{
  "best_metric": 0.7337000966072083,
  "best_model_checkpoint": "/home/ubuntu/utah/babylm-24/src/evaluation/results/finetune/DebertaV2-Base-10M_babylm-A/mnli/checkpoint-147264",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 147264,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010185788787483703,
      "grad_norm": 3.972818613052368,
      "learning_rate": 2.9938885267275102e-05,
      "loss": 1.1058,
      "step": 500
    },
    {
      "epoch": 0.020371577574967405,
      "grad_norm": 3.13022780418396,
      "learning_rate": 2.9877770534550196e-05,
      "loss": 1.0388,
      "step": 1000
    },
    {
      "epoch": 0.03055736636245111,
      "grad_norm": 5.194879055023193,
      "learning_rate": 2.9816655801825297e-05,
      "loss": 0.999,
      "step": 1500
    },
    {
      "epoch": 0.04074315514993481,
      "grad_norm": 6.858974933624268,
      "learning_rate": 2.975554106910039e-05,
      "loss": 0.9637,
      "step": 2000
    },
    {
      "epoch": 0.050928943937418515,
      "grad_norm": 7.074331760406494,
      "learning_rate": 2.9694426336375492e-05,
      "loss": 0.9697,
      "step": 2500
    },
    {
      "epoch": 0.06111473272490222,
      "grad_norm": 5.386518478393555,
      "learning_rate": 2.9633311603650586e-05,
      "loss": 0.9472,
      "step": 3000
    },
    {
      "epoch": 0.07130052151238592,
      "grad_norm": 5.582085609436035,
      "learning_rate": 2.9572196870925687e-05,
      "loss": 0.9273,
      "step": 3500
    },
    {
      "epoch": 0.08148631029986962,
      "grad_norm": 10.966410636901855,
      "learning_rate": 2.951108213820078e-05,
      "loss": 0.9316,
      "step": 4000
    },
    {
      "epoch": 0.09167209908735333,
      "grad_norm": 7.840672492980957,
      "learning_rate": 2.9449967405475882e-05,
      "loss": 0.9237,
      "step": 4500
    },
    {
      "epoch": 0.10185788787483703,
      "grad_norm": 7.442521095275879,
      "learning_rate": 2.9388852672750976e-05,
      "loss": 0.918,
      "step": 5000
    },
    {
      "epoch": 0.11204367666232073,
      "grad_norm": 8.379261016845703,
      "learning_rate": 2.9327737940026077e-05,
      "loss": 0.8798,
      "step": 5500
    },
    {
      "epoch": 0.12222946544980444,
      "grad_norm": 6.880769729614258,
      "learning_rate": 2.926662320730117e-05,
      "loss": 0.9045,
      "step": 6000
    },
    {
      "epoch": 0.13241525423728814,
      "grad_norm": 5.965817451477051,
      "learning_rate": 2.9205508474576272e-05,
      "loss": 0.8941,
      "step": 6500
    },
    {
      "epoch": 0.14260104302477183,
      "grad_norm": 9.67165756225586,
      "learning_rate": 2.914439374185137e-05,
      "loss": 0.878,
      "step": 7000
    },
    {
      "epoch": 0.15278683181225555,
      "grad_norm": 6.4934892654418945,
      "learning_rate": 2.9083279009126467e-05,
      "loss": 0.8723,
      "step": 7500
    },
    {
      "epoch": 0.16297262059973924,
      "grad_norm": 11.987072944641113,
      "learning_rate": 2.9022164276401565e-05,
      "loss": 0.8913,
      "step": 8000
    },
    {
      "epoch": 0.17315840938722293,
      "grad_norm": 9.218280792236328,
      "learning_rate": 2.8961049543676662e-05,
      "loss": 0.868,
      "step": 8500
    },
    {
      "epoch": 0.18334419817470665,
      "grad_norm": 6.697582244873047,
      "learning_rate": 2.8899934810951763e-05,
      "loss": 0.8618,
      "step": 9000
    },
    {
      "epoch": 0.19352998696219034,
      "grad_norm": 8.035455703735352,
      "learning_rate": 2.8838820078226858e-05,
      "loss": 0.8641,
      "step": 9500
    },
    {
      "epoch": 0.20371577574967406,
      "grad_norm": 6.890414237976074,
      "learning_rate": 2.877770534550196e-05,
      "loss": 0.8574,
      "step": 10000
    },
    {
      "epoch": 0.21390156453715775,
      "grad_norm": 9.001678466796875,
      "learning_rate": 2.8716590612777053e-05,
      "loss": 0.8547,
      "step": 10500
    },
    {
      "epoch": 0.22408735332464147,
      "grad_norm": 7.271191120147705,
      "learning_rate": 2.8655475880052154e-05,
      "loss": 0.8332,
      "step": 11000
    },
    {
      "epoch": 0.23427314211212516,
      "grad_norm": 9.849822998046875,
      "learning_rate": 2.8594361147327248e-05,
      "loss": 0.8447,
      "step": 11500
    },
    {
      "epoch": 0.24445893089960888,
      "grad_norm": 15.07829761505127,
      "learning_rate": 2.853324641460235e-05,
      "loss": 0.8251,
      "step": 12000
    },
    {
      "epoch": 0.25464471968709257,
      "grad_norm": 6.886602878570557,
      "learning_rate": 2.8472131681877446e-05,
      "loss": 0.8444,
      "step": 12500
    },
    {
      "epoch": 0.2648305084745763,
      "grad_norm": 9.94825267791748,
      "learning_rate": 2.8411016949152544e-05,
      "loss": 0.8504,
      "step": 13000
    },
    {
      "epoch": 0.27501629726205995,
      "grad_norm": 8.0740966796875,
      "learning_rate": 2.834990221642764e-05,
      "loss": 0.8182,
      "step": 13500
    },
    {
      "epoch": 0.28520208604954367,
      "grad_norm": 10.770129203796387,
      "learning_rate": 2.828878748370274e-05,
      "loss": 0.8373,
      "step": 14000
    },
    {
      "epoch": 0.2953878748370274,
      "grad_norm": 9.214078903198242,
      "learning_rate": 2.8227672750977836e-05,
      "loss": 0.8089,
      "step": 14500
    },
    {
      "epoch": 0.3055736636245111,
      "grad_norm": 14.231884956359863,
      "learning_rate": 2.8166558018252934e-05,
      "loss": 0.8337,
      "step": 15000
    },
    {
      "epoch": 0.31575945241199477,
      "grad_norm": 9.379185676574707,
      "learning_rate": 2.810544328552803e-05,
      "loss": 0.8315,
      "step": 15500
    },
    {
      "epoch": 0.3259452411994785,
      "grad_norm": 8.370843887329102,
      "learning_rate": 2.804432855280313e-05,
      "loss": 0.808,
      "step": 16000
    },
    {
      "epoch": 0.3361310299869622,
      "grad_norm": 7.201444149017334,
      "learning_rate": 2.7983213820078227e-05,
      "loss": 0.808,
      "step": 16500
    },
    {
      "epoch": 0.34631681877444587,
      "grad_norm": 10.696725845336914,
      "learning_rate": 2.7922099087353324e-05,
      "loss": 0.813,
      "step": 17000
    },
    {
      "epoch": 0.3565026075619296,
      "grad_norm": 6.316204071044922,
      "learning_rate": 2.7860984354628422e-05,
      "loss": 0.8213,
      "step": 17500
    },
    {
      "epoch": 0.3666883963494133,
      "grad_norm": 14.780086517333984,
      "learning_rate": 2.779986962190352e-05,
      "loss": 0.8165,
      "step": 18000
    },
    {
