{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 72.72103004291846,
  "global_step": 5600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.3,
      "learning_rate": 0.0001,
      "loss": 0.0529,
      "step": 100
    },
    {
      "epoch": 1.3,
      "eval_loss": 0.3332812786102295,
      "eval_runtime": 188.2511,
      "eval_samples_per_second": 16.558,
      "eval_steps_per_second": 2.072,
      "eval_wer": 0.2371088585279859,
      "step": 100
    },
    {
      "epoch": 2.59,
      "learning_rate": 9.98699609882965e-05,
      "loss": 0.0471,
      "step": 200
    },
    {
      "epoch": 2.59,
      "eval_loss": 0.35051167011260986,
      "eval_runtime": 136.2299,
      "eval_samples_per_second": 22.88,
      "eval_steps_per_second": 2.863,
      "eval_wer": 0.23723477932380532,
      "step": 200
    },
    {
      "epoch": 3.89,
      "learning_rate": 9.973992197659299e-05,
      "loss": 0.0463,
      "step": 300
    },
    {
      "epoch": 3.89,
      "eval_loss": 0.3505023121833801,
      "eval_runtime": 142.4796,
      "eval_samples_per_second": 21.877,
      "eval_steps_per_second": 2.737,
      "eval_wer": 0.23931247245482593,
      "step": 300
    },
    {
      "epoch": 5.19,
      "learning_rate": 9.960988296488946e-05,
      "loss": 0.0461,
      "step": 400
    },
    {
      "epoch": 5.19,
      "eval_loss": 0.4004528224468231,
      "eval_runtime": 143.9556,
      "eval_samples_per_second": 21.653,
      "eval_steps_per_second": 2.709,
      "eval_wer": 0.2405716804130202,
      "step": 400
    },
    {
      "epoch": 6.49,
      "learning_rate": 9.947984395318596e-05,
      "loss": 0.0442,
      "step": 500
    },
    {
      "epoch": 6.49,
      "eval_loss": 0.4172374904155731,
      "eval_runtime": 151.5028,
      "eval_samples_per_second": 20.574,
      "eval_steps_per_second": 2.574,
      "eval_wer": 0.24863061134546371,
      "step": 500
    },
    {
      "epoch": 7.79,
      "learning_rate": 9.934980494148245e-05,
      "loss": 0.0431,
      "step": 600
    },
    {
      "epoch": 7.79,
      "eval_loss": 0.3723874092102051,
      "eval_runtime": 155.4666,
      "eval_samples_per_second": 20.049,
      "eval_steps_per_second": 2.509,
      "eval_wer": 0.24006799722974248,
      "step": 600
    },
    {
      "epoch": 9.09,
      "learning_rate": 9.921976592977894e-05,
      "loss": 0.0413,
      "step": 700
    },
    {
      "epoch": 9.09,
      "eval_loss": 0.38365304470062256,
      "eval_runtime": 157.1121,
      "eval_samples_per_second": 19.839,
      "eval_steps_per_second": 2.482,
      "eval_wer": 0.24466410627715168,
      "step": 700
    },
    {
      "epoch": 10.39,
      "learning_rate": 9.908972691807543e-05,
      "loss": 0.0417,
      "step": 800
    },
    {
      "epoch": 10.39,
      "eval_loss": 0.37991076707839966,
      "eval_runtime": 160.431,
      "eval_samples_per_second": 19.429,
      "eval_steps_per_second": 2.431,
      "eval_wer": 0.23786438330290247,
      "step": 800
    },
    {
      "epoch": 11.68,
      "learning_rate": 9.89596879063719e-05,
      "loss": 0.0423,
      "step": 900
    },
    {
      "epoch": 11.68,
      "eval_loss": 0.41009148955345154,
      "eval_runtime": 161.1546,
      "eval_samples_per_second": 19.342,
      "eval_steps_per_second": 2.42,
      "eval_wer": 0.24485298747088083,
      "step": 900
    },
    {
      "epoch": 12.98,
      "learning_rate": 9.88296488946684e-05,
