{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 775,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06462035541195477,
      "grad_norm": 2.068004846572876,
      "learning_rate": 5.76923076923077e-06,
      "loss": 0.9769228935241699,
      "step": 10
    },
    {
      "epoch": 0.12924071082390953,
      "grad_norm": 0.9047558307647705,
      "learning_rate": 1.217948717948718e-05,
      "loss": 0.7594649791717529,
      "step": 20
    },
    {
      "epoch": 0.1938610662358643,
      "grad_norm": 0.4503297805786133,
      "learning_rate": 1.858974358974359e-05,
      "loss": 0.5557493209838867,
      "step": 30
    },
    {
      "epoch": 0.25848142164781907,
      "grad_norm": 0.37129801511764526,
      "learning_rate": 2.5e-05,
      "loss": 0.4763153076171875,
      "step": 40
    },
    {
      "epoch": 0.32310177705977383,
      "grad_norm": 0.2953853905200958,
      "learning_rate": 3.141025641025641e-05,
      "loss": 0.4166147708892822,
      "step": 50
    },
    {
      "epoch": 0.3877221324717286,
      "grad_norm": 0.29160749912261963,
      "learning_rate": 3.782051282051282e-05,
      "loss": 0.3905663013458252,
      "step": 60
    },
    {
      "epoch": 0.45234248788368336,
      "grad_norm": 0.27939939498901367,
      "learning_rate": 4.423076923076923e-05,
      "loss": 0.3649837732315063,
      "step": 70
    },
    {
      "epoch": 0.5169628432956381,
      "grad_norm": 0.3130805492401123,
      "learning_rate": 4.999974605277764e-05,
      "loss": 0.3609442949295044,
      "step": 80
    },
    {
      "epoch": 0.5815831987075929,
      "grad_norm": 0.3720378577709198,
      "learning_rate": 4.996927862815294e-05,
      "loss": 0.3543627977371216,
      "step": 90
    },
    {
      "epoch": 0.6462035541195477,
      "grad_norm": 0.2810651361942291,
      "learning_rate": 4.988809267347141e-05,
      "loss": 0.345255446434021,
      "step": 100
    },
    {
      "epoch": 0.7108239095315024,
      "grad_norm": 0.3126155734062195,
      "learning_rate": 4.975635309667227e-05,
      "loss": 0.33807756900787356,
      "step": 110
    },
    {
      "epoch": 0.7754442649434572,
      "grad_norm": 0.3109033703804016,
      "learning_rate": 4.957432749209755e-05,
      "loss": 0.3394132614135742,
      "step": 120
    },
    {
      "epoch": 0.840064620355412,
      "grad_norm": 0.3006167411804199,
      "learning_rate": 4.934238559694448e-05,
      "loss": 0.3244359016418457,
      "step": 130
    },
    {
      "epoch": 0.9046849757673667,
      "grad_norm": 0.3104219138622284,
      "learning_rate": 4.90609985402415e-05,
      "loss": 0.3221057176589966,
      "step": 140
    },
    {
      "epoch": 0.9693053311793215,
      "grad_norm": 0.330458402633667,
      "learning_rate": 4.873073788587338e-05,
      "loss": 0.3329624176025391,
      "step": 150
    },
    {
      "epoch": 1.0323101777059773,
      "grad_norm": 0.2666167616844177,
      "learning_rate": 4.835227447159939e-05,
      "loss": 0.3191133737564087,
      "step": 160
    },
    {
      "epoch": 1.0969305331179322,
      "grad_norm": 0.3217558264732361,
      "learning_rate": 4.7926377046422575e-05,
      "loss": 0.3055490255355835,
      "step": 170
    },
    {
      "epoch": 1.1615508885298869,
      "grad_norm": 0.31330329179763794,
      "learning_rate": 4.7453910709078145e-05,
      "loss": 0.31946792602539065,
      "step": 180
    },
    {
      "epoch": 1.2261712439418417,
      "grad_norm": 0.3254242539405823,
      "learning_rate": 4.69358351508126e-05,
      "loss": 0.3053389549255371,
      "step": 190
    },
    {
      "epoch": 1.2907915993537964,
      "grad_norm": 0.3096122145652771,
