{
  "best_global_step": 5000,
  "best_metric": 0.8924599885940552,
  "best_model_checkpoint": "/kaggle/working/tts_processed/checkpoint-5000",
  "epoch": 24.752475247524753,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.12376237623762376,
      "grad_norm": 363.43377685546875,
      "learning_rate": 4.8e-08,
      "loss": 10.89256103515625,
      "step": 25
    },
    {
      "epoch": 0.24752475247524752,
      "grad_norm": 336.6938171386719,
      "learning_rate": 9.8e-08,
      "loss": 10.47541259765625,
      "step": 50
    },
    {
      "epoch": 0.3712871287128713,
      "grad_norm": 255.0070343017578,
      "learning_rate": 1.4799999999999998e-07,
      "loss": 10.08507568359375,
      "step": 75
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 247.54541015625,
      "learning_rate": 1.98e-07,
      "loss": 9.282888793945313,
      "step": 100
    },
    {
      "epoch": 0.6188118811881188,
      "grad_norm": 124.208740234375,
      "learning_rate": 2.48e-07,
      "loss": 8.717689208984375,
      "step": 125
    },
    {
      "epoch": 0.7425742574257426,
      "grad_norm": 194.5498809814453,
      "learning_rate": 2.98e-07,
      "loss": 8.212156372070313,
      "step": 150
    },
    {
      "epoch": 0.8663366336633663,
      "grad_norm": 143.2777557373047,
      "learning_rate": 3.4799999999999994e-07,
      "loss": 7.8293896484375,
      "step": 175
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 55.50076675415039,
      "learning_rate": 3.98e-07,
      "loss": 7.3809716796875,
      "step": 200
    },
    {
      "epoch": 1.113861386138614,
      "grad_norm": 69.43047332763672,
      "learning_rate": 4.48e-07,
      "loss": 6.937587890625,
      "step": 225
    },
    {
      "epoch": 1.2376237623762376,
      "grad_norm": 49.69321060180664,
      "learning_rate": 4.979999999999999e-07,
      "loss": 6.8110888671875,
      "step": 250
    },
    {
      "epoch": 1.3613861386138613,
      "grad_norm": 26.72427749633789,
      "learning_rate": 5.48e-07,
      "loss": 6.522671508789062,
      "step": 275
    },
    {
      "epoch": 1.4851485148514851,
      "grad_norm": 51.36345672607422,
      "learning_rate": 5.979999999999999e-07,
      "loss": 6.398966064453125,
      "step": 300
    },
    {
      "epoch": 1.608910891089109,
      "grad_norm": 38.85439682006836,
      "learning_rate": 6.48e-07,
      "loss": 6.198662109375,
      "step": 325
    },
    {
      "epoch": 1.7326732673267327,
      "grad_norm": 27.508665084838867,
      "learning_rate": 6.979999999999999e-07,
      "loss": 6.075740356445312,
      "step": 350
    },
    {
      "epoch": 1.8564356435643563,
      "grad_norm": 36.8642463684082,
      "learning_rate": 7.48e-07,
      "loss": 6.0834722900390625,
      "step": 375
    },
    {
      "epoch": 1.9801980198019802,
      "grad_norm": 27.07192039489746,
      "learning_rate": 7.98e-07,
      "loss": 5.9166015625,
      "step": 400
    },
    {
      "epoch": 2.103960396039604,
      "grad_norm": 22.465463638305664,
      "learning_rate": 8.48e-07,
      "loss": 5.795056762695313,
      "step": 425
    },
    {
      "epoch": 2.227722772277228,
      "grad_norm": 34.86723709106445,
      "learning_rate": 8.98e-07,
      "loss": 5.741965942382812,
      "step": 450
    },
    {
      "epoch": 2.3514851485148514,
      "grad_norm": 22.37932777404785,
      "learning_rate": 9.479999999999999e-07,
      "loss": 5.716455688476563,
      "step": 475
    },
    {
      "epoch": 2.4752475247524752,
      "grad_norm": 29.54213523864746,
      "learning_rate": 9.98e-07,
      "loss": 5.599214477539062,
      "step": 500
    },
    {
      "epoch": 2.4752475247524752,
      "eval_loss": 1.2534884214401245,
      "eval_runtime": 82.4141,
      "eval_samples_per_second": 19.608,
      "eval_steps_per_second": 4.902,
      "step": 500
    },
    {
      "epoch": 2.599009900990099,
      "grad_norm": 21.54619026184082,
      "learning_rate": 9.999953592789052e-07,
      "loss": 5.526549072265625,
      "step": 525
    },
    {
      "epoch": 2.7227722772277225,
      "grad_norm": 31.9982967376709,
      "learning_rate": 9.999806557001091e-07,
      "loss": 5.38673828125,
      "step": 550
    },
    {
      "epoch": 2.8465346534653464,
      "grad_norm": 19.33993148803711,
      "learning_rate": 9.999558815029513e-07,
      "loss": 5.291504516601562,
      "step": 575
    },
    {
      "epoch": 2.9702970297029703,
      "grad_norm": 18.838518142700195,
      "learning_rate": 9.999210371864342e-07,
      "loss": 5.345247192382812,
      "step": 600
    },
    {
      "epoch": 3.094059405940594,
      "grad_norm": 19.941740036010742,
      "learning_rate": 9.998761234523925e-07,
      "loss": 5.081980285644531,
      "step": 625
    },
    {
      "epoch": 3.217821782178218,
      "grad_norm": 24.84575080871582,
      "learning_rate": 9.998211412054793e-07,