      "loss": 0.0425,
      "step": 1000
    },
    {
      "epoch": 12.98,
      "eval_loss": 0.38417309522628784,
      "eval_runtime": 162.4688,
      "eval_samples_per_second": 19.185,
      "eval_steps_per_second": 2.4,
      "eval_wer": 0.23931247245482593,
      "step": 1000
    },
    {
      "epoch": 14.28,
      "learning_rate": 9.869960988296489e-05,
      "loss": 0.0413,
      "step": 1100
    },
    {
      "epoch": 14.28,
      "eval_loss": 0.37902718782424927,
      "eval_runtime": 164.4062,
      "eval_samples_per_second": 18.959,
      "eval_steps_per_second": 2.372,
      "eval_wer": 0.24491594786879053,
      "step": 1100
    },
    {
      "epoch": 15.58,
      "learning_rate": 9.856957087126138e-05,
      "loss": 0.0416,
      "step": 1200
    },
    {
      "epoch": 15.58,
      "eval_loss": 0.38784804940223694,
      "eval_runtime": 163.9531,
      "eval_samples_per_second": 19.012,
      "eval_steps_per_second": 2.379,
      "eval_wer": 0.23931247245482593,
      "step": 1200
    },
    {
      "epoch": 16.88,
      "learning_rate": 9.844083224967491e-05,
      "loss": 0.0436,
      "step": 1300
    },
    {
      "epoch": 16.88,
      "eval_loss": 0.36406952142715454,
      "eval_runtime": 162.4063,
      "eval_samples_per_second": 19.193,
      "eval_steps_per_second": 2.401,
      "eval_wer": 0.23817918529245105,
      "step": 1300
    },
    {
      "epoch": 18.18,
      "learning_rate": 9.83107932379714e-05,
      "loss": 0.0424,
      "step": 1400
    },
    {
      "epoch": 18.18,
      "eval_loss": 0.3773825764656067,
      "eval_runtime": 161.8281,
      "eval_samples_per_second": 19.261,
      "eval_steps_per_second": 2.41,
      "eval_wer": 0.2359126109677013,
      "step": 1400
    },
    {
      "epoch": 19.48,
      "learning_rate": 9.818075422626789e-05,
      "loss": 0.0379,
      "step": 1500
    },
    {
      "epoch": 19.48,
      "eval_loss": 0.39104992151260376,
      "eval_runtime": 153.8281,
      "eval_samples_per_second": 20.263,
      "eval_steps_per_second": 2.535,
      "eval_wer": 0.23396083863250017,
      "step": 1500
    },
    {
      "epoch": 20.77,
      "learning_rate": 9.805071521456437e-05,
      "loss": 0.0401,
      "step": 1600
    },
    {
      "epoch": 20.77,
      "eval_loss": 0.4012731611728668,
      "eval_runtime": 166.1564,
      "eval_samples_per_second": 18.759,
      "eval_steps_per_second": 2.347,
      "eval_wer": 0.23786438330290247,
      "step": 1600
    },
    {
      "epoch": 22.08,
      "learning_rate": 9.792067620286086e-05,
      "loss": 0.0397,
      "step": 1700
    },
    {
      "epoch": 22.08,
      "eval_loss": 0.3917081952095032,
      "eval_runtime": 151.5157,
      "eval_samples_per_second": 20.572,
      "eval_steps_per_second": 2.574,
      "eval_wer": 0.23981615563810363,
      "step": 1700
    },
    {
      "epoch": 23.37,
      "learning_rate": 9.779063719115735e-05,
      "loss": 0.0399,
      "step": 1800
    },
    {
      "epoch": 23.37,
      "eval_loss": 0.41352856159210205,
      "eval_runtime": 169.9898,
      "eval_samples_per_second": 18.336,
      "eval_steps_per_second": 2.294,
      "eval_wer": 0.24151608638166594,
      "step": 1800
    },
    {
      "epoch": 24.67,
      "learning_rate": 9.766059817945384e-05,
      "loss": 0.0407,
      "step": 1900
    },
    {
      "epoch": 24.67,
      "eval_loss": 0.3818851113319397,
      "eval_runtime": 176.1576,