      "learning_rate": 4.637320270602314e-05,
      "loss": 0.30810558795928955,
      "step": 200
    },
    {
      "epoch": 1.3554119547657513,
      "grad_norm": 0.30546557903289795,
      "learning_rate": 4.576715621471678e-05,
      "loss": 0.31534519195556643,
      "step": 210
    },
    {
      "epoch": 1.420032310177706,
      "grad_norm": 0.32512548565864563,
      "learning_rate": 4.511892670113104e-05,
      "loss": 0.3046459913253784,
      "step": 220
    },
    {
      "epoch": 1.4846526655896608,
      "grad_norm": 0.32627010345458984,
      "learning_rate": 4.442983087323167e-05,
      "loss": 0.2983341455459595,
      "step": 230
    },
    {
      "epoch": 1.5492730210016155,
      "grad_norm": 0.3522598147392273,
      "learning_rate": 4.370126844816628e-05,
      "loss": 0.3133318185806274,
      "step": 240
    },
    {
      "epoch": 1.6138933764135701,
      "grad_norm": 0.30818620324134827,
      "learning_rate": 4.293471930910668e-05,
      "loss": 0.30533685684204104,
      "step": 250
    },
    {
      "epoch": 1.678513731825525,
      "grad_norm": 0.2774396538734436,
      "learning_rate": 4.213174049925498e-05,
      "loss": 0.2984048843383789,
      "step": 260
    },
    {
      "epoch": 1.7431340872374799,
      "grad_norm": 0.2943349778652191,
      "learning_rate": 4.129396305911934e-05,
      "loss": 0.2998037338256836,
      "step": 270
    },
    {
      "epoch": 1.8077544426494345,
      "grad_norm": 0.30737221240997314,
      "learning_rate": 4.042308871348371e-05,
      "loss": 0.3016969680786133,
      "step": 280
    },
    {
      "epoch": 1.8723747980613892,
      "grad_norm": 0.3105849325656891,
      "learning_rate": 3.9520886414800936e-05,
      "loss": 0.3014758348464966,
      "step": 290
    },
    {
      "epoch": 1.936995153473344,
      "grad_norm": 0.29112061858177185,
      "learning_rate": 3.858918875003053e-05,
      "loss": 0.297167444229126,
      "step": 300
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3953491747379303,
      "learning_rate": 3.7629888218219774e-05,
      "loss": 0.29504544734954835,
      "step": 310
    },
    {
      "epoch": 2.0646203554119547,
      "grad_norm": 0.29763659834861755,
      "learning_rate": 3.664493338638898e-05,
      "loss": 0.28955745697021484,
      "step": 320
    },
    {
      "epoch": 2.1292407108239093,
      "grad_norm": 0.3380239009857178,
      "learning_rate": 3.5636324931529554e-05,
      "loss": 0.28567984104156496,
      "step": 330
    },
    {
      "epoch": 2.1938610662358644,
      "grad_norm": 0.31489691138267517,
      "learning_rate": 3.46061115767542e-05,
      "loss": 0.28932580947875974,
      "step": 340
    },
    {
      "epoch": 2.258481421647819,
      "grad_norm": 0.3106003403663635,
      "learning_rate": 3.3556385929854164e-05,
      "loss": 0.2879155158996582,
      "step": 350
    },
    {
      "epoch": 2.3231017770597737,
      "grad_norm": 0.2789410948753357,
      "learning_rate": 3.248928023271624e-05,
      "loss": 0.2938661813735962,
      "step": 360
    },
    {
      "epoch": 2.387722132471729,
      "grad_norm": 0.2844054102897644,
      "learning_rate": 3.140696203023348e-05,
      "loss": 0.2907016515731812,
      "step": 370
    },
    {
      "epoch": 2.4523424878836835,
      "grad_norm": 0.3297322392463684,
      "learning_rate": 3.0311629767507232e-05,
      "loss": 0.28118228912353516,
      "step": 380
    },
    {
      "epoch": 2.516962843295638,
      "grad_norm": 0.30098527669906616,
      "learning_rate": 2.920550832428337e-05,
      "loss": 0.2846884489059448,
      "step": 390
    },
    {
      "epoch": 2.581583198707593,
      "grad_norm": 0.31197914481163025,