      "loss": 5.025005493164063,
      "step": 650
    },
    {
      "epoch": 3.3415841584158414,
      "grad_norm": 24.532045364379883,
      "learning_rate": 9.99756091553148e-07,
      "loss": 5.004798889160156,
      "step": 675
    },
    {
      "epoch": 3.4653465346534653,
      "grad_norm": 21.20513916015625,
      "learning_rate": 9.996809758056298e-07,
      "loss": 4.870877685546875,
      "step": 700
    },
    {
      "epoch": 3.589108910891089,
      "grad_norm": 16.589248657226562,
      "learning_rate": 9.995957954759072e-07,
      "loss": 4.95923095703125,
      "step": 725
    },
    {
      "epoch": 3.7128712871287126,
      "grad_norm": 26.863059997558594,
      "learning_rate": 9.995005522796837e-07,
      "loss": 4.81072265625,
      "step": 750
    },
    {
      "epoch": 3.8366336633663365,
      "grad_norm": 14.973981857299805,
      "learning_rate": 9.993952481353493e-07,
      "loss": 4.8048876953125,
      "step": 775
    },
    {
      "epoch": 3.9603960396039604,
      "grad_norm": 18.018110275268555,
      "learning_rate": 9.992798851639418e-07,
      "loss": 4.770014343261718,
      "step": 800
    },
    {
      "epoch": 4.084158415841584,
      "grad_norm": 15.32919979095459,
      "learning_rate": 9.991544656891043e-07,
      "loss": 4.663901672363282,
      "step": 825
    },
    {
      "epoch": 4.207920792079208,
      "grad_norm": 14.925799369812012,
      "learning_rate": 9.990189922370373e-07,
      "loss": 4.614352111816406,
      "step": 850
    },
    {
      "epoch": 4.3316831683168315,
      "grad_norm": 14.5320405960083,
      "learning_rate": 9.988734675364496e-07,
      "loss": 4.607881774902344,
      "step": 875
    },
    {
      "epoch": 4.455445544554456,
      "grad_norm": 17.06329345703125,
      "learning_rate": 9.987178945185019e-07,
      "loss": 4.582254028320312,
      "step": 900
    },
    {
      "epoch": 4.579207920792079,
      "grad_norm": 16.860803604125977,
      "learning_rate": 9.985522763167486e-07,
      "loss": 4.6025848388671875,
      "step": 925
    },
    {
      "epoch": 4.702970297029703,
      "grad_norm": 12.743483543395996,
      "learning_rate": 9.983766162670739e-07,
      "loss": 4.603981628417968,
      "step": 950
    },
    {
      "epoch": 4.826732673267327,
      "grad_norm": 11.648216247558594,
      "learning_rate": 9.981909179076255e-07,
      "loss": 4.5333544921875,
      "step": 975
    },
    {
      "epoch": 4.9504950495049505,
      "grad_norm": 12.38879108428955,
      "learning_rate": 9.979951849787429e-07,
      "loss": 4.538429565429688,
      "step": 1000
    },
    {
      "epoch": 4.9504950495049505,
      "eval_loss": 1.0112695693969727,
      "eval_runtime": 82.132,
      "eval_samples_per_second": 19.676,
      "eval_steps_per_second": 4.919,
      "step": 1000
    },
    {
      "epoch": 5.074257425742574,
      "grad_norm": 39.850196838378906,
      "learning_rate": 9.977894214228815e-07,
      "loss": 4.55607666015625,
      "step": 1025
    },
    {
      "epoch": 5.198019801980198,
      "grad_norm": 13.423649787902832,
      "learning_rate": 9.975736313845345e-07,
      "loss": 4.43293212890625,
      "step": 1050
    },
    {
      "epoch": 5.321782178217822,
      "grad_norm": 15.34720516204834,
      "learning_rate": 9.97347819210148e-07,
      "loss": 4.4848849487304685,
      "step": 1075
    },
    {
      "epoch": 5.445544554455446,
      "grad_norm": 10.895915031433105,
      "learning_rate": 9.971119894480344e-07,
      "loss": 4.418946228027344,
      "step": 1100
    },
    {
      "epoch": 5.569306930693069,
      "grad_norm": 13.77745532989502,
      "learning_rate": 9.96866146848281e-07,
      "loss": 4.675699768066406,
      "step": 1125
    },
    {
      "epoch": 5.693069306930693,
      "grad_norm": 21.53974723815918,
      "learning_rate": 9.96610296362653e-07,
      "loss": 4.483356628417969,
      "step": 1150
    },
    {
      "epoch": 5.816831683168317,
      "grad_norm": 14.47573184967041,
      "learning_rate": 9.963444431444948e-07,
      "loss": 4.555756530761719,
      "step": 1175
    },
    {
      "epoch": 5.9405940594059405,
      "grad_norm": 11.701032638549805,
      "learning_rate": 9.960685925486265e-07,
      "loss": 4.419072265625,
      "step": 1200
    },
    {
      "epoch": 6.064356435643564,
      "grad_norm": 17.197729110717773,
      "learning_rate": 9.957827501312352e-07,
      "loss": 4.375239562988281,
      "step": 1225
    },
    {
      "epoch": 6.188118811881188,
      "grad_norm": 12.936083793640137,
      "learning_rate": 9.954869216497633e-07,
      "loss": 4.3699853515625,
      "step": 1250
    },
    {
      "epoch": 6.311881188118812,
      "grad_norm": 15.114112854003906,
      "learning_rate": 9.95181113062793e-07,
      "loss": 4.41375732421875,
      "step": 1275
    },
    {
      "epoch": 6.435643564356436,