      "epoch": 0.376874185136897,
      "grad_norm": 11.34945011138916,
      "learning_rate": 2.773875488917862e-05,
      "loss": 0.7943,
      "step": 18500
    },
    {
      "epoch": 0.3870599739243807,
      "grad_norm": 7.136452674865723,
      "learning_rate": 2.7677640156453718e-05,
      "loss": 0.8017,
      "step": 19000
    },
    {
      "epoch": 0.3972457627118644,
      "grad_norm": 6.209742069244385,
      "learning_rate": 2.7616525423728815e-05,
      "loss": 0.7941,
      "step": 19500
    },
    {
      "epoch": 0.4074315514993481,
      "grad_norm": 8.923270225524902,
      "learning_rate": 2.7555410691003913e-05,
      "loss": 0.7826,
      "step": 20000
    },
    {
      "epoch": 0.41761734028683184,
      "grad_norm": 6.371203899383545,
      "learning_rate": 2.749429595827901e-05,
      "loss": 0.7938,
      "step": 20500
    },
    {
      "epoch": 0.4278031290743155,
      "grad_norm": 8.679354667663574,
      "learning_rate": 2.7433181225554108e-05,
      "loss": 0.7936,
      "step": 21000
    },
    {
      "epoch": 0.4379889178617992,
      "grad_norm": 5.962934970855713,
      "learning_rate": 2.7372066492829205e-05,
      "loss": 0.7924,
      "step": 21500
    },
    {
      "epoch": 0.44817470664928294,
      "grad_norm": 11.835200309753418,
      "learning_rate": 2.7310951760104303e-05,
      "loss": 0.7894,
      "step": 22000
    },
    {
      "epoch": 0.4583604954367666,
      "grad_norm": 9.62806510925293,
      "learning_rate": 2.72498370273794e-05,
      "loss": 0.7963,
      "step": 22500
    },
    {
      "epoch": 0.4685462842242503,
      "grad_norm": 7.6390509605407715,
      "learning_rate": 2.7188722294654498e-05,
      "loss": 0.783,
      "step": 23000
    },
    {
      "epoch": 0.47873207301173404,
      "grad_norm": 5.826275825500488,
      "learning_rate": 2.7127607561929596e-05,
      "loss": 0.7859,
      "step": 23500
    },
    {
      "epoch": 0.48891786179921776,
      "grad_norm": 5.199416637420654,
      "learning_rate": 2.7066492829204693e-05,
      "loss": 0.7989,
      "step": 24000
    },
    {
      "epoch": 0.4991036505867014,
      "grad_norm": 12.38355541229248,
      "learning_rate": 2.7005378096479794e-05,
      "loss": 0.7864,
      "step": 24500
    },
    {
      "epoch": 0.5092894393741851,
      "grad_norm": 7.700937271118164,
      "learning_rate": 2.694426336375489e-05,
      "loss": 0.7747,
      "step": 25000
    },
    {
      "epoch": 0.5194752281616688,
      "grad_norm": 10.8720703125,
      "learning_rate": 2.688314863102999e-05,
      "loss": 0.7974,
      "step": 25500
    },
    {
      "epoch": 0.5296610169491526,
      "grad_norm": 6.260867595672607,
      "learning_rate": 2.6822033898305083e-05,
      "loss": 0.7914,
      "step": 26000
    },
    {
      "epoch": 0.5398468057366362,
      "grad_norm": 11.113471984863281,
      "learning_rate": 2.6760919165580184e-05,
      "loss": 0.782,
      "step": 26500
    },
    {
      "epoch": 0.5500325945241199,
      "grad_norm": 7.497796058654785,
      "learning_rate": 2.6699804432855282e-05,
      "loss": 0.7642,
      "step": 27000
    },
    {
      "epoch": 0.5602183833116037,
      "grad_norm": 6.1969170570373535,
      "learning_rate": 2.663868970013038e-05,
      "loss": 0.7693,
      "step": 27500
    },
    {
      "epoch": 0.5704041720990873,
      "grad_norm": 9.965928077697754,
      "learning_rate": 2.6577574967405477e-05,
      "loss": 0.7681,
      "step": 28000
    },
    {
      "epoch": 0.5805899608865711,
      "grad_norm": 5.052292346954346,
      "learning_rate": 2.6516460234680575e-05,
      "loss": 0.7664,
      "step": 28500
    },
    {
      "epoch": 0.5907757496740548,
      "grad_norm": 6.7289276123046875,
      "learning_rate": 2.6455345501955672e-05,
      "loss": 0.7445,
      "step": 29000
    },
    {
      "epoch": 0.6009615384615384,
      "grad_norm": 8.478078842163086,
      "learning_rate": 2.639423076923077e-05,
      "loss": 0.7859,
      "step": 29500
    },
    {
      "epoch": 0.6111473272490222,
      "grad_norm": 6.2336106300354,
      "learning_rate": 2.6333116036505867e-05,
      "loss": 0.7598,
      "step": 30000
    },
    {
      "epoch": 0.6213331160365059,
      "grad_norm": 11.82295036315918,
      "learning_rate": 2.6272001303780965e-05,
      "loss": 0.7507,
      "step": 30500
    },
    {
      "epoch": 0.6315189048239895,
      "grad_norm": 8.545003890991211,
      "learning_rate": 2.6210886571056066e-05,
      "loss": 0.7769,
      "step": 31000
    },
    {
      "epoch": 0.6417046936114733,
      "grad_norm": 9.849065780639648,
      "learning_rate": 2.614977183833116e-05,
      "loss": 0.7781,
      "step": 31500
    },
    {
      "epoch": 0.651890482398957,
      "grad_norm": 7.641623020172119,
      "learning_rate": 2.608865710560626e-05,
      "loss": 0.7789,
      "step": 32000
    },
    {
      "epoch": 0.6620762711864406,
      "grad_norm": 16.397785186767578,
      "learning_rate": 2.6027542372881355e-05,
      "loss": 0.7542,
      "step": 32500
    },
    {
      "epoch": 0.6722620599739244,
      "grad_norm": 11.119671821594238,
      "learning_rate": 2.5966427640156456e-05,
      "loss": 0.7716,
      "step": 33000
    },
    {
      "epoch": 0.6824478487614081,
      "grad_norm": 12.466675758361816,
      "learning_rate": 2.590531290743155e-05,
      "loss": 0.7743,
      "step": 33500
    },
    {
      "epoch": 0.6926336375488917,
      "grad_norm": 8.151625633239746,
      "learning_rate": 2.584419817470665e-05,
      "loss": 0.7581,
      "step": 34000
    },
    {
      "epoch": 0.7028194263363755,
      "grad_norm": 7.743143558502197,
      "learning_rate": 2.5783083441981745e-05,
      "loss": 0.7599,
      "step": 34500
    },
    {
      "epoch": 0.7130052151238592,
      "grad_norm": 5.249680995941162,
      "learning_rate": 2.5721968709256846e-05,
      "loss": 0.7483,
      "step": 35000
    },
    {
      "epoch": 0.7231910039113429,
      "grad_norm": 6.738178730010986,
      "learning_rate": 2.566085397653194e-05,
      "loss": 0.7781,
      "step": 35500
    },
    {
      "epoch": 0.7333767926988266,
      "grad_norm": 6.7923102378845215,
      "learning_rate": 2.559973924380704e-05,
      "loss": 0.7464,
      "step": 36000
    },
    {
      "epoch": 0.7435625814863103,
      "grad_norm": 15.575695991516113,
      "learning_rate": 2.553862451108214e-05,
      "loss": 0.7571,
      "step": 36500
    },
    {
      "epoch": 0.753748370273794,
      "grad_norm": 13.128719329833984,
      "learning_rate": 2.5477509778357236e-05,
      "loss": 0.7625,