      "learning_rate": 2.809084449569365e-05,
      "loss": 0.28516063690185545,
      "step": 400
    },
    {
      "epoch": 2.6462035541195474,
      "grad_norm": 0.30557820200920105,
      "learning_rate": 2.6969902428481557e-05,
      "loss": 0.2813657283782959,
      "step": 410
    },
    {
      "epoch": 2.7108239095315025,
      "grad_norm": 0.3514881730079651,
      "learning_rate": 2.584495902198307e-05,
      "loss": 0.2873976469039917,
      "step": 420
    },
    {
      "epoch": 2.775444264943457,
      "grad_norm": 0.3010100722312927,
      "learning_rate": 2.4718299303203742e-05,
      "loss": 0.2792309045791626,
      "step": 430
    },
    {
      "epoch": 2.840064620355412,
      "grad_norm": 0.33192241191864014,
      "learning_rate": 2.3592211785386642e-05,
      "loss": 0.2784327745437622,
      "step": 440
    },
    {
      "epoch": 2.904684975767367,
      "grad_norm": 0.3093332350254059,
      "learning_rate": 2.246898381949894e-05,
      "loss": 0.28588788509368895,
      "step": 450
    },
    {
      "epoch": 2.9693053311793216,
      "grad_norm": 0.33140891790390015,
      "learning_rate": 2.1350896948079226e-05,
      "loss": 0.2833216428756714,
      "step": 460
    },
    {
      "epoch": 3.0323101777059773,
      "grad_norm": 0.31609705090522766,
      "learning_rate": 2.0240222270883288e-05,
      "loss": 0.28030478954315186,
      "step": 470
    },
    {
      "epoch": 3.096930533117932,
      "grad_norm": 0.2810538113117218,
      "learning_rate": 1.913921583174154e-05,
      "loss": 0.27578275203704833,
      "step": 480
    },
    {
      "epoch": 3.161550888529887,
      "grad_norm": 0.3165380358695984,
      "learning_rate": 1.805011403599862e-05,
      "loss": 0.27012245655059813,
      "step": 490
    },
    {
      "epoch": 3.2261712439418417,
      "grad_norm": 0.3220997452735901,
      "learning_rate": 1.697512910784352e-05,
      "loss": 0.2735466957092285,
      "step": 500
    },
    {
      "epoch": 3.2907915993537964,
      "grad_norm": 0.3499109447002411,
      "learning_rate": 1.5916444596757218e-05,
      "loss": 0.27160613536834716,
      "step": 510
    },
    {
      "epoch": 3.355411954765751,
      "grad_norm": 0.3451310694217682,
      "learning_rate": 1.4876210942205604e-05,
      "loss": 0.27280707359313966,
      "step": 520
    },
    {
      "epoch": 3.420032310177706,
      "grad_norm": 0.33574041724205017,
      "learning_rate": 1.3856541105586545e-05,
      "loss": 0.27128801345825193,
      "step": 530
    },
    {
      "epoch": 3.484652665589661,
      "grad_norm": 0.3565731346607208,
      "learning_rate": 1.2859506278303979e-05,
      "loss": 0.2736222743988037,
      "step": 540
    },
    {
      "epoch": 3.5492730210016155,
      "grad_norm": 0.32914265990257263,
      "learning_rate": 1.1887131674686596e-05,
      "loss": 0.27256240844726565,
      "step": 550
    },
    {
      "epoch": 3.61389337641357,
      "grad_norm": 0.31808099150657654,
      "learning_rate": 1.0941392418297026e-05,
      "loss": 0.27215673923492434,
      "step": 560
    },
    {
      "epoch": 3.678513731825525,
      "grad_norm": 0.3377521336078644,
      "learning_rate": 1.0024209529987149e-05,
      "loss": 0.2722803592681885,
      "step": 570
    },
    {
      "epoch": 3.74313408723748,
      "grad_norm": 0.33442333340644836,
      "learning_rate": 9.137446025848914e-06,
      "loss": 0.27033517360687254,
      "step": 580
    },
    {
      "epoch": 3.8077544426494345,
      "grad_norm": 0.32362648844718933,
      "learning_rate": 8.282903132986526e-06,
      "loss": 0.268932580947876,
      "step": 590
    },
    {
      "epoch": 3.872374798061389,