      "eval_samples_per_second": 17.694,
      "eval_steps_per_second": 2.214,
      "eval_wer": 0.23717181892589562,
      "step": 1900
    },
    {
      "epoch": 25.97,
      "learning_rate": 9.753055916775033e-05,
      "loss": 0.0392,
      "step": 2000
    },
    {
      "epoch": 25.97,
      "eval_loss": 0.3882431983947754,
      "eval_runtime": 167.0577,
      "eval_samples_per_second": 18.658,
      "eval_steps_per_second": 2.335,
      "eval_wer": 0.23786438330290247,
      "step": 2000
    },
    {
      "epoch": 27.27,
      "learning_rate": 9.740052015604681e-05,
      "loss": 0.0381,
      "step": 2100
    },
    {
      "epoch": 27.27,
      "eval_loss": 0.3893887996673584,
      "eval_runtime": 171.6954,
      "eval_samples_per_second": 18.154,
      "eval_steps_per_second": 2.271,
      "eval_wer": 0.23943839325064534,
      "step": 2100
    },
    {
      "epoch": 28.57,
      "learning_rate": 9.72704811443433e-05,
      "loss": 0.0401,
      "step": 2200
    },
    {
      "epoch": 28.57,
      "eval_loss": 0.3673301041126251,
      "eval_runtime": 174.7303,
      "eval_samples_per_second": 17.839,
      "eval_steps_per_second": 2.232,
      "eval_wer": 0.2357866901718819,
      "step": 2200
    },
    {
      "epoch": 29.86,
      "learning_rate": 9.71404421326398e-05,
      "loss": 0.0391,
      "step": 2300
    },
    {
      "epoch": 29.86,
      "eval_loss": 0.3780101239681244,
      "eval_runtime": 157.4375,
      "eval_samples_per_second": 19.798,
      "eval_steps_per_second": 2.477,
      "eval_wer": 0.2354089277844236,
      "step": 2300
    },
    {
      "epoch": 31.17,
      "learning_rate": 9.701040312093628e-05,
      "loss": 0.0377,
      "step": 2400
    },
    {
      "epoch": 31.17,
      "eval_loss": 0.3910522758960724,
      "eval_runtime": 159.5625,
      "eval_samples_per_second": 19.535,
      "eval_steps_per_second": 2.444,
      "eval_wer": 0.23654221494679847,
      "step": 2400
    },
    {
      "epoch": 32.46,
      "learning_rate": 9.688036410923278e-05,
      "loss": 0.0378,
      "step": 2500
    },
    {
      "epoch": 32.46,
      "eval_loss": 0.3806402087211609,
      "eval_runtime": 170.7978,
      "eval_samples_per_second": 18.25,
      "eval_steps_per_second": 2.283,
      "eval_wer": 0.23918655165900649,
      "step": 2500
    },
    {
      "epoch": 33.76,
      "learning_rate": 9.675032509752925e-05,
      "loss": 0.0347,
      "step": 2600
    },
    {
      "epoch": 33.76,
      "eval_loss": 0.3844529390335083,
      "eval_runtime": 160.3906,
      "eval_samples_per_second": 19.434,
      "eval_steps_per_second": 2.432,
      "eval_wer": 0.23515708619278475,
      "step": 2600
    },
    {
      "epoch": 35.06,
      "learning_rate": 9.662028608582574e-05,
      "loss": 0.0369,
      "step": 2700
    },
    {
      "epoch": 35.06,
      "eval_loss": 0.39440667629241943,
      "eval_runtime": 167.7032,
      "eval_samples_per_second": 18.586,
      "eval_steps_per_second": 2.326,
      "eval_wer": 0.23931247245482593,
      "step": 2700
    },
    {
      "epoch": 36.36,
      "learning_rate": 9.649024707412224e-05,
      "loss": 0.0375,
      "step": 2800
    },
    {
      "epoch": 36.36,
      "eval_loss": 0.37219446897506714,
      "eval_runtime": 165.2969,
      "eval_samples_per_second": 18.857,
      "eval_steps_per_second": 2.359,
      "eval_wer": 0.23012025436000755,
      "step": 2800
    },
    {
      "epoch": 37.66,