      "step": 37000
    },
    {
      "epoch": 0.7639341590612777,
      "grad_norm": 4.938621997833252,
      "learning_rate": 2.5416395045632337e-05,
      "loss": 0.7477,
      "step": 37500
    },
    {
      "epoch": 0.7741199478487614,
      "grad_norm": 8.46985912322998,
      "learning_rate": 2.535528031290743e-05,
      "loss": 0.7419,
      "step": 38000
    },
    {
      "epoch": 0.7843057366362451,
      "grad_norm": 18.881208419799805,
      "learning_rate": 2.5294165580182532e-05,
      "loss": 0.7403,
      "step": 38500
    },
    {
      "epoch": 0.7944915254237288,
      "grad_norm": 9.986156463623047,
      "learning_rate": 2.5233050847457626e-05,
      "loss": 0.7508,
      "step": 39000
    },
    {
      "epoch": 0.8046773142112125,
      "grad_norm": 11.625401496887207,
      "learning_rate": 2.5171936114732727e-05,
      "loss": 0.7426,
      "step": 39500
    },
    {
      "epoch": 0.8148631029986962,
      "grad_norm": 9.926522254943848,
      "learning_rate": 2.511082138200782e-05,
      "loss": 0.7365,
      "step": 40000
    },
    {
      "epoch": 0.8250488917861799,
      "grad_norm": 8.104277610778809,
      "learning_rate": 2.5049706649282922e-05,
      "loss": 0.7532,
      "step": 40500
    },
    {
      "epoch": 0.8352346805736637,
      "grad_norm": 5.898177623748779,
      "learning_rate": 2.4988591916558017e-05,
      "loss": 0.7657,
      "step": 41000
    },
    {
      "epoch": 0.8454204693611473,
      "grad_norm": 6.856070041656494,
      "learning_rate": 2.4927477183833118e-05,
      "loss": 0.7438,
      "step": 41500
    },
    {
      "epoch": 0.855606258148631,
      "grad_norm": 11.806628227233887,
      "learning_rate": 2.4866362451108212e-05,
      "loss": 0.7348,
      "step": 42000
    },
    {
      "epoch": 0.8657920469361148,
      "grad_norm": 12.08478832244873,
      "learning_rate": 2.4805247718383313e-05,
      "loss": 0.7277,
      "step": 42500
    },
    {
      "epoch": 0.8759778357235984,
      "grad_norm": 5.707648277282715,
      "learning_rate": 2.474413298565841e-05,
      "loss": 0.7458,
      "step": 43000
    },
    {
      "epoch": 0.8861636245110821,
      "grad_norm": 5.36006498336792,
      "learning_rate": 2.4683018252933508e-05,
      "loss": 0.7446,
      "step": 43500
    },
    {
      "epoch": 0.8963494132985659,
      "grad_norm": 6.603504180908203,
      "learning_rate": 2.4621903520208605e-05,
      "loss": 0.7278,
      "step": 44000
    },
    {
      "epoch": 0.9065352020860495,
      "grad_norm": 7.789953231811523,
      "learning_rate": 2.4560788787483703e-05,
      "loss": 0.7375,
      "step": 44500
    },
    {
      "epoch": 0.9167209908735332,
      "grad_norm": 11.424116134643555,
      "learning_rate": 2.4499674054758804e-05,
      "loss": 0.7297,
      "step": 45000
    },
    {
      "epoch": 0.926906779661017,
      "grad_norm": 6.7079620361328125,
      "learning_rate": 2.4438559322033898e-05,
      "loss": 0.7411,
      "step": 45500
    },
    {
      "epoch": 0.9370925684485006,
      "grad_norm": 11.931846618652344,
      "learning_rate": 2.4377444589309e-05,
      "loss": 0.749,
      "step": 46000
    },
    {
      "epoch": 0.9472783572359843,
      "grad_norm": 8.87076473236084,
      "learning_rate": 2.4316329856584093e-05,
      "loss": 0.7312,
      "step": 46500
    },
    {
      "epoch": 0.9574641460234681,
      "grad_norm": 12.787999153137207,
      "learning_rate": 2.4255215123859194e-05,
      "loss": 0.7244,
      "step": 47000
    },
    {
      "epoch": 0.9676499348109517,
      "grad_norm": 6.917387962341309,
      "learning_rate": 2.4194100391134288e-05,
      "loss": 0.733,
      "step": 47500
    },
    {
      "epoch": 0.9778357235984355,
      "grad_norm": 7.564793109893799,
      "learning_rate": 2.413298565840939e-05,
      "loss": 0.7253,
      "step": 48000
    },
    {
      "epoch": 0.9880215123859192,
      "grad_norm": 7.665174961090088,
      "learning_rate": 2.4071870925684483e-05,
      "loss": 0.7191,
      "step": 48500
    },
    {
      "epoch": 0.9982073011734028,
      "grad_norm": 9.235818862915039,
      "learning_rate": 2.4010756192959584e-05,
      "loss": 0.7288,
      "step": 49000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6841890811920166,
      "eval_loss": 0.7356013059616089,
      "eval_runtime": 8.0037,
      "eval_samples_per_second": 613.213,
      "eval_steps_per_second": 76.714,
      "step": 49088
    },
    {
      "epoch": 1.0083930899608866,
      "grad_norm": 9.232441902160645,
      "learning_rate": 2.3949641460234682e-05,
      "loss": 0.666,
      "step": 49500
    },
    {
      "epoch": 1.0185788787483703,
      "grad_norm": 5.098124980926514,
      "learning_rate": 2.388852672750978e-05,
      "loss": 0.6762,
      "step": 50000
    },
    {
      "epoch": 1.028764667535854,
      "grad_norm": 6.507536888122559,
      "learning_rate": 2.3827411994784877e-05,
      "loss": 0.6509,
      "step": 50500
    },
    {
      "epoch": 1.0389504563233376,
      "grad_norm": 12.516643524169922,
      "learning_rate": 2.3766297262059974e-05,
      "loss": 0.6569,
      "step": 51000
    },
    {
      "epoch": 1.0491362451108215,
      "grad_norm": 11.644853591918945,
      "learning_rate": 2.3705182529335072e-05,
      "loss": 0.6498,
      "step": 51500
    },
    {
      "epoch": 1.0593220338983051,
      "grad_norm": 15.303723335266113,
      "learning_rate": 2.364406779661017e-05,
      "loss": 0.6565,
      "step": 52000
    },
    {
      "epoch": 1.0695078226857888,
      "grad_norm": 11.212811470031738,
      "learning_rate": 2.3582953063885267e-05,
      "loss": 0.6492,
      "step": 52500
    },
    {
      "epoch": 1.0796936114732725,
      "grad_norm": 7.695071220397949,
      "learning_rate": 2.3521838331160365e-05,
      "loss": 0.6618,
      "step": 53000
    },
    {
      "epoch": 1.0898794002607561,
      "grad_norm": 8.234328269958496,
      "learning_rate": 2.3460723598435462e-05,
      "loss": 0.6709,
      "step": 53500
    },
    {
      "epoch": 1.1000651890482398,
      "grad_norm": 12.574545860290527,
      "learning_rate": 2.339960886571056e-05,
      "loss": 0.6518,
      "step": 54000
    },
    {
      "epoch": 1.1102509778357237,
      "grad_norm": 5.914109706878662,
      "learning_rate": 2.333849413298566e-05,
      "loss": 0.6494,
      "step": 54500
    },
    {
      "epoch": 1.1204367666232073,
      "grad_norm": 11.49263858795166,
      "learning_rate": 2.3277379400260758e-05,
      "loss": 0.6723,
      "step": 55000
    },
    {
      "epoch": 1.130622555410691,
