": 1.4956837874808391e-05,
+      "loss": 1.116,
+      "step": 6160
+    },
+    {
+      "epoch": 1.8968501211491866,
+      "grad_norm": 1.5105503797531128,
+      "learning_rate": 1.4919967027748306e-05,
+      "loss": 1.1457,
+      "step": 6165
+    },
+    {
+      "epoch": 1.8983885235183262,
+      "grad_norm": 1.2878894805908203,
+      "learning_rate": 1.4883122343979822e-05,
+      "loss": 1.1197,
+      "step": 6170
+    },
+    {
+      "epoch": 1.8999269258874658,
+      "grad_norm": 1.2365055084228516,
+      "learning_rate": 1.4846303919135355e-05,
+      "loss": 1.1429,
+      "step": 6175
+    },
+    {
+      "epoch": 1.9014653282566054,
+      "grad_norm": 2.5517539978027344,
+      "learning_rate": 1.4809511848779217e-05,
+      "loss": 1.1013,
+      "step": 6180
+    },
+    {
+      "epoch": 1.9030037306257452,
+      "grad_norm": 1.3758244514465332,
+      "learning_rate": 1.4772746228407289e-05,
+      "loss": 1.0995,
+      "step": 6185
+    },
+    {
+      "epoch": 1.9045421329948848,
+      "grad_norm": 1.0151314735412598,
+      "learning_rate": 1.4736007153446801e-05,
+      "loss": 1.0569,
+      "step": 6190
+    },
+    {
+      "epoch": 1.9060805353640244,
+      "grad_norm": 0.9583563208580017,
+      "learning_rate": 1.4699294719256091e-05,
+      "loss": 1.1816,
+      "step": 6195
+    },
+    {
+      "epoch": 1.9076189377331643,
+      "grad_norm": 1.0372930765151978,
+      "learning_rate": 1.466260902112433e-05,
+      "loss": 1.1818,
+      "step": 6200
+    },
+    {
+      "epoch": 1.9091573401023036,
+      "grad_norm": 1.4027732610702515,
+      "learning_rate": 1.4625950154271317e-05,
+      "loss": 1.3359,
+      "step": 6205
+    },
+    {
+      "epoch": 1.9106957424714435,
+      "grad_norm": 1.8428670167922974,
+      "learning_rate": 1.4589318213847197e-05,
+      "loss": 1.1615,
+      "step": 6210
+    },
+    {
+      "epoch": 1.912234144840583,
+      "grad_norm": 2.2269155979156494,
+      "learning_rate": 1.4552713294932226e-05,
+      "loss": 1.1065,
+      "step": 6215
+    },
+    {
+      "epoch": 1.9137725472097227,
+      "grad_norm": 1.9540749788284302,
+      "learning_rate": 1.4516135492536539e-05,
+      "loss": 1.0822,
+      "step": 6220
+    },
+    {
+      "epoch": 1.9153109495788625,
+      "grad_norm": 1.690127968788147,
+      "learning_rate": 1.447958490159987e-05,
+      "loss": 1.1871,
+      "step": 6225
+    },
+    {
+      "epoch": 1.9168493519480019,
+      "grad_norm": 1.6063660383224487,
+      "learning_rate": 1.444306161699131e-05,
+      "loss": 1.2001,
+      "step": 6230
+    },
+    {
+      "epoch": 1.9183877543171417,
+      "grad_norm": 1.3402923345565796,
+      "learning_rate": 1.4406565733509126e-05,
+      "loss": 1.0825,
+      "step": 6235
+    },
+    {
+      "epoch": 1.9199261566862813,
+      "grad_norm": 1.2019435167312622,
+      "learning_rate": 1.4370097345880407e-05,
+      "loss": 1.2244,
+      "step": 6240
+    },
+    {
+      "epoch": 1.921464559055421,
+      "grad_norm": 0.9290767312049866,
+      "learning_rate": 1.433365654876091e-05,
+      "loss": 1.2298,
+      "step": 6245
+    },
+    {
+      "epoch": 1.9230029614245607,
+      "grad_norm": 1.7722569704055786,
+      "learning_rate": 1.4297243436734797e-05,
+      "loss": 1.2057,
+      "step": 6250
+    },
+    {
+      "epoch": 1.9245413637937,
+      "grad_norm": 1.9733870029449463,
+      "learning_rate": 1.4260858104314297e-05,
+      "loss": 1.1947,
+      "step": 6255
+    },
+    {
+      "epoch": 1.92607976616284,
+      "grad_norm": 1.5969133377075195,
+      "learning_rate": 1.422450064593961e-05,
+      "loss": 1.1204,
+      "step": 6260
+    },
+    {
+      "epoch": 1.9276181685319795,
+      "grad_norm": 1.4854793548583984,
+      "learning_rate": 1.4188171155978566e-05,
+      "loss": 1.2412,
+      "step": 6265
+    },
+    {
+      "epoch": 1.9291565709011191,
+      "grad_norm": 1.1663974523544312,
+      "learning_rate": 1.4151869728726378e-05,
+      "loss": 1.2189,
+      "step": 6270
+    },
+    {
+      "epoch": 1.930694973270259,
+      "grad_norm": 1.081315279006958,
+      "learning_rate": 1.4115596458405459e-05,
+      "loss": 1.2044,
+      "step": 6275
+    },
+    {
+      "epoch": 1.9322333756393983,
+      "grad_norm": 1.7983838319778442,
+      "learning_rate": 1.4079351439165106e-05,
+      "loss": 1.187,
+      "step": 6280
+    },
+    {
+      "epoch": 1.9337717780085382,
+      "grad_norm": 1.329074740409851,
+      "learning_rate": 1.4043134765081297e-05,
+      "loss": 1.1532,
+      "step": 6285
+    },
+    {
+      "epoch": 1.9353101803776778,
+      "grad_norm": 1.1615062952041626,
+      "learning_rate": 1.4006946530156462e-05,
+      "loss": 1.1056,
+      "step": 6290
+    },
+    {
+      "epoch": 1.9368485827468174,
+      "grad_norm": 1.0058648586273193,
+      "learning_rate": 1.397078682831917e-05,
+      "loss": 1.1568,
+      "step": 6295
+    },
+    {
+      "epoch": 1.9383869851159572,
+      "grad_norm": 1.2599167823791504,
+      "learning_rate": 1.3934655753423976e-05,
+      "loss": 1.1126,
+      "step": 6300
+    },
+    {
+      "epoch": 1.9399253874850966,
+      "grad_norm": 1.3756744861602783,
+      "learning_rate": 1.389855339925113e-05,
+      "loss": 1.0468,
+      "step": 6305
+    },
+    {
+      "epoch": 1.9414637898542364,
+      "grad_norm": 1.7275495529174805,
+      "learning_rate": 1.386247985950628e-05,
+      "loss": 1.1714,
+      "step": 6310
+    },
+    {
+      "epoch": 1.943002192223376,
+      "grad_norm": 1.5819746255874634,
+      "learning_rate": 1.3826435227820344e-05,
+      "loss": 1.1858,
+      "step": 6315
+    },
+    {
+      "epoch": 1.9445405945925156,
+      "grad_norm": 1.3510098457336426,
+      "learning_rate": 1.3790419597749199e-05,
+      "loss": 1.1779,
+      "step": 6320
+    },
+    {
+      "epoch": 1.9460789969616554,
+      "grad_norm": 1.3834389448165894,
+      "learning_rate": 1.3754433062773409e-05,
+      "loss": 1.1468,
+      "step": 6325
+    },
+    {
+      "epoch": 1.9476173993307948,
+      "grad_norm": 1.2175872325897217,
+      "learning_rate": 1.3718475716298073e-05,
+      "loss": 1.1402,
+      "step": 6330
+    },
+    {
+      "epoch": 1.9491558016999346,
+      "grad_norm": 0.9370445609092712,
+      "learning_rate": 1.368254765165249e-05,
+      "loss": 1.0989,
+      "step": 6335
+    },
+    {
+      "epoch": 1.9506942040690742,
+      "grad_norm": 2.2398204803466797,
+      "learning_rate": 1.3646648962089965e-05,
+      "loss": 1.17,
+      "step": 6340
+    },
+    {
+      "epoch": 1.9522326064382138,
+      "grad_norm": 1.014602780342102,
+      "learning_rate": 1.3610779740787571e-05,
+      "loss": 1.1466,
+      "step": 6345
+    },
+    {
+      "epoch": 1.9537710088073537,
+      "grad_norm": 1.669202446937561,
+      "learning_rate": 1.3574940080845875e-05,
+      "loss": 1.2066,
+      "step": 6350
+    },
+    {
+      "epoch": 1.9553094111764933,
+      "grad_norm": 1.477262258529663,
+      "learning_rate": 1.3539130075288731e-05,
+      "loss": 1.0777,
+      "step": 6355
+    },
+    {
+      "epoch": 1.9568478135456329,
+      "grad_norm": 2.8099093437194824,
+      "learning_rate": 1.3503349817063047e-05,
+      "loss": 1.2807,
+      "step": 6360
+    },
+    {
+      "epoch": 1.9583862159147725,
+      "grad_norm": 1.3674906492233276,
+      "learning_rate": 1.3467599399038445e-05,
+      "loss": 1.1896,
+      "step": 6365
+    },
+    {
+      "epoch": 1.959924618283912,
+      "grad_norm": 1.12281334400177,
+      "learning_rate": 1.3431878914007167e-05,
+      "loss": 1.1885,
+      "step": 6370
+    },
+    {
+      "epoch": 1.961463020653052,
+      "grad_norm": 1.3317580223083496,
+      "learning_rate": 1.3396188454683745e-05,
+      "loss": 1.1905,
+      "step": 6375
+    },
+    {
+      "epoch": 1.9630014230221915,
+      "grad_norm": 1.4334850311279297,
+      "learning_rate": 1.3360528113704751e-05,
+      "loss": 1.1287,
+      "step": 6380
+    },
+    {
+      "epoch": 1.964539825391331,
+      "grad_norm": 1.2335008382797241,
+      "learning_rate": 1.332489798362862e-05,
+      "loss": 1.2626,
+      "step": 6385
+    },
+    {
+      "epoch": 1.9660782277604707,
+      "grad_norm": 2.188669204711914,
+      "learning_rate": 1.3289298156935348e-05,
+      "loss": 1.0611,
+      "step": 6390
+    },
+    {
+      "epoch": 1.9676166301296103,
+      "grad_norm": 2.0799977779388428,
+      "learning_rate": 1.3253728726026276e-05,
+      "loss": 1.1448,
+      "step": 6395
+    },
+    {
+      "epoch": 1.9691550324987501,
+      "grad_norm": 1.1249916553497314,
+      "learning_rate": 1.321818978322387e-05,
+      "loss": 1.0762,
+      "step": 6400
+    },
+    {
+      "epoch": 1.9706934348678897,
+      "grad_norm": 1.3204679489135742,
+      "learning_rate": 1.3182681420771453e-05,
+      "loss": 1.0773,
+      "step": 6405
+    },
+    {
+      "epoch": 1.9722318372370293,
+      "grad_norm": 1.5374983549118042,
+      "learning_rate": 1.3147203730832963e-05,
+      "loss": 1.0408,
+      "step": 6410
+    },
+    {
+      "epoch": 1.9737702396061692,
+      "grad_norm": 1.7096116542816162,
+      "learning_rate": 1.3111756805492752e-05,
+      "loss": 1.1046,
+      "step": 6415
+    },
+    {
+      "epoch": 1.9753086419753085,
+      "grad_norm": 1.8964815139770508,
+      "learning_rate": 1.3076340736755293e-05,
+      "loss": 1.1332,
+      "step": 6420
+    },
+    {
+      "epoch": 1.9768470443444484,
+      "grad_norm": 1.614093542098999,
+      "learning_rate": 1.304095561654498e-05,
+      "loss": 1.1041,
+      "step": 6425
+    },
+    {
+      "epoch": 1.978385446713588,
+      "grad_norm": 1.7104381322860718,
+      "learning_rate": 1.3005601536705889e-05,
+      "loss": 1.3129,
+      "step": 6430
+    },
+    {
+      "epoch": 1.9799238490827276,
+      "grad_norm": 1.825789451599121,
+      "learning_rate": 1.2970278589001505e-05,
+      "loss": 1.2037,
+      "step": 6435
+    },
+    {
+      "epoch": 1.9814622514518674,
+      "grad_norm": 0.9708835482597351,
+      "learning_rate": 1.293498686511454e-05,
+      "loss": 1.1983,
+      "step": 6440
+    },
+    {
+      "epoch": 1.9830006538210068,
+      "grad_norm": 1.194024920463562,
+      "learning_rate": 1.2899726456646635e-05,
+      "loss": 1.1271,
+      "step": 6445
+    },
+    {
+      "epoch": 1.9845390561901466,
+      "grad_norm": 1.3245621919631958,
+      "learning_rate": 1.2864497455118152e-05,
+      "loss": 1.1675,
+      "step": 6450
+    },
+    {
+      "epoch": 1.9860774585592862,
+      "grad_norm": 1.3149336576461792,
+      "learning_rate": 1.2829299951967954e-05,
+      "loss": 1.0849,
+      "step": 6455
+    },
+    {
+      "epoch": 1.9876158609284258,
+      "grad_norm": 1.097424030303955,
+      "learning_rate": 1.2794134038553141e-05,
+      "loss": 1.0848,
+      "step": 6460
+    },
+    {
+      "epoch": 1.9891542632975656,
+      "grad_norm": 1.0103280544281006,
+      "learning_rate": 1.2758999806148813e-05,
+      "loss": 1.1289,
+      "step": 6465
+    },
+    {
+      "epoch": 1.990692665666705,
+      "grad_norm": 1.243175983428955,
+      "learning_rate": 1.2723897345947828e-05,
+      "loss": 1.1198,
+      "step": 6470
+    },
+    {
+      "epoch": 1.9922310680358448,
+      "grad_norm": 2.570831298828125,
+      "learning_rate": 1.2688826749060611e-05,
+      "loss": 1.1318,
+      "step": 6475
+    },
+    {
+      "epoch": 1.9937694704049844,
+      "grad_norm": 1.352293610572815,
+      "learning_rate": 1.2653788106514852e-05,
+      "loss": 1.2472,
+      "step": 6480
+    },
+    {
+      "epoch": 1.995307872774124,
+      "grad_norm": 2.3871512413024902,
+      "learning_rate": 1.2618781509255332e-05,
+      "loss": 1.2089,
+      "step": 6485
+    },
+    {
+      "epoch": 1.9968462751432638,
+      "grad_norm": 0.9660901427268982,
+      "learning_rate": 1.2583807048143617e-05,
+      "loss": 1.2677,
+      "step": 6490
+    },
+    {
+      "epoch": 1.9983846775124032,
+      "grad_norm": 1.012502670288086,
+      "learning_rate": 1.2548864813957909e-05,
+      "loss": 1.214,
+      "step": 6495
+    },
+    {
+      "epoch": 1.999923079881543,
+      "grad_norm": 2.1729748249053955,
+      "learning_rate": 1.2513954897392727e-05,
+      "loss": 1.1746,
+      "step": 6500
+    },
+    {
+      "epoch": 2.001461482250683,
+      "grad_norm": 1.1099797487258911,
+      "learning_rate": 1.2479077389058708e-05,
+      "loss": 1.0032,
+      "step": 6505
+    },
+    {
+      "epoch": 2.0029998846198223,
+      "grad_norm": 1.2987159490585327,
+      "learning_rate": 1.2444232379482398e-05,
+      "loss": 1.0929,
+      "step": 6510
+    },
+    {
+      "epoch": 2.004538286988962,
+      "grad_norm": 1.4184848070144653,
+      "learning_rate": 1.2409419959105981e-05,
+      "loss": 1.1196,
+      "step": 6515
+    },
+    {
+      "epoch": 2.0060766893581015,
+      "grad_norm": 1.006828784942627,
+      "learning_rate": 1.237464021828704e-05,
+      "loss": 1.1062,
+      "step": 6520
+    },
+    {
+      "epoch": 2.0076150917272413,
+      "grad_norm": 1.3644285202026367,
+      "learning_rate": 1.233989324729834e-05,
+      "loss": 1.1301,
+      "step": 6525
+    },
+    {
+      "epoch": 2.009153494096381,
+      "grad_norm": 0.9252627491950989,
+      "learning_rate": 1.2305179136327608e-05,
+      "loss": 1.2008,
+      "step": 6530
+    },
+    {
+      "epoch": 2.0106918964655205,
+      "grad_norm": 2.844708204269409,
+      "learning_rate": 1.2270497975477253e-05,
+      "loss": 1.1736,
+      "step": 6535
+    },
+    {
+      "epoch": 2.0122302988346603,
+      "grad_norm": 1.0435899496078491,
+      "learning_rate": 1.2235849854764194e-05,
+      "loss": 1.1997,
+      "step": 6540
+    },
+    {
+      "epoch": 2.0137687012037997,
+      "grad_norm": 1.7181456089019775,
+      "learning_rate": 1.2201234864119554e-05,
+      "loss": 1.0245,
+      "step": 6545
+    },
+    {
+      "epoch": 2.0153071035729395,
+      "grad_norm": 2.6238577365875244,
+      "learning_rate": 1.2166653093388506e-05,
+      "loss": 1.0753,
+      "step": 6550
+    },
+    {
+      "epoch": 2.0168455059420793,
+      "grad_norm": 1.2975784540176392,
+      "learning_rate": 1.2132104632329963e-05,
+      "loss": 1.1651,
+      "step": 6555
+    },
+    {
+      "epoch": 2.0183839083112187,
+      "grad_norm": 1.737640380859375,
+      "learning_rate": 1.2097589570616394e-05,
+      "loss": 1.1217,
+      "step": 6560
+    },
+    {
+      "epoch": 2.0199223106803585,
+      "grad_norm": 1.2361128330230713,
+      "learning_rate": 1.2063107997833581e-05,
+      "loss": 1.1589,
+      "step": 6565
+    },
+    {
+      "epoch": 2.021460713049498,
+      "grad_norm": 1.1839301586151123,
+      "learning_rate": 1.2028660003480399e-05,
+      "loss": 1.1789,
+      "step": 6570
+    },
+    {
+      "epoch": 2.0229991154186377,
+      "grad_norm": 1.2682217359542847,
+      "learning_rate": 1.1994245676968538e-05,
+      "loss": 1.2168,
+      "step": 6575
+    },
+    {
+      "epoch": 2.0245375177877776,
+      "grad_norm": 1.4984492063522339,
+      "learning_rate": 1.1959865107622307e-05,
+      "loss": 1.0566,
+      "step": 6580
+    },
+    {
+      "epoch": 2.026075920156917,
+      "grad_norm": 1.7460529804229736,
+      "learning_rate": 1.1925518384678421e-05,
+      "loss": 1.0522,
+      "step": 6585
+    },
+    {
+      "epoch": 2.027614322526057,
+      "grad_norm": 1.5842326879501343,
+      "learning_rate": 1.1891205597285712e-05,
+      "loss": 1.0596,
+      "step": 6590
+    },
+    {
+      "epoch": 2.029152724895196,
+      "grad_norm": 1.3402278423309326,
+      "learning_rate": 1.1856926834504963e-05,
+      "loss": 1.2102,
+      "step": 6595
+    },
+    {
+      "epoch": 2.030691127264336,
+      "grad_norm": 1.3944041728973389,
+      "learning_rate": 1.1822682185308612e-05,
+      "loss": 1.1225,
+      "step": 6600
+    },
+    {
+      "epoch": 2.032229529633476,
+      "grad_norm": 1.0951045751571655,
+      "learning_rate": 1.1788471738580581e-05,
+      "loss": 1.0554,
+      "step": 6605
+    },
+    {
+      "epoch": 2.033767932002615,
