{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 250,
  "global_step": 558,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.3969754904508591,
      "epoch": 0.03587443946188341,
      "grad_norm": 0.62109375,
      "learning_rate": 5.294117647058824e-05,
      "loss": 2.051294708251953,
      "mean_token_accuracy": 0.5748387612402439,
      "num_tokens": 92395.0,
      "step": 10
    },
    {
      "entropy": 1.1848952442407608,
      "epoch": 0.07174887892376682,
      "grad_norm": 0.34375,
      "learning_rate": 9.999662790086726e-05,
      "loss": 1.152494525909424,
      "mean_token_accuracy": 0.7453485190868377,
      "num_tokens": 204363.0,
      "step": 20
    },
    {
      "entropy": 0.6448489271104336,
      "epoch": 0.10762331838565023,
      "grad_norm": 0.275390625,
      "learning_rate": 9.987865218277088e-05,
      "loss": 0.67830491065979,
      "mean_token_accuracy": 0.8489482343196869,
      "num_tokens": 307781.0,
      "step": 30
    },
    {
      "entropy": 0.5587928161025048,
      "epoch": 0.14349775784753363,
      "grad_norm": 0.33203125,
      "learning_rate": 9.959252607444427e-05,
      "loss": 0.6177556991577149,
      "mean_token_accuracy": 0.8661479398608207,
      "num_tokens": 419690.0,
      "step": 40
    },
    {
      "entropy": 0.5453228108584881,
      "epoch": 0.17937219730941703,
      "grad_norm": 0.2578125,
      "learning_rate": 9.913921416122937e-05,
      "loss": 0.6175318241119385,
      "mean_token_accuracy": 0.863138884305954,
      "num_tokens": 518165.0,
      "step": 50
    },
    {
      "entropy": 0.5307354614138603,
      "epoch": 0.21524663677130046,
      "grad_norm": 0.2021484375,
      "learning_rate": 9.852024464350382e-05,
      "loss": 0.5555857181549072,
      "mean_token_accuracy": 0.8736034348607064,
      "num_tokens": 632573.0,
      "step": 60
    },
    {
      "entropy": 0.5804678507149219,
      "epoch": 0.25112107623318386,
      "grad_norm": 0.26953125,
      "learning_rate": 9.773770418482792e-05,
      "loss": 0.6112886905670166,
      "mean_token_accuracy": 0.8588371470570564,
      "num_tokens": 746882.0,
      "step": 70
    },
    {
      "entropy": 0.4803255133330822,
      "epoch": 0.28699551569506726,
      "grad_norm": 0.1953125,
      "learning_rate": 9.679423087740625e-05,
      "loss": 0.5199193000793457,
      "mean_token_accuracy": 0.875262650847435,
      "num_tokens": 860710.0,
      "step": 80
    },
    {
      "entropy": 0.4950253814458847,
      "epoch": 0.32286995515695066,
      "grad_norm": 0.2734375,
      "learning_rate": 9.569300534857875e-05,
      "loss": 0.5125812530517578,
      "mean_token_accuracy": 0.876356054842472,
      "num_tokens": 973776.0,
      "step": 90
    },
    {
      "entropy": 0.5185210466384887,
      "epoch": 0.35874439461883406,
      "grad_norm": 0.25390625,
      "learning_rate": 9.443774003832332e-05,
      "loss": 0.5369254112243652,
      "mean_token_accuracy": 0.8700352162122726,
      "num_tokens": 1081833.0,
      "step": 100
    },
    {
      "entropy": 0.47246704772114756,
      "epoch": 0.39461883408071746,
      "grad_norm": 0.337890625,
      "learning_rate": 9.303266668391679e-05,
      "loss": 0.4889512062072754,
      "mean_token_accuracy": 0.8794563993811607,
      "num_tokens": 1195496.0,