      "grad_norm": 17.72418785095215,
      "learning_rate": 9.948653305299256e-07,
      "loss": 4.430754089355469,
      "step": 1300
    },
    {
      "epoch": 6.5594059405940595,
      "grad_norm": 18.342178344726562,
      "learning_rate": 9.945395804116578e-07,
      "loss": 4.4040380859375,
      "step": 1325
    },
    {
      "epoch": 6.683168316831683,
      "grad_norm": 12.649271965026855,
      "learning_rate": 9.94203869269254e-07,
      "loss": 4.3536474609375,
      "step": 1350
    },
    {
      "epoch": 6.806930693069307,
      "grad_norm": 32.92485809326172,
      "learning_rate": 9.938582038646131e-07,
      "loss": 4.349139709472656,
      "step": 1375
    },
    {
      "epoch": 6.930693069306931,
      "grad_norm": 11.99206829071045,
      "learning_rate": 9.935025911601334e-07,
      "loss": 4.317200317382812,
      "step": 1400
    },
    {
      "epoch": 7.054455445544554,
      "grad_norm": 15.798395156860352,
      "learning_rate": 9.931370383185716e-07,
      "loss": 4.390576782226563,
      "step": 1425
    },
    {
      "epoch": 7.178217821782178,
      "grad_norm": 12.485000610351562,
      "learning_rate": 9.92761552702899e-07,
      "loss": 4.415244750976562,
      "step": 1450
    },
    {
      "epoch": 7.301980198019802,
      "grad_norm": 14.873079299926758,
      "learning_rate": 9.923761418761526e-07,
      "loss": 4.3524868774414065,
      "step": 1475
    },
    {
      "epoch": 7.425742574257426,
      "grad_norm": 13.601946830749512,
      "learning_rate": 9.919808136012837e-07,
      "loss": 4.294336547851563,
      "step": 1500
    },
    {
      "epoch": 7.425742574257426,
      "eval_loss": 0.9704130291938782,
      "eval_runtime": 86.4979,
      "eval_samples_per_second": 18.683,
      "eval_steps_per_second": 4.671,
      "step": 1500
    },
    {
      "epoch": 7.5495049504950495,
      "grad_norm": 15.53011417388916,
      "learning_rate": 9.915755758410008e-07,
      "loss": 4.254639587402344,
      "step": 1525
    },
    {
      "epoch": 7.673267326732673,
      "grad_norm": 11.795248985290527,
      "learning_rate": 9.91160436757609e-07,
      "loss": 4.307816467285156,
      "step": 1550
    },
    {
      "epoch": 7.797029702970297,
      "grad_norm": 15.00704574584961,
      "learning_rate": 9.907354047128464e-07,
      "loss": 4.349948120117188,
      "step": 1575
    },
    {
      "epoch": 7.920792079207921,
      "grad_norm": 13.17168140411377,
      "learning_rate": 9.903004882677155e-07,
      "loss": 4.2257516479492185,
      "step": 1600
    },
    {
      "epoch": 8.044554455445544,
      "grad_norm": 13.770243644714355,
      "learning_rate": 9.898556961823103e-07,
      "loss": 4.310787048339844,
      "step": 1625
    },
    {
      "epoch": 8.168316831683168,
      "grad_norm": 13.734089851379395,
      "learning_rate": 9.894010374156398e-07,
      "loss": 4.148223266601563,
      "step": 1650
    },
    {
      "epoch": 8.292079207920793,
      "grad_norm": 15.90095329284668,
      "learning_rate": 9.889365211254485e-07,
      "loss": 4.3510107421875,
      "step": 1675
    },
    {
      "epoch": 8.415841584158416,
      "grad_norm": 12.133323669433594,
      "learning_rate": 9.884621566680306e-07,
      "loss": 4.337423400878906,
      "step": 1700
    },
    {
      "epoch": 8.53960396039604,
      "grad_norm": 50.35639190673828,
      "learning_rate": 9.87977953598043e-07,
      "loss": 4.3044869995117185,
      "step": 1725
    },
    {
      "epoch": 8.663366336633663,
      "grad_norm": 12.457736015319824,
      "learning_rate": 9.874839216683112e-07,
      "loss": 4.3085205078125,
      "step": 1750
    },
    {
      "epoch": 8.787128712871286,
      "grad_norm": 14.483731269836426,
      "learning_rate": 9.869800708296346e-07,
      "loss": 4.1831143188476565,
      "step": 1775
    },
    {
      "epoch": 8.910891089108912,
      "grad_norm": 11.174741744995117,
      "learning_rate": 9.864664112305846e-07,
      "loss": 4.162017822265625,
      "step": 1800
    },
    {
      "epoch": 9.034653465346535,
      "grad_norm": 10.312405586242676,
      "learning_rate": 9.859429532173012e-07,
      "loss": 4.320256958007812,
      "step": 1825
    },
    {
      "epoch": 9.158415841584159,
      "grad_norm": 13.814125061035156,
      "learning_rate": 9.85409707333284e-07,
      "loss": 4.314511413574219,
      "step": 1850
    },
    {
      "epoch": 9.282178217821782,
      "grad_norm": 11.276887893676758,
      "learning_rate": 9.848666843191804e-07,
      "loss": 4.327219543457031,
      "step": 1875
    },
    {
      "epoch": 9.405940594059405,
      "grad_norm": 13.48475170135498,
      "learning_rate": 9.84313895112568e-07,
      "loss": 4.2580517578125,
      "step": 1900
    },
    {
      "epoch": 9.52970297029703,
      "grad_norm": 13.70500659942627,
      "learning_rate": 9.837513508477365e-07,
      "loss": 4.269812316894531,