      "grad_norm": 11.228019714355469,
      "learning_rate": 2.3216264667535856e-05,
      "loss": 0.6328,
      "step": 55500
    },
    {
      "epoch": 1.1408083441981747,
      "grad_norm": 14.478355407714844,
      "learning_rate": 2.3155149934810953e-05,
      "loss": 0.667,
      "step": 56000
    },
    {
      "epoch": 1.1509941329856583,
      "grad_norm": 18.817468643188477,
      "learning_rate": 2.309403520208605e-05,
      "loss": 0.6674,
      "step": 56500
    },
    {
      "epoch": 1.161179921773142,
      "grad_norm": 12.483678817749023,
      "learning_rate": 2.303292046936115e-05,
      "loss": 0.6677,
      "step": 57000
    },
    {
      "epoch": 1.1713657105606259,
      "grad_norm": 7.133495807647705,
      "learning_rate": 2.2971805736636246e-05,
      "loss": 0.642,
      "step": 57500
    },
    {
      "epoch": 1.1815514993481095,
      "grad_norm": 7.099282741546631,
      "learning_rate": 2.2910691003911343e-05,
      "loss": 0.664,
      "step": 58000
    },
    {
      "epoch": 1.1917372881355932,
      "grad_norm": 9.188867568969727,
      "learning_rate": 2.284957627118644e-05,
      "loss": 0.6703,
      "step": 58500
    },
    {
      "epoch": 1.2019230769230769,
      "grad_norm": 5.470687389373779,
      "learning_rate": 2.278846153846154e-05,
      "loss": 0.6575,
      "step": 59000
    },
    {
      "epoch": 1.2121088657105605,
      "grad_norm": 8.013532638549805,
      "learning_rate": 2.2727346805736636e-05,
      "loss": 0.6465,
      "step": 59500
    },
    {
      "epoch": 1.2222946544980444,
      "grad_norm": 8.846539497375488,
      "learning_rate": 2.2666232073011734e-05,
      "loss": 0.6682,
      "step": 60000
    },
    {
      "epoch": 1.232480443285528,
      "grad_norm": 11.2225341796875,
      "learning_rate": 2.260511734028683e-05,
      "loss": 0.653,
      "step": 60500
    },
    {
      "epoch": 1.2426662320730117,
      "grad_norm": 12.363032341003418,
      "learning_rate": 2.254400260756193e-05,
      "loss": 0.6807,
      "step": 61000
    },
    {
      "epoch": 1.2528520208604954,
      "grad_norm": 8.01014232635498,
      "learning_rate": 2.248288787483703e-05,
      "loss": 0.6625,
      "step": 61500
    },
    {
      "epoch": 1.263037809647979,
      "grad_norm": 20.960845947265625,
      "learning_rate": 2.2421773142112124e-05,
      "loss": 0.6524,
      "step": 62000
    },
    {
      "epoch": 1.2732235984354627,
      "grad_norm": 13.620718002319336,
      "learning_rate": 2.2360658409387225e-05,
      "loss": 0.6625,
      "step": 62500
    },
    {
      "epoch": 1.2834093872229466,
      "grad_norm": 18.55965232849121,
      "learning_rate": 2.2299543676662322e-05,
      "loss": 0.6434,
      "step": 63000
    },
    {
      "epoch": 1.2935951760104303,
      "grad_norm": 12.919173240661621,
      "learning_rate": 2.223842894393742e-05,
      "loss": 0.652,
      "step": 63500
    },
    {
      "epoch": 1.303780964797914,
      "grad_norm": 19.561386108398438,
      "learning_rate": 2.2177314211212517e-05,
      "loss": 0.6498,
      "step": 64000
    },
    {
      "epoch": 1.3139667535853976,
      "grad_norm": 10.529995918273926,
      "learning_rate": 2.2116199478487615e-05,
      "loss": 0.666,
      "step": 64500
    },
    {
      "epoch": 1.3241525423728815,
      "grad_norm": 13.268050193786621,
      "learning_rate": 2.2055084745762713e-05,
      "loss": 0.6499,
      "step": 65000
    },
    {
      "epoch": 1.3343383311603652,
      "grad_norm": 6.3072285652160645,
      "learning_rate": 2.199397001303781e-05,
      "loss": 0.6798,
      "step": 65500
    },
    {
      "epoch": 1.3445241199478488,
      "grad_norm": 5.304599761962891,
      "learning_rate": 2.1932855280312908e-05,
      "loss": 0.6559,
      "step": 66000
    },
    {
      "epoch": 1.3547099087353325,
      "grad_norm": 13.698792457580566,
      "learning_rate": 2.1871740547588005e-05,
      "loss": 0.6441,
      "step": 66500
    },
    {
      "epoch": 1.3648956975228161,
      "grad_norm": 8.951632499694824,
      "learning_rate": 2.1810625814863103e-05,
      "loss": 0.6564,
      "step": 67000
    },
    {
      "epoch": 1.3750814863102998,
      "grad_norm": 5.84067440032959,
      "learning_rate": 2.17495110821382e-05,
      "loss": 0.6686,
      "step": 67500
    },
    {
      "epoch": 1.3852672750977835,
      "grad_norm": 6.442080497741699,
      "learning_rate": 2.16883963494133e-05,
      "loss": 0.6517,
      "step": 68000
    },
    {
      "epoch": 1.3954530638852674,
      "grad_norm": 5.514212608337402,
      "learning_rate": 2.1627281616688395e-05,
      "loss": 0.6664,
      "step": 68500
    },
    {
      "epoch": 1.405638852672751,
      "grad_norm": 14.150158882141113,
      "learning_rate": 2.1566166883963496e-05,
      "loss": 0.6561,
      "step": 69000
    },
    {
      "epoch": 1.4158246414602347,
      "grad_norm": 17.196884155273438,
      "learning_rate": 2.150505215123859e-05,
      "loss": 0.6513,
      "step": 69500
    },
    {
      "epoch": 1.4260104302477183,
      "grad_norm": 6.181870937347412,
      "learning_rate": 2.144393741851369e-05,
      "loss": 0.65,
      "step": 70000
    },
    {
      "epoch": 1.436196219035202,
      "grad_norm": 14.865707397460938,
      "learning_rate": 2.1382822685788786e-05,
      "loss": 0.6596,
      "step": 70500
    },
    {
      "epoch": 1.4463820078226859,
      "grad_norm": 15.808574676513672,
      "learning_rate": 2.1321707953063886e-05,
      "loss": 0.6559,
      "step": 71000
    },
    {
      "epoch": 1.4565677966101696,
      "grad_norm": 16.76003074645996,
      "learning_rate": 2.1260593220338984e-05,
      "loss": 0.6467,
      "step": 71500
    },
    {
      "epoch": 1.4667535853976532,
      "grad_norm": 21.466825485229492,
      "learning_rate": 2.119947848761408e-05,
      "loss": 0.6579,
      "step": 72000
    },
    {
      "epoch": 1.4769393741851369,
      "grad_norm": 18.81052589416504,
      "learning_rate": 2.113836375488918e-05,
      "loss": 0.6699,
      "step": 72500
    },
    {
      "epoch": 1.4871251629726205,
      "grad_norm": 10.334957122802734,
      "learning_rate": 2.1077249022164277e-05,
      "loss": 0.655,
      "step": 73000
    },
    {
      "epoch": 1.4973109517601042,
      "grad_norm": 14.836852073669434,
      "learning_rate": 2.1016134289439378e-05,
      "loss": 0.66,
      "step": 73500
    },
    {
      "epoch": 1.5074967405475879,
      "grad_norm": 21.439233779907227,
      "learning_rate": 2.0955019556714472e-05,
      "loss": 0.6486,
      "step": 74000