      "grad_norm": 0.33127161860466003,
      "learning_rate": 7.462316630796834e-06,
      "loss": 0.274311900138855,
      "step": 600
    },
    {
      "epoch": 3.936995153473344,
      "grad_norm": 0.35842862725257874,
      "learning_rate": 6.677353325189409e-06,
      "loss": 0.2703368902206421,
      "step": 610
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4013536870479584,
      "learning_rate": 5.929607662908282e-06,
      "loss": 0.27242910861968994,
      "step": 620
    },
    {
      "epoch": 4.064620355411955,
      "grad_norm": 0.3234418034553528,
      "learning_rate": 5.220598492832171e-06,
      "loss": 0.2654104232788086,
      "step": 630
    },
    {
      "epoch": 4.129240710823909,
      "grad_norm": 0.3155248761177063,
      "learning_rate": 4.551765980832059e-06,
      "loss": 0.26429879665374756,
      "step": 640
    },
    {
      "epoch": 4.193861066235864,
      "grad_norm": 0.3309212327003479,
      "learning_rate": 3.924468684452523e-06,
      "loss": 0.2637316226959229,
      "step": 650
    },
    {
      "epoch": 4.258481421647819,
      "grad_norm": 0.3667273223400116,
      "learning_rate": 3.339980793358982e-06,
      "loss": 0.2668980360031128,
      "step": 660
    },
    {
      "epoch": 4.323101777059774,
      "grad_norm": 0.3333350718021393,
      "learning_rate": 2.7994895411561333e-06,
      "loss": 0.2676853179931641,
      "step": 670
    },
    {
      "epoch": 4.387722132471729,
      "grad_norm": 0.32394319772720337,
      "learning_rate": 2.3040927938347528e-06,
      "loss": 0.26267526149749754,
      "step": 680
    },
    {
      "epoch": 4.4523424878836835,
      "grad_norm": 0.3705294132232666,
      "learning_rate": 1.854796819745383e-06,
      "loss": 0.26965837478637694,
      "step": 690
    },
    {
      "epoch": 4.516962843295638,
      "grad_norm": 0.3425300121307373,
      "learning_rate": 1.4525142456284779e-06,
      "loss": 0.2649491310119629,
      "step": 700
    },
    {
      "epoch": 4.581583198707593,
      "grad_norm": 0.32138586044311523,
      "learning_rate": 1.098062202852984e-06,
      "loss": 0.264300274848938,
      "step": 710
    },
    {
      "epoch": 4.646203554119547,
      "grad_norm": 0.34855136275291443,
      "learning_rate": 7.921606676285776e-07,
      "loss": 0.2624504804611206,
      "step": 720
    },
    {
      "epoch": 4.710823909531502,
      "grad_norm": 0.33398354053497314,
      "learning_rate": 5.354309985631605e-07,
      "loss": 0.2657909870147705,
      "step": 730
    },
    {
      "epoch": 4.775444264943458,
      "grad_norm": 0.32164713740348816,
      "learning_rate": 3.283946745360755e-07,
      "loss": 0.2667935609817505,
      "step": 740
    },
    {
      "epoch": 4.840064620355412,
      "grad_norm": 0.3328765332698822,
      "learning_rate": 1.714722354507592e-07,
      "loss": 0.26322848796844484,
      "step": 750
    },
    {
      "epoch": 4.904684975767367,
      "grad_norm": 0.3057484030723572,
      "learning_rate": 6.498242801838694e-08,
      "loss": 0.26327192783355713,
      "step": 760
    },
    {
      "epoch": 4.969305331179322,
      "grad_norm": 0.3382052183151245,
      "learning_rate": 9.141558307687592e-09,
      "loss": 0.2605318546295166,
      "step": 770
    },
    {
      "epoch": 5.0,
      "step": 775,
      "total_flos": 7.509104074275226e+17,
      "train_loss": 0.31352297429115544,
      "train_runtime": 25009.3076,
      "train_samples_per_second": 0.99,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 10,
  "max_steps": 775,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.509104074275226e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}