      "step": 1925
    },
    {
      "epoch": 9.653465346534654,
      "grad_norm": 12.701525688171387,
      "learning_rate": 9.831790628554612e-07,
      "loss": 4.257364196777344,
      "step": 1950
    },
    {
      "epoch": 9.777227722772277,
      "grad_norm": 11.086860656738281,
      "learning_rate": 9.82597042662776e-07,
      "loss": 4.120067443847656,
      "step": 1975
    },
    {
      "epoch": 9.900990099009901,
      "grad_norm": 14.811179161071777,
      "learning_rate": 9.820053019927408e-07,
      "loss": 4.13207275390625,
      "step": 2000
    },
    {
      "epoch": 9.900990099009901,
      "eval_loss": 0.9480290412902832,
      "eval_runtime": 81.9519,
      "eval_samples_per_second": 19.719,
      "eval_steps_per_second": 4.93,
      "step": 2000
    },
    {
      "epoch": 10.024752475247524,
      "grad_norm": 13.266257286071777,
      "learning_rate": 9.814038527642059e-07,
      "loss": 4.281947021484375,
      "step": 2025
    },
    {
      "epoch": 10.148514851485148,
      "grad_norm": 13.618546485900879,
      "learning_rate": 9.80792707091571e-07,
      "loss": 4.166450500488281,
      "step": 2050
    },
    {
      "epoch": 10.272277227722773,
      "grad_norm": 12.208344459533691,
      "learning_rate": 9.801718772845422e-07,
      "loss": 4.151670227050781,
      "step": 2075
    },
    {
      "epoch": 10.396039603960396,
      "grad_norm": 12.220300674438477,
      "learning_rate": 9.795413758478835e-07,
      "loss": 4.250581359863281,
      "step": 2100
    },
    {
      "epoch": 10.51980198019802,
      "grad_norm": 10.937910079956055,
      "learning_rate": 9.789012154811646e-07,
      "loss": 4.219160461425782,
      "step": 2125
    },
    {
      "epoch": 10.643564356435643,
      "grad_norm": 12.94577407836914,
      "learning_rate": 9.78251409078506e-07,
      "loss": 4.256142272949218,
      "step": 2150
    },
    {
      "epoch": 10.767326732673267,
      "grad_norm": 12.101743698120117,
      "learning_rate": 9.77591969728319e-07,
      "loss": 4.145942077636719,
      "step": 2175
    },
    {
      "epoch": 10.891089108910892,
      "grad_norm": 15.079996109008789,
      "learning_rate": 9.769229107130408e-07,
      "loss": 4.2266848754882815,
      "step": 2200
    },
    {
      "epoch": 11.014851485148515,
      "grad_norm": 13.785998344421387,
      "learning_rate": 9.762442455088694e-07,
      "loss": 4.173150634765625,
      "step": 2225
    },
    {
      "epoch": 11.138613861386139,
      "grad_norm": 33.18384552001953,
      "learning_rate": 9.7555598778549e-07,
      "loss": 4.193008728027344,
      "step": 2250
    },
    {
      "epoch": 11.262376237623762,
      "grad_norm": 17.419424057006836,
      "learning_rate": 9.748581514058005e-07,
      "loss": 4.254207153320312,
      "step": 2275
    },
    {
      "epoch": 11.386138613861386,
      "grad_norm": 11.025071144104004,
      "learning_rate": 9.741507504256326e-07,
      "loss": 4.130965270996094,
      "step": 2300
    },
    {
      "epoch": 11.509900990099009,
      "grad_norm": 11.500884056091309,
      "learning_rate": 9.734337990934681e-07,
      "loss": 4.12019775390625,
      "step": 2325
    },
    {
      "epoch": 11.633663366336634,
      "grad_norm": 13.27956771850586,
      "learning_rate": 9.727073118501523e-07,
      "loss": 4.216612854003906,
      "step": 2350
    },
    {
      "epoch": 11.757425742574258,
      "grad_norm": 16.143634796142578,
      "learning_rate": 9.719713033286028e-07,
      "loss": 4.163922424316406,
      "step": 2375
    },
    {
      "epoch": 11.881188118811881,
      "grad_norm": 14.004642486572266,
      "learning_rate": 9.712257883535152e-07,
      "loss": 4.1760952758789065,
      "step": 2400
    },
    {
      "epoch": 12.004950495049505,
      "grad_norm": 11.351595878601074,
      "learning_rate": 9.704707819410634e-07,
      "loss": 4.229373168945313,
      "step": 2425
    },
    {
      "epoch": 12.128712871287128,
      "grad_norm": 9.855143547058105,
      "learning_rate": 9.697062992985993e-07,
      "loss": 4.096666259765625,
      "step": 2450
    },
    {
      "epoch": 12.252475247524753,
      "grad_norm": 11.369580268859863,
      "learning_rate": 9.689323558243444e-07,
      "loss": 4.218254089355469,
      "step": 2475
    },
    {
      "epoch": 12.376237623762377,
      "grad_norm": 9.929821014404297,
      "learning_rate": 9.681489671070802e-07,
      "loss": 4.054186401367187,
      "step": 2500
    },
    {
      "epoch": 12.376237623762377,
      "eval_loss": 0.9306696653366089,
      "eval_runtime": 82.6001,
      "eval_samples_per_second": 19.564,
      "eval_steps_per_second": 4.891,
      "step": 2500
    },
    {
      "epoch": 12.5,
      "grad_norm": 23.94037628173828,
      "learning_rate": 9.673561489258352e-07,
      "loss": 4.108751220703125,
      "step": 2525
    },
    {