    },
    {
      "epoch": 1.5176825293350718,
      "grad_norm": 21.09993553161621,
      "learning_rate": 2.0893904823989573e-05,
      "loss": 0.6432,
      "step": 74500
    },
    {
      "epoch": 1.5278683181225554,
      "grad_norm": 8.926955223083496,
      "learning_rate": 2.0832790091264667e-05,
      "loss": 0.6418,
      "step": 75000
    },
    {
      "epoch": 1.538054106910039,
      "grad_norm": 3.551163911819458,
      "learning_rate": 2.0771675358539768e-05,
      "loss": 0.6418,
      "step": 75500
    },
    {
      "epoch": 1.548239895697523,
      "grad_norm": 10.46932315826416,
      "learning_rate": 2.0710560625814862e-05,
      "loss": 0.657,
      "step": 76000
    },
    {
      "epoch": 1.5584256844850066,
      "grad_norm": 10.022995948791504,
      "learning_rate": 2.0649445893089963e-05,
      "loss": 0.6564,
      "step": 76500
    },
    {
      "epoch": 1.5686114732724903,
      "grad_norm": 15.299884796142578,
      "learning_rate": 2.0588331160365057e-05,
      "loss": 0.6535,
      "step": 77000
    },
    {
      "epoch": 1.578797262059974,
      "grad_norm": 11.149444580078125,
      "learning_rate": 2.0527216427640158e-05,
      "loss": 0.6409,
      "step": 77500
    },
    {
      "epoch": 1.5889830508474576,
      "grad_norm": 4.011179447174072,
      "learning_rate": 2.0466101694915252e-05,
      "loss": 0.6299,
      "step": 78000
    },
    {
      "epoch": 1.5991688396349413,
      "grad_norm": 13.744772911071777,
      "learning_rate": 2.0404986962190353e-05,
      "loss": 0.6511,
      "step": 78500
    },
    {
      "epoch": 1.609354628422425,
      "grad_norm": 11.129698753356934,
      "learning_rate": 2.0343872229465447e-05,
      "loss": 0.6349,
      "step": 79000
    },
    {
      "epoch": 1.6195404172099086,
      "grad_norm": 16.331953048706055,
      "learning_rate": 2.0282757496740548e-05,
      "loss": 0.6517,
      "step": 79500
    },
    {
      "epoch": 1.6297262059973925,
      "grad_norm": 12.816121101379395,
      "learning_rate": 2.0221642764015646e-05,
      "loss": 0.6409,
      "step": 80000
    },
    {
      "epoch": 1.6399119947848761,
      "grad_norm": 13.881726264953613,
      "learning_rate": 2.0160528031290743e-05,
      "loss": 0.6606,
      "step": 80500
    },
    {
      "epoch": 1.6500977835723598,
      "grad_norm": 14.215683937072754,
      "learning_rate": 2.0099413298565844e-05,
      "loss": 0.657,
      "step": 81000
    },
    {
      "epoch": 1.6602835723598437,
      "grad_norm": 19.557342529296875,
      "learning_rate": 2.003829856584094e-05,
      "loss": 0.6451,
      "step": 81500
    },
    {
      "epoch": 1.6704693611473274,
      "grad_norm": 17.83816909790039,
      "learning_rate": 1.997718383311604e-05,
      "loss": 0.6397,
      "step": 82000
    },
    {
      "epoch": 1.680655149934811,
      "grad_norm": 8.789198875427246,
      "learning_rate": 1.9916069100391133e-05,
      "loss": 0.6514,
      "step": 82500
    },
    {
      "epoch": 1.6908409387222947,
      "grad_norm": 11.361543655395508,
      "learning_rate": 1.9854954367666234e-05,
      "loss": 0.6466,
      "step": 83000
    },
    {
      "epoch": 1.7010267275097783,
      "grad_norm": 15.630517959594727,
      "learning_rate": 1.979383963494133e-05,
      "loss": 0.6316,
      "step": 83500
    },
    {
      "epoch": 1.711212516297262,
      "grad_norm": 10.55432415008545,
      "learning_rate": 1.973272490221643e-05,
      "loss": 0.6533,
      "step": 84000
    },
    {
      "epoch": 1.7213983050847457,
      "grad_norm": 8.428609848022461,
      "learning_rate": 1.9671610169491524e-05,
      "loss": 0.6478,
      "step": 84500
    },
    {
      "epoch": 1.7315840938722293,
      "grad_norm": 4.304576873779297,
      "learning_rate": 1.9610495436766625e-05,
      "loss": 0.6449,
      "step": 85000
    },
    {
      "epoch": 1.7417698826597132,
      "grad_norm": 21.92937469482422,
      "learning_rate": 1.9549380704041722e-05,
      "loss": 0.6425,
      "step": 85500
    },
    {
      "epoch": 1.7519556714471969,
      "grad_norm": 19.813392639160156,
      "learning_rate": 1.948826597131682e-05,
      "loss": 0.6238,
      "step": 86000
    },
    {
      "epoch": 1.7621414602346805,
      "grad_norm": 6.831646919250488,
      "learning_rate": 1.9427151238591917e-05,
      "loss": 0.6553,
      "step": 86500
    },
    {
      "epoch": 1.7723272490221644,
      "grad_norm": 11.058158874511719,
      "learning_rate": 1.9366036505867015e-05,
      "loss": 0.6528,
      "step": 87000
    },
    {
      "epoch": 1.782513037809648,
      "grad_norm": 7.01440954208374,
      "learning_rate": 1.9304921773142112e-05,
      "loss": 0.6506,
      "step": 87500
    },
    {
      "epoch": 1.7926988265971318,
      "grad_norm": 4.963765621185303,
      "learning_rate": 1.924380704041721e-05,
      "loss": 0.6412,
      "step": 88000
    },
    {
      "epoch": 1.8028846153846154,
      "grad_norm": 38.63767623901367,
      "learning_rate": 1.9182692307692307e-05,
      "loss": 0.6629,
      "step": 88500
    },
    {
      "epoch": 1.813070404172099,
      "grad_norm": 11.423843383789062,
      "learning_rate": 1.9121577574967405e-05,
      "loss": 0.673,
      "step": 89000
    },
    {
      "epoch": 1.8232561929595827,
      "grad_norm": 11.73025894165039,
      "learning_rate": 1.9060462842242506e-05,
      "loss": 0.625,
      "step": 89500
    },
    {
      "epoch": 1.8334419817470664,
      "grad_norm": 11.493837356567383,
      "learning_rate": 1.89993481095176e-05,
      "loss": 0.6658,
      "step": 90000
    },
    {
      "epoch": 1.84362777053455,
      "grad_norm": 12.196702003479004,
      "learning_rate": 1.89382333767927e-05,
      "loss": 0.6451,
      "step": 90500
    },
    {
      "epoch": 1.8538135593220337,
      "grad_norm": 9.690689086914062,
      "learning_rate": 1.8877118644067795e-05,
      "loss": 0.6282,
      "step": 91000
    },
    {
      "epoch": 1.8639993481095176,
      "grad_norm": 13.679101943969727,
      "learning_rate": 1.8816003911342896e-05,
      "loss": 0.6328,
      "step": 91500
    },
    {
      "epoch": 1.8741851368970013,
      "grad_norm": 14.264899253845215,
      "learning_rate": 1.8754889178617994e-05,
      "loss": 0.6355,
      "step": 92000
    },
    {
      "epoch": 1.8843709256844852,
      "grad_norm": 9.039682388305664,
      "learning_rate": 1.869377444589309e-05,
      "loss": 0.6471,
      "step": 92500
    },
    {
      "epoch": 1.8945567144719688,
      "grad_norm": 9.861899375915527,
      "learning_rate": 1.863265971316819e-05,