      "step": 110
    },
    {
      "entropy": 0.4534280378371477,
      "epoch": 0.4304932735426009,
      "grad_norm": 0.33984375,
      "learning_rate": 9.148252205394665e-05,
      "loss": 0.4708422660827637,
      "mean_token_accuracy": 0.8855773583054543,
      "num_tokens": 1299697.0,
      "step": 120
    },
    {
      "entropy": 0.42982515431940554,
      "epoch": 0.4663677130044843,
      "grad_norm": 0.275390625,
      "learning_rate": 8.979253197976633e-05,
      "loss": 0.44471211433410646,
      "mean_token_accuracy": 0.8913678646087646,
      "num_tokens": 1401611.0,
      "step": 130
    },
    {
      "entropy": 0.42483637742698194,
      "epoch": 0.5022421524663677,
      "grad_norm": 0.2060546875,
      "learning_rate": 8.796839373822721e-05,
      "loss": 0.43284382820129397,
      "mean_token_accuracy": 0.8964968830347061,
      "num_tokens": 1511714.0,
      "step": 140
    },
    {
      "entropy": 0.38399339467287064,
      "epoch": 0.5381165919282511,
      "grad_norm": 0.2236328125,
      "learning_rate": 8.60162568450785e-05,
      "loss": 0.416139554977417,
      "mean_token_accuracy": 0.8991150856018066,
      "num_tokens": 1613398.0,
      "step": 150
    },
    {
      "entropy": 0.41539575792849065,
      "epoch": 0.5739910313901345,
      "grad_norm": 0.240234375,
      "learning_rate": 8.394270232378419e-05,
      "loss": 0.4283764839172363,
      "mean_token_accuracy": 0.8929989218711853,
      "num_tokens": 1710736.0,
      "step": 160
    },
    {
      "entropy": 0.47740858122706414,
      "epoch": 0.6098654708520179,
      "grad_norm": 0.263671875,
      "learning_rate": 8.175472051964552e-05,
      "loss": 0.4906760692596436,
      "mean_token_accuracy": 0.8809969320893287,
      "num_tokens": 1815452.0,
      "step": 170
    },
    {
      "entropy": 0.4214516717940569,
      "epoch": 0.6457399103139013,
      "grad_norm": 0.279296875,
      "learning_rate": 7.945968753402216e-05,
      "loss": 0.44007339477539065,
      "mean_token_accuracy": 0.8928113669157028,
      "num_tokens": 1917703.0,
      "step": 180
    },
    {
      "entropy": 0.4015789397060871,
      "epoch": 0.6816143497757847,
      "grad_norm": 0.28125,
      "learning_rate": 7.70653403580964e-05,
      "loss": 0.4222086429595947,
      "mean_token_accuracy": 0.8981553256511688,
      "num_tokens": 2025309.0,
      "step": 190
    },
    {
      "entropy": 0.4304059199988842,
      "epoch": 0.7174887892376681,
      "grad_norm": 0.283203125,
      "learning_rate": 7.457975079000954e-05,
      "loss": 0.4446873188018799,
      "mean_token_accuracy": 0.8913767769932747,
      "num_tokens": 2126377.0,
      "step": 200
    },
    {
      "entropy": 0.40952816382050516,
      "epoch": 0.7533632286995515,
      "grad_norm": 0.259765625,
      "learning_rate": 7.201129822330041e-05,
      "loss": 0.4252801418304443,
      "mean_token_accuracy": 0.8945473477244377,
      "num_tokens": 2239100.0,
      "step": 210
    },
    {
      "entropy": 0.414412359893322,
      "epoch": 0.7892376681614349,
      "grad_norm": 0.28515625,
      "learning_rate": 6.936864139838158e-05,
      "loss": 0.4217662811279297,
      "mean_token_accuracy": 0.8939862757921219,
      "num_tokens": 2335750.0,
      "step": 220
    },
    {