      "epoch": 12.623762376237623,
      "grad_norm": 10.571653366088867,
      "learning_rate": 9.665539172495656e-07,
      "loss": 4.077666320800781,
      "step": 2550
    },
    {
      "epoch": 12.747524752475247,
      "grad_norm": 12.341602325439453,
      "learning_rate": 9.657422882368347e-07,
      "loss": 4.187429504394531,
      "step": 2575
    },
    {
      "epoch": 12.871287128712872,
      "grad_norm": 9.840188026428223,
      "learning_rate": 9.64921278235487e-07,
      "loss": 4.135216369628906,
      "step": 2600
    },
    {
      "epoch": 12.995049504950495,
      "grad_norm": 11.124446868896484,
      "learning_rate": 9.640909037823194e-07,
      "loss": 4.087601318359375,
      "step": 2625
    },
    {
      "epoch": 13.118811881188119,
      "grad_norm": 11.180383682250977,
      "learning_rate": 9.63251181602747e-07,
      "loss": 4.0938818359375,
      "step": 2650
    },
    {
      "epoch": 13.242574257425742,
      "grad_norm": 12.539691925048828,
      "learning_rate": 9.624021286104674e-07,
      "loss": 4.085841369628906,
      "step": 2675
    },
    {
      "epoch": 13.366336633663366,
      "grad_norm": 11.089624404907227,
      "learning_rate": 9.615437619071198e-07,
      "loss": 4.057616577148438,
      "step": 2700
    },
    {
      "epoch": 13.490099009900991,
      "grad_norm": 10.118108749389648,
      "learning_rate": 9.6067609878194e-07,
      "loss": 4.123738403320313,
      "step": 2725
    },
    {
      "epoch": 13.613861386138614,
      "grad_norm": 13.25112247467041,
      "learning_rate": 9.597991567114124e-07,
      "loss": 4.035686645507813,
      "step": 2750
    },
    {
      "epoch": 13.737623762376238,
      "grad_norm": 10.743168830871582,
      "learning_rate": 9.589129533589185e-07,
      "loss": 4.089259643554687,
      "step": 2775
    },
    {
      "epoch": 13.861386138613861,
      "grad_norm": 13.900023460388184,
      "learning_rate": 9.580175065743798e-07,
      "loss": 4.086239929199219,
      "step": 2800
    },
    {
      "epoch": 13.985148514851485,
      "grad_norm": 14.214149475097656,
      "learning_rate": 9.571128343939005e-07,
      "loss": 4.133528747558594,
      "step": 2825
    },
    {
      "epoch": 14.108910891089108,
      "grad_norm": 13.158830642700195,
      "learning_rate": 9.561989550394016e-07,
      "loss": 4.137518005371094,
      "step": 2850
    },
    {
      "epoch": 14.232673267326733,
      "grad_norm": 13.256293296813965,
      "learning_rate": 9.552758869182557e-07,
      "loss": 4.049294128417968,
      "step": 2875
    },
    {
      "epoch": 14.356435643564357,
      "grad_norm": 10.833374977111816,
      "learning_rate": 9.543436486229161e-07,
      "loss": 4.0441677856445315,
      "step": 2900
    },
    {
      "epoch": 14.48019801980198,
      "grad_norm": 12.035900115966797,
      "learning_rate": 9.534022589305411e-07,
      "loss": 4.099019775390625,
      "step": 2925
    },
    {
      "epoch": 14.603960396039604,
      "grad_norm": 11.711991310119629,
      "learning_rate": 9.524517368026168e-07,
      "loss": 4.064933166503907,
      "step": 2950
    },
    {
      "epoch": 14.727722772277227,
      "grad_norm": 20.470388412475586,
      "learning_rate": 9.514921013845754e-07,
      "loss": 4.110359802246093,
      "step": 2975
    },
    {
      "epoch": 14.851485148514852,
      "grad_norm": 12.352797508239746,
      "learning_rate": 9.505233720054085e-07,
      "loss": 4.122251892089844,
      "step": 3000
    },
    {
      "epoch": 14.851485148514852,
      "eval_loss": 0.920803427696228,
      "eval_runtime": 86.0706,
      "eval_samples_per_second": 18.775,
      "eval_steps_per_second": 4.694,
      "step": 3000
    },
    {
      "epoch": 14.975247524752476,
      "grad_norm": 12.874964714050293,
      "learning_rate": 9.495455681772791e-07,
      "loss": 4.0896240234375,
      "step": 3025
    },
    {
      "epoch": 15.099009900990099,
      "grad_norm": 12.206281661987305,
      "learning_rate": 9.485587095951275e-07,
      "loss": 4.052577819824219,
      "step": 3050
    },
    {
      "epoch": 15.222772277227723,
      "grad_norm": 11.373126983642578,
      "learning_rate": 9.475628161362747e-07,
      "loss": 4.107190551757813,
      "step": 3075
    },
    {
      "epoch": 15.346534653465346,
      "grad_norm": 10.882997512817383,
      "learning_rate": 9.465579078600232e-07,
      "loss": 4.058759155273438,
      "step": 3100
    },
    {
      "epoch": 15.47029702970297,
      "grad_norm": 14.810848236083984,
      "learning_rate": 9.455440050072511e-07,
      "loss": 4.046248168945312,
      "step": 3125
    },
    {
      "epoch": 15.594059405940595,
      "grad_norm": 13.74386215209961,
      "learning_rate": 9.44521128000006e-07,
      "loss": 4.134948120117188,
      "step": 3150
    },
    {
      "epoch": 15.717821782178218,
      "grad_norm": 14.098150253295898,