      "loss": 0.6549,
      "step": 93000
    },
    {
      "epoch": 1.9047425032594525,
      "grad_norm": 14.948457717895508,
      "learning_rate": 1.8571544980443286e-05,
      "loss": 0.6436,
      "step": 93500
    },
    {
      "epoch": 1.9149282920469362,
      "grad_norm": 15.613487243652344,
      "learning_rate": 1.8510430247718384e-05,
      "loss": 0.6355,
      "step": 94000
    },
    {
      "epoch": 1.9251140808344198,
      "grad_norm": 14.105792045593262,
      "learning_rate": 1.844931551499348e-05,
      "loss": 0.6345,
      "step": 94500
    },
    {
      "epoch": 1.9352998696219035,
      "grad_norm": 7.004807472229004,
      "learning_rate": 1.838820078226858e-05,
      "loss": 0.6456,
      "step": 95000
    },
    {
      "epoch": 1.9454856584093871,
      "grad_norm": 10.506879806518555,
      "learning_rate": 1.8327086049543677e-05,
      "loss": 0.6303,
      "step": 95500
    },
    {
      "epoch": 1.9556714471968708,
      "grad_norm": 11.595491409301758,
      "learning_rate": 1.8265971316818774e-05,
      "loss": 0.6509,
      "step": 96000
    },
    {
      "epoch": 1.9658572359843545,
      "grad_norm": 11.29542064666748,
      "learning_rate": 1.820485658409387e-05,
      "loss": 0.6273,
      "step": 96500
    },
    {
      "epoch": 1.9760430247718384,
      "grad_norm": 4.132607460021973,
      "learning_rate": 1.814374185136897e-05,
      "loss": 0.643,
      "step": 97000
    },
    {
      "epoch": 1.986228813559322,
      "grad_norm": 9.745738983154297,
      "learning_rate": 1.8082627118644067e-05,
      "loss": 0.6656,
      "step": 97500
    },
    {
      "epoch": 1.996414602346806,
      "grad_norm": 4.289452075958252,
      "learning_rate": 1.8021512385919164e-05,
      "loss": 0.6367,
      "step": 98000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7218826413154602,
      "eval_loss": 0.6872764825820923,
      "eval_runtime": 7.6175,
      "eval_samples_per_second": 644.307,
      "eval_steps_per_second": 80.604,
      "step": 98176
    },
    {
      "epoch": 2.0066003911342896,
      "grad_norm": 19.260536193847656,
      "learning_rate": 1.7960397653194265e-05,
      "loss": 0.5568,
      "step": 98500
    },
    {
      "epoch": 2.0167861799217732,
      "grad_norm": 8.347018241882324,
      "learning_rate": 1.7899282920469363e-05,
      "loss": 0.5344,
      "step": 99000
    },
    {
      "epoch": 2.026971968709257,
      "grad_norm": 12.238682746887207,
      "learning_rate": 1.783816818774446e-05,
      "loss": 0.5685,
      "step": 99500
    },
    {
      "epoch": 2.0371577574967406,
      "grad_norm": 18.42786979675293,
      "learning_rate": 1.7777053455019558e-05,
      "loss": 0.5539,
      "step": 100000
    },
    {
      "epoch": 2.047343546284224,
      "grad_norm": 25.627717971801758,
      "learning_rate": 1.7715938722294655e-05,
      "loss": 0.5393,
      "step": 100500
    },
    {
      "epoch": 2.057529335071708,
      "grad_norm": 15.942806243896484,
      "learning_rate": 1.7654823989569753e-05,
      "loss": 0.5191,
      "step": 101000
    },
    {
      "epoch": 2.0677151238591915,
      "grad_norm": 7.591663360595703,
      "learning_rate": 1.759370925684485e-05,
      "loss": 0.5428,
      "step": 101500
    },
    {
      "epoch": 2.077900912646675,
      "grad_norm": 7.183356285095215,
      "learning_rate": 1.7532594524119948e-05,
      "loss": 0.5475,
      "step": 102000
    },
    {
      "epoch": 2.088086701434159,
      "grad_norm": 19.916288375854492,
      "learning_rate": 1.7471479791395046e-05,
      "loss": 0.5586,
      "step": 102500
    },
    {
      "epoch": 2.098272490221643,
      "grad_norm": 9.07480239868164,
      "learning_rate": 1.7410365058670143e-05,
      "loss": 0.5564,
      "step": 103000
    },
    {
      "epoch": 2.1084582790091266,
      "grad_norm": 7.647058010101318,
      "learning_rate": 1.734925032594524e-05,
      "loss": 0.543,
      "step": 103500
    },
    {
      "epoch": 2.1186440677966103,
      "grad_norm": 8.871644973754883,
      "learning_rate": 1.728813559322034e-05,
      "loss": 0.5623,
      "step": 104000
    },
    {
      "epoch": 2.128829856584094,
      "grad_norm": 9.516389846801758,
      "learning_rate": 1.7227020860495436e-05,
      "loss": 0.5464,
      "step": 104500
    },
    {
      "epoch": 2.1390156453715776,
      "grad_norm": 13.605746269226074,
      "learning_rate": 1.7165906127770537e-05,
      "loss": 0.5425,
      "step": 105000
    },
    {
      "epoch": 2.1492014341590613,
      "grad_norm": 14.016572952270508,
      "learning_rate": 1.710479139504563e-05,
      "loss": 0.5496,
      "step": 105500
    },
    {
      "epoch": 2.159387222946545,
      "grad_norm": 14.719120979309082,
      "learning_rate": 1.7043676662320732e-05,
      "loss": 0.5591,
      "step": 106000
    },
    {
      "epoch": 2.1695730117340286,
      "grad_norm": 14.922130584716797,
      "learning_rate": 1.6982561929595826e-05,
      "loss": 0.5462,
      "step": 106500
    },
    {
      "epoch": 2.1797588005215123,
      "grad_norm": 10.74787425994873,
      "learning_rate": 1.6921447196870927e-05,
      "loss": 0.5699,
      "step": 107000
    },
    {
      "epoch": 2.189944589308996,
      "grad_norm": 7.978420257568359,
      "learning_rate": 1.6860332464146024e-05,
      "loss": 0.5449,
      "step": 107500
    },
    {
      "epoch": 2.2001303780964796,
      "grad_norm": 15.361347198486328,
      "learning_rate": 1.6799217731421122e-05,
      "loss": 0.5443,
      "step": 108000
    },
    {
      "epoch": 2.2103161668839633,
      "grad_norm": 6.552661418914795,
      "learning_rate": 1.673810299869622e-05,
      "loss": 0.5663,
      "step": 108500
    },
    {
      "epoch": 2.2205019556714474,
      "grad_norm": 6.750521659851074,
      "learning_rate": 1.6676988265971317e-05,
      "loss": 0.5488,
      "step": 109000
    },
    {
      "epoch": 2.230687744458931,
      "grad_norm": 11.149799346923828,
      "learning_rate": 1.6615873533246415e-05,
      "loss": 0.5412,
      "step": 109500
    },
    {
      "epoch": 2.2408735332464147,
      "grad_norm": 19.362638473510742,
      "learning_rate": 1.6554758800521512e-05,
      "loss": 0.5505,
      "step": 110000
    },
    {
      "epoch": 2.2510593220338984,
      "grad_norm": 11.263615608215332,
      "learning_rate": 1.6493644067796613e-05,
      "loss": 0.5591,
      "step": 110500
    },
    {
      "epoch": 2.261245110821382,
      "grad_norm": 8.90456771850586,
      "learning_rate": 1.6432529335071707e-05,
      "loss": 0.5491,