+      "grad_norm": 0.8644259572029114,
+      "learning_rate": 1.1754295583116004e-05,
+      "loss": 1.1563,
+      "step": 6610
+    },
+    {
+      "epoch": 2.035306334371755,
+      "grad_norm": 0.9207926392555237,
+      "learning_rate": 1.1720153807620999e-05,
+      "loss": 1.2274,
+      "step": 6615
+    },
+    {
+      "epoch": 2.0368447367408944,
+      "grad_norm": 0.9534634947776794,
+      "learning_rate": 1.168604650071247e-05,
+      "loss": 1.1378,
+      "step": 6620
+    },
+    {
+      "epoch": 2.038383139110034,
+      "grad_norm": 1.1268069744110107,
+      "learning_rate": 1.1651973750917854e-05,
+      "loss": 1.1663,
+      "step": 6625
+    },
+    {
+      "epoch": 2.039921541479174,
+      "grad_norm": 1.331639051437378,
+      "learning_rate": 1.1617935646674885e-05,
+      "loss": 1.1084,
+      "step": 6630
+    },
+    {
+      "epoch": 2.0414599438483134,
+      "grad_norm": 1.2026249170303345,
+      "learning_rate": 1.1583932276331358e-05,
+      "loss": 1.2261,
+      "step": 6635
+    },
+    {
+      "epoch": 2.0429983462174532,
+      "grad_norm": 0.9286217093467712,
+      "learning_rate": 1.154996372814495e-05,
+      "loss": 1.0844,
+      "step": 6640
+    },
+    {
+      "epoch": 2.0445367485865926,
+      "grad_norm": 1.787848711013794,
+      "learning_rate": 1.1516030090282914e-05,
+      "loss": 1.2545,
+      "step": 6645
+    },
+    {
+      "epoch": 2.0460751509557324,
+      "grad_norm": 1.1162513494491577,
+      "learning_rate": 1.1482131450821937e-05,
+      "loss": 1.1272,
+      "step": 6650
+    },
+    {
+      "epoch": 2.0476135533248723,
+      "grad_norm": 1.1817599534988403,
+      "learning_rate": 1.1448267897747818e-05,
+      "loss": 1.256,
+      "step": 6655
+    },
+    {
+      "epoch": 2.0491519556940117,
+      "grad_norm": 1.7543742656707764,
+      "learning_rate": 1.1414439518955334e-05,
+      "loss": 1.1881,
+      "step": 6660
+    },
+    {
+      "epoch": 2.0506903580631515,
+      "grad_norm": 1.54066801071167,
+      "learning_rate": 1.1380646402247927e-05,
+      "loss": 1.1412,
+      "step": 6665
+    },
+    {
+      "epoch": 2.052228760432291,
+      "grad_norm": 1.0868725776672363,
+      "learning_rate": 1.1346888635337522e-05,
+      "loss": 1.02,
+      "step": 6670
+    },
+    {
+      "epoch": 2.0537671628014307,
+      "grad_norm": 1.416038155555725,
+      "learning_rate": 1.1313166305844306e-05,
+      "loss": 1.0248,
+      "step": 6675
+    },
+    {
+      "epoch": 2.0553055651705705,
+      "grad_norm": 1.1514451503753662,
+      "learning_rate": 1.1279479501296492e-05,
+      "loss": 1.13,
+      "step": 6680
+    },
+    {
+      "epoch": 2.05684396753971,
+      "grad_norm": 1.4153822660446167,
+      "learning_rate": 1.1245828309130061e-05,
+      "loss": 1.121,
+      "step": 6685
+    },
+    {
+      "epoch": 2.0583823699088497,
+      "grad_norm": 1.6066848039627075,
+      "learning_rate": 1.1212212816688558e-05,
+      "loss": 1.0533,
+      "step": 6690
+    },
+    {
+      "epoch": 2.059920772277989,
+      "grad_norm": 1.7452136278152466,
+      "learning_rate": 1.1178633111222909e-05,
+      "loss": 1.1413,
+      "step": 6695
+    },
+    {
+      "epoch": 2.061459174647129,
+      "grad_norm": 0.9885658025741577,
+      "learning_rate": 1.1145089279891102e-05,
+      "loss": 1.2047,
+      "step": 6700
+    },
+    {
+      "epoch": 2.0629975770162687,
+      "grad_norm": 4.310669898986816,
+      "learning_rate": 1.1111581409758043e-05,
+      "loss": 1.1422,
+      "step": 6705
+    },
+    {
+      "epoch": 2.064535979385408,
+      "grad_norm": 1.114450454711914,
+      "learning_rate": 1.107810958779531e-05,
+      "loss": 1.1967,
+      "step": 6710
+    },
+    {
+      "epoch": 2.066074381754548,
+      "grad_norm": 1.012830376625061,
+      "learning_rate": 1.1044673900880858e-05,
+      "loss": 1.1861,
+      "step": 6715
+    },
+    {
+      "epoch": 2.0676127841236878,
+      "grad_norm": 1.9352067708969116,
+      "learning_rate": 1.101127443579891e-05,
+      "loss": 1.1204,
+      "step": 6720
+    },
+    {
+      "epoch": 2.069151186492827,
+      "grad_norm": 1.8013137578964233,
+      "learning_rate": 1.0977911279239663e-05,
+      "loss": 1.1936,
+      "step": 6725
+    },
+    {
+      "epoch": 2.070689588861967,
+      "grad_norm": 1.516996145248413,
+      "learning_rate": 1.0944584517799045e-05,
+      "loss": 1.1004,
+      "step": 6730
+    },
+    {
+      "epoch": 2.0722279912311063,
+      "grad_norm": 1.0563892126083374,
+      "learning_rate": 1.091129423797855e-05,
+      "loss": 1.1286,
+      "step": 6735
+    },
+    {
+      "epoch": 2.073766393600246,
+      "grad_norm": 1.5256750583648682,
+      "learning_rate": 1.0878040526184965e-05,
+      "loss": 1.1727,
+      "step": 6740
+    },
+    {
+      "epoch": 2.075304795969386,
+      "grad_norm": 1.3976362943649292,
+      "learning_rate": 1.0844823468730158e-05,
+      "loss": 1.1221,
+      "step": 6745
+    },
+    {
+      "epoch": 2.0768431983385254,
+      "grad_norm": 1.1188265085220337,
+      "learning_rate": 1.081164315183088e-05,
+      "loss": 1.1319,
+      "step": 6750
+    },
+    {
+      "epoch": 2.078381600707665,
+      "grad_norm": 2.3145363330841064,
+      "learning_rate": 1.0778499661608491e-05,
+      "loss": 1.1622,
+      "step": 6755
+    },
+    {
+      "epoch": 2.0799200030768046,
+      "grad_norm": 2.635695457458496,
+      "learning_rate": 1.0745393084088789e-05,
+      "loss": 1.2318,
+      "step": 6760
+    },
+    {
+      "epoch": 2.0814584054459444,
+      "grad_norm": 1.1798350811004639,
+      "learning_rate": 1.0712323505201773e-05,
+      "loss": 0.9846,
+      "step": 6765
+    },
+    {
+      "epoch": 2.0829968078150842,
+      "grad_norm": 1.4065446853637695,
+      "learning_rate": 1.0679291010781362e-05,
+      "loss": 1.1356,
+      "step": 6770
+    },
+    {
+      "epoch": 2.0845352101842236,
+      "grad_norm": 1.7471407651901245,
+      "learning_rate": 1.0646295686565259e-05,
+      "loss": 1.08,
+      "step": 6775
+    },
+    {
+      "epoch": 2.0860736125533634,
+      "grad_norm": 1.3534942865371704,
+      "learning_rate": 1.0613337618194691e-05,
+      "loss": 1.1235,
+      "step": 6780
+    },
+    {
+      "epoch": 2.087612014922503,
+      "grad_norm": 0.9599131941795349,
+      "learning_rate": 1.0580416891214162e-05,
+      "loss": 1.1837,
+      "step": 6785
+    },
+    {
+      "epoch": 2.0891504172916426,
+      "grad_norm": 1.544646143913269,
+      "learning_rate": 1.0547533591071285e-05,
+      "loss": 1.2486,
+      "step": 6790
+    },
+    {
+      "epoch": 2.0906888196607825,
+      "grad_norm": 1.2420892715454102,
+      "learning_rate": 1.0514687803116499e-05,
+      "loss": 1.2542,
+      "step": 6795
+    },
+    {
+      "epoch": 2.092227222029922,
+      "grad_norm": 1.1411772966384888,
+      "learning_rate": 1.0481879612602882e-05,
+      "loss": 1.1346,
+      "step": 6800
+    },
+    {
+      "epoch": 2.0937656243990617,
+      "grad_norm": 1.7670328617095947,
+      "learning_rate": 1.0449109104685958e-05,
+      "loss": 1.0802,
+      "step": 6805
+    },
+    {
+      "epoch": 2.095304026768201,
+      "grad_norm": 2.038987636566162,
+      "learning_rate": 1.0416376364423396e-05,
+      "loss": 1.1316,
+      "step": 6810
+    },
+    {
+      "epoch": 2.096842429137341,
+      "grad_norm": 1.5804378986358643,
+      "learning_rate": 1.0383681476774876e-05,
+      "loss": 1.172,
+      "step": 6815
+    },
+    {
+      "epoch": 2.0983808315064807,
+      "grad_norm": 1.1744779348373413,
+      "learning_rate": 1.035102452660183e-05,
+      "loss": 1.1978,
+      "step": 6820
+    },
+    {
+      "epoch": 2.09991923387562,
+      "grad_norm": 1.132908582687378,
+      "learning_rate": 1.031840559866717e-05,
+      "loss": 1.1058,
+      "step": 6825
+    },
+    {
+      "epoch": 2.10145763624476,
+      "grad_norm": 1.5223757028579712,
+      "learning_rate": 1.0285824777635172e-05,
+      "loss": 1.0153,
+      "step": 6830
+    },
+    {
+      "epoch": 2.1029960386138993,
+      "grad_norm": 1.6681894063949585,
+      "learning_rate": 1.0253282148071198e-05,
+      "loss": 1.1409,
+      "step": 6835
+    },
+    {
+      "epoch": 2.104534440983039,
+      "grad_norm": 1.0178622007369995,
+      "learning_rate": 1.022077779444145e-05,
+      "loss": 1.3659,
+      "step": 6840
+    },
+    {
+      "epoch": 2.106072843352179,
+      "grad_norm": 1.5879106521606445,
+      "learning_rate": 1.0188311801112823e-05,
+      "loss": 1.0986,
+      "step": 6845
+    },
+    {
+      "epoch": 2.1076112457213183,
+      "grad_norm": 1.4189927577972412,
+      "learning_rate": 1.0155884252352616e-05,
+      "loss": 1.1785,
+      "step": 6850
+    },
+    {
+      "epoch": 2.109149648090458,
+      "grad_norm": 1.4003583192825317,
+      "learning_rate": 1.0123495232328342e-05,
+      "loss": 1.1677,
+      "step": 6855
+    },
+    {
+      "epoch": 2.1106880504595975,
+      "grad_norm": 1.5137903690338135,
+      "learning_rate": 1.009114482510754e-05,
+      "loss": 1.0557,
+      "step": 6860
+    },
+    {
+      "epoch": 2.1122264528287373,
+      "grad_norm": 1.2828454971313477,
+      "learning_rate": 1.0058833114657493e-05,
+      "loss": 0.9958,
+      "step": 6865
+    },
+    {
+      "epoch": 2.113764855197877,
+      "grad_norm": 1.0032036304473877,
+      "learning_rate": 1.0026560184845066e-05,
+      "loss": 1.0089,
+      "step": 6870
+    },
+    {
+      "epoch": 2.1153032575670165,
+      "grad_norm": 1.6723458766937256,
+      "learning_rate": 9.994326119436478e-06,
+      "loss": 1.0826,
+      "step": 6875
+    },
+    {
+      "epoch": 2.1168416599361564,
+      "grad_norm": 1.6122030019760132,
+      "learning_rate": 9.962131002097022e-06,
+      "loss": 1.0821,
+      "step": 6880
+    },
+    {
+      "epoch": 2.1183800623052957,
+      "grad_norm": 1.042712926864624,
+      "learning_rate": 9.929974916390953e-06,
+      "loss": 1.0976,
+      "step": 6885
+    },
+    {
+      "epoch": 2.1199184646744356,
+      "grad_norm": 1.353058099746704,
+      "learning_rate": 9.897857945781196e-06,
+      "loss": 1.2036,
+      "step": 6890
+    },
+    {
+      "epoch": 2.1214568670435754,
+      "grad_norm": 1.3776549100875854,
+      "learning_rate": 9.865780173629147e-06,
+      "loss": 1.1875,
+      "step": 6895
+    },
+    {
+      "epoch": 2.1229952694127148,
+      "grad_norm": 1.416298747062683,
+      "learning_rate": 9.833741683194475e-06,
+      "loss": 1.1701,
+      "step": 6900
+    },
+    {
+      "epoch": 2.1245336717818546,
+      "grad_norm": 3.549651861190796,
+      "learning_rate": 9.801742557634872e-06,
+      "loss": 1.1895,
+      "step": 6905
+    },
+    {
+      "epoch": 2.1260720741509944,
+      "grad_norm": 1.4298040866851807,
+      "learning_rate": 9.76978288000586e-06,
+      "loss": 1.2195,
+      "step": 6910
+    },
+    {
+      "epoch": 2.127610476520134,
+      "grad_norm": 1.3986989259719849,
+      "learning_rate": 9.73786273326059e-06,
+      "loss": 1.2749,
+      "step": 6915
+    },
+    {
+      "epoch": 2.1291488788892736,
+      "grad_norm": 1.6647675037384033,
+      "learning_rate": 9.70598220024958e-06,
+      "loss": 1.1431,
+      "step": 6920
+    },
+    {
+      "epoch": 2.130687281258413,
+      "grad_norm": 1.683068037033081,
+      "learning_rate": 9.674141363720554e-06,
+      "loss": 1.1656,
+      "step": 6925
+    },
+    {
+      "epoch": 2.132225683627553,
+      "grad_norm": 1.4037578105926514,
+      "learning_rate": 9.642340306318203e-06,
+      "loss": 1.1445,
+      "step": 6930
+    },
+    {
+      "epoch": 2.1337640859966926,
+      "grad_norm": 3.2454349994659424,
+      "learning_rate": 9.61057911058393e-06,
+      "loss": 1.1231,
+      "step": 6935
+    },
+    {
+      "epoch": 2.135302488365832,
+      "grad_norm": 1.0862977504730225,
+      "learning_rate": 9.578857858955715e-06,
+      "loss": 1.0971,
+      "step": 6940
+    },
+    {
+      "epoch": 2.136840890734972,
+      "grad_norm": 1.2978265285491943,
+      "learning_rate": 9.547176633767857e-06,
+      "loss": 1.2674,
+      "step": 6945
+    },
+    {
+      "epoch": 2.1383792931041112,
+      "grad_norm": 1.3811675310134888,
+      "learning_rate": 9.515535517250737e-06,
+      "loss": 1.1261,
+      "step": 6950
+    },
+    {
+      "epoch": 2.139917695473251,
+      "grad_norm": 1.3552972078323364,
+      "learning_rate": 9.483934591530668e-06,
+      "loss": 0.9968,
+      "step": 6955
+    },
+    {
+      "epoch": 2.141456097842391,
+      "grad_norm": 1.5487886667251587,
+      "learning_rate": 9.452373938629619e-06,
+      "loss": 1.1008,
+      "step": 6960
+    },
+    {
+      "epoch": 2.1429945002115303,
+      "grad_norm": 1.354445219039917,
+      "learning_rate": 9.420853640465025e-06,
+      "loss": 1.11,
+      "step": 6965
+    },
+    {
+      "epoch": 2.14453290258067,
+      "grad_norm": 1.501561164855957,
+      "learning_rate": 9.389373778849612e-06,
+      "loss": 1.2466,
+      "step": 6970
+    },
+    {
+      "epoch": 2.1460713049498095,
+      "grad_norm": 1.2629624605178833,
+      "learning_rate": 9.357934435491106e-06,
+      "loss": 1.2212,
+      "step": 6975
+    },
+    {
+      "epoch": 2.1476097073189493,
+      "grad_norm": 2.1113383769989014,
+      "learning_rate": 9.3265356919921e-06,
+      "loss": 1.1319,
+      "step": 6980
+    },
+    {
+      "epoch": 2.149148109688089,
+      "grad_norm": 1.405587077140808,
+      "learning_rate": 9.295177629849802e-06,
+      "loss": 1.2019,
+      "step": 6985
+    },
+    {
+      "epoch": 2.1506865120572285,
+      "grad_norm": 1.0090694427490234,
+      "learning_rate": 9.26386033045582e-06,
+      "loss": 1.2154,
+      "step": 6990
+    },
+    {
+      "epoch": 2.1522249144263683,
+      "grad_norm": 1.7210766077041626,
+      "learning_rate": 9.232583875095949e-06,
+      "loss": 1.1205,
+      "step": 6995
+    },
+    {
+      "epoch": 2.1537633167955077,
+      "grad_norm": 3.97908616065979,
+      "learning_rate": 9.201348344950001e-06,
+      "loss": 1.0795,
+      "step": 7000
+    },
+    {
+      "epoch": 2.1553017191646475,
+      "grad_norm": 1.465118646621704,
+      "learning_rate": 9.170153821091537e-06,
+      "loss": 1.1504,
+      "step": 7005
+    },
+    {
+      "epoch": 2.1568401215337873,
+      "grad_norm": 1.2877440452575684,
+      "learning_rate": 9.13900038448771e-06,
+      "loss": 1.2082,
+      "step": 7010
+    },
+    {
+      "epoch": 2.1583785239029267,
+      "grad_norm": 1.5617296695709229,
+      "learning_rate": 9.107888115999002e-06,
+      "loss": 1.1485,
+      "step": 7015
+    },
+    {
+      "epoch": 2.1599169262720666,
+      "grad_norm": 1.6196231842041016,
+      "learning_rate": 9.07681709637905e-06,
+      "loss": 1.1636,
+      "step": 7020
+    },
+    {
+      "epoch": 2.161455328641206,
+      "grad_norm": 1.7071741819381714,
+      "learning_rate": 9.045787406274437e-06,
+      "loss": 1.2166,
+      "step": 7025
+    },
+    {
+      "epoch": 2.1629937310103458,
+      "grad_norm": 1.6276894807815552,
+      "learning_rate": 9.014799126224471e-06,
+      "loss": 1.0885,
+      "step": 7030
+    },
+    {
+      "epoch": 2.1645321333794856,
+      "grad_norm": 1.0166163444519043,
+      "learning_rate": 8.983852336660959e-06,
+      "loss": 1.203,
+      "step": 7035
+    },
+    {
+      "epoch": 2.166070535748625,
+      "grad_norm": 1.7075228691101074,
+      "learning_rate": 8.952947117908047e-06,
+      "loss": 1.0762,
+      "step": 7040
+    },
+    {
+      "epoch": 2.167608938117765,
+      "grad_norm": 1.1821060180664062,
+      "learning_rate": 8.922083550181959e-06,
+      "loss": 1.0562,
+      "step": 7045
+    },
+    {
+      "epoch": 2.169147340486904,
+      "grad_norm": 1.4123872518539429,
+      "learning_rate": 8.891261713590807e-06,
+      "loss": 1.1853,
+      "step": 7050
+    },
+    {
+      "epoch": 2.170685742856044,
+      "grad_norm": 1.1683062314987183,
+      "learning_rate": 8.860481688134417e-06,
+      "loss": 1.1117,
+      "step": 7055
+    },
+    {
+      "epoch": 2.172224145225184,
+      "grad_norm": 2.6672306060791016,