      "learning_rate": 9.434892974410932e-07,
      "loss": 3.9966055297851564,
      "step": 3175
    },
    {
      "epoch": 15.841584158415841,
      "grad_norm": 10.933466911315918,
      "learning_rate": 9.424485341136599e-07,
      "loss": 4.099876708984375,
      "step": 3200
    },
    {
      "epoch": 15.965346534653465,
      "grad_norm": 9.829030990600586,
      "learning_rate": 9.41398858980778e-07,
      "loss": 3.9940283203125,
      "step": 3225
    },
    {
      "epoch": 16.08910891089109,
      "grad_norm": 10.526725769042969,
      "learning_rate": 9.403402931850207e-07,
      "loss": 4.016290588378906,
      "step": 3250
    },
    {
      "epoch": 16.212871287128714,
      "grad_norm": 9.416884422302246,
      "learning_rate": 9.392728580480371e-07,
      "loss": 4.113368835449219,
      "step": 3275
    },
    {
      "epoch": 16.336633663366335,
      "grad_norm": 11.515923500061035,
      "learning_rate": 9.381965750701228e-07,
      "loss": 4.077259216308594,
      "step": 3300
    },
    {
      "epoch": 16.46039603960396,
      "grad_norm": 12.539885520935059,
      "learning_rate": 9.37111465929787e-07,
      "loss": 3.9970358276367186,
      "step": 3325
    },
    {
      "epoch": 16.584158415841586,
      "grad_norm": 11.099287986755371,
      "learning_rate": 9.360175524833151e-07,
      "loss": 4.0322119140625,
      "step": 3350
    },
    {
      "epoch": 16.707920792079207,
      "grad_norm": 12.52721118927002,
      "learning_rate": 9.349148567643294e-07,
      "loss": 4.074447937011719,
      "step": 3375
    },
    {
      "epoch": 16.831683168316832,
      "grad_norm": 9.58317756652832,
      "learning_rate": 9.338034009833445e-07,
      "loss": 4.086669921875,
      "step": 3400
    },
    {
      "epoch": 16.955445544554454,
      "grad_norm": 11.126392364501953,
      "learning_rate": 9.326832075273207e-07,
      "loss": 4.079499206542969,
      "step": 3425
    },
    {
      "epoch": 17.07920792079208,
      "grad_norm": 11.195257186889648,
      "learning_rate": 9.315542989592124e-07,
      "loss": 4.001406555175781,
      "step": 3450
    },
    {
      "epoch": 17.202970297029704,
      "grad_norm": 11.249403953552246,
      "learning_rate": 9.304166980175141e-07,
      "loss": 4.042621765136719,
      "step": 3475
    },
    {
      "epoch": 17.326732673267326,
      "grad_norm": 22.27759552001953,
      "learning_rate": 9.292704276158021e-07,
      "loss": 4.025510559082031,
      "step": 3500
    },
    {
      "epoch": 17.326732673267326,
      "eval_loss": 0.9127047657966614,
      "eval_runtime": 81.0553,
      "eval_samples_per_second": 19.937,
      "eval_steps_per_second": 4.984,
      "step": 3500
    },
    {
      "epoch": 17.45049504950495,
      "grad_norm": 10.390947341918945,
      "learning_rate": 9.281155108422731e-07,
      "loss": 4.0325146484375,
      "step": 3525
    },
    {
      "epoch": 17.574257425742573,
      "grad_norm": 11.612289428710938,
      "learning_rate": 9.269519709592794e-07,
      "loss": 4.141344909667969,
      "step": 3550
    },
    {
      "epoch": 17.698019801980198,
      "grad_norm": 15.810525894165039,
      "learning_rate": 9.257798314028599e-07,
      "loss": 4.035428771972656,
      "step": 3575
    },
    {
      "epoch": 17.821782178217823,
      "grad_norm": 17.175111770629883,
      "learning_rate": 9.245991157822681e-07,
      "loss": 4.041367492675781,
      "step": 3600
    },
    {
      "epoch": 17.945544554455445,
      "grad_norm": 10.926711082458496,
      "learning_rate": 9.234098478794969e-07,
      "loss": 4.121484069824219,
      "step": 3625
    },
    {
      "epoch": 18.06930693069307,
      "grad_norm": 11.38853931427002,
      "learning_rate": 9.222120516487997e-07,
      "loss": 3.9692178344726563,
      "step": 3650
    },
    {
      "epoch": 18.193069306930692,
      "grad_norm": 11.953936576843262,
      "learning_rate": 9.210057512162072e-07,
      "loss": 4.020705261230469,
      "step": 3675
    },
    {
      "epoch": 18.316831683168317,
      "grad_norm": 13.49695110321045,
      "learning_rate": 9.197909708790419e-07,
      "loss": 4.005714721679688,
      "step": 3700
    },
    {
      "epoch": 18.440594059405942,
      "grad_norm": 12.650299072265625,
      "learning_rate": 9.185677351054289e-07,
      "loss": 4.006802062988282,
      "step": 3725
    },
    {
      "epoch": 18.564356435643564,
      "grad_norm": 14.999792098999023,
      "learning_rate": 9.173360685338026e-07,
      "loss": 3.983961486816406,
      "step": 3750
    },
    {
      "epoch": 18.68811881188119,
      "grad_norm": 8.753050804138184,
      "learning_rate": 9.160959959724109e-07,
      "loss": 4.121242980957032,
      "step": 3775
    },
    {
      "epoch": 18.81188118811881,
      "grad_norm": 11.767355918884277,
      "learning_rate": 9.148475423988149e-07,