      "entropy": 0.45669897869229314,
      "epoch": 0.8251121076233184,
      "grad_norm": 0.294921875,
      "learning_rate": 6.666068921228433e-05,
      "loss": 0.46158409118652344,
      "mean_token_accuracy": 0.8881383046507836,
      "num_tokens": 2451761.0,
      "step": 230
    },
    {
      "entropy": 0.4027619417756796,
      "epoch": 0.8609865470852018,
      "grad_norm": 0.314453125,
      "learning_rate": 6.389657068507819e-05,
      "loss": 0.41291356086730957,
      "mean_token_accuracy": 0.8973603308200836,
      "num_tokens": 2553271.0,
      "step": 240
    },
    {
      "entropy": 0.38662399984896184,
      "epoch": 0.8968609865470852,
      "grad_norm": 0.279296875,
      "learning_rate": 6.108560418421397e-05,
      "loss": 0.38063457012176516,
      "mean_token_accuracy": 0.9039406031370163,
      "num_tokens": 2656981.0,
      "step": 250
    },
    {
      "epoch": 0.8968609865470852,
      "eval_entropy": 0.4262866288738907,
      "eval_loss": 0.5014730095863342,
      "eval_mean_token_accuracy": 0.8819858220519212,
      "eval_num_tokens": 2656981.0,
      "eval_runtime": 278.3421,
      "eval_samples_per_second": 5.421,
      "eval_steps_per_second": 1.358,
      "step": 250
    },
    {
      "entropy": 0.38649687841534613,
      "epoch": 0.9327354260089686,
      "grad_norm": 0.3046875,
      "learning_rate": 5.8237266010541046e-05,
      "loss": 0.4081888675689697,
      "mean_token_accuracy": 0.8996688678860665,
      "num_tokens": 2759594.0,
      "step": 260
    },
    {
      "entropy": 0.3772343125194311,
      "epoch": 0.968609865470852,
      "grad_norm": 0.244140625,
      "learning_rate": 5.536115845190105e-05,
      "loss": 0.3852355480194092,
      "mean_token_accuracy": 0.902637392282486,
      "num_tokens": 2859621.0,
      "step": 270
    },
    {
      "entropy": 0.42768477018062884,
      "epoch": 1.0035874439461883,
      "grad_norm": 0.2373046875,
      "learning_rate": 5.2466977411995567e-05,
      "loss": 0.4248982906341553,
      "mean_token_accuracy": 0.8919965258011451,
      "num_tokens": 2972688.0,
      "step": 280
    },
    {
      "entropy": 0.3386215832084417,
      "epoch": 1.0394618834080718,
      "grad_norm": 0.30859375,
      "learning_rate": 4.9564479723657075e-05,
      "loss": 0.3396793603897095,
      "mean_token_accuracy": 0.9132053166627884,
      "num_tokens": 3067583.0,
      "step": 290
    },
    {
      "entropy": 0.3685107693076134,
      "epoch": 1.075336322869955,
      "grad_norm": 0.322265625,
      "learning_rate": 4.6663450256716226e-05,
      "loss": 0.3796958923339844,
      "mean_token_accuracy": 0.9038235887885093,
      "num_tokens": 3172296.0,
      "step": 300
    },
    {
      "entropy": 0.35373203866183756,
      "epoch": 1.1112107623318386,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.3773668931351055e-05,
      "loss": 0.35225601196289064,
      "mean_token_accuracy": 0.9102007254958153,
      "num_tokens": 3276803.0,
      "step": 310
    },
    {
      "entropy": 0.3816000144928694,
      "epoch": 1.1470852017937219,
      "grad_norm": 0.328125,
      "learning_rate": 4.0904877748122436e-05,
      "loss": 0.38906025886535645,
      "mean_token_accuracy": 0.9023444890975952,
      "num_tokens": 3391132.0,
      "step": 320
    },