      "step": 111000
    },
    {
      "epoch": 2.2714308996088657,
      "grad_norm": 11.962569236755371,
      "learning_rate": 1.6371414602346808e-05,
      "loss": 0.556,
      "step": 111500
    },
    {
      "epoch": 2.2816166883963493,
      "grad_norm": 8.397544860839844,
      "learning_rate": 1.6310299869621902e-05,
      "loss": 0.5561,
      "step": 112000
    },
    {
      "epoch": 2.291802477183833,
      "grad_norm": 16.376155853271484,
      "learning_rate": 1.6249185136897003e-05,
      "loss": 0.5435,
      "step": 112500
    },
    {
      "epoch": 2.3019882659713167,
      "grad_norm": 9.14609432220459,
      "learning_rate": 1.6188070404172097e-05,
      "loss": 0.5457,
      "step": 113000
    },
    {
      "epoch": 2.3121740547588003,
      "grad_norm": 26.53936767578125,
      "learning_rate": 1.61269556714472e-05,
      "loss": 0.5295,
      "step": 113500
    },
    {
      "epoch": 2.322359843546284,
      "grad_norm": 17.222721099853516,
      "learning_rate": 1.6065840938722293e-05,
      "loss": 0.5177,
      "step": 114000
    },
    {
      "epoch": 2.332545632333768,
      "grad_norm": 9.585100173950195,
      "learning_rate": 1.6004726205997394e-05,
      "loss": 0.5419,
      "step": 114500
    },
    {
      "epoch": 2.3427314211212518,
      "grad_norm": 9.392489433288574,
      "learning_rate": 1.5943611473272488e-05,
      "loss": 0.5296,
      "step": 115000
    },
    {
      "epoch": 2.3529172099087354,
      "grad_norm": 15.146162986755371,
      "learning_rate": 1.588249674054759e-05,
      "loss": 0.5518,
      "step": 115500
    },
    {
      "epoch": 2.363102998696219,
      "grad_norm": 15.631641387939453,
      "learning_rate": 1.582138200782269e-05,
      "loss": 0.5584,
      "step": 116000
    },
    {
      "epoch": 2.3732887874837028,
      "grad_norm": 17.875492095947266,
      "learning_rate": 1.5760267275097784e-05,
      "loss": 0.5601,
      "step": 116500
    },
    {
      "epoch": 2.3834745762711864,
      "grad_norm": 7.7471184730529785,
      "learning_rate": 1.5699152542372885e-05,
      "loss": 0.549,
      "step": 117000
    },
    {
      "epoch": 2.39366036505867,
      "grad_norm": 14.712841987609863,
      "learning_rate": 1.563803780964798e-05,
      "loss": 0.5293,
      "step": 117500
    },
    {
      "epoch": 2.4038461538461537,
      "grad_norm": 9.229011535644531,
      "learning_rate": 1.557692307692308e-05,
      "loss": 0.5483,
      "step": 118000
    },
    {
      "epoch": 2.4140319426336374,
      "grad_norm": 11.47548770904541,
      "learning_rate": 1.5515808344198174e-05,
      "loss": 0.5529,
      "step": 118500
    },
    {
      "epoch": 2.424217731421121,
      "grad_norm": 29.980873107910156,
      "learning_rate": 1.5454693611473275e-05,
      "loss": 0.5489,
      "step": 119000
    },
    {
      "epoch": 2.4344035202086047,
      "grad_norm": 10.478185653686523,
      "learning_rate": 1.539357887874837e-05,
      "loss": 0.5595,
      "step": 119500
    },
    {
      "epoch": 2.444589308996089,
      "grad_norm": 12.978096008300781,
      "learning_rate": 1.533246414602347e-05,
      "loss": 0.5612,
      "step": 120000
    },
    {
      "epoch": 2.4547750977835725,
      "grad_norm": 17.834806442260742,
      "learning_rate": 1.5271349413298564e-05,
      "loss": 0.5412,
      "step": 120500
    },
    {
      "epoch": 2.464960886571056,
      "grad_norm": 25.001754760742188,
      "learning_rate": 1.5210234680573665e-05,
      "loss": 0.5673,
      "step": 121000
    },
    {
      "epoch": 2.47514667535854,
      "grad_norm": 27.953767776489258,
      "learning_rate": 1.5149119947848761e-05,
      "loss": 0.5497,
      "step": 121500
    },
    {
      "epoch": 2.4853324641460235,
      "grad_norm": 9.370855331420898,
      "learning_rate": 1.508800521512386e-05,
      "loss": 0.5344,
      "step": 122000
    },
    {
      "epoch": 2.495518252933507,
      "grad_norm": 18.109821319580078,
      "learning_rate": 1.5026890482398956e-05,
      "loss": 0.5572,
      "step": 122500
    },
    {
      "epoch": 2.505704041720991,
      "grad_norm": 20.435270309448242,
      "learning_rate": 1.4965775749674055e-05,
      "loss": 0.5631,
      "step": 123000
    },
    {
      "epoch": 2.5158898305084745,
      "grad_norm": 10.656272888183594,
      "learning_rate": 1.4904661016949153e-05,
      "loss": 0.5344,
      "step": 123500
    },
    {
      "epoch": 2.526075619295958,
      "grad_norm": 18.796552658081055,
      "learning_rate": 1.484354628422425e-05,
      "loss": 0.5464,
      "step": 124000
    },
    {
      "epoch": 2.5362614080834422,
      "grad_norm": 9.630306243896484,
      "learning_rate": 1.4782431551499348e-05,
      "loss": 0.547,
      "step": 124500
    },
    {
      "epoch": 2.5464471968709255,
      "grad_norm": 6.3599467277526855,
      "learning_rate": 1.4721316818774445e-05,
      "loss": 0.5577,
      "step": 125000
    },
    {
      "epoch": 2.5566329856584096,
      "grad_norm": 8.967144966125488,
      "learning_rate": 1.4660202086049545e-05,
      "loss": 0.5524,
      "step": 125500
    },
    {
      "epoch": 2.5668187744458932,
      "grad_norm": 16.690263748168945,
      "learning_rate": 1.4599087353324642e-05,
      "loss": 0.5546,
      "step": 126000
    },
    {
      "epoch": 2.577004563233377,
      "grad_norm": 17.713790893554688,
      "learning_rate": 1.453797262059974e-05,
      "loss": 0.5686,
      "step": 126500
    },
    {
      "epoch": 2.5871903520208606,
      "grad_norm": 7.970760822296143,
      "learning_rate": 1.4476857887874837e-05,
      "loss": 0.5273,
      "step": 127000
    },
    {
      "epoch": 2.5973761408083442,
      "grad_norm": 8.869507789611816,
      "learning_rate": 1.4415743155149935e-05,
      "loss": 0.5472,
      "step": 127500
    },
    {
      "epoch": 2.607561929595828,
      "grad_norm": 17.247589111328125,
      "learning_rate": 1.4354628422425032e-05,
      "loss": 0.5444,
      "step": 128000
    },
    {
      "epoch": 2.6177477183833116,
      "grad_norm": 16.415138244628906,
      "learning_rate": 1.429351368970013e-05,
      "loss": 0.5445,
      "step": 128500
    },
    {
      "epoch": 2.627933507170795,
      "grad_norm": 10.408729553222656,
      "learning_rate": 1.4232398956975227e-05,
      "loss": 0.5667,
      "step": 129000
    },
    {
      "epoch": 2.638119295958279,
      "grad_norm": 53.77888488769531,
      "learning_rate": 1.4171284224250327e-05,
      "loss": 0.547,
      "step": 129500
    },
    {
      "epoch": 2.648305084745763,