      "loss": 3.988392333984375,
      "step": 3800
    },
    {
      "epoch": 18.935643564356436,
      "grad_norm": 13.041923522949219,
      "learning_rate": 9.135907329593869e-07,
      "loss": 4.037374877929688,
      "step": 3825
    },
    {
      "epoch": 19.059405940594058,
      "grad_norm": 17.08255386352539,
      "learning_rate": 9.123255929688023e-07,
      "loss": 4.015922241210937,
      "step": 3850
    },
    {
      "epoch": 19.183168316831683,
      "grad_norm": 10.86015796661377,
      "learning_rate": 9.110521479095312e-07,
      "loss": 4.031682739257812,
      "step": 3875
    },
    {
      "epoch": 19.306930693069308,
      "grad_norm": 14.082168579101562,
      "learning_rate": 9.097704234313243e-07,
      "loss": 3.973893737792969,
      "step": 3900
    },
    {
      "epoch": 19.43069306930693,
      "grad_norm": 12.722586631774902,
      "learning_rate": 9.084804453506966e-07,
      "loss": 4.083378295898438,
      "step": 3925
    },
    {
      "epoch": 19.554455445544555,
      "grad_norm": 12.841690063476562,
      "learning_rate": 9.071822396504072e-07,
      "loss": 4.002433166503907,
      "step": 3950
    },
    {
      "epoch": 19.678217821782177,
      "grad_norm": 11.057418823242188,
      "learning_rate": 9.058758324789359e-07,
      "loss": 4.0087548828125,
      "step": 3975
    },
    {
      "epoch": 19.801980198019802,
      "grad_norm": 10.621847152709961,
      "learning_rate": 9.045612501499573e-07,
      "loss": 3.928231201171875,
      "step": 4000
    },
    {
      "epoch": 19.801980198019802,
      "eval_loss": 0.9033195376396179,
      "eval_runtime": 84.5771,
      "eval_samples_per_second": 19.107,
      "eval_steps_per_second": 4.777,
      "step": 4000
    },
    {
      "epoch": 19.925742574257427,
      "grad_norm": 10.820324897766113,
      "learning_rate": 9.032385191418093e-07,
      "loss": 4.026563720703125,
      "step": 4025
    },
    {
      "epoch": 20.04950495049505,
      "grad_norm": 12.08041763305664,
      "learning_rate": 9.019076660969609e-07,
      "loss": 3.98918212890625,
      "step": 4050
    },
    {
      "epoch": 20.173267326732674,
      "grad_norm": 12.175297737121582,
      "learning_rate": 9.005687178214758e-07,
      "loss": 3.962005615234375,
      "step": 4075
    },
    {
      "epoch": 20.297029702970296,
      "grad_norm": 15.371024131774902,
      "learning_rate": 8.992217012844713e-07,
      "loss": 3.970758972167969,
      "step": 4100
    },
    {
      "epoch": 20.42079207920792,
      "grad_norm": 12.102559089660645,
      "learning_rate": 8.978666436175761e-07,
      "loss": 3.9590899658203127,
      "step": 4125
    },
    {
      "epoch": 20.544554455445546,
      "grad_norm": 11.632902145385742,
      "learning_rate": 8.965035721143834e-07,
      "loss": 3.9882553100585936,
      "step": 4150
    },
    {
      "epoch": 20.668316831683168,
      "grad_norm": 14.134031295776367,
      "learning_rate": 8.951325142299014e-07,
      "loss": 3.931231689453125,
      "step": 4175
    },
    {
      "epoch": 20.792079207920793,
      "grad_norm": 11.716127395629883,
      "learning_rate": 8.937534975799996e-07,
      "loss": 3.9607308959960936,
      "step": 4200
    },
    {
      "epoch": 20.915841584158414,
      "grad_norm": 11.062108993530273,
      "learning_rate": 8.923665499408535e-07,
      "loss": 3.955765380859375,
      "step": 4225
    },
    {
      "epoch": 21.03960396039604,
      "grad_norm": 10.680591583251953,
      "learning_rate": 8.909716992483847e-07,
      "loss": 3.8945526123046874,
      "step": 4250
    },
    {
      "epoch": 21.163366336633665,
      "grad_norm": 13.237275123596191,
      "learning_rate": 8.895689735976983e-07,
      "loss": 3.956689758300781,
      "step": 4275
    },
    {
      "epoch": 21.287128712871286,
      "grad_norm": 17.047996520996094,
      "learning_rate": 8.881584012425165e-07,
      "loss": 3.9637225341796873,
      "step": 4300
    },
    {
      "epoch": 21.41089108910891,
      "grad_norm": 13.230228424072266,
      "learning_rate": 8.867400105946103e-07,
      "loss": 3.951360168457031,
      "step": 4325
    },
    {
      "epoch": 21.534653465346533,
      "grad_norm": 12.366623878479004,
      "learning_rate": 8.853138302232271e-07,
      "loss": 3.9974639892578123,
      "step": 4350
    },
    {
      "epoch": 21.65841584158416,
      "grad_norm": 13.02001667022705,
      "learning_rate": 8.838798888545143e-07,
      "loss": 3.9881216430664064,
      "step": 4375
    },
    {
      "epoch": 21.782178217821784,
      "grad_norm": 12.130973815917969,
      "learning_rate": 8.824382153709421e-07,
      "loss": 3.9237442016601562,
      "step": 4400
    },
    {
      "epoch": 21.905940594059405,
      "grad_norm": 11.393623352050781,
      "learning_rate": 8.809888388107209e-07,
      "loss": 4.023025207519531,
      "step": 4425
    },