+      "learning_rate": 8.829743553704056e-06,
+      "loss": 1.1283,
+      "step": 7060
+    },
+    {
+      "epoch": 2.173762547594323,
+      "grad_norm": 1.9148149490356445,
+      "learning_rate": 8.799047390082296e-06,
+      "loss": 1.1936,
+      "step": 7065
+    },
+    {
+      "epoch": 2.175300949963463,
+      "grad_norm": 1.1535136699676514,
+      "learning_rate": 8.768393276942743e-06,
+      "loss": 1.1856,
+      "step": 7070
+    },
+    {
+      "epoch": 2.1768393523326024,
+      "grad_norm": 2.045712471008301,
+      "learning_rate": 8.737781293849864e-06,
+      "loss": 1.2375,
+      "step": 7075
+    },
+    {
+      "epoch": 2.178377754701742,
+      "grad_norm": 1.2342534065246582,
+      "learning_rate": 8.707211520258782e-06,
+      "loss": 1.186,
+      "step": 7080
+    },
+    {
+      "epoch": 2.179916157070882,
+      "grad_norm": 1.2086020708084106,
+      "learning_rate": 8.676684035515076e-06,
+      "loss": 1.1079,
+      "step": 7085
+    },
+    {
+      "epoch": 2.1814545594400214,
+      "grad_norm": 1.6078280210494995,
+      "learning_rate": 8.646198918854526e-06,
+      "loss": 1.0061,
+      "step": 7090
+    },
+    {
+      "epoch": 2.1829929618091612,
+      "grad_norm": 1.091389775276184,
+      "learning_rate": 8.61575624940298e-06,
+      "loss": 1.0278,
+      "step": 7095
+    },
+    {
+      "epoch": 2.1845313641783006,
+      "grad_norm": 1.2044703960418701,
+      "learning_rate": 8.585356106176094e-06,
+      "loss": 1.1275,
+      "step": 7100
+    },
+    {
+      "epoch": 2.1860697665474405,
+      "grad_norm": 1.288192629814148,
+      "learning_rate": 8.55499856807913e-06,
+      "loss": 1.0621,
+      "step": 7105
+    },
+    {
+      "epoch": 2.1876081689165803,
+      "grad_norm": 1.0985409021377563,
+      "learning_rate": 8.524683713906805e-06,
+      "loss": 1.0334,
+      "step": 7110
+    },
+    {
+      "epoch": 2.1891465712857197,
+      "grad_norm": 1.8316186666488647,
+      "learning_rate": 8.49441162234301e-06,
+      "loss": 1.1259,
+      "step": 7115
+    },
+    {
+      "epoch": 2.1906849736548595,
+      "grad_norm": 1.233383059501648,
+      "learning_rate": 8.464182371960668e-06,
+      "loss": 1.213,
+      "step": 7120
+    },
+    {
+      "epoch": 2.192223376023999,
+      "grad_norm": 1.8326971530914307,
+      "learning_rate": 8.433996041221492e-06,
+      "loss": 1.165,
+      "step": 7125
+    },
+    {
+      "epoch": 2.1937617783931387,
+      "grad_norm": 1.156266450881958,
+      "learning_rate": 8.403852708475792e-06,
+      "loss": 1.1034,
+      "step": 7130
+    },
+    {
+      "epoch": 2.1953001807622785,
+      "grad_norm": 1.1190317869186401,
+      "learning_rate": 8.373752451962286e-06,
+      "loss": 1.0357,
+      "step": 7135
+    },
+    {
+      "epoch": 2.196838583131418,
+      "grad_norm": 1.4066323041915894,
+      "learning_rate": 8.34369534980789e-06,
+      "loss": 1.0494,
+      "step": 7140
+    },
+    {
+      "epoch": 2.1983769855005577,
+      "grad_norm": 1.2038296461105347,
+      "learning_rate": 8.31368148002748e-06,
+      "loss": 1.1883,
+      "step": 7145
+    },
+    {
+      "epoch": 2.199915387869697,
+      "grad_norm": 1.0969994068145752,
+      "learning_rate": 8.283710920523763e-06,
+      "loss": 1.0538,
+      "step": 7150
+    },
+    {
+      "epoch": 2.201453790238837,
+      "grad_norm": 1.6629050970077515,
+      "learning_rate": 8.253783749086993e-06,
+      "loss": 1.1396,
+      "step": 7155
+    },
+    {
+      "epoch": 2.2029921926079767,
+      "grad_norm": 1.3252410888671875,
+      "learning_rate": 8.223900043394825e-06,
+      "loss": 1.2732,
+      "step": 7160
+    },
+    {
+      "epoch": 2.204530594977116,
+      "grad_norm": 1.4837875366210938,
+      "learning_rate": 8.194059881012105e-06,
+      "loss": 1.0282,
+      "step": 7165
+    },
+    {
+      "epoch": 2.206068997346256,
+      "grad_norm": 1.2127585411071777,
+      "learning_rate": 8.164263339390635e-06,
+      "loss": 1.067,
+      "step": 7170
+    },
+    {
+      "epoch": 2.2076073997153958,
+      "grad_norm": 1.0039737224578857,
+      "learning_rate": 8.13451049586903e-06,
+      "loss": 1.1353,
+      "step": 7175
+    },
+    {
+      "epoch": 2.209145802084535,
+      "grad_norm": 2.2187910079956055,
+      "learning_rate": 8.104801427672456e-06,
+      "loss": 1.0704,
+      "step": 7180
+    },
+    {
+      "epoch": 2.210684204453675,
+      "grad_norm": 1.5790579319000244,
+      "learning_rate": 8.07513621191246e-06,
+      "loss": 1.1664,
+      "step": 7185
+    },
+    {
+      "epoch": 2.2122226068228144,
+      "grad_norm": 0.9343917369842529,
+      "learning_rate": 8.045514925586784e-06,
+      "loss": 1.2461,
+      "step": 7190
+    },
+    {
+      "epoch": 2.213761009191954,
+      "grad_norm": 1.3252272605895996,
+      "learning_rate": 8.015937645579148e-06,
+      "loss": 1.0664,
+      "step": 7195
+    },
+    {
+      "epoch": 2.215299411561094,
+      "grad_norm": 1.1168420314788818,
+      "learning_rate": 7.986404448659023e-06,
+      "loss": 1.145,
+      "step": 7200
+    },
+    {
+      "epoch": 2.2168378139302334,
+      "grad_norm": 3.980158567428589,
+      "learning_rate": 7.956915411481505e-06,
+      "loss": 1.099,
+      "step": 7205
+    },
+    {
+      "epoch": 2.218376216299373,
+      "grad_norm": 1.854601502418518,
+      "learning_rate": 7.927470610587028e-06,
+      "loss": 1.2413,
+      "step": 7210
+    },
+    {
+      "epoch": 2.2199146186685126,
+      "grad_norm": 1.069262981414795,
+      "learning_rate": 7.898070122401224e-06,
+      "loss": 1.2822,
+      "step": 7215
+    },
+    {
+      "epoch": 2.2214530210376524,
+      "grad_norm": 1.6758251190185547,
+      "learning_rate": 7.868714023234727e-06,
+      "loss": 1.0832,
+      "step": 7220
+    },
+    {
+      "epoch": 2.2229914234067922,
+      "grad_norm": 1.4776010513305664,
+      "learning_rate": 7.839402389282924e-06,
+      "loss": 1.1282,
+      "step": 7225
+    },
+    {
+      "epoch": 2.2245298257759316,
+      "grad_norm": 1.4931747913360596,
+      "learning_rate": 7.810135296625818e-06,
+      "loss": 1.2089,
+      "step": 7230
+    },
+    {
+      "epoch": 2.2260682281450714,
+      "grad_norm": 1.7104672193527222,
+      "learning_rate": 7.78091282122779e-06,
+      "loss": 1.0814,
+      "step": 7235
+    },
+    {
+      "epoch": 2.227606630514211,
+      "grad_norm": 1.155410647392273,
+      "learning_rate": 7.751735038937405e-06,
+      "loss": 1.1818,
+      "step": 7240
+    },
+    {
+      "epoch": 2.2291450328833506,
+      "grad_norm": 1.6542633771896362,
+      "learning_rate": 7.722602025487243e-06,
+      "loss": 1.2211,
+      "step": 7245
+    },
+    {
+      "epoch": 2.2306834352524905,
+      "grad_norm": 1.7440063953399658,
+      "learning_rate": 7.693513856493684e-06,
+      "loss": 1.0336,
+      "step": 7250
+    },
+    {
+      "epoch": 2.23222183762163,
+      "grad_norm": 1.1944167613983154,
+      "learning_rate": 7.6644706074567e-06,
+      "loss": 1.1502,
+      "step": 7255
+    },
+    {
+      "epoch": 2.2337602399907697,
+      "grad_norm": 1.1156004667282104,
+      "learning_rate": 7.63547235375966e-06,
+      "loss": 1.1147,
+      "step": 7260
+    },
+    {
+      "epoch": 2.235298642359909,
+      "grad_norm": 1.6988039016723633,
+      "learning_rate": 7.6065191706691795e-06,
+      "loss": 1.1423,
+      "step": 7265
+    },
+    {
+      "epoch": 2.236837044729049,
+      "grad_norm": 1.2307072877883911,
+      "learning_rate": 7.577611133334858e-06,
+      "loss": 1.2132,
+      "step": 7270
+    },
+    {
+      "epoch": 2.2383754470981887,
+      "grad_norm": 2.0237016677856445,
+      "learning_rate": 7.54874831678914e-06,
+      "loss": 0.992,
+      "step": 7275
+    },
+    {
+      "epoch": 2.239913849467328,
+      "grad_norm": 1.5955995321273804,
+      "learning_rate": 7.519930795947072e-06,
+      "loss": 1.1199,
+      "step": 7280
+    },
+    {
+      "epoch": 2.241452251836468,
+      "grad_norm": 2.980849504470825,
+      "learning_rate": 7.491158645606167e-06,
+      "loss": 1.2796,
+      "step": 7285
+    },
+    {
+      "epoch": 2.2429906542056073,
+      "grad_norm": 1.1380635499954224,
+      "learning_rate": 7.462431940446135e-06,
+      "loss": 1.1254,
+      "step": 7290
+    },
+    {
+      "epoch": 2.244529056574747,
+      "grad_norm": 1.3550338745117188,
+      "learning_rate": 7.433750755028773e-06,
+      "loss": 1.12,
+      "step": 7295
+    },
+    {
+      "epoch": 2.246067458943887,
+      "grad_norm": 1.9419124126434326,
+      "learning_rate": 7.40511516379769e-06,
+      "loss": 1.1136,
+      "step": 7300
+    },
+    {
+      "epoch": 2.2476058613130263,
+      "grad_norm": 2.2289304733276367,
+      "learning_rate": 7.376525241078189e-06,
+      "loss": 1.0981,
+      "step": 7305
+    },
+    {
+      "epoch": 2.249144263682166,
+      "grad_norm": 1.1924974918365479,
+      "learning_rate": 7.347981061077011e-06,
+      "loss": 0.9997,
+      "step": 7310
+    },
+    {
+      "epoch": 2.250682666051306,
+      "grad_norm": 2.636819362640381,
+      "learning_rate": 7.319482697882168e-06,
+      "loss": 1.1327,
+      "step": 7315
+    },
+    {
+      "epoch": 2.2522210684204453,
+      "grad_norm": 1.150176763534546,
+      "learning_rate": 7.291030225462781e-06,
+      "loss": 1.1047,
+      "step": 7320
+    },
+    {
+      "epoch": 2.253759470789585,
+      "grad_norm": 2.1936254501342773,
+      "learning_rate": 7.262623717668821e-06,
+      "loss": 1.0588,
+      "step": 7325
+    },
+    {
+      "epoch": 2.2552978731587245,
+      "grad_norm": 1.0446133613586426,
+      "learning_rate": 7.2342632482309825e-06,
+      "loss": 1.0701,
+      "step": 7330
+    },
+    {
+      "epoch": 2.2568362755278644,
+      "grad_norm": 1.2231892347335815,
+      "learning_rate": 7.205948890760464e-06,
+      "loss": 1.134,
+      "step": 7335
+    },
+    {
+      "epoch": 2.258374677897004,
+      "grad_norm": 1.346924066543579,
+      "learning_rate": 7.177680718748767e-06,
+      "loss": 1.0544,
+      "step": 7340
+    },
+    {
+      "epoch": 2.2599130802661436,
+      "grad_norm": 1.9672859907150269,
+      "learning_rate": 7.149458805567505e-06,
+      "loss": 1.1593,
+      "step": 7345
+    },
+    {
+      "epoch": 2.2614514826352834,
+      "grad_norm": 2.7677721977233887,
+      "learning_rate": 7.1212832244682585e-06,
+      "loss": 1.183,
+      "step": 7350
+    },
+    {
+      "epoch": 2.2629898850044228,
+      "grad_norm": 1.1894586086273193,
+      "learning_rate": 7.093154048582313e-06,
+      "loss": 1.0514,
+      "step": 7355
+    },
+    {
+      "epoch": 2.2645282873735626,
+      "grad_norm": 1.667694091796875,
+      "learning_rate": 7.065071350920538e-06,
+      "loss": 1.0868,
+      "step": 7360
+    },
+    {
+      "epoch": 2.2660666897427024,
+      "grad_norm": 1.0725141763687134,
+      "learning_rate": 7.037035204373147e-06,
+      "loss": 1.154,
+      "step": 7365
+    },
+    {
+      "epoch": 2.267605092111842,
+      "grad_norm": 1.3035621643066406,
+      "learning_rate": 7.009045681709522e-06,
+      "loss": 1.135,
+      "step": 7370
+    },
+    {
+      "epoch": 2.2691434944809816,
+      "grad_norm": 1.461717963218689,
+      "learning_rate": 6.981102855578062e-06,
+      "loss": 1.0843,
+      "step": 7375
+    },
+    {
+      "epoch": 2.270681896850121,
+      "grad_norm": 1.022071361541748,
+      "learning_rate": 6.953206798505918e-06,
+      "loss": 1.0135,
+      "step": 7380
+    },
+    {
+      "epoch": 2.272220299219261,
+      "grad_norm": 1.2068042755126953,
+      "learning_rate": 6.925357582898886e-06,
+      "loss": 1.1527,
+      "step": 7385
+    },
+    {
+      "epoch": 2.2737587015884007,
+      "grad_norm": 1.2736159563064575,
+      "learning_rate": 6.8975552810411765e-06,
+      "loss": 1.1415,
+      "step": 7390
+    },
+    {
+      "epoch": 2.27529710395754,
+      "grad_norm": 1.3243377208709717,
+      "learning_rate": 6.869799965095214e-06,
+      "loss": 1.1465,
+      "step": 7395
+    },
+    {
+      "epoch": 2.27683550632668,
+      "grad_norm": 1.1751413345336914,
+      "learning_rate": 6.842091707101473e-06,
+      "loss": 1.157,
+      "step": 7400
+    },
+    {
+      "epoch": 2.2783739086958192,
+      "grad_norm": 1.1726213693618774,
+      "learning_rate": 6.814430578978309e-06,
+      "loss": 1.1401,
+      "step": 7405
+    },
+    {
+      "epoch": 2.279912311064959,
+      "grad_norm": 1.083336353302002,
+      "learning_rate": 6.786816652521719e-06,
+      "loss": 1.1267,
+      "step": 7410
+    },
+    {
+      "epoch": 2.281450713434099,
+      "grad_norm": 3.8451592922210693,
+      "learning_rate": 6.759249999405212e-06,
+      "loss": 1.1394,
+      "step": 7415
+    },
+    {
+      "epoch": 2.2829891158032383,
+      "grad_norm": 1.711990237236023,
+      "learning_rate": 6.73173069117958e-06,
+      "loss": 1.1841,
+      "step": 7420
+    },
+    {
+      "epoch": 2.284527518172378,
+      "grad_norm": 1.154262900352478,
+      "learning_rate": 6.704258799272722e-06,
+      "loss": 1.0895,
+      "step": 7425
+    },
+    {
+      "epoch": 2.2860659205415175,
+      "grad_norm": 1.3660565614700317,
+      "learning_rate": 6.676834394989495e-06,
+      "loss": 1.1304,
+      "step": 7430
+    },
+    {
+      "epoch": 2.2876043229106573,
+      "grad_norm": 1.9204599857330322,
+      "learning_rate": 6.649457549511459e-06,
+      "loss": 1.033,
+      "step": 7435
+    },
+    {
+      "epoch": 2.289142725279797,
+      "grad_norm": 1.165299654006958,
+      "learning_rate": 6.622128333896768e-06,
+      "loss": 1.0643,
+      "step": 7440
+    },
+    {
+      "epoch": 2.2906811276489365,
+      "grad_norm": 1.4679332971572876,
+      "learning_rate": 6.594846819079939e-06,
+      "loss": 1.0537,
+      "step": 7445
+    },
+    {
+      "epoch": 2.2922195300180763,
+      "grad_norm": 1.5946282148361206,
+      "learning_rate": 6.56761307587167e-06,
+      "loss": 1.146,
+      "step": 7450
+    },
+    {
+      "epoch": 2.2937579323872157,
+      "grad_norm": 1.4443016052246094,
+      "learning_rate": 6.540427174958661e-06,
+      "loss": 1.1959,
+      "step": 7455
+    },
+    {
+      "epoch": 2.2952963347563555,
+      "grad_norm": 0.886721670627594,
+      "learning_rate": 6.513289186903463e-06,
+      "loss": 1.1575,
+      "step": 7460
+    },
+    {
+      "epoch": 2.2968347371254954,
+      "grad_norm": 1.1904209852218628,
+      "learning_rate": 6.486199182144229e-06,
+      "loss": 1.0784,
+      "step": 7465
+    },
+    {
+      "epoch": 2.2983731394946347,
+      "grad_norm": 1.3473881483078003,
+      "learning_rate": 6.459157230994603e-06,
+      "loss": 1.1037,
+      "step": 7470
+    },
+    {
+      "epoch": 2.2999115418637746,
+      "grad_norm": 1.1764485836029053,
+      "learning_rate": 6.432163403643482e-06,
+      "loss": 1.0831,
+      "step": 7475
+    },
+    {
+      "epoch": 2.301449944232914,
+      "grad_norm": 1.0809344053268433,
+      "learning_rate": 6.405217770154853e-06,
+      "loss": 1.2457,
+      "step": 7480
+    },
+    {
+      "epoch": 2.3029883466020538,
+      "grad_norm": 1.5625346899032593,
+      "learning_rate": 6.378320400467636e-06,
+      "loss": 1.2338,
+      "step": 7485
+    },
+    {
+      "epoch": 2.3045267489711936,
+      "grad_norm": 0.9581547379493713,
+      "learning_rate": 6.3514713643954475e-06,
+      "loss": 1.1475,
+      "step": 7490
+    },
+    {
+      "epoch": 2.306065151340333,
+      "grad_norm": 2.769169569015503,
+      "learning_rate": 6.324670731626478e-06,
+      "loss": 1.2225,
+      "step": 7495
+    },
+    {
+      "epoch": 2.307603553709473,
+      "grad_norm": 1.2585043907165527,
+      "learning_rate": 6.297918571723288e-06,
+      "loss": 1.0701,
+      "step": 7500
+    },
+    {
+      "epoch": 2.309141956078612,
+      "grad_norm": 1.8813979625701904,
+      "learning_rate": 6.271214954122581e-06,
+      "loss": 1.1267,
+      "step": 7505
+    },
+    {
+      "epoch": 2.310680358447752,
+      "grad_norm": 1.1062171459197998,
+      "learning_rate": 6.244559948135109e-06,