    {
      "entropy": 0.3066167201846838,
      "epoch": 1.1829596412556054,
      "grad_norm": 0.294921875,
      "learning_rate": 3.806674794584374e-05,
      "loss": 0.30766735076904295,
      "mean_token_accuracy": 0.921451187133789,
      "num_tokens": 3497920.0,
      "step": 330
    },
    {
      "entropy": 0.3733608294278383,
      "epoch": 1.2188340807174889,
      "grad_norm": 0.365234375,
      "learning_rate": 3.5268847398001766e-05,
      "loss": 0.39584877490997317,
      "mean_token_accuracy": 0.9041571974754333,
      "num_tokens": 3599625.0,
      "step": 340
    },
    {
      "entropy": 0.3333755239844322,
      "epoch": 1.2547085201793722,
      "grad_norm": 0.30078125,
      "learning_rate": 3.252060835764181e-05,
      "loss": 0.3276608943939209,
      "mean_token_accuracy": 0.9138269037008285,
      "num_tokens": 3707539.0,
      "step": 350
    },
    {
      "entropy": 0.335588826239109,
      "epoch": 1.2905829596412555,
      "grad_norm": 0.248046875,
      "learning_rate": 2.9831295659454838e-05,
      "loss": 0.3360959529876709,
      "mean_token_accuracy": 0.9134533792734146,
      "num_tokens": 3813716.0,
      "step": 360
    },
    {
      "entropy": 0.32520635724067687,
      "epoch": 1.326457399103139,
      "grad_norm": 0.337890625,
      "learning_rate": 2.7209975486263807e-05,
      "loss": 0.3224537134170532,
      "mean_token_accuracy": 0.9162669718265534,
      "num_tokens": 3924180.0,
      "step": 370
    },
    {
      "entropy": 0.3270633514970541,
      "epoch": 1.3623318385650225,
      "grad_norm": 0.283203125,
      "learning_rate": 2.466548480520296e-05,
      "loss": 0.323980975151062,
      "mean_token_accuracy": 0.9139964297413826,
      "num_tokens": 4026780.0,
      "step": 380
    },
    {
      "entropy": 0.3498925618827343,
      "epoch": 1.3982062780269058,
      "grad_norm": 0.30078125,
      "learning_rate": 2.220640157662679e-05,
      "loss": 0.37245047092437744,
      "mean_token_accuracy": 0.9063089922070503,
      "num_tokens": 4130194.0,
      "step": 390
    },
    {
      "entropy": 0.33601379320025443,
      "epoch": 1.4340807174887893,
      "grad_norm": 0.353515625,
      "learning_rate": 1.9841015836179734e-05,
      "loss": 0.32559800148010254,
      "mean_token_accuracy": 0.9136791557073594,
      "num_tokens": 4236882.0,
      "step": 400
    },
    {
      "entropy": 0.322609855979681,
      "epoch": 1.4699551569506726,
      "grad_norm": 0.36328125,
      "learning_rate": 1.7577301747514922e-05,
      "loss": 0.33389127254486084,
      "mean_token_accuracy": 0.9131381779909133,
      "num_tokens": 4338348.0,
      "step": 410
    },
    {
      "entropy": 0.3596556928008795,
      "epoch": 1.505829596412556,
      "grad_norm": 0.369140625,
      "learning_rate": 1.5422890719877e-05,
      "loss": 0.36695308685302735,
      "mean_token_accuracy": 0.9074400275945663,
      "num_tokens": 4439178.0,
      "step": 420
    },
    {
      "entropy": 0.3173448100686073,
      "epoch": 1.5417040358744396,
      "grad_norm": 0.34375,
      "learning_rate": 1.338504568117518e-05,
      "loss": 0.3288587093353271,
      "mean_token_accuracy": 0.9189538955688477,
      "num_tokens": 4548339.0,
      "step": 430
    },
    {
      "entropy": 0.3539650794118643,