    {
      "epoch": 22.02970297029703,
      "grad_norm": 10.824515342712402,
      "learning_rate": 8.795317883672165e-07,
      "loss": 3.9387527465820313,
      "step": 4450
    },
    {
      "epoch": 22.153465346534652,
      "grad_norm": 20.875307083129883,
      "learning_rate": 8.780670933883622e-07,
      "loss": 3.97408203125,
      "step": 4475
    },
    {
      "epoch": 22.277227722772277,
      "grad_norm": 17.839189529418945,
      "learning_rate": 8.765947833760677e-07,
      "loss": 4.018715209960938,
      "step": 4500
    },
    {
      "epoch": 22.277227722772277,
      "eval_loss": 0.905015230178833,
      "eval_runtime": 84.5389,
      "eval_samples_per_second": 19.115,
      "eval_steps_per_second": 4.779,
      "step": 4500
    },
    {
      "epoch": 22.400990099009903,
      "grad_norm": 13.415910720825195,
      "learning_rate": 8.751148879856249e-07,
      "loss": 3.94878173828125,
      "step": 4525
    },
    {
      "epoch": 22.524752475247524,
      "grad_norm": 11.05270767211914,
      "learning_rate": 8.736274370251103e-07,
      "loss": 3.9672274780273438,
      "step": 4550
    },
    {
      "epoch": 22.64851485148515,
      "grad_norm": 11.800627708435059,
      "learning_rate": 8.721324604547849e-07,
      "loss": 3.927026672363281,
      "step": 4575
    },
    {
      "epoch": 22.77227722772277,
      "grad_norm": 10.663934707641602,
      "learning_rate": 8.706299883864909e-07,
      "loss": 3.9203482055664063,
      "step": 4600
    },
    {
      "epoch": 22.896039603960396,
      "grad_norm": 16.535402297973633,
      "learning_rate": 8.691200510830444e-07,
      "loss": 3.9108685302734374,
      "step": 4625
    },
    {
      "epoch": 23.019801980198018,
      "grad_norm": 9.973243713378906,
      "learning_rate": 8.676026789576269e-07,
      "loss": 3.960736083984375,
      "step": 4650
    },
    {
      "epoch": 23.143564356435643,
      "grad_norm": 21.756301879882812,
      "learning_rate": 8.660779025731721e-07,
      "loss": 3.991618957519531,
      "step": 4675
    },
    {
      "epoch": 23.26732673267327,
      "grad_norm": 12.650042533874512,
      "learning_rate": 8.645457526417503e-07,
      "loss": 3.9313955688476563,
      "step": 4700
    },
    {
      "epoch": 23.39108910891089,
      "grad_norm": 10.731987953186035,
      "learning_rate": 8.630062600239496e-07,
      "loss": 3.90642333984375,
      "step": 4725
    },
    {
      "epoch": 23.514851485148515,
      "grad_norm": 13.895561218261719,
      "learning_rate": 8.614594557282553e-07,
      "loss": 3.871036376953125,
      "step": 4750
    },
    {
      "epoch": 23.638613861386137,
      "grad_norm": 15.030277252197266,
      "learning_rate": 8.599053709104241e-07,
      "loss": 3.9301937866210936,
      "step": 4775
    },
    {
      "epoch": 23.762376237623762,
      "grad_norm": 12.624570846557617,
      "learning_rate": 8.583440368728572e-07,
      "loss": 3.915058898925781,
      "step": 4800
    },
    {
      "epoch": 23.886138613861387,
      "grad_norm": 11.839984893798828,
      "learning_rate": 8.567754850639699e-07,
      "loss": 3.924931640625,
      "step": 4825
    },
    {
      "epoch": 24.00990099009901,
      "grad_norm": 13.87601089477539,
      "learning_rate": 8.551997470775576e-07,
      "loss": 3.9027645874023436,
      "step": 4850
    },
    {
      "epoch": 24.133663366336634,
      "grad_norm": 11.400012969970703,
      "learning_rate": 8.536168546521602e-07,
      "loss": 3.9296466064453126,
      "step": 4875
    },
    {
      "epoch": 24.257425742574256,
      "grad_norm": 11.829928398132324,
      "learning_rate": 8.520268396704222e-07,
      "loss": 3.993326416015625,
      "step": 4900
    },
    {
      "epoch": 24.38118811881188,
      "grad_norm": 11.399077415466309,
      "learning_rate": 8.504297341584508e-07,
      "loss": 3.891920471191406,
      "step": 4925
    },
    {
      "epoch": 24.504950495049506,
      "grad_norm": 13.8989839553833,
      "learning_rate": 8.488255702851707e-07,
      "loss": 3.9781057739257815,
      "step": 4950
    },
    {
      "epoch": 24.628712871287128,
      "grad_norm": 12.6453218460083,
      "learning_rate": 8.472143803616764e-07,
      "loss": 3.9407168579101564,
      "step": 4975
    },
    {
      "epoch": 24.752475247524753,
      "grad_norm": 9.551429748535156,
      "learning_rate": 8.45596196840581e-07,
      "loss": 3.9068170166015626,
      "step": 5000
    },
    {
      "epoch": 24.752475247524753,
      "eval_loss": 0.8924599885940552,
      "eval_runtime": 85.1496,
      "eval_samples_per_second": 18.978,
      "eval_steps_per_second": 4.745,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 18000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 90,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1465432462475584e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}