+      "loss": 1.0784,
+      "step": 7510
+    },
+    {
+      "epoch": 2.312218760816892,
+      "grad_norm": 1.0726068019866943,
+      "learning_rate": 6.217953622945449e-06,
+      "loss": 1.2017,
+      "step": 7515
+    },
+    {
+      "epoch": 2.313757163186031,
+      "grad_norm": 1.0642775297164917,
+      "learning_rate": 6.191396047611794e-06,
+      "loss": 1.1995,
+      "step": 7520
+    },
+    {
+      "epoch": 2.315295565555171,
+      "grad_norm": 1.322033405303955,
+      "learning_rate": 6.164887291065838e-06,
+      "loss": 1.0586,
+      "step": 7525
+    },
+    {
+      "epoch": 2.3168339679243104,
+      "grad_norm": 1.146404504776001,
+      "learning_rate": 6.138427422112539e-06,
+      "loss": 1.1833,
+      "step": 7530
+    },
+    {
+      "epoch": 2.3183723702934502,
+      "grad_norm": 1.8734060525894165,
+      "learning_rate": 6.1120165094299655e-06,
+      "loss": 1.0803,
+      "step": 7535
+    },
+    {
+      "epoch": 2.31991077266259,
+      "grad_norm": 1.196945309638977,
+      "learning_rate": 6.085654621569137e-06,
+      "loss": 1.168,
+      "step": 7540
+    },
+    {
+      "epoch": 2.3214491750317294,
+      "grad_norm": 0.9148879647254944,
+      "learning_rate": 6.0593418269538045e-06,
+      "loss": 1.0812,
+      "step": 7545
+    },
+    {
+      "epoch": 2.3229875774008693,
+      "grad_norm": 1.2411441802978516,
+      "learning_rate": 6.0330781938803034e-06,
+      "loss": 1.0235,
+      "step": 7550
+    },
+    {
+      "epoch": 2.3245259797700086,
+      "grad_norm": 1.3500548601150513,
+      "learning_rate": 6.006863790517392e-06,
+      "loss": 1.0834,
+      "step": 7555
+    },
+    {
+      "epoch": 2.3260643821391485,
+      "grad_norm": 1.3045114278793335,
+      "learning_rate": 5.980698684905989e-06,
+      "loss": 1.0431,
+      "step": 7560
+    },
+    {
+      "epoch": 2.3276027845082883,
+      "grad_norm": 1.2521401643753052,
+      "learning_rate": 5.954582944959111e-06,
+      "loss": 1.1774,
+      "step": 7565
+    },
+    {
+      "epoch": 2.3291411868774277,
+      "grad_norm": 1.590803861618042,
+      "learning_rate": 5.928516638461639e-06,
+      "loss": 1.2343,
+      "step": 7570
+    },
+    {
+      "epoch": 2.3306795892465675,
+      "grad_norm": 1.66427481174469,
+      "learning_rate": 5.902499833070119e-06,
+      "loss": 1.0343,
+      "step": 7575
+    },
+    {
+      "epoch": 2.332217991615707,
+      "grad_norm": 1.8136968612670898,
+      "learning_rate": 5.876532596312645e-06,
+      "loss": 1.2398,
+      "step": 7580
+    },
+    {
+      "epoch": 2.3337563939848467,
+      "grad_norm": 1.165752649307251,
+      "learning_rate": 5.850614995588627e-06,
+      "loss": 0.9754,
+      "step": 7585
+    },
+    {
+      "epoch": 2.3352947963539865,
+      "grad_norm": 2.509021043777466,
+      "learning_rate": 5.824747098168651e-06,
+      "loss": 1.0585,
+      "step": 7590
+    },
+    {
+      "epoch": 2.336833198723126,
+      "grad_norm": 1.7842833995819092,
+      "learning_rate": 5.798928971194301e-06,
+      "loss": 1.0558,
+      "step": 7595
+    },
+    {
+      "epoch": 2.3383716010922657,
+      "grad_norm": 1.0677943229675293,
+      "learning_rate": 5.773160681677983e-06,
+      "loss": 1.2275,
+      "step": 7600
+    },
+    {
+      "epoch": 2.339910003461405,
+      "grad_norm": 1.1144322156906128,
+      "learning_rate": 5.747442296502725e-06,
+      "loss": 1.3907,
+      "step": 7605
+    },
+    {
+      "epoch": 2.341448405830545,
+      "grad_norm": 2.3197555541992188,
+      "learning_rate": 5.721773882422057e-06,
+      "loss": 1.2227,
+      "step": 7610
+    },
+    {
+      "epoch": 2.3429868081996847,
+      "grad_norm": 1.0514782667160034,
+      "learning_rate": 5.69615550605978e-06,
+      "loss": 1.2214,
+      "step": 7615
+    },
+    {
+      "epoch": 2.344525210568824,
+      "grad_norm": 1.129257082939148,
+      "learning_rate": 5.6705872339098186e-06,
+      "loss": 1.1048,
+      "step": 7620
+    },
+    {
+      "epoch": 2.346063612937964,
+      "grad_norm": 2.5535309314727783,
+      "learning_rate": 5.645069132336078e-06,
+      "loss": 1.0361,
+      "step": 7625
+    },
+    {
+      "epoch": 2.3476020153071033,
+      "grad_norm": 1.601874828338623,
+      "learning_rate": 5.6196012675722055e-06,
+      "loss": 1.1273,
+      "step": 7630
+    },
+    {
+      "epoch": 2.349140417676243,
+      "grad_norm": 1.2392939329147339,
+      "learning_rate": 5.594183705721484e-06,
+      "loss": 0.9944,
+      "step": 7635
+    },
+    {
+      "epoch": 2.350678820045383,
+      "grad_norm": 1.375126600265503,
+      "learning_rate": 5.568816512756633e-06,
+      "loss": 1.1266,
+      "step": 7640
+    },
+    {
+      "epoch": 2.3522172224145224,
+      "grad_norm": 1.151104211807251,
+      "learning_rate": 5.5434997545196015e-06,
+      "loss": 1.2082,
+      "step": 7645
+    },
+    {
+      "epoch": 2.353755624783662,
+      "grad_norm": 1.6542187929153442,
+      "learning_rate": 5.5182334967214725e-06,
+      "loss": 1.1022,
+      "step": 7650
+    },
+    {
+      "epoch": 2.355294027152802,
+      "grad_norm": 3.7500414848327637,
+      "learning_rate": 5.493017804942238e-06,
+      "loss": 1.2594,
+      "step": 7655
+    },
+    {
+      "epoch": 2.3568324295219414,
+      "grad_norm": 0.9335038661956787,
+      "learning_rate": 5.467852744630633e-06,
+      "loss": 1.0596,
+      "step": 7660
+    },
+    {
+      "epoch": 2.358370831891081,
+      "grad_norm": 1.5884743928909302,
+      "learning_rate": 5.4427383811039985e-06,
+      "loss": 1.1742,
+      "step": 7665
+    },
+    {
+      "epoch": 2.3599092342602206,
+      "grad_norm": 1.1601238250732422,
+      "learning_rate": 5.417674779548062e-06,
+      "loss": 1.2038,
+      "step": 7670
+    },
+    {
+      "epoch": 2.3614476366293604,
+      "grad_norm": 1.1803358793258667,
+      "learning_rate": 5.39266200501681e-06,
+      "loss": 1.1756,
+      "step": 7675
+    },
+    {
+      "epoch": 2.3629860389985002,
+      "grad_norm": 1.501497507095337,
+      "learning_rate": 5.367700122432315e-06,
+      "loss": 1.1091,
+      "step": 7680
+    },
+    {
+      "epoch": 2.3645244413676396,
+      "grad_norm": 1.0719640254974365,
+      "learning_rate": 5.342789196584527e-06,
+      "loss": 1.1683,
+      "step": 7685
+    },
+    {
+      "epoch": 2.3660628437367794,
+      "grad_norm": 1.4727404117584229,
+      "learning_rate": 5.317929292131163e-06,
+      "loss": 1.1414,
+      "step": 7690
+    },
+    {
+      "epoch": 2.367601246105919,
+      "grad_norm": 1.195420742034912,
+      "learning_rate": 5.293120473597515e-06,
+      "loss": 1.1998,
+      "step": 7695
+    },
+    {
+      "epoch": 2.3691396484750586,
+      "grad_norm": 1.4516286849975586,
+      "learning_rate": 5.268362805376237e-06,
+      "loss": 1.1199,
+      "step": 7700
+    },
+    {
+      "epoch": 2.3706780508441985,
+      "grad_norm": 1.0601075887680054,
+      "learning_rate": 5.243656351727258e-06,
+      "loss": 1.1662,
+      "step": 7705
+    },
+    {
+      "epoch": 2.372216453213338,
+      "grad_norm": 1.7174220085144043,
+      "learning_rate": 5.219001176777574e-06,
+      "loss": 1.1581,
+      "step": 7710
+    },
+    {
+      "epoch": 2.3737548555824777,
+      "grad_norm": 1.3260307312011719,
+      "learning_rate": 5.194397344521065e-06,
+      "loss": 1.0319,
+      "step": 7715
+    },
+    {
+      "epoch": 2.375293257951617,
+      "grad_norm": 1.4584814310073853,
+      "learning_rate": 5.16984491881837e-06,
+      "loss": 1.1525,
+      "step": 7720
+    },
+    {
+      "epoch": 2.376831660320757,
+      "grad_norm": 1.018962025642395,
+      "learning_rate": 5.145343963396682e-06,
+      "loss": 1.2017,
+      "step": 7725
+    },
+    {
+      "epoch": 2.3783700626898967,
+      "grad_norm": 1.7327367067337036,
+      "learning_rate": 5.120894541849599e-06,
+      "loss": 1.1709,
+      "step": 7730
+    },
+    {
+      "epoch": 2.379908465059036,
+      "grad_norm": 1.5583730936050415,
+      "learning_rate": 5.096496717636984e-06,
+      "loss": 1.2094,
+      "step": 7735
+    },
+    {
+      "epoch": 2.381446867428176,
+      "grad_norm": 1.2519557476043701,
+      "learning_rate": 5.072150554084745e-06,
+      "loss": 1.0785,
+      "step": 7740
+    },
+    {
+      "epoch": 2.3829852697973157,
+      "grad_norm": 1.1944867372512817,
+      "learning_rate": 5.04785611438473e-06,
+      "loss": 1.2777,
+      "step": 7745
+    },
+    {
+      "epoch": 2.384523672166455,
+      "grad_norm": 1.240370512008667,
+      "learning_rate": 5.023613461594512e-06,
+      "loss": 1.2303,
+      "step": 7750
+    },
+    {
+      "epoch": 2.386062074535595,
+      "grad_norm": 1.117368459701538,
+      "learning_rate": 4.999422658637254e-06,
+      "loss": 1.0655,
+      "step": 7755
+    },
+    {
+      "epoch": 2.3876004769047343,
+      "grad_norm": 1.7489937543869019,
+      "learning_rate": 4.9752837683015505e-06,
+      "loss": 1.203,
+      "step": 7760
+    },
+    {
+      "epoch": 2.389138879273874,
+      "grad_norm": 1.187821388244629,
+      "learning_rate": 4.95119685324125e-06,
+      "loss": 1.1294,
+      "step": 7765
+    },
+    {
+      "epoch": 2.390677281643014,
+      "grad_norm": 1.0779712200164795,
+      "learning_rate": 4.927161975975284e-06,
+      "loss": 1.1303,
+      "step": 7770
+    },
+    {
+      "epoch": 2.3922156840121533,
+      "grad_norm": 1.066462516784668,
+      "learning_rate": 4.903179198887536e-06,
+      "loss": 1.2634,
+      "step": 7775
+    },
+    {
+      "epoch": 2.393754086381293,
+      "grad_norm": 1.1906163692474365,
+      "learning_rate": 4.879248584226645e-06,
+      "loss": 1.0123,
+      "step": 7780
+    },
+    {
+      "epoch": 2.3952924887504325,
+      "grad_norm": 1.7448662519454956,
+      "learning_rate": 4.85537019410586e-06,
+      "loss": 1.165,
+      "step": 7785
+    },
+    {
+      "epoch": 2.3968308911195724,
+      "grad_norm": 1.0964466333389282,
+      "learning_rate": 4.831544090502896e-06,
+      "loss": 1.1348,
+      "step": 7790
+    },
+    {
+      "epoch": 2.398369293488712,
+      "grad_norm": 1.2719080448150635,
+      "learning_rate": 4.807770335259726e-06,
+      "loss": 1.0279,
+      "step": 7795
+    },
+    {
+      "epoch": 2.3999076958578516,
+      "grad_norm": 1.6911888122558594,
+      "learning_rate": 4.784048990082484e-06,
+      "loss": 1.2083,
+      "step": 7800
+    },
+    {
+      "epoch": 2.4014460982269914,
+      "grad_norm": 1.3245162963867188,
+      "learning_rate": 4.760380116541246e-06,
+      "loss": 1.0125,
+      "step": 7805
+    },
+    {
+      "epoch": 2.402984500596131,
+      "grad_norm": 1.16487455368042,
+      "learning_rate": 4.736763776069897e-06,
+      "loss": 1.1643,
+      "step": 7810
+    },
+    {
+      "epoch": 2.4045229029652706,
+      "grad_norm": 1.4798593521118164,
+      "learning_rate": 4.713200029965978e-06,
+      "loss": 1.1108,
+      "step": 7815
+    },
+    {
+      "epoch": 2.4060613053344104,
+      "grad_norm": 1.7683199644088745,
+      "learning_rate": 4.689688939390521e-06,
+      "loss": 1.1573,
+      "step": 7820
+    },
+    {
+      "epoch": 2.40759970770355,
+      "grad_norm": 1.2997316122055054,
+      "learning_rate": 4.666230565367874e-06,
+      "loss": 1.132,
+      "step": 7825
+    },
+    {
+      "epoch": 2.4091381100726896,
+      "grad_norm": 1.1322225332260132,
+      "learning_rate": 4.642824968785572e-06,
+      "loss": 1.1534,
+      "step": 7830
+    },
+    {
+      "epoch": 2.410676512441829,
+      "grad_norm": 1.4464402198791504,
+      "learning_rate": 4.619472210394154e-06,
+      "loss": 1.1766,
+      "step": 7835
+    },
+    {
+      "epoch": 2.412214914810969,
+      "grad_norm": 2.9878416061401367,
+      "learning_rate": 4.596172350807004e-06,
+      "loss": 1.1984,
+      "step": 7840
+    },
+    {
+      "epoch": 2.4137533171801087,
+      "grad_norm": 1.0388379096984863,
+      "learning_rate": 4.572925450500232e-06,
+      "loss": 1.108,
+      "step": 7845
+    },
+    {
+      "epoch": 2.415291719549248,
+      "grad_norm": 1.252841591835022,
+      "learning_rate": 4.549731569812457e-06,
+      "loss": 1.1951,
+      "step": 7850
+    },
+    {
+      "epoch": 2.416830121918388,
+      "grad_norm": 1.5610461235046387,
+      "learning_rate": 4.526590768944713e-06,
+      "loss": 1.0645,
+      "step": 7855
+    },
+    {
+      "epoch": 2.4183685242875272,
+      "grad_norm": 4.270869255065918,
+      "learning_rate": 4.5035031079602445e-06,
+      "loss": 1.1204,
+      "step": 7860
+    },
+    {
+      "epoch": 2.419906926656667,
+      "grad_norm": 1.0736466646194458,
+      "learning_rate": 4.480468646784364e-06,
+      "loss": 1.0063,
+      "step": 7865
+    },
+    {
+      "epoch": 2.421445329025807,
+      "grad_norm": 1.0907769203186035,
+      "learning_rate": 4.457487445204311e-06,
+      "loss": 1.1522,
+      "step": 7870
+    },
+    {
+      "epoch": 2.4229837313949463,
+      "grad_norm": 2.1617796421051025,
+      "learning_rate": 4.434559562869098e-06,
+      "loss": 1.2111,
+      "step": 7875
+    },
+    {
+      "epoch": 2.424522133764086,
+      "grad_norm": 1.529151201248169,
+      "learning_rate": 4.411685059289314e-06,
+      "loss": 1.1406,
+      "step": 7880
+    },
+    {
+      "epoch": 2.4260605361332255,
+      "grad_norm": 2.3434321880340576,
+      "learning_rate": 4.388863993837031e-06,
+      "loss": 1.1284,
+      "step": 7885
+    },
+    {
+      "epoch": 2.4275989385023653,
+      "grad_norm": 1.4387180805206299,
+      "learning_rate": 4.366096425745597e-06,
+      "loss": 1.2728,
+      "step": 7890
+    },
+    {
+      "epoch": 2.429137340871505,
+      "grad_norm": 1.0508075952529907,
+      "learning_rate": 4.343382414109512e-06,
+      "loss": 1.2069,
+      "step": 7895
+    },
+    {
+      "epoch": 2.4306757432406445,
+      "grad_norm": 1.2803845405578613,
+      "learning_rate": 4.320722017884274e-06,
+      "loss": 1.1109,
+      "step": 7900
+    },
+    {
+      "epoch": 2.4322141456097843,
+      "grad_norm": 1.1172047853469849,
+      "learning_rate": 4.2981152958862155e-06,
+      "loss": 1.1645,
+      "step": 7905
+    },
+    {
+      "epoch": 2.4337525479789237,
+      "grad_norm": 1.058237075805664,
+      "learning_rate": 4.275562306792352e-06,
+      "loss": 1.1356,
+      "step": 7910
+    },
+    {
+      "epoch": 2.4352909503480635,
+      "grad_norm": 1.0330730676651,
+      "learning_rate": 4.253063109140224e-06,
+      "loss": 1.1087,
+      "step": 7915
+    },
+    {
+      "epoch": 2.4368293527172034,
+      "grad_norm": 1.2216901779174805,
+      "learning_rate": 4.2306177613277765e-06,
+      "loss": 1.0055,
+      "step": 7920
+    },
+    {
+      "epoch": 2.4383677550863427,
+      "grad_norm": 1.5149664878845215,
+      "learning_rate": 4.208226321613154e-06,
+      "loss": 1.1642,
+      "step": 7925
+    },
+    {
+      "epoch": 2.4399061574554826,
+      "grad_norm": 1.3599375486373901,
+      "learning_rate": 4.185888848114614e-06,
+      "loss": 0.9859,
+      "step": 7930
+    },
+    {
+      "epoch": 2.441444559824622,
+      "grad_norm": 1.1671425104141235,
+      "learning_rate": 4.163605398810305e-06,
+      "loss": 1.1014,
+      "step": 7935
+    },
+    {
+      "epoch": 2.4429829621937618,
+      "grad_norm": 1.6030899286270142,
+      "learning_rate": 4.141376031538186e-06,
+      "loss": 1.1734,
+      "step": 7940
+    },
+    {
+      "epoch": 2.4445213645629016,
+      "grad_norm": 1.209281086921692,
+      "learning_rate": 4.1192008039958235e-06,
+      "loss": 1.0406,
+      "step": 7945
+    },
+    {
+      "epoch": 2.446059766932041,
+      "grad_norm": 2.271026372909546,
+      "learning_rate": 4.097079773740256e-06,
+      "loss": 1.1752,
+      "step": 7950
+    },
+    {
+      "epoch": 2.447598169301181,
+      "grad_norm": 0.9315564632415771,
+      "learning_rate": 4.075012998187866e-06,
+      "loss": 1.1408,
+      "step": 7955
+    },
+    {
+      "epoch": 2.44913657167032,
+      "grad_norm": 1.7139698266983032,
+      "learning_rate": 4.053000534614218e-06,
+      "loss": 1.1057,
+      "step": 7960