      "learning_rate": 9.636020806241873e-05,
      "loss": 0.0363,
      "step": 2900
    },
    {
      "epoch": 37.66,
      "eval_loss": 0.3664211332798004,
      "eval_runtime": 167.8076,
      "eval_samples_per_second": 18.575,
      "eval_steps_per_second": 2.324,
      "eval_wer": 0.23112762072656298,
      "step": 2900
    },
    {
      "epoch": 38.95,
      "learning_rate": 9.623016905071522e-05,
      "loss": 0.034,
      "step": 3000
    },
    {
      "epoch": 38.95,
      "eval_loss": 0.392531156539917,
      "eval_runtime": 173.8435,
      "eval_samples_per_second": 17.93,
      "eval_steps_per_second": 2.243,
      "eval_wer": 0.23081281873701442,
      "step": 3000
    },
    {
      "epoch": 40.26,
      "learning_rate": 9.610013003901171e-05,
      "loss": 0.0331,
      "step": 3100
    },
    {
      "epoch": 40.26,
      "eval_loss": 0.3893636465072632,
      "eval_runtime": 167.8125,
      "eval_samples_per_second": 18.574,
      "eval_steps_per_second": 2.324,
      "eval_wer": 0.2323868286847573,
      "step": 3100
    },
    {
      "epoch": 41.55,
      "learning_rate": 9.59700910273082e-05,
      "loss": 0.0339,
      "step": 3200
    },
    {
      "epoch": 41.55,
      "eval_loss": 0.39426469802856445,
      "eval_runtime": 168.7188,
      "eval_samples_per_second": 18.475,
      "eval_steps_per_second": 2.312,
      "eval_wer": 0.233331234653403,
      "step": 3200
    },
    {
      "epoch": 42.85,
      "learning_rate": 9.584005201560469e-05,
      "loss": 0.033,
      "step": 3300
    },
    {
      "epoch": 42.85,
      "eval_loss": 0.3865768313407898,
      "eval_runtime": 161.5781,
      "eval_samples_per_second": 19.291,
      "eval_steps_per_second": 2.414,
      "eval_wer": 0.2350311653969653,
      "step": 3300
    },
    {
      "epoch": 44.15,
      "learning_rate": 9.571001300390118e-05,
      "loss": 0.0334,
      "step": 3400
    },
    {
      "epoch": 44.15,
      "eval_loss": 0.38849422335624695,
      "eval_runtime": 166.5778,
      "eval_samples_per_second": 18.712,
      "eval_steps_per_second": 2.341,
      "eval_wer": 0.2320720266952087,
      "step": 3400
    },
    {
      "epoch": 45.45,
      "learning_rate": 9.557997399219767e-05,
      "loss": 0.0331,
      "step": 3500
    },
    {
      "epoch": 45.45,
      "eval_loss": 0.41077303886413574,
      "eval_runtime": 165.0937,
      "eval_samples_per_second": 18.88,
      "eval_steps_per_second": 2.362,
      "eval_wer": 0.232009066297299,
      "step": 3500
    },
    {
      "epoch": 46.75,
      "learning_rate": 9.544993498049415e-05,
      "loss": 0.0326,
      "step": 3600
    },
    {
      "epoch": 46.75,
      "eval_loss": 0.3925323784351349,
      "eval_runtime": 159.9375,
      "eval_samples_per_second": 19.489,
      "eval_steps_per_second": 2.438,
      "eval_wer": 0.2414531259837562,
      "step": 3600
    },
    {
      "epoch": 48.05,
      "learning_rate": 9.531989596879064e-05,
      "loss": 0.036,
      "step": 3700
    },
    {
      "epoch": 48.05,
      "eval_loss": 0.3832598030567169,
      "eval_runtime": 171.125,
      "eval_samples_per_second": 18.215,
      "eval_steps_per_second": 2.279,
      "eval_wer": 0.2387458288736385,
      "step": 3700
    },
    {
      "epoch": 49.35,
      "learning_rate": 9.518985695708713e-05,
      "loss": 0.0346,
      "step": 3800
    },
    {
      "epoch": 49.35,
      "eval_loss": 0.3931749761104584,