      "grad_norm": 12.244620323181152,
      "learning_rate": 1.4110169491525424e-05,
      "loss": 0.5428,
      "step": 130000
    },
    {
      "epoch": 2.658490873533246,
      "grad_norm": 11.80124568939209,
      "learning_rate": 1.4049054758800522e-05,
      "loss": 0.5565,
      "step": 130500
    },
    {
      "epoch": 2.6686766623207303,
      "grad_norm": 13.250449180603027,
      "learning_rate": 1.398794002607562e-05,
      "loss": 0.5423,
      "step": 131000
    },
    {
      "epoch": 2.678862451108214,
      "grad_norm": 15.81285572052002,
      "learning_rate": 1.3926825293350719e-05,
      "loss": 0.5621,
      "step": 131500
    },
    {
      "epoch": 2.6890482398956976,
      "grad_norm": 22.584035873413086,
      "learning_rate": 1.3865710560625816e-05,
      "loss": 0.5521,
      "step": 132000
    },
    {
      "epoch": 2.6992340286831813,
      "grad_norm": 23.865680694580078,
      "learning_rate": 1.3804595827900914e-05,
      "loss": 0.551,
      "step": 132500
    },
    {
      "epoch": 2.709419817470665,
      "grad_norm": 18.28876304626465,
      "learning_rate": 1.3743481095176011e-05,
      "loss": 0.5451,
      "step": 133000
    },
    {
      "epoch": 2.7196056062581486,
      "grad_norm": 16.23969078063965,
      "learning_rate": 1.3682366362451109e-05,
      "loss": 0.5392,
      "step": 133500
    },
    {
      "epoch": 2.7297913950456323,
      "grad_norm": 14.673959732055664,
      "learning_rate": 1.3621251629726206e-05,
      "loss": 0.5451,
      "step": 134000
    },
    {
      "epoch": 2.739977183833116,
      "grad_norm": 8.019514083862305,
      "learning_rate": 1.3560136897001304e-05,
      "loss": 0.5568,
      "step": 134500
    },
    {
      "epoch": 2.7501629726205996,
      "grad_norm": 13.495898246765137,
      "learning_rate": 1.3499022164276401e-05,
      "loss": 0.5734,
      "step": 135000
    },
    {
      "epoch": 2.7603487614080837,
      "grad_norm": 7.548976421356201,
      "learning_rate": 1.3437907431551499e-05,
      "loss": 0.557,
      "step": 135500
    },
    {
      "epoch": 2.770534550195567,
      "grad_norm": 20.049760818481445,
      "learning_rate": 1.3376792698826597e-05,
      "loss": 0.5658,
      "step": 136000
    },
    {
      "epoch": 2.780720338983051,
      "grad_norm": 9.346122741699219,
      "learning_rate": 1.3315677966101694e-05,
      "loss": 0.535,
      "step": 136500
    },
    {
      "epoch": 2.7909061277705347,
      "grad_norm": 15.080660820007324,
      "learning_rate": 1.3254563233376792e-05,
      "loss": 0.5538,
      "step": 137000
    },
    {
      "epoch": 2.8010919165580184,
      "grad_norm": 11.485374450683594,
      "learning_rate": 1.3193448500651891e-05,
      "loss": 0.5419,
      "step": 137500
    },
    {
      "epoch": 2.811277705345502,
      "grad_norm": 12.089446067810059,
      "learning_rate": 1.3132333767926988e-05,
      "loss": 0.5651,
      "step": 138000
    },
    {
      "epoch": 2.8214634941329857,
      "grad_norm": 5.466490268707275,
      "learning_rate": 1.3071219035202088e-05,
      "loss": 0.5217,
      "step": 138500
    },
    {
      "epoch": 2.8316492829204694,
      "grad_norm": 12.89148235321045,
      "learning_rate": 1.3010104302477185e-05,
      "loss": 0.5477,
      "step": 139000
    },
    {
      "epoch": 2.841835071707953,
      "grad_norm": 5.610709190368652,
      "learning_rate": 1.2948989569752283e-05,
      "loss": 0.5377,
      "step": 139500
    },
    {
      "epoch": 2.8520208604954367,
      "grad_norm": 26.2186222076416,
      "learning_rate": 1.288787483702738e-05,
      "loss": 0.5281,
      "step": 140000
    },
    {
      "epoch": 2.8622066492829203,
      "grad_norm": 2.1066462993621826,
      "learning_rate": 1.2826760104302478e-05,
      "loss": 0.5387,
      "step": 140500
    },
    {
      "epoch": 2.872392438070404,
      "grad_norm": 13.6646728515625,
      "learning_rate": 1.2765645371577575e-05,
      "loss": 0.5416,
      "step": 141000
    },
    {
      "epoch": 2.8825782268578877,
      "grad_norm": 14.357284545898438,
      "learning_rate": 1.2704530638852673e-05,
      "loss": 0.5504,
      "step": 141500
    },
    {
      "epoch": 2.8927640156453718,
      "grad_norm": 14.08674144744873,
      "learning_rate": 1.264341590612777e-05,
      "loss": 0.5691,
      "step": 142000
    },
    {
      "epoch": 2.9029498044328554,
      "grad_norm": 20.460561752319336,
      "learning_rate": 1.2582301173402868e-05,
      "loss": 0.5624,
      "step": 142500
    },
    {
      "epoch": 2.913135593220339,
      "grad_norm": 12.731225967407227,
      "learning_rate": 1.2521186440677966e-05,
      "loss": 0.5439,
      "step": 143000
    },
    {
      "epoch": 2.9233213820078228,
      "grad_norm": 3.781522274017334,
      "learning_rate": 1.2460071707953065e-05,
      "loss": 0.5348,
      "step": 143500
    },
    {
      "epoch": 2.9335071707953064,
      "grad_norm": 14.624204635620117,
      "learning_rate": 1.2398956975228162e-05,
      "loss": 0.5589,
      "step": 144000
    },
    {
      "epoch": 2.94369295958279,
      "grad_norm": 18.551095962524414,
      "learning_rate": 1.233784224250326e-05,
      "loss": 0.532,
      "step": 144500
    },
    {
      "epoch": 2.9538787483702738,
      "grad_norm": 17.094831466674805,
      "learning_rate": 1.2276727509778357e-05,
      "loss": 0.558,
      "step": 145000
    },
    {
      "epoch": 2.9640645371577574,
      "grad_norm": 12.306907653808594,
      "learning_rate": 1.2215612777053455e-05,
      "loss": 0.542,
      "step": 145500
    },
    {
      "epoch": 2.974250325945241,
      "grad_norm": 12.134025573730469,
      "learning_rate": 1.2154498044328553e-05,
      "loss": 0.5429,
      "step": 146000
    },
    {
      "epoch": 2.9844361147327247,
      "grad_norm": 22.295795440673828,
      "learning_rate": 1.209338331160365e-05,
      "loss": 0.5632,
      "step": 146500
    },
    {
      "epoch": 2.9946219035202084,
      "grad_norm": 17.749858856201172,
      "learning_rate": 1.2032268578878748e-05,
      "loss": 0.5402,
      "step": 147000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7337000966072083,
      "eval_loss": 0.7322831153869629,
      "eval_runtime": 8.3496,
      "eval_samples_per_second": 587.813,
      "eval_steps_per_second": 73.536,
      "step": 147264
    }
  ],
  "logging_steps": 500,
  "max_steps": 245440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.001
      },
      "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": 6.950698126852608e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}