+    },
+    {
+      "epoch": 2.45067497403946,
+      "grad_norm": 1.6565955877304077,
+      "learning_rate": 4.03104244015389e-06,
+      "loss": 1.1457,
+      "step": 7965
+    },
+    {
+      "epoch": 2.4522133764086,
+      "grad_norm": 1.3629543781280518,
+      "learning_rate": 4.0091387718003415e-06,
+      "loss": 1.2844,
+      "step": 7970
+    },
+    {
+      "epoch": 2.453751778777739,
+      "grad_norm": 2.2979674339294434,
+      "learning_rate": 3.987289586405785e-06,
+      "loss": 1.1268,
+      "step": 7975
+    },
+    {
+      "epoch": 2.455290181146879,
+      "grad_norm": 1.8692574501037598,
+      "learning_rate": 3.9654949406809995e-06,
+      "loss": 1.1367,
+      "step": 7980
+    },
+    {
+      "epoch": 2.4568285835160184,
+      "grad_norm": 1.706198811531067,
+      "learning_rate": 3.94375489119522e-06,
+      "loss": 1.1304,
+      "step": 7985
+    },
+    {
+      "epoch": 2.4583669858851582,
+      "grad_norm": 1.2476152181625366,
+      "learning_rate": 3.922069494375963e-06,
+      "loss": 1.1083,
+      "step": 7990
+    },
+    {
+      "epoch": 2.459905388254298,
+      "grad_norm": 1.4470798969268799,
+      "learning_rate": 3.900438806508885e-06,
+      "loss": 1.1529,
+      "step": 7995
+    },
+    {
+      "epoch": 2.4614437906234374,
+      "grad_norm": 1.186686635017395,
+      "learning_rate": 3.878862883737666e-06,
+      "loss": 1.1808,
+      "step": 8000
+    },
+    {
+      "epoch": 2.4629821929925773,
+      "grad_norm": 0.927186906337738,
+      "learning_rate": 3.857341782063812e-06,
+      "loss": 1.1583,
+      "step": 8005
+    },
+    {
+      "epoch": 2.4645205953617166,
+      "grad_norm": 1.1961028575897217,
+      "learning_rate": 3.835875557346552e-06,
+      "loss": 1.0924,
+      "step": 8010
+    },
+    {
+      "epoch": 2.4660589977308565,
+      "grad_norm": 1.1408125162124634,
+      "learning_rate": 3.814464265302692e-06,
+      "loss": 1.1786,
+      "step": 8015
+    },
+    {
+      "epoch": 2.4675974000999963,
+      "grad_norm": 1.9700957536697388,
+      "learning_rate": 3.7931079615064284e-06,
+      "loss": 1.0253,
+      "step": 8020
+    },
+    {
+      "epoch": 2.4691358024691357,
+      "grad_norm": 1.8638861179351807,
+      "learning_rate": 3.7718067013892465e-06,
+      "loss": 1.2051,
+      "step": 8025
+    },
+    {
+      "epoch": 2.4706742048382755,
+      "grad_norm": 1.2263269424438477,
+      "learning_rate": 3.7505605402397753e-06,
+      "loss": 1.1933,
+      "step": 8030
+    },
+    {
+      "epoch": 2.472212607207415,
+      "grad_norm": 1.856704831123352,
+      "learning_rate": 3.7293695332036027e-06,
+      "loss": 1.0577,
+      "step": 8035
+    },
+    {
+      "epoch": 2.4737510095765547,
+      "grad_norm": 1.062137246131897,
+      "learning_rate": 3.7082337352831923e-06,
+      "loss": 1.1723,
+      "step": 8040
+    },
+    {
+      "epoch": 2.4752894119456945,
+      "grad_norm": 2.7450308799743652,
+      "learning_rate": 3.6871532013376896e-06,
+      "loss": 1.3035,
+      "step": 8045
+    },
+    {
+      "epoch": 2.476827814314834,
+      "grad_norm": 1.344294786453247,
+      "learning_rate": 3.666127986082796e-06,
+      "loss": 1.2075,
+      "step": 8050
+    },
+    {
+      "epoch": 2.4783662166839737,
+      "grad_norm": 1.329000473022461,
+      "learning_rate": 3.645158144090649e-06,
+      "loss": 1.0978,
+      "step": 8055
+    },
+    {
+      "epoch": 2.479904619053113,
+      "grad_norm": 1.112028956413269,
+      "learning_rate": 3.624243729789642e-06,
+      "loss": 1.1551,
+      "step": 8060
+    },
+    {
+      "epoch": 2.481443021422253,
+      "grad_norm": 0.9248641133308411,
+      "learning_rate": 3.603384797464318e-06,
+      "loss": 1.0623,
+      "step": 8065
+    },
+    {
+      "epoch": 2.4829814237913927,
+      "grad_norm": 1.3773822784423828,
+      "learning_rate": 3.582581401255211e-06,
+      "loss": 1.146,
+      "step": 8070
+    },
+    {
+      "epoch": 2.484519826160532,
+      "grad_norm": 0.9391597509384155,
+      "learning_rate": 3.561833595158698e-06,
+      "loss": 1.1969,
+      "step": 8075
+    },
+    {
+      "epoch": 2.486058228529672,
+      "grad_norm": 1.0027233362197876,
+      "learning_rate": 3.5411414330268676e-06,
+      "loss": 0.9617,
+      "step": 8080
+    },
+    {
+      "epoch": 2.487596630898812,
+      "grad_norm": 1.8053239583969116,
+      "learning_rate": 3.5205049685674035e-06,
+      "loss": 1.0424,
+      "step": 8085
+    },
+    {
+      "epoch": 2.489135033267951,
+      "grad_norm": 1.6183481216430664,
+      "learning_rate": 3.4999242553433954e-06,
+      "loss": 1.0828,
+      "step": 8090
+    },
+    {
+      "epoch": 2.490673435637091,
+      "grad_norm": 1.1142714023590088,
+      "learning_rate": 3.4793993467732518e-06,
+      "loss": 1.2526,
+      "step": 8095
+    },
+    {
+      "epoch": 2.4922118380062304,
+      "grad_norm": 2.0687549114227295,
+      "learning_rate": 3.458930296130519e-06,
+      "loss": 1.0448,
+      "step": 8100
+    },
+    {
+      "epoch": 2.49375024037537,
+      "grad_norm": 1.1878471374511719,
+      "learning_rate": 3.4385171565437606e-06,
+      "loss": 1.1272,
+      "step": 8105
+    },
+    {
+      "epoch": 2.49528864274451,
+      "grad_norm": 1.824986219406128,
+      "learning_rate": 3.418159980996441e-06,
+      "loss": 1.0827,
+      "step": 8110
+    },
+    {
+      "epoch": 2.4968270451136494,
+      "grad_norm": 1.2641549110412598,
+      "learning_rate": 3.3978588223267383e-06,
+      "loss": 1.0358,
+      "step": 8115
+    },
+    {
+      "epoch": 2.498365447482789,
+      "grad_norm": 1.385815143585205,
+      "learning_rate": 3.3776137332274553e-06,
+      "loss": 1.1586,
+      "step": 8120
+    },
+    {
+      "epoch": 2.4999038498519286,
+      "grad_norm": 1.2434886693954468,
+      "learning_rate": 3.3574247662458645e-06,
+      "loss": 1.0912,
+      "step": 8125
+    },
+    {
+      "epoch": 2.5014422522210684,
+      "grad_norm": 1.060655951499939,
+      "learning_rate": 3.3372919737835574e-06,
+      "loss": 1.2092,
+      "step": 8130
+    },
+    {
+      "epoch": 2.502980654590208,
+      "grad_norm": 1.700639247894287,
+      "learning_rate": 3.317215408096322e-06,
+      "loss": 1.0629,
+      "step": 8135
+    },
+    {
+      "epoch": 2.5045190569593476,
+      "grad_norm": 1.9962120056152344,
+      "learning_rate": 3.297195121294022e-06,
+      "loss": 1.1027,
+      "step": 8140
+    },
+    {
+      "epoch": 2.5060574593284874,
+      "grad_norm": 0.9801268577575684,
+      "learning_rate": 3.2772311653404276e-06,
+      "loss": 1.2366,
+      "step": 8145
+    },
+    {
+      "epoch": 2.5075958616976273,
+      "grad_norm": 1.4176025390625,
+      "learning_rate": 3.257323592053116e-06,
+      "loss": 1.1483,
+      "step": 8150
+    },
+    {
+      "epoch": 2.5091342640667667,
+      "grad_norm": 0.9962260127067566,
+      "learning_rate": 3.2374724531033044e-06,
+      "loss": 1.0985,
+      "step": 8155
+    },
+    {
+      "epoch": 2.5106726664359065,
+      "grad_norm": 1.4499822854995728,
+      "learning_rate": 3.2176778000157367e-06,
+      "loss": 1.0903,
+      "step": 8160
+    },
+    {
+      "epoch": 2.512211068805046,
+      "grad_norm": 2.288475751876831,
+      "learning_rate": 3.1979396841685577e-06,
+      "loss": 1.1369,
+      "step": 8165
+    },
+    {
+      "epoch": 2.5137494711741857,
+      "grad_norm": 1.5029224157333374,
+      "learning_rate": 3.17825815679314e-06,
+      "loss": 1.2038,
+      "step": 8170
+    },
+    {
+      "epoch": 2.5152878735433255,
+      "grad_norm": 1.0565627813339233,
+      "learning_rate": 3.1586332689740037e-06,
+      "loss": 1.1077,
+      "step": 8175
+    },
+    {
+      "epoch": 2.516826275912465,
+      "grad_norm": 2.1373684406280518,
+      "learning_rate": 3.1390650716486474e-06,
+      "loss": 1.0097,
+      "step": 8180
+    },
+    {
+      "epoch": 2.5183646782816047,
+      "grad_norm": 1.037817358970642,
+      "learning_rate": 3.119553615607426e-06,
+      "loss": 1.0125,
+      "step": 8185
+    },
+    {
+      "epoch": 2.519903080650744,
+      "grad_norm": 1.1087840795516968,
+      "learning_rate": 3.1000989514934105e-06,
+      "loss": 1.0558,
+      "step": 8190
+    },
+    {
+      "epoch": 2.521441483019884,
+      "grad_norm": 1.8037227392196655,
+      "learning_rate": 3.0807011298022852e-06,
+      "loss": 1.1562,
+      "step": 8195
+    },
+    {
+      "epoch": 2.5229798853890237,
+      "grad_norm": 1.823003888130188,
+      "learning_rate": 3.061360200882174e-06,
+      "loss": 1.1018,
+      "step": 8200
+    },
+    {
+      "epoch": 2.524518287758163,
+      "grad_norm": 1.0306799411773682,
+      "learning_rate": 3.0420762149335565e-06,
+      "loss": 1.0944,
+      "step": 8205
+    },
+    {
+      "epoch": 2.526056690127303,
+      "grad_norm": 1.979805827140808,
+      "learning_rate": 3.022849222009097e-06,
+      "loss": 1.1025,
+      "step": 8210
+    },
+    {
+      "epoch": 2.5275950924964423,
+      "grad_norm": 1.0842509269714355,
+      "learning_rate": 3.0036792720135266e-06,
+      "loss": 1.1262,
+      "step": 8215
+    },
+    {
+      "epoch": 2.529133494865582,
+      "grad_norm": 1.5210621356964111,
+      "learning_rate": 2.9845664147035326e-06,
+      "loss": 1.1311,
+      "step": 8220
+    },
+    {
+      "epoch": 2.530671897234722,
+      "grad_norm": 1.3788422346115112,
+      "learning_rate": 2.965510699687615e-06,
+      "loss": 1.1261,
+      "step": 8225
+    },
+    {
+      "epoch": 2.5322102996038613,
+      "grad_norm": 1.1447571516036987,
+      "learning_rate": 2.9465121764259447e-06,
+      "loss": 1.2294,
+      "step": 8230
+    },
+    {
+      "epoch": 2.533748701973001,
+      "grad_norm": 1.024839162826538,
+      "learning_rate": 2.927570894230261e-06,
+      "loss": 1.1547,
+      "step": 8235
+    },
+    {
+      "epoch": 2.5352871043421406,
+      "grad_norm": 1.3132535219192505,
+      "learning_rate": 2.908686902263724e-06,
+      "loss": 1.0528,
+      "step": 8240
+    },
+    {
+      "epoch": 2.5368255067112804,
+      "grad_norm": 1.7579002380371094,
+      "learning_rate": 2.889860249540788e-06,
+      "loss": 1.2613,
+      "step": 8245
+    },
+    {
+      "epoch": 2.53836390908042,
+      "grad_norm": 2.2410359382629395,
+      "learning_rate": 2.8710909849270994e-06,
+      "loss": 1.0307,
+      "step": 8250
+    },
+    {
+      "epoch": 2.5399023114495596,
+      "grad_norm": 2.274693489074707,
+      "learning_rate": 2.852379157139329e-06,
+      "loss": 1.217,
+      "step": 8255
+    },
+    {
+      "epoch": 2.5414407138186994,
+      "grad_norm": 1.411793828010559,
+      "learning_rate": 2.8337248147450757e-06,
+      "loss": 1.0502,
+      "step": 8260
+    },
+    {
+      "epoch": 2.542979116187839,
+      "grad_norm": 1.5521130561828613,
+      "learning_rate": 2.815128006162751e-06,
+      "loss": 1.3229,
+      "step": 8265
+    },
+    {
+      "epoch": 2.5445175185569786,
+      "grad_norm": 0.9623503684997559,
+      "learning_rate": 2.7965887796613884e-06,
+      "loss": 1.1239,
+      "step": 8270
+    },
+    {
+      "epoch": 2.5460559209261184,
+      "grad_norm": 2.273772716522217,
+      "learning_rate": 2.7781071833606065e-06,
+      "loss": 1.063,
+      "step": 8275
+    },
+    {
+      "epoch": 2.547594323295258,
+      "grad_norm": 2.1022188663482666,
+      "learning_rate": 2.7596832652304283e-06,
+      "loss": 1.1397,
+      "step": 8280
+    },
+    {
+      "epoch": 2.5491327256643976,
+      "grad_norm": 1.1091395616531372,
+      "learning_rate": 2.7413170730911597e-06,
+      "loss": 1.1236,
+      "step": 8285
+    },
+    {
+      "epoch": 2.550671128033537,
+      "grad_norm": 2.22455096244812,
+      "learning_rate": 2.7230086546132907e-06,
+      "loss": 1.0664,
+      "step": 8290
+    },
+    {
+      "epoch": 2.552209530402677,
+      "grad_norm": 1.0935211181640625,
+      "learning_rate": 2.70475805731735e-06,
+      "loss": 1.1093,
+      "step": 8295
+    },
+    {
+      "epoch": 2.5537479327718167,
+      "grad_norm": 1.054491639137268,
+      "learning_rate": 2.6865653285737757e-06,
+      "loss": 1.1725,
+      "step": 8300
+    },
+    {
+      "epoch": 2.555286335140956,
+      "grad_norm": 1.5011531114578247,
+      "learning_rate": 2.668430515602832e-06,
+      "loss": 1.0883,
+      "step": 8305
+    },
+    {
+      "epoch": 2.556824737510096,
+      "grad_norm": 1.2729263305664062,
+      "learning_rate": 2.6503536654744338e-06,
+      "loss": 1.1141,
+      "step": 8310
+    },
+    {
+      "epoch": 2.5583631398792352,
+      "grad_norm": 1.7168303728103638,
+      "learning_rate": 2.6323348251080626e-06,
+      "loss": 1.1521,
+      "step": 8315
+    },
+    {
+      "epoch": 2.559901542248375,
+      "grad_norm": 1.7469420433044434,
+      "learning_rate": 2.6143740412726435e-06,
+      "loss": 1.0427,
+      "step": 8320
+    },
+    {
+      "epoch": 2.561439944617515,
+      "grad_norm": 1.154000997543335,
+      "learning_rate": 2.596471360586378e-06,
+      "loss": 1.1774,
+      "step": 8325
+    },
+    {
+      "epoch": 2.5629783469866543,
+      "grad_norm": 1.1220345497131348,
+      "learning_rate": 2.5786268295166892e-06,
+      "loss": 1.1262,
+      "step": 8330
+    },
+    {
+      "epoch": 2.564516749355794,
+      "grad_norm": 1.7687299251556396,
+      "learning_rate": 2.5608404943800622e-06,
+      "loss": 1.2676,
+      "step": 8335
+    },
+    {
+      "epoch": 2.5660551517249335,
+      "grad_norm": 1.3014109134674072,
+      "learning_rate": 2.5431124013419237e-06,
+      "loss": 1.0543,
+      "step": 8340
+    },
+    {
+      "epoch": 2.5675935540940733,
+      "grad_norm": 3.2140731811523438,
+      "learning_rate": 2.525442596416541e-06,
+      "loss": 1.1211,
+      "step": 8345
+    },
+    {
+      "epoch": 2.569131956463213,
+      "grad_norm": 1.5560851097106934,
+      "learning_rate": 2.5078311254668834e-06,
+      "loss": 1.1792,
+      "step": 8350
+    },
+    {
+      "epoch": 2.5706703588323525,
+      "grad_norm": 1.287881851196289,
+      "learning_rate": 2.490278034204502e-06,
+      "loss": 1.1749,
+      "step": 8355
+    },
+    {
+      "epoch": 2.5722087612014923,
+      "grad_norm": 1.4491333961486816,
+      "learning_rate": 2.4727833681894437e-06,
+      "loss": 1.079,
+      "step": 8360
+    },
+    {
+      "epoch": 2.5737471635706317,
+      "grad_norm": 1.628779411315918,
+      "learning_rate": 2.4553471728300885e-06,
+      "loss": 1.1926,
+      "step": 8365
+    },
+    {
+      "epoch": 2.5752855659397715,
+      "grad_norm": 1.1830757856369019,
+      "learning_rate": 2.4379694933830634e-06,
+      "loss": 1.1329,
+      "step": 8370
+    },
+    {
+      "epoch": 2.5768239683089114,
+      "grad_norm": 1.2473928928375244,
+      "learning_rate": 2.4206503749531236e-06,
+      "loss": 1.0945,
+      "step": 8375
+    },
+    {
+      "epoch": 2.5783623706780507,
+      "grad_norm": 0.9560319781303406,
+      "learning_rate": 2.4033898624929884e-06,
+      "loss": 1.2291,
+      "step": 8380
+    },
+    {
+      "epoch": 2.5799007730471906,
+      "grad_norm": 1.598518967628479,
+      "learning_rate": 2.386188000803302e-06,
+      "loss": 1.113,
+      "step": 8385
+    },
+    {
+      "epoch": 2.58143917541633,
+      "grad_norm": 3.0550365447998047,
+      "learning_rate": 2.3690448345324634e-06,
+      "loss": 1.2072,
+      "step": 8390
+    },
+    {
+      "epoch": 2.5829775777854698,
+      "grad_norm": 2.201508045196533,
+      "learning_rate": 2.351960408176518e-06,
+      "loss": 1.092,
+      "step": 8395
+    },
+    {
+      "epoch": 2.5845159801546096,
+      "grad_norm": 1.3211005926132202,
+      "learning_rate": 2.3349347660790582e-06,
+      "loss": 1.2229,
+      "step": 8400
+    },
+    {
+      "epoch": 2.586054382523749,
+      "grad_norm": 1.2909151315689087,
+      "learning_rate": 2.317967952431094e-06,
+      "loss": 1.0366,
+      "step": 8405
+    },
+    {
+      "epoch": 2.587592784892889,
+      "grad_norm": 1.7140899896621704,
+      "learning_rate": 2.3010600112709364e-06,
+      "loss": 1.1082,
+      "step": 8410
+    },