      "eval_runtime": 166.4063,
      "eval_samples_per_second": 18.731,
      "eval_steps_per_second": 2.344,
      "eval_wer": 0.23263867027639615,
      "step": 3800
    },
    {
      "epoch": 50.64,
      "learning_rate": 9.505981794538362e-05,
      "loss": 0.0349,
      "step": 3900
    },
    {
      "epoch": 50.64,
      "eval_loss": 0.3744593858718872,
      "eval_runtime": 157.9687,
      "eval_samples_per_second": 19.732,
      "eval_steps_per_second": 2.469,
      "eval_wer": 0.2354089277844236,
      "step": 3900
    },
    {
      "epoch": 51.94,
      "learning_rate": 9.492977893368012e-05,
      "loss": 0.034,
      "step": 4000
    },
    {
      "epoch": 51.94,
      "eval_loss": 0.39862367510795593,
      "eval_runtime": 155.8749,
      "eval_samples_per_second": 19.997,
      "eval_steps_per_second": 2.502,
      "eval_wer": 0.2328275514701253,
      "step": 4000
    },
    {
      "epoch": 53.24,
      "learning_rate": 9.479973992197659e-05,
      "loss": 0.0334,
      "step": 4100
    },
    {
      "epoch": 53.24,
      "eval_loss": 0.39466869831085205,
      "eval_runtime": 162.0467,
      "eval_samples_per_second": 19.235,
      "eval_steps_per_second": 2.407,
      "eval_wer": 0.23352011584713214,
      "step": 4100
    },
    {
      "epoch": 54.54,
      "learning_rate": 9.466970091027308e-05,
      "loss": 0.0325,
      "step": 4200
    },
    {
      "epoch": 54.54,
      "eval_loss": 0.39441126585006714,
      "eval_runtime": 162.4844,
      "eval_samples_per_second": 19.183,
      "eval_steps_per_second": 2.4,
      "eval_wer": 0.23616445255934018,
      "step": 4200
    },
    {
      "epoch": 55.84,
      "learning_rate": 9.453966189856957e-05,
      "loss": 0.0308,
      "step": 4300
    },
    {
      "epoch": 55.84,
      "eval_loss": 0.39918699860572815,
      "eval_runtime": 155.5625,
      "eval_samples_per_second": 20.037,
      "eval_steps_per_second": 2.507,
      "eval_wer": 0.23477932380532646,
      "step": 4300
    },
    {
      "epoch": 57.14,
      "learning_rate": 9.440962288686607e-05,
      "loss": 0.0316,
      "step": 4400
    },
    {
      "epoch": 57.14,
      "eval_loss": 0.39010030031204224,
      "eval_runtime": 163.0782,
      "eval_samples_per_second": 19.114,
      "eval_steps_per_second": 2.391,
      "eval_wer": 0.229805452370459,
      "step": 4400
    },
    {
      "epoch": 58.44,
      "learning_rate": 9.427958387516256e-05,
      "loss": 0.0308,
      "step": 4500
    },
    {
      "epoch": 58.44,
      "eval_loss": 0.404751718044281,
      "eval_runtime": 161.3906,
      "eval_samples_per_second": 19.313,
      "eval_steps_per_second": 2.416,
      "eval_wer": 0.23194610589938927,
      "step": 4500
    },
    {
      "epoch": 59.73,
      "learning_rate": 9.414954486345903e-05,
      "loss": 0.028,
      "step": 4600
    },
    {
      "epoch": 59.73,
      "eval_loss": 0.41103655099868774,
      "eval_runtime": 154.3905,
      "eval_samples_per_second": 20.189,
      "eval_steps_per_second": 2.526,
      "eval_wer": 0.2277277592394384,
      "step": 4600
    },
    {
      "epoch": 61.04,
      "learning_rate": 9.401950585175553e-05,
      "loss": 0.0299,
      "step": 4700
    },
    {
      "epoch": 61.04,
      "eval_loss": 0.40084338188171387,
      "eval_runtime": 168.5468,
      "eval_samples_per_second": 18.493,
      "eval_steps_per_second": 2.314,
      "eval_wer": 0.23295347226594473,