+    {
+      "epoch": 2.589131187262028,
+      "grad_norm": 1.139414668083191,
+      "learning_rate": 2.2842109864841034e-06,
+      "loss": 1.1882,
+      "step": 8415
+    },
+    {
+      "epoch": 2.590669589631168,
+      "grad_norm": 1.4526281356811523,
+      "learning_rate": 2.2674209218031787e-06,
+      "loss": 1.1233,
+      "step": 8420
+    },
+    {
+      "epoch": 2.592207992000308,
+      "grad_norm": 1.918116569519043,
+      "learning_rate": 2.25068986080772e-06,
+      "loss": 1.2028,
+      "step": 8425
+    },
+    {
+      "epoch": 2.593746394369447,
+      "grad_norm": 1.6036014556884766,
+      "learning_rate": 2.2340178469241467e-06,
+      "loss": 1.2098,
+      "step": 8430
+    },
+    {
+      "epoch": 2.595284796738587,
+      "grad_norm": 1.7327996492385864,
+      "learning_rate": 2.2174049234255895e-06,
+      "loss": 1.095,
+      "step": 8435
+    },
+    {
+      "epoch": 2.5968231991077264,
+      "grad_norm": 2.0552759170532227,
+      "learning_rate": 2.2008511334318306e-06,
+      "loss": 1.2627,
+      "step": 8440
+    },
+    {
+      "epoch": 2.5983616014768662,
+      "grad_norm": 1.393501877784729,
+      "learning_rate": 2.184356519909167e-06,
+      "loss": 1.0631,
+      "step": 8445
+    },
+    {
+      "epoch": 2.599900003846006,
+      "grad_norm": 3.89595890045166,
+      "learning_rate": 2.1679211256702884e-06,
+      "loss": 1.0974,
+      "step": 8450
+    },
+    {
+      "epoch": 2.6014384062151454,
+      "grad_norm": 1.317922592163086,
+      "learning_rate": 2.1515449933741854e-06,
+      "loss": 1.1896,
+      "step": 8455
+    },
+    {
+      "epoch": 2.6029768085842853,
+      "grad_norm": 1.5205051898956299,
+      "learning_rate": 2.135228165526032e-06,
+      "loss": 1.2268,
+      "step": 8460
+    },
+    {
+      "epoch": 2.6045152109534246,
+      "grad_norm": 1.2494088411331177,
+      "learning_rate": 2.118970684477062e-06,
+      "loss": 1.1769,
+      "step": 8465
+    },
+    {
+      "epoch": 2.6060536133225645,
+      "grad_norm": 1.1973810195922852,
+      "learning_rate": 2.1027725924244903e-06,
+      "loss": 1.1078,
+      "step": 8470
+    },
+    {
+      "epoch": 2.6075920156917043,
+      "grad_norm": 1.2799108028411865,
+      "learning_rate": 2.0866339314113662e-06,
+      "loss": 1.2161,
+      "step": 8475
+    },
+    {
+      "epoch": 2.6091304180608437,
+      "grad_norm": 1.1334717273712158,
+      "learning_rate": 2.0705547433264943e-06,
+      "loss": 1.1708,
+      "step": 8480
+    },
+    {
+      "epoch": 2.6106688204299835,
+      "grad_norm": 1.3745144605636597,
+      "learning_rate": 2.0545350699043174e-06,
+      "loss": 1.1895,
+      "step": 8485
+    },
+    {
+      "epoch": 2.612207222799123,
+      "grad_norm": 1.3669915199279785,
+      "learning_rate": 2.0385749527247837e-06,
+      "loss": 1.2637,
+      "step": 8490
+    },
+    {
+      "epoch": 2.6137456251682627,
+      "grad_norm": 1.3522690534591675,
+      "learning_rate": 2.0226744332132812e-06,
+      "loss": 1.1513,
+      "step": 8495
+    },
+    {
+      "epoch": 2.6152840275374025,
+      "grad_norm": 1.3111488819122314,
+      "learning_rate": 2.0068335526405023e-06,
+      "loss": 1.0883,
+      "step": 8500
+    },
+    {
+      "epoch": 2.616822429906542,
+      "grad_norm": 1.5023269653320312,
+      "learning_rate": 1.9910523521223355e-06,
+      "loss": 1.0835,
+      "step": 8505
+    },
+    {
+      "epoch": 2.6183608322756817,
+      "grad_norm": 1.2174749374389648,
+      "learning_rate": 1.975330872619782e-06,
+      "loss": 1.2002,
+      "step": 8510
+    },
+    {
+      "epoch": 2.619899234644821,
+      "grad_norm": 1.0910028219223022,
+      "learning_rate": 1.95966915493882e-06,
+      "loss": 1.2196,
+      "step": 8515
+    },
+    {
+      "epoch": 2.621437637013961,
+      "grad_norm": 1.2541165351867676,
+      "learning_rate": 1.9440672397303127e-06,
+      "loss": 1.0791,
+      "step": 8520
+    },
+    {
+      "epoch": 2.6229760393831008,
+      "grad_norm": 1.4798517227172852,
+      "learning_rate": 1.928525167489914e-06,
+      "loss": 1.2197,
+      "step": 8525
+    },
+    {
+      "epoch": 2.62451444175224,
+      "grad_norm": 2.8591833114624023,
+      "learning_rate": 1.913042978557944e-06,
+      "loss": 1.0815,
+      "step": 8530
+    },
+    {
+      "epoch": 2.62605284412138,
+      "grad_norm": 1.3371671438217163,
+      "learning_rate": 1.8976207131192914e-06,
+      "loss": 1.1042,
+      "step": 8535
+    },
+    {
+      "epoch": 2.6275912464905193,
+      "grad_norm": 1.2439570426940918,
+      "learning_rate": 1.8822584112033082e-06,
+      "loss": 1.0295,
+      "step": 8540
+    },
+    {
+      "epoch": 2.629129648859659,
+      "grad_norm": 1.8023666143417358,
+      "learning_rate": 1.8669561126837236e-06,
+      "loss": 1.1148,
+      "step": 8545
+    },
+    {
+      "epoch": 2.630668051228799,
+      "grad_norm": 1.4807002544403076,
+      "learning_rate": 1.8517138572784976e-06,
+      "loss": 1.1129,
+      "step": 8550
+    },
+    {
+      "epoch": 2.632206453597939,
+      "grad_norm": 1.9293781518936157,
+      "learning_rate": 1.836531684549772e-06,
+      "loss": 1.1038,
+      "step": 8555
+    },
+    {
+      "epoch": 2.633744855967078,
+      "grad_norm": 0.8880111575126648,
+      "learning_rate": 1.821409633903723e-06,
+      "loss": 1.2579,
+      "step": 8560
+    },
+    {
+      "epoch": 2.6352832583362176,
+      "grad_norm": 1.4220179319381714,
+      "learning_rate": 1.8063477445904835e-06,
+      "loss": 1.2608,
+      "step": 8565
+    },
+    {
+      "epoch": 2.6368216607053574,
+      "grad_norm": 1.7923526763916016,
+      "learning_rate": 1.7913460557040351e-06,
+      "loss": 1.212,
+      "step": 8570
+    },
+    {
+      "epoch": 2.638360063074497,
+      "grad_norm": 1.4313488006591797,
+      "learning_rate": 1.776404606182097e-06,
+      "loss": 1.1288,
+      "step": 8575
+    },
+    {
+      "epoch": 2.639898465443637,
+      "grad_norm": 1.1481549739837646,
+      "learning_rate": 1.7615234348060449e-06,
+      "loss": 1.1998,
+      "step": 8580
+    },
+    {
+      "epoch": 2.6414368678127764,
+      "grad_norm": 2.2744739055633545,
+      "learning_rate": 1.7467025802007987e-06,
+      "loss": 1.1055,
+      "step": 8585
+    },
+    {
+      "epoch": 2.642975270181916,
+      "grad_norm": 1.0382848978042603,
+      "learning_rate": 1.7319420808347142e-06,
+      "loss": 1.1111,
+      "step": 8590
+    },
+    {
+      "epoch": 2.6445136725510556,
+      "grad_norm": 1.4756346940994263,
+      "learning_rate": 1.717241975019493e-06,
+      "loss": 1.1661,
+      "step": 8595
+    },
+    {
+      "epoch": 2.6460520749201955,
+      "grad_norm": 0.9779161214828491,
+      "learning_rate": 1.7026023009100944e-06,
+      "loss": 0.9952,
+      "step": 8600
+    },
+    {
+      "epoch": 2.6475904772893353,
+      "grad_norm": 1.9879071712493896,
+      "learning_rate": 1.688023096504604e-06,
+      "loss": 1.2905,
+      "step": 8605
+    },
+    {
+      "epoch": 2.6491288796584747,
+      "grad_norm": 1.308909296989441,
+      "learning_rate": 1.673504399644174e-06,
+      "loss": 1.0405,
+      "step": 8610
+    },
+    {
+      "epoch": 2.6506672820276145,
+      "grad_norm": 1.6702406406402588,
+      "learning_rate": 1.6590462480128882e-06,
+      "loss": 1.101,
+      "step": 8615
+    },
+    {
+      "epoch": 2.652205684396754,
+      "grad_norm": 1.1122729778289795,
+      "learning_rate": 1.644648679137703e-06,
+      "loss": 1.0497,
+      "step": 8620
+    },
+    {
+      "epoch": 2.6537440867658937,
+      "grad_norm": 1.0254093408584595,
+      "learning_rate": 1.630311730388312e-06,
+      "loss": 1.2324,
+      "step": 8625
+    },
+    {
+      "epoch": 2.6552824891350335,
+      "grad_norm": 2.15950608253479,
+      "learning_rate": 1.6160354389770649e-06,
+      "loss": 1.1059,
+      "step": 8630
+    },
+    {
+      "epoch": 2.656820891504173,
+      "grad_norm": 1.1212074756622314,
+      "learning_rate": 1.6018198419588793e-06,
+      "loss": 1.1365,
+      "step": 8635
+    },
+    {
+      "epoch": 2.6583592938733127,
+      "grad_norm": 0.9412396550178528,
+      "learning_rate": 1.5876649762311458e-06,
+      "loss": 1.1834,
+      "step": 8640
+    },
+    {
+      "epoch": 2.659897696242452,
+      "grad_norm": 1.0666898488998413,
+      "learning_rate": 1.5735708785336033e-06,
+      "loss": 1.1222,
+      "step": 8645
+    },
+    {
+      "epoch": 2.661436098611592,
+      "grad_norm": 0.9774393439292908,
+      "learning_rate": 1.559537585448273e-06,
+      "loss": 1.0845,
+      "step": 8650
+    },
+    {
+      "epoch": 2.6629745009807317,
+      "grad_norm": 2.3956403732299805,
+      "learning_rate": 1.5455651333993626e-06,
+      "loss": 1.1634,
+      "step": 8655
+    },
+    {
+      "epoch": 2.664512903349871,
+      "grad_norm": 1.4635967016220093,
+      "learning_rate": 1.5316535586531483e-06,
+      "loss": 1.2237,
+      "step": 8660
+    },
+    {
+      "epoch": 2.666051305719011,
+      "grad_norm": 2.2025790214538574,
+      "learning_rate": 1.5178028973179104e-06,
+      "loss": 1.1956,
+      "step": 8665
+    },
+    {
+      "epoch": 2.6675897080881503,
+      "grad_norm": 1.1397591829299927,
+      "learning_rate": 1.504013185343811e-06,
+      "loss": 1.183,
+      "step": 8670
+    },
+    {
+      "epoch": 2.66912811045729,
+      "grad_norm": 1.7163399457931519,
+      "learning_rate": 1.4902844585228282e-06,
+      "loss": 1.0752,
+      "step": 8675
+    },
+    {
+      "epoch": 2.67066651282643,
+      "grad_norm": 1.7945340871810913,
+      "learning_rate": 1.476616752488641e-06,
+      "loss": 1.1313,
+      "step": 8680
+    },
+    {
+      "epoch": 2.6722049151955694,
+      "grad_norm": 1.072609543800354,
+      "learning_rate": 1.4630101027165444e-06,
+      "loss": 0.9894,
+      "step": 8685
+    },
+    {
+      "epoch": 2.673743317564709,
+      "grad_norm": 1.3113356828689575,
+      "learning_rate": 1.4494645445233658e-06,
+      "loss": 1.1927,
+      "step": 8690
+    },
+    {
+      "epoch": 2.6752817199338486,
+      "grad_norm": 2.284916639328003,
+      "learning_rate": 1.4359801130673616e-06,
+      "loss": 1.206,
+      "step": 8695
+    },
+    {
+      "epoch": 2.6768201223029884,
+      "grad_norm": 1.0177814960479736,
+      "learning_rate": 1.4225568433481329e-06,
+      "loss": 1.1056,
+      "step": 8700
+    },
+    {
+      "epoch": 2.678358524672128,
+      "grad_norm": 1.0384098291397095,
+      "learning_rate": 1.4091947702065262e-06,
+      "loss": 1.1469,
+      "step": 8705
+    },
+    {
+      "epoch": 2.6798969270412676,
+      "grad_norm": 2.263399362564087,
+      "learning_rate": 1.3958939283245543e-06,
+      "loss": 1.038,
+      "step": 8710
+    },
+    {
+      "epoch": 2.6814353294104074,
+      "grad_norm": 2.890347480773926,
+      "learning_rate": 1.382654352225296e-06,
+      "loss": 1.1278,
+      "step": 8715
+    },
+    {
+      "epoch": 2.682973731779547,
+      "grad_norm": 1.9222828149795532,
+      "learning_rate": 1.3694760762728215e-06,
+      "loss": 1.1225,
+      "step": 8720
+    },
+    {
+      "epoch": 2.6845121341486866,
+      "grad_norm": 1.1517970561981201,
+      "learning_rate": 1.3563591346720804e-06,
+      "loss": 1.1967,
+      "step": 8725
+    },
+    {
+      "epoch": 2.6860505365178264,
+      "grad_norm": 1.7945805788040161,
+      "learning_rate": 1.3433035614688338e-06,
+      "loss": 1.1659,
+      "step": 8730
+    },
+    {
+      "epoch": 2.687588938886966,
+      "grad_norm": 0.9292943477630615,
+      "learning_rate": 1.3303093905495528e-06,
+      "loss": 1.2293,
+      "step": 8735
+    },
+    {
+      "epoch": 2.6891273412561056,
+      "grad_norm": 1.3209861516952515,
+      "learning_rate": 1.3173766556413393e-06,
+      "loss": 1.0543,
+      "step": 8740
+    },
+    {
+      "epoch": 2.690665743625245,
+      "grad_norm": 1.1976464986801147,
+      "learning_rate": 1.3045053903118303e-06,
+      "loss": 1.0793,
+      "step": 8745
+    },
+    {
+      "epoch": 2.692204145994385,
+      "grad_norm": 1.2206952571868896,
+      "learning_rate": 1.2916956279691223e-06,
+      "loss": 1.1523,
+      "step": 8750
+    },
+    {
+      "epoch": 2.6937425483635247,
+      "grad_norm": 1.2255494594573975,
+      "learning_rate": 1.2789474018616714e-06,
+      "loss": 1.0986,
+      "step": 8755
+    },
+    {
+      "epoch": 2.695280950732664,
+      "grad_norm": 2.243601083755493,
+      "learning_rate": 1.26626074507821e-06,
+      "loss": 1.0889,
+      "step": 8760
+    },
+    {
+      "epoch": 2.696819353101804,
+      "grad_norm": 1.0789637565612793,
+      "learning_rate": 1.2536356905476748e-06,
+      "loss": 1.1933,
+      "step": 8765
+    },
+    {
+      "epoch": 2.6983577554709433,
+      "grad_norm": 1.564278483390808,
+      "learning_rate": 1.2410722710390954e-06,
+      "loss": 1.1819,
+      "step": 8770
+    },
+    {
+      "epoch": 2.699896157840083,
+      "grad_norm": 1.5282732248306274,
+      "learning_rate": 1.2285705191615426e-06,
+      "loss": 1.2118,
+      "step": 8775
+    },
+    {
+      "epoch": 2.701434560209223,
+      "grad_norm": 1.438302993774414,
+      "learning_rate": 1.21613046736401e-06,
+      "loss": 1.2306,
+      "step": 8780
+    },
+    {
+      "epoch": 2.7029729625783623,
+      "grad_norm": 1.248544454574585,
+      "learning_rate": 1.203752147935347e-06,
+      "loss": 1.2009,
+      "step": 8785
+    },
+    {
+      "epoch": 2.704511364947502,
+      "grad_norm": 1.5993555784225464,
+      "learning_rate": 1.1914355930041837e-06,
+      "loss": 1.0794,
+      "step": 8790
+    },
+    {
+      "epoch": 2.7060497673166415,
+      "grad_norm": 1.6044913530349731,
+      "learning_rate": 1.179180834538826e-06,
+      "loss": 1.1427,
+      "step": 8795
+    },
+    {
+      "epoch": 2.7075881696857813,
+      "grad_norm": 1.4509234428405762,
+      "learning_rate": 1.166987904347186e-06,
+      "loss": 1.123,
+      "step": 8800
+    },
+    {
+      "epoch": 2.709126572054921,
+      "grad_norm": 1.0035152435302734,
+      "learning_rate": 1.1548568340767036e-06,
+      "loss": 1.1457,
+      "step": 8805
+    },
+    {
+      "epoch": 2.7106649744240605,
+      "grad_norm": 1.530730962753296,
+      "learning_rate": 1.142787655214253e-06,
+      "loss": 1.1163,
+      "step": 8810
+    },
+    {
+      "epoch": 2.7122033767932003,
+      "grad_norm": 1.7669676542282104,
+      "learning_rate": 1.1307803990860594e-06,
+      "loss": 1.0995,
+      "step": 8815
+    },
+    {
+      "epoch": 2.7137417791623397,
+      "grad_norm": 1.4253056049346924,
+      "learning_rate": 1.1188350968576372e-06,
+      "loss": 1.1652,
+      "step": 8820
+    },
+    {
+      "epoch": 2.7152801815314795,
+      "grad_norm": 1.14907705783844,
+      "learning_rate": 1.1069517795336825e-06,
+      "loss": 1.1864,
+      "step": 8825
+    },
+    {
+      "epoch": 2.7168185839006194,
+      "grad_norm": 1.3076614141464233,
+      "learning_rate": 1.0951304779580146e-06,
+      "loss": 1.0647,
+      "step": 8830
+    },
+    {
+      "epoch": 2.7183569862697587,
+      "grad_norm": 1.654503583908081,
+      "learning_rate": 1.0833712228134952e-06,
+      "loss": 1.2105,
+      "step": 8835
+    },
+    {
+      "epoch": 2.7198953886388986,
+      "grad_norm": 1.136584758758545,
+      "learning_rate": 1.0716740446219175e-06,
+      "loss": 1.167,
+      "step": 8840
+    },
+    {
+      "epoch": 2.721433791008038,
+      "grad_norm": 1.5034282207489014,
+      "learning_rate": 1.0600389737439681e-06,
+      "loss": 1.1643,
+      "step": 8845
+    },
+    {
+      "epoch": 2.7229721933771778,
+      "grad_norm": 1.1271815299987793,
+      "learning_rate": 1.0484660403791314e-06,
+      "loss": 1.2436,
+      "step": 8850
+    },
+    {
+      "epoch": 2.7245105957463176,
+      "grad_norm": 1.1455885171890259,
+      "learning_rate": 1.0369552745656013e-06,
+      "loss": 1.0879,
+      "step": 8855
+    },
+    {
+      "epoch": 2.726048998115457,
+      "grad_norm": 1.2253069877624512,
+      "learning_rate": 1.025506706180221e-06,
+      "loss": 1.206,
+      "step": 8860
+    },
+    {
+      "epoch": 2.727587400484597,
+      "grad_norm": 1.2960996627807617,
+      "learning_rate": 1.0141203649383924e-06,
+      "loss": 1.2533,
+      "step": 8865
+    },
+    {
+      "epoch": 2.729125802853736,
+      "grad_norm": 1.177007794380188,
+      "learning_rate": 1.0027962803939944e-06,
+      "loss": 1.1394,
+      "step": 8870
+    },
+    {
+      "epoch": 2.730664205222876,
+      "grad_norm": 1.0711764097213745,
+      "learning_rate": 9.91534481939338e-07,
+      "loss": 1.1713,
+      "step": 8875
+    },
+    {
+      "epoch": 2.732202607592016,
+      "grad_norm": 1.1851369142532349,
+      "learning_rate": 9.80334998805041e-07,
+      "loss": 1.1108,
+      "step": 8880
+    },
+    {
+      "epoch": 2.733741009961155,
+      "grad_norm": 1.4155503511428833,
+      "learning_rate": 9.691978600599977e-07,
+      "loss": 1.191,
+      "step": 8885
+    },
+    {
+      "epoch": 2.735279412330295,
+      "grad_norm": 1.3459392786026,
+      "learning_rate": 9.581230946112824e-07,
+      "loss": 0.9247,
+      "step": 8890
+    },
+    {
+      "epoch": 2.7368178146994344,
+      "grad_norm": 1.4829347133636475,
+      "learning_rate": 9.471107312040567e-07,
+      "loss": 1.0897,
+      "step": 8895
+    },
+    {
+      "epoch": 2.7383562170685742,
+      "grad_norm": 1.6139320135116577,
+      "learning_rate": 9.361607984215342e-07,
+      "loss": 1.1201,
+      "step": 8900
+    },
+    {
+      "epoch": 2.739894619437714,
+      "grad_norm": 1.500117540359497,
+      "learning_rate": 9.25273324684886e-07,
+      "loss": 1.2296,
+      "step": 8905
+    },
+    {
+      "epoch": 2.7414330218068534,
+      "grad_norm": 0.9885631203651428,
+      "learning_rate": 9.144483382531571e-07,
+      "loss": 1.1622,
+      "step": 8910
+    },
+    {
+      "epoch": 2.7429714241759933,
+      "grad_norm": 1.3149123191833496,
+      "learning_rate": 9.036858672232057e-07,
+      "loss": 1.1528,
+      "step": 8915
+    },
+    {
+      "epoch": 2.7445098265451326,
+      "grad_norm": 1.1396979093551636,
+      "learning_rate": 8.929859395296364e-07,
+      "loss": 1.1723,
+      "step": 8920
+    },
+    {
+      "epoch": 2.7460482289142725,
+      "grad_norm": 1.0069197416305542,
+      "learning_rate": 8.823485829447003e-07,
+      "loss": 1.1311,
+      "step": 8925
+    },
+    {
+      "epoch": 2.7475866312834123,
+      "grad_norm": 1.2922357320785522,
+      "learning_rate": 8.717738250782675e-07,
+      "loss": 1.2323,
+      "step": 8930
+    },
+    {
+      "epoch": 2.7491250336525517,
+      "grad_norm": 1.268516182899475,
+      "learning_rate": 8.612616933777046e-07,
+      "loss": 1.1243,
+      "step": 8935
+    },
+    {
+      "epoch": 2.7506634360216915,
+      "grad_norm": 1.552514910697937,
+      "learning_rate": 8.508122151278442e-07,
+      "loss": 1.0926,
+      "step": 8940
+    },
+    {
+      "epoch": 2.752201838390831,
+      "grad_norm": 1.1461021900177002,
+      "learning_rate": 8.404254174509019e-07,
+      "loss": 1.1677,
+      "step": 8945
+    },
+    {
+      "epoch": 2.7537402407599707,
+      "grad_norm": 1.04787278175354,
+      "learning_rate": 8.301013273063791e-07,
+      "loss": 1.1723,
+      "step": 8950
+    },
+    {
+      "epoch": 2.7552786431291105,
+      "grad_norm": 2.725139617919922,
+      "learning_rate": 8.198399714910404e-07,
+      "loss": 1.183,
+      "step": 8955
+    },
+    {
+      "epoch": 2.75681704549825,
+      "grad_norm": 3.641418218612671,
+      "learning_rate": 8.096413766388117e-07,
+      "loss": 1.0918,
+      "step": 8960
+    },
+    {
+      "epoch": 2.7583554478673897,
+      "grad_norm": 1.5253046751022339,
+      "learning_rate": 7.995055692207127e-07,
+      "loss": 1.2304,
+      "step": 8965
+    },
+    {
+      "epoch": 2.759893850236529,
+      "grad_norm": 1.2496076822280884,
+      "learning_rate": 7.894325755448073e-07,
+      "loss": 1.1795,
+      "step": 8970
+    },
+    {
+      "epoch": 2.761432252605669,
+      "grad_norm": 1.7881182432174683,
+      "learning_rate": 7.794224217561152e-07,
+      "loss": 1.0852,
+      "step": 8975
+    },
+    {
+      "epoch": 2.7629706549748088,
+      "grad_norm": 2.052133083343506,
+      "learning_rate": 7.694751338365447e-07,
+      "loss": 1.1222,
+      "step": 8980
+    },
+    {
+      "epoch": 2.7645090573439486,
+      "grad_norm": 1.0607606172561646,
+      "learning_rate": 7.595907376048512e-07,
+      "loss": 1.0621,
+      "step": 8985
+    },
+    {
+      "epoch": 2.766047459713088,
+      "grad_norm": 1.3713186979293823,
+      "learning_rate": 7.497692587165345e-07,
+      "loss": 1.2151,
+      "step": 8990
+    },
+    {
+      "epoch": 2.7675858620822273,
+      "grad_norm": 1.383420467376709,
+      "learning_rate": 7.40010722663792e-07,
+      "loss": 1.0939,
+      "step": 8995
+    },
+    {
+      "epoch": 2.769124264451367,
+      "grad_norm": 0.9419933557510376,
+      "learning_rate": 7.303151547754627e-07,
+      "loss": 1.1962,
+      "step": 9000
+    },
+    {
+      "epoch": 2.770662666820507,
+      "grad_norm": 1.4444811344146729,
+      "learning_rate": 7.20682580216922e-07,
+      "loss": 1.1294,
+      "step": 9005
+    },
+    {
+      "epoch": 2.772201069189647,
+      "grad_norm": 1.7534838914871216,
+      "learning_rate": 7.111130239900677e-07,
+      "loss": 1.1511,
+      "step": 9010
+    },
+    {
+      "epoch": 2.773739471558786,
+      "grad_norm": 1.5162113904953003,
+      "learning_rate": 7.016065109332226e-07,
+      "loss": 1.0526,
+      "step": 9015
+    },
+    {
+      "epoch": 2.7752778739279256,
+      "grad_norm": 1.8117800951004028,
+      "learning_rate": 6.921630657210659e-07,
+      "loss": 1.0708,
+      "step": 9020
+    },
+    {
+      "epoch": 2.7768162762970654,
+      "grad_norm": 1.098140001296997,
+      "learning_rate": 6.827827128645992e-07,
+      "loss": 1.0874,
+      "step": 9025
+    },
+    {
+      "epoch": 2.7783546786662052,
+      "grad_norm": 0.9571955800056458,
+      "learning_rate": 6.734654767110521e-07,
+      "loss": 1.1287,
+      "step": 9030
+    },
+    {
+      "epoch": 2.779893081035345,
+      "grad_norm": 2.1803581714630127,
+      "learning_rate": 6.6421138144383e-07,
+      "loss": 1.2326,
+      "step": 9035
+    },
+    {
+      "epoch": 2.7814314834044844,
+      "grad_norm": 1.0323264598846436,
+      "learning_rate": 6.550204510824609e-07,
+      "loss": 1.1212,
+      "step": 9040
+    },
+    {
+      "epoch": 2.7829698857736243,
+      "grad_norm": 2.1114048957824707,
+      "learning_rate": 6.458927094825179e-07,
+      "loss": 1.2275,
+      "step": 9045
+    },
+    {
+      "epoch": 2.7845082881427636,
+      "grad_norm": 1.30685293674469,
+      "learning_rate": 6.368281803355691e-07,
+      "loss": 1.1677,
+      "step": 9050
+    },
+    {
+      "epoch": 2.7860466905119035,
+      "grad_norm": 1.824419379234314,
+      "learning_rate": 6.27826887169114e-07,
+      "loss": 1.057,
+      "step": 9055
+    },
+    {
+      "epoch": 2.7875850928810433,
+      "grad_norm": 0.9725929498672485,
+      "learning_rate": 6.188888533465053e-07,
+      "loss": 1.166,
+      "step": 9060
+    },
+    {
+      "epoch": 2.7891234952501827,
+      "grad_norm": 1.6305102109909058,
+      "learning_rate": 6.100141020669137e-07,
+      "loss": 1.166,
+      "step": 9065
+    },
+    {
+      "epoch": 2.7906618976193225,
+      "grad_norm": 1.65754234790802,
+      "learning_rate": 6.012026563652573e-07,
+      "loss": 1.2142,
+      "step": 9070
+    },
+    {
+      "epoch": 2.792200299988462,
+      "grad_norm": 1.303271770477295,
+      "learning_rate": 5.924545391121361e-07,
+      "loss": 1.2034,
+      "step": 9075
+    },
+    {
+      "epoch": 2.7937387023576017,
+      "grad_norm": 2.9017386436462402,
+      "learning_rate": 5.837697730137814e-07,
+      "loss": 1.0679,
+      "step": 9080
+    },
+    {
+      "epoch": 2.7952771047267415,
+      "grad_norm": 1.4329947233200073,
+      "learning_rate": 5.751483806119923e-07,
+      "loss": 1.2121,
+      "step": 9085
+    },
+    {
+      "epoch": 2.796815507095881,
+      "grad_norm": 0.9450810551643372,
+      "learning_rate": 5.665903842840714e-07,
+      "loss": 1.0761,
+      "step": 9090
+    },
+    {
+      "epoch": 2.7983539094650207,
+      "grad_norm": 2.1919000148773193,
+      "learning_rate": 5.580958062427866e-07,
+      "loss": 1.1133,
+      "step": 9095
+    },
+    {
+      "epoch": 2.79989231183416,
+      "grad_norm": 1.355247139930725,
+      "learning_rate": 5.496646685362844e-07,
+      "loss": 1.0562,
+      "step": 9100
+    },
+    {
+      "epoch": 2.8014307142033,
+      "grad_norm": 1.137313723564148,
+      "learning_rate": 5.412969930480599e-07,
+      "loss": 1.1331,
+      "step": 9105
+    },
+    {
+      "epoch": 2.8029691165724397,
+      "grad_norm": 3.3919472694396973,
+      "learning_rate": 5.329928014968843e-07,
+      "loss": 1.1539,
+      "step": 9110
+    },
+    {
+      "epoch": 2.804507518941579,
+      "grad_norm": 1.3040916919708252,
+      "learning_rate": 5.247521154367552e-07,
+      "loss": 1.1187,
+      "step": 9115
+    },
+    {
+      "epoch": 2.806045921310719,
+      "grad_norm": 1.5795443058013916,
+      "learning_rate": 5.165749562568323e-07,
+      "loss": 1.1518,
+      "step": 9120
+    },
+    {
+      "epoch": 2.8075843236798583,
+      "grad_norm": 1.486302137374878,
+      "learning_rate": 5.084613451813935e-07,
+      "loss": 1.0955,
+      "step": 9125
+    },
+    {
+      "epoch": 2.809122726048998,
+      "grad_norm": 1.4122854471206665,
+      "learning_rate": 5.00411303269771e-07,
+      "loss": 1.1842,
+      "step": 9130
+    },
+    {
+      "epoch": 2.810661128418138,
+      "grad_norm": 1.3084039688110352,
+      "learning_rate": 4.924248514163038e-07,
+      "loss": 1.1416,
+      "step": 9135
+    },
+    {
+      "epoch": 2.8121995307872774,
+      "grad_norm": 1.42378830909729,
+      "learning_rate": 4.845020103502712e-07,
+      "loss": 1.1191,
+      "step": 9140
+    },
+    {
+      "epoch": 2.813737933156417,
+      "grad_norm": 1.8624348640441895,
+      "learning_rate": 4.766428006358542e-07,
+      "loss": 1.021,
+      "step": 9145
+    },
+    {
+      "epoch": 2.8152763355255566,
+      "grad_norm": 1.3880897760391235,
+      "learning_rate": 4.688472426720714e-07,
+      "loss": 1.0401,
+      "step": 9150
+    },
+    {
+      "epoch": 2.8168147378946964,
+      "grad_norm": 1.7636202573776245,
+      "learning_rate": 4.611153566927373e-07,
+      "loss": 1.1469,
+      "step": 9155
+    },
+    {
+      "epoch": 2.818353140263836,
+      "grad_norm": 1.6327768564224243,
+      "learning_rate": 4.534471627663878e-07,
+      "loss": 1.1372,
+      "step": 9160
+    },
+    {
+      "epoch": 2.8198915426329756,
+      "grad_norm": 1.5780202150344849,
+      "learning_rate": 4.4584268079625735e-07,
+      "loss": 1.1779,
+      "step": 9165
+    },
+    {
+      "epoch": 2.8214299450021154,
+      "grad_norm": 1.1375956535339355,
+      "learning_rate": 4.3830193052020186e-07,
+      "loss": 1.11,
+      "step": 9170
+    },
+    {
+      "epoch": 2.822968347371255,
+      "grad_norm": 1.3757110834121704,
+      "learning_rate": 4.308249315106649e-07,
+      "loss": 1.0576,
+      "step": 9175
+    },
+    {
+      "epoch": 2.8245067497403946,
+      "grad_norm": 1.2449181079864502,
+      "learning_rate": 4.234117031746143e-07,
+      "loss": 1.1202,
+      "step": 9180
+    },
+    {
+      "epoch": 2.8260451521095344,
+      "grad_norm": 1.2886162996292114,
+      "learning_rate": 4.1606226475350287e-07,
+      "loss": 1.2508,
+      "step": 9185
+    },
+    {
+      "epoch": 2.827583554478674,
+      "grad_norm": 2.6857354640960693,
+      "learning_rate": 4.087766353232103e-07,
+      "loss": 1.1784,
+      "step": 9190
+    },
+    {
+      "epoch": 2.8291219568478136,
+      "grad_norm": 1.6899442672729492,
+      "learning_rate": 4.015548337939962e-07,
+      "loss": 1.2262,
+      "step": 9195
+    },
+    {
+      "epoch": 2.830660359216953,
+      "grad_norm": 1.3128477334976196,
+      "learning_rate": 3.943968789104496e-07,
+      "loss": 1.186,
+      "step": 9200
+    },
+    {
+      "epoch": 2.832198761586093,
+      "grad_norm": 1.329758882522583,
+      "learning_rate": 3.87302789251448e-07,
+      "loss": 1.1431,
+      "step": 9205
+    },
+    {
+      "epoch": 2.8337371639552327,
+      "grad_norm": 1.1432725191116333,
+      "learning_rate": 3.8027258323010127e-07,
+      "loss": 1.0308,
+      "step": 9210
+    },
+    {
+      "epoch": 2.835275566324372,
+      "grad_norm": 1.351478099822998,
+      "learning_rate": 3.733062790936964e-07,
+      "loss": 1.0939,
+      "step": 9215
+    },
+    {
+      "epoch": 2.836813968693512,
+      "grad_norm": 1.2677690982818604,
+      "learning_rate": 3.6640389492367534e-07,
+      "loss": 1.1051,
+      "step": 9220
+    },
+    {
+      "epoch": 2.8383523710626513,
+      "grad_norm": 2.572026014328003,
+      "learning_rate": 3.5956544863555983e-07,
+      "loss": 1.0794,
+      "step": 9225
+    },
+    {
+      "epoch": 2.839890773431791,
+      "grad_norm": 1.2987449169158936,
+      "learning_rate": 3.5279095797892127e-07,
+      "loss": 1.2363,
+      "step": 9230
+    },
+    {
+      "epoch": 2.841429175800931,
+      "grad_norm": 1.7580738067626953,
+      "learning_rate": 3.460804405373302e-07,
+      "loss": 1.0583,
+      "step": 9235
+    },
+    {
+      "epoch": 2.8429675781700703,
+      "grad_norm": 1.4407984018325806,
+      "learning_rate": 3.394339137283098e-07,
+      "loss": 1.2122,
+      "step": 9240
+    },
+    {
+      "epoch": 2.84450598053921,
+      "grad_norm": 1.2053810358047485,
+      "learning_rate": 3.328513948032991e-07,
+      "loss": 1.1871,
+      "step": 9245
+    },
+    {
+      "epoch": 2.8460443829083495,
+      "grad_norm": 1.9730299711227417,
+      "learning_rate": 3.263329008475924e-07,
+      "loss": 1.0696,
+      "step": 9250
+    },
+    {
+      "epoch": 2.8475827852774893,
+      "grad_norm": 0.805873453617096,
+      "learning_rate": 3.1987844878030307e-07,
+      "loss": 1.0279,
+      "step": 9255
+    },
+    {
+      "epoch": 2.849121187646629,
+      "grad_norm": 1.1469902992248535,
+      "learning_rate": 3.1348805535432735e-07,
+      "loss": 1.1212,
+      "step": 9260
+    },
+    {
+      "epoch": 2.8506595900157685,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 3.071617371562946e-07,
+      "loss": 1.1438,
+      "step": 9265
+    },
+    {
+      "epoch": 2.8521979923849083,
+      "grad_norm": 1.4344321489334106,
+      "learning_rate": 3.0089951060651156e-07,
+      "loss": 1.2395,
+      "step": 9270
+    },
+    {
+      "epoch": 2.8537363947540477,
+      "grad_norm": 1.1882743835449219,
+      "learning_rate": 2.947013919589431e-07,
+      "loss": 1.0578,
+      "step": 9275
+    },
+    {
+      "epoch": 2.8552747971231875,
+      "grad_norm": 1.8451581001281738,
+      "learning_rate": 2.88567397301151e-07,
+      "loss": 1.1106,
+      "step": 9280
+    },
+    {
+      "epoch": 2.8568131994923274,
+      "grad_norm": 1.0893770456314087,
+      "learning_rate": 2.824975425542664e-07,
+      "loss": 1.0286,
+      "step": 9285
+    },
+    {
+      "epoch": 2.8583516018614668,
+      "grad_norm": 1.2297017574310303,
+      "learning_rate": 2.764918434729369e-07,
+      "loss": 1.1635,
+      "step": 9290
+    },
+    {
+      "epoch": 2.8598900042306066,
+      "grad_norm": 1.1600873470306396,
+      "learning_rate": 2.7055031564529043e-07,
+      "loss": 1.1557,
+      "step": 9295
+    },
+    {
+      "epoch": 2.861428406599746,
+      "grad_norm": 1.1898655891418457,
+      "learning_rate": 2.646729744928966e-07,
+      "loss": 1.0724,
+      "step": 9300
+    },
+    {
+      "epoch": 2.862966808968886,
+      "grad_norm": 1.7865312099456787,
+      "learning_rate": 2.588598352707278e-07,
+      "loss": 1.1252,
+      "step": 9305
+    },
+    {
+      "epoch": 2.8645052113380256,
+      "grad_norm": 1.6626031398773193,
+      "learning_rate": 2.531109130671061e-07,
+      "loss": 1.1687,
+      "step": 9310
+    },
+    {
+      "epoch": 2.866043613707165,