      "step": 4700
    },
    {
      "epoch": 62.33,
      "learning_rate": 9.388946684005202e-05,
      "loss": 0.033,
      "step": 4800
    },
    {
      "epoch": 62.33,
      "eval_loss": 0.3985605835914612,
      "eval_runtime": 168.5159,
      "eval_samples_per_second": 18.497,
      "eval_steps_per_second": 2.314,
      "eval_wer": 0.23490524460114587,
      "step": 4800
    },
    {
      "epoch": 63.63,
      "learning_rate": 9.375942782834851e-05,
      "loss": 0.0345,
      "step": 4900
    },
    {
      "epoch": 63.63,
      "eval_loss": 0.3911936581134796,
      "eval_runtime": 158.9531,
      "eval_samples_per_second": 19.61,
      "eval_steps_per_second": 2.454,
      "eval_wer": 0.23559780897815275,
      "step": 4900
    },
    {
      "epoch": 64.93,
      "learning_rate": 9.3629388816645e-05,
      "loss": 0.0322,
      "step": 5000
    },
    {
      "epoch": 64.93,
      "eval_loss": 0.3985706567764282,
      "eval_runtime": 175.1875,
      "eval_samples_per_second": 17.792,
      "eval_steps_per_second": 2.226,
      "eval_wer": 0.231568343511931,
      "step": 5000
    },
    {
      "epoch": 66.23,
      "learning_rate": 9.349934980494148e-05,
      "loss": 0.0313,
      "step": 5100
    },
    {
      "epoch": 66.23,
      "eval_loss": 0.4163081645965576,
      "eval_runtime": 179.4796,
      "eval_samples_per_second": 17.367,
      "eval_steps_per_second": 2.173,
      "eval_wer": 0.23049801674746584,
      "step": 5100
    },
    {
      "epoch": 67.53,
      "learning_rate": 9.336931079323797e-05,
      "loss": 0.0315,
      "step": 5200
    },
    {
      "epoch": 67.53,
      "eval_loss": 0.3985958993434906,
      "eval_runtime": 170.0289,
      "eval_samples_per_second": 18.332,
      "eval_steps_per_second": 2.294,
      "eval_wer": 0.23528300698860416,
      "step": 5200
    },
    {
      "epoch": 68.82,
      "learning_rate": 9.323927178153446e-05,
      "loss": 0.0294,
      "step": 5300
    },
    {
      "epoch": 68.82,
      "eval_loss": 0.40526434779167175,
      "eval_runtime": 177.9803,
      "eval_samples_per_second": 17.513,
      "eval_steps_per_second": 2.191,
      "eval_wer": 0.2295536107788201,
      "step": 5300
    },
    {
      "epoch": 70.13,
      "learning_rate": 9.310923276983095e-05,
      "loss": 0.0302,
      "step": 5400
    },
    {
      "epoch": 70.13,
      "eval_loss": 0.3818342089653015,
      "eval_runtime": 177.545,
      "eval_samples_per_second": 17.556,
      "eval_steps_per_second": 2.197,
      "eval_wer": 0.23131650192029213,
      "step": 5400
    },
    {
      "epoch": 71.42,
      "learning_rate": 9.297919375812744e-05,
      "loss": 0.0318,
      "step": 5500
    },
    {
      "epoch": 71.42,
      "eval_loss": 0.3933159410953522,
      "eval_runtime": 168.7969,
      "eval_samples_per_second": 18.466,
      "eval_steps_per_second": 2.31,
      "eval_wer": 0.23301643266385444,
      "step": 5500
    },
    {
      "epoch": 72.72,
      "learning_rate": 9.284915474642393e-05,
      "loss": 0.0289,
      "step": 5600
    },
    {
      "epoch": 72.72,
      "eval_loss": 0.38790163397789,
      "eval_runtime": 167.7031,
      "eval_samples_per_second": 18.586,
      "eval_steps_per_second": 2.326,
      "eval_wer": 0.23396083863250017,
      "step": 5600
    }
  ],
  "max_steps": 77000,
  "num_train_epochs": 1000,
  "total_flos": 1.80069386799743e+20,
  "trial_name": null,
  "trial_params": null
}