+      "grad_norm": 1.245299220085144,
+      "learning_rate": 2.474262228036872e-07,
+      "loss": 1.0814,
+      "step": 9315
+    },
+    {
+      "epoch": 2.867582016076305,
+      "grad_norm": 1.5507864952087402,
+      "learning_rate": 2.418057792354045e-07,
+      "loss": 1.085,
+      "step": 9320
+    },
+    {
+      "epoch": 2.869120418445444,
+      "grad_norm": 2.039008140563965,
+      "learning_rate": 2.3624959695043302e-07,
+      "loss": 1.0728,
+      "step": 9325
+    },
+    {
+      "epoch": 2.870658820814584,
+      "grad_norm": 1.1104451417922974,
+      "learning_rate": 2.3075769037015638e-07,
+      "loss": 1.1025,
+      "step": 9330
+    },
+    {
+      "epoch": 2.872197223183724,
+      "grad_norm": 1.1640777587890625,
+      "learning_rate": 2.2533007374912485e-07,
+      "loss": 1.2473,
+      "step": 9335
+    },
+    {
+      "epoch": 2.873735625552863,
+      "grad_norm": 1.0820621252059937,
+      "learning_rate": 2.1996676117502224e-07,
+      "loss": 1.1593,
+      "step": 9340
+    },
+    {
+      "epoch": 2.875274027922003,
+      "grad_norm": 1.981393575668335,
+      "learning_rate": 2.146677665686325e-07,
+      "loss": 1.2071,
+      "step": 9345
+    },
+    {
+      "epoch": 2.8768124302911424,
+      "grad_norm": 1.2725703716278076,
+      "learning_rate": 2.094331036837871e-07,
+      "loss": 1.179,
+      "step": 9350
+    },
+    {
+      "epoch": 2.8783508326602822,
+      "grad_norm": 1.1721535921096802,
+      "learning_rate": 2.0426278610735094e-07,
+      "loss": 1.1551,
+      "step": 9355
+    },
+    {
+      "epoch": 2.879889235029422,
+      "grad_norm": 1.6702649593353271,
+      "learning_rate": 1.9915682725917262e-07,
+      "loss": 0.9862,
+      "step": 9360
+    },
+    {
+      "epoch": 2.8814276373985614,
+      "grad_norm": 0.7977102994918823,
+      "learning_rate": 1.9411524039205376e-07,
+      "loss": 1.1702,
+      "step": 9365
+    },
+    {
+      "epoch": 2.8829660397677013,
+      "grad_norm": 1.0486433506011963,
+      "learning_rate": 1.891380385917213e-07,
+      "loss": 1.1837,
+      "step": 9370
+    },
+    {
+      "epoch": 2.8845044421368407,
+      "grad_norm": 1.5188802480697632,
+      "learning_rate": 1.842252347767748e-07,
+      "loss": 1.153,
+      "step": 9375
+    },
+    {
+      "epoch": 2.8860428445059805,
+      "grad_norm": 1.3773759603500366,
+      "learning_rate": 1.7937684169867797e-07,
+      "loss": 1.1127,
+      "step": 9380
+    },
+    {
+      "epoch": 2.8875812468751203,
+      "grad_norm": 1.7607942819595337,
+      "learning_rate": 1.7459287194170615e-07,
+      "loss": 1.1165,
+      "step": 9385
+    },
+    {
+      "epoch": 2.8891196492442597,
+      "grad_norm": 1.1703269481658936,
+      "learning_rate": 1.6987333792292115e-07,
+      "loss": 1.2549,
+      "step": 9390
+    },
+    {
+      "epoch": 2.8906580516133995,
+      "grad_norm": 1.1491749286651611,
+      "learning_rate": 1.6521825189213526e-07,
+      "loss": 1.1517,
+      "step": 9395
+    },
+    {
+      "epoch": 2.892196453982539,
+      "grad_norm": 1.7843077182769775,
+      "learning_rate": 1.6062762593188896e-07,
+      "loss": 1.0008,
+      "step": 9400
+    },
+    {
+      "epoch": 2.8937348563516787,
+      "grad_norm": 1.4365227222442627,
+      "learning_rate": 1.5610147195740943e-07,
+      "loss": 1.0936,
+      "step": 9405
+    },
+    {
+      "epoch": 2.8952732587208185,
+      "grad_norm": 1.1041733026504517,
+      "learning_rate": 1.5163980171658542e-07,
+      "loss": 1.171,
+      "step": 9410
+    },
+    {
+      "epoch": 2.896811661089958,
+      "grad_norm": 0.9337441921234131,
+      "learning_rate": 1.472426267899285e-07,
+      "loss": 1.1666,
+      "step": 9415
+    },
+    {
+      "epoch": 2.8983500634590977,
+      "grad_norm": 1.0685206651687622,
+      "learning_rate": 1.4290995859055633e-07,
+      "loss": 1.164,
+      "step": 9420
+    },
+    {
+      "epoch": 2.899888465828237,
+      "grad_norm": 1.5505434274673462,
+      "learning_rate": 1.386418083641483e-07,
+      "loss": 1.2611,
+      "step": 9425
+    },
+    {
+      "epoch": 2.901426868197377,
+      "grad_norm": 1.558901071548462,
+      "learning_rate": 1.3443818718893442e-07,
+      "loss": 1.1891,
+      "step": 9430
+    },
+    {
+      "epoch": 2.9029652705665168,
+      "grad_norm": 1.5380712747573853,
+      "learning_rate": 1.3029910597564532e-07,
+      "loss": 1.1009,
+      "step": 9435
+    },
+    {
+      "epoch": 2.9045036729356566,
+      "grad_norm": 1.8693783283233643,
+      "learning_rate": 1.2622457546749567e-07,
+      "loss": 1.1941,
+      "step": 9440
+    },
+    {
+      "epoch": 2.906042075304796,
+      "grad_norm": 1.0699331760406494,
+      "learning_rate": 1.2221460624016466e-07,
+      "loss": 1.1932,
+      "step": 9445
+    },
+    {
+      "epoch": 2.9075804776739353,
+      "grad_norm": 1.01662278175354,
+      "learning_rate": 1.1826920870174895e-07,
+      "loss": 1.2584,
+      "step": 9450
+    },
+    {
+      "epoch": 2.909118880043075,
+      "grad_norm": 1.0580191612243652,
+      "learning_rate": 1.143883930927514e-07,
+      "loss": 1.1569,
+      "step": 9455
+    },
+    {
+      "epoch": 2.910657282412215,
+      "grad_norm": 1.4499502182006836,
+      "learning_rate": 1.1057216948604509e-07,
+      "loss": 1.1326,
+      "step": 9460
+    },
+    {
+      "epoch": 2.912195684781355,
+      "grad_norm": 1.1739625930786133,
+      "learning_rate": 1.068205477868539e-07,
+      "loss": 1.011,
+      "step": 9465
+    },
+    {
+      "epoch": 2.913734087150494,
+      "grad_norm": 1.2261017560958862,
+      "learning_rate": 1.0313353773271917e-07,
+      "loss": 1.1797,
+      "step": 9470
+    },
+    {
+      "epoch": 2.915272489519634,
+      "grad_norm": 1.1943910121917725,
+      "learning_rate": 9.951114889348855e-08,
+      "loss": 1.1561,
+      "step": 9475
+    },
+    {
+      "epoch": 2.9168108918887734,
+      "grad_norm": 1.6418399810791016,
+      "learning_rate": 9.595339067127174e-08,
+      "loss": 1.084,
+      "step": 9480
+    },
+    {
+      "epoch": 2.9183492942579132,
+      "grad_norm": 1.0940558910369873,
+      "learning_rate": 9.24602723004292e-08,
+      "loss": 1.109,
+      "step": 9485
+    },
+    {
+      "epoch": 2.919887696627053,
+      "grad_norm": 1.4907678365707397,
+      "learning_rate": 8.903180284755008e-08,
+      "loss": 1.0309,
+      "step": 9490
+    },
+    {
+      "epoch": 2.9214260989961924,
+      "grad_norm": 1.6203380823135376,
+      "learning_rate": 8.566799121141334e-08,
+      "loss": 1.1081,
+      "step": 9495
+    },
+    {
+      "epoch": 2.9229645013653323,
+      "grad_norm": 1.3921715021133423,
+      "learning_rate": 8.23688461229849e-08,
+      "loss": 1.1672,
+      "step": 9500
+    },
+    {
+      "epoch": 2.9245029037344716,
+      "grad_norm": 1.6819571256637573,
+      "learning_rate": 7.913437614538166e-08,
+      "loss": 1.0484,
+      "step": 9505
+    },
+    {
+      "epoch": 2.9260413061036115,
+      "grad_norm": 1.9692211151123047,
+      "learning_rate": 7.596458967384922e-08,
+      "loss": 1.2351,
+      "step": 9510
+    },
+    {
+      "epoch": 2.9275797084727513,
+      "grad_norm": 1.2511972188949585,
+      "learning_rate": 7.285949493574806e-08,
+      "loss": 1.2732,
+      "step": 9515
+    },
+    {
+      "epoch": 2.9291181108418907,
+      "grad_norm": 0.920283854007721,
+      "learning_rate": 6.98190999905285e-08,
+      "loss": 1.1274,
+      "step": 9520
+    },
+    {
+      "epoch": 2.9306565132110305,
+      "grad_norm": 1.1254204511642456,
+      "learning_rate": 6.684341272970018e-08,
+      "loss": 1.023,
+      "step": 9525
+    },
+    {
+      "epoch": 2.93219491558017,
+      "grad_norm": 0.9248682856559753,
+      "learning_rate": 6.393244087683215e-08,
+      "loss": 1.2624,
+      "step": 9530
+    },
+    {
+      "epoch": 2.9337333179493097,
+      "grad_norm": 1.0681920051574707,
+      "learning_rate": 6.108619198751109e-08,
+      "loss": 1.1346,
+      "step": 9535
+    },
+    {
+      "epoch": 2.9352717203184495,
+      "grad_norm": 1.8738933801651,
+      "learning_rate": 5.8304673449338653e-08,
+      "loss": 1.1768,
+      "step": 9540
+    },
+    {
+      "epoch": 2.936810122687589,
+      "grad_norm": 1.5245164632797241,
+      "learning_rate": 5.558789248190366e-08,
+      "loss": 1.143,
+      "step": 9545
+    },
+    {
+      "epoch": 2.9383485250567287,
+      "grad_norm": 1.1693507432937622,
+      "learning_rate": 5.293585613675989e-08,
+      "loss": 1.1751,
+      "step": 9550
+    },
+    {
+      "epoch": 2.939886927425868,
+      "grad_norm": 1.6646199226379395,
+      "learning_rate": 5.034857129741777e-08,
+      "loss": 1.122,
+      "step": 9555
+    },
+    {
+      "epoch": 2.941425329795008,
+      "grad_norm": 1.3896692991256714,
+      "learning_rate": 4.782604467931939e-08,
+      "loss": 1.1138,
+      "step": 9560
+    },
+    {
+      "epoch": 2.9429637321641477,
+      "grad_norm": 1.3115090131759644,
+      "learning_rate": 4.5368282829827415e-08,
+      "loss": 1.1148,
+      "step": 9565
+    },
+    {
+      "epoch": 2.944502134533287,
+      "grad_norm": 1.0006541013717651,
+      "learning_rate": 4.2975292128200064e-08,
+      "loss": 1.2622,
+      "step": 9570
+    },
+    {
+      "epoch": 2.946040536902427,
+      "grad_norm": 1.6213256120681763,
+      "learning_rate": 4.064707878557728e-08,
+      "loss": 1.2263,
+      "step": 9575
+    },
+    {
+      "epoch": 2.9475789392715663,
+      "grad_norm": 4.496711254119873,
+      "learning_rate": 3.838364884496681e-08,
+      "loss": 1.2043,
+      "step": 9580
+    },
+    {
+      "epoch": 2.949117341640706,
+      "grad_norm": 1.6494616270065308,
+      "learning_rate": 3.618500818123039e-08,
+      "loss": 1.1806,
+      "step": 9585
+    },
+    {
+      "epoch": 2.950655744009846,
+      "grad_norm": 1.9413509368896484,
+      "learning_rate": 3.405116250106144e-08,
+      "loss": 1.2,
+      "step": 9590
+    },
+    {
+      "epoch": 2.9521941463789854,
+      "grad_norm": 2.1733412742614746,
+      "learning_rate": 3.1982117342979624e-08,
+      "loss": 1.0731,
+      "step": 9595
+    },
+    {
+      "epoch": 2.953732548748125,
+      "grad_norm": 1.8069325685501099,
+      "learning_rate": 2.9977878077305785e-08,
+      "loss": 1.2088,
+      "step": 9600
+    },
+    {
+      "epoch": 2.9552709511172646,
+      "grad_norm": 2.179551362991333,
+      "learning_rate": 2.8038449906153673e-08,
+      "loss": 1.082,
+      "step": 9605
+    },
+    {
+      "epoch": 2.9568093534864044,
+      "grad_norm": 1.2581409215927124,
+      "learning_rate": 2.6163837863418806e-08,
+      "loss": 1.076,
+      "step": 9610
+    },
+    {
+      "epoch": 2.958347755855544,
+      "grad_norm": 1.4859249591827393,
+      "learning_rate": 2.4354046814764607e-08,
+      "loss": 1.085,
+      "step": 9615
+    },
+    {
+      "epoch": 2.9598861582246836,
+      "grad_norm": 1.2801562547683716,
+      "learning_rate": 2.260908145760299e-08,
+      "loss": 1.1067,
+      "step": 9620
+    },
+    {
+      "epoch": 2.9614245605938234,
+      "grad_norm": 1.3596243858337402,
+      "learning_rate": 2.0928946321091547e-08,
+      "loss": 1.2352,
+      "step": 9625
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 1.4657814502716064,
+      "learning_rate": 1.931364576611139e-08,
+      "loss": 1.1577,
+      "step": 9630
+    },
+    {
+      "epoch": 2.9645013653321026,
+      "grad_norm": 1.2036793231964111,
+      "learning_rate": 1.7763183985269883e-08,
+      "loss": 1.1403,
+      "step": 9635
+    },
+    {
+      "epoch": 2.9660397677012424,
+      "grad_norm": 2.143611192703247,
+      "learning_rate": 1.6277565002875696e-08,
+      "loss": 1.0896,
+      "step": 9640
+    },
+    {
+      "epoch": 2.967578170070382,
+      "grad_norm": 1.2435050010681152,
+      "learning_rate": 1.4856792674936004e-08,
+      "loss": 1.2366,
+      "step": 9645
+    },
+    {
+      "epoch": 2.9691165724395217,
+      "grad_norm": 1.358459711074829,
+      "learning_rate": 1.3500870689145407e-08,
+      "loss": 1.2478,
+      "step": 9650
+    },
+    {
+      "epoch": 2.970654974808661,
+      "grad_norm": 1.11142897605896,
+      "learning_rate": 1.2209802564877582e-08,
+      "loss": 1.2521,
+      "step": 9655
+    },
+    {
+      "epoch": 2.972193377177801,
+      "grad_norm": 1.6148406267166138,
+      "learning_rate": 1.0983591653168645e-08,
+      "loss": 1.275,
+      "step": 9660
+    },
+    {
+      "epoch": 2.9737317795469407,
+      "grad_norm": 0.9854974150657654,
+      "learning_rate": 9.822241136722699e-09,
+      "loss": 1.0939,
+      "step": 9665
+    },
+    {
+      "epoch": 2.97527018191608,
+      "grad_norm": 1.0898098945617676,
+      "learning_rate": 8.72575402988407e-09,
+      "loss": 1.1264,
+      "step": 9670
+    },
+    {
+      "epoch": 2.97680858428522,
+      "grad_norm": 1.4026813507080078,
+      "learning_rate": 7.694133178653973e-09,
+      "loss": 1.0411,
+      "step": 9675
+    },
+    {
+      "epoch": 2.9783469866543593,
+      "grad_norm": 1.993453025817871,
+      "learning_rate": 6.727381260657195e-09,
+      "loss": 1.2356,
+      "step": 9680
+    },
+    {
+      "epoch": 2.979885389023499,
+      "grad_norm": 1.4184306859970093,
+      "learning_rate": 5.825500785150428e-09,
+      "loss": 1.139,
+      "step": 9685
+    },
+    {
+      "epoch": 2.981423791392639,
+      "grad_norm": 1.8053029775619507,
+      "learning_rate": 4.988494093022267e-09,
+      "loss": 1.1943,
+      "step": 9690
+    },
+    {
+      "epoch": 2.9829621937617783,
+      "grad_norm": 1.1999495029449463,
+      "learning_rate": 4.216363356765452e-09,
+      "loss": 1.2553,
+      "step": 9695
+    },
+    {
+      "epoch": 2.984500596130918,
+      "grad_norm": 1.530155062675476,
+      "learning_rate": 3.5091105804907487e-09,
+      "loss": 1.1807,
+      "step": 9700
+    },
+    {
+      "epoch": 2.9860389985000575,
+      "grad_norm": 1.3728301525115967,
+      "learning_rate": 2.8667375999102964e-09,
+      "loss": 1.2475,
+      "step": 9705
+    },
+    {
+      "epoch": 2.9875774008691973,
+      "grad_norm": 1.2360360622406006,
+      "learning_rate": 2.2892460823403794e-09,
+      "loss": 0.998,
+      "step": 9710
+    },
+    {
+      "epoch": 2.989115803238337,
+      "grad_norm": 1.8913969993591309,
+      "learning_rate": 1.7766375266931035e-09,
+      "loss": 1.0908,
+      "step": 9715
+    },
+    {
+      "epoch": 2.9906542056074765,
+      "grad_norm": 0.986847460269928,
+      "learning_rate": 1.328913263473619e-09,
+      "loss": 1.07,
+      "step": 9720
+    },
+    {
+      "epoch": 2.9921926079766163,
+      "grad_norm": 1.6557801961898804,
+      "learning_rate": 9.460744547745704e-10,
+      "loss": 1.1596,
+      "step": 9725
+    },
+    {
+      "epoch": 2.9937310103457557,
+      "grad_norm": 2.2507128715515137,
+      "learning_rate": 6.281220942733201e-10,
+      "loss": 1.0748,
+      "step": 9730
+    },
+    {
+      "epoch": 2.9952694127148956,
+      "grad_norm": 2.129380702972412,
+      "learning_rate": 3.750570072375004e-10,
+      "loss": 1.3533,
+      "step": 9735
+    },
+    {
+      "epoch": 2.9968078150840354,
+      "grad_norm": 1.2016565799713135,
+      "learning_rate": 1.8687985050558355e-10,
+      "loss": 1.0902,
+      "step": 9740
+    },
+    {
+      "epoch": 2.9983462174531748,
+      "grad_norm": 1.1874008178710938,
+      "learning_rate": 6.359111250908711e-11,
+      "loss": 1.1668,
+      "step": 9745
+    },
+    {
+      "epoch": 2.9998846198223146,
+      "grad_norm": 1.8799904584884644,
+      "learning_rate": 5.191113247593471e-12,
+      "loss": 1.1766,
+      "step": 9750
+    },
+    {
+      "epoch": 2.9998846198223146,
+      "step": 9750,
+      "total_flos": 7.742111986089984e+17,
+      "train_loss": 1.1629726183475593,
+      "train_runtime": 20329.8435,
+      "train_samples_per_second": 7.674,
+      "train_steps_per_second": 0.48
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 9750,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 7.742111986089984e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}