      "epoch": 1.577578475336323,
      "grad_norm": 0.3046875,
      "learning_rate": 1.1470636593276779e-05,
      "loss": 0.35024569034576414,
      "mean_token_accuracy": 0.9065869942307472,
      "num_tokens": 4660945.0,
      "step": 440
    },
    {
      "entropy": 0.33447692096233367,
      "epoch": 1.6134529147982062,
      "grad_norm": 0.30078125,
      "learning_rate": 9.686117292063501e-06,
      "loss": 0.3201354742050171,
      "mean_token_accuracy": 0.9138964533805847,
      "num_tokens": 4760382.0,
      "step": 450
    },
    {
      "entropy": 0.3287306733429432,
      "epoch": 1.6493273542600897,
      "grad_norm": 0.296875,
      "learning_rate": 8.037503730327634e-06,
      "loss": 0.32874948978424073,
      "mean_token_accuracy": 0.9138829991221428,
      "num_tokens": 4874370.0,
      "step": 460
    },
    {
      "entropy": 0.3220908358693123,
      "epoch": 1.6852017937219732,
      "grad_norm": 0.279296875,
      "learning_rate": 6.530353696854791e-06,
      "loss": 0.3312650442123413,
      "mean_token_accuracy": 0.9188773602247238,
      "num_tokens": 4996174.0,
      "step": 470
    },
    {
      "entropy": 0.34412881508469584,
      "epoch": 1.7210762331838565,
      "grad_norm": 0.3046875,
      "learning_rate": 5.169748080064651e-06,
      "loss": 0.338838791847229,
      "mean_token_accuracy": 0.9112714245915413,
      "num_tokens": 5105558.0,
      "step": 480
    },
    {
      "entropy": 0.3259685214608908,
      "epoch": 1.7569506726457398,
      "grad_norm": 0.3359375,
      "learning_rate": 3.960273739373044e-06,
      "loss": 0.3209350824356079,
      "mean_token_accuracy": 0.9152405560016632,
      "num_tokens": 5220771.0,
      "step": 490
    },
    {
      "entropy": 0.3517936758697033,
      "epoch": 1.7928251121076233,
      "grad_norm": 0.294921875,
      "learning_rate": 2.9060080420194648e-06,
      "loss": 0.35311083793640136,
      "mean_token_accuracy": 0.9091673374176026,
      "num_tokens": 5326562.0,
      "step": 500
    },
    {
      "epoch": 1.7928251121076233,
      "eval_entropy": 0.3918723243648413,
      "eval_loss": 0.5272403955459595,
      "eval_mean_token_accuracy": 0.8795070842144981,
      "eval_num_tokens": 5326562.0,
      "eval_runtime": 277.68,
      "eval_samples_per_second": 5.434,
      "eval_steps_per_second": 1.361,
      "step": 500
    },
    {
      "entropy": 0.3449239183217287,
      "epoch": 1.8286995515695068,
      "grad_norm": 0.2890625,
      "learning_rate": 2.0105051174893363e-06,
      "loss": 0.3280923366546631,
      "mean_token_accuracy": 0.9127265498042106,
      "num_tokens": 5439381.0,
      "step": 510
    },
    {
      "entropy": 0.32186402194201946,
      "epoch": 1.86457399103139,
      "grad_norm": 0.404296875,
      "learning_rate": 1.2767838758698912e-06,
      "loss": 0.32121551036834717,
      "mean_token_accuracy": 0.9174392357468605,
      "num_tokens": 5533378.0,
      "step": 520
    },
    {
      "entropy": 0.3244446087628603,
      "epoch": 1.9004484304932734,
      "grad_norm": 0.29296875,
      "learning_rate": 7.073178305320727e-07,
      "loss": 0.310935378074646,
      "mean_token_accuracy": 0.917025238275528,
      "num_tokens": 5637745.0,
      "step": 530
    },
    {
      "entropy": 0.3581434443593025,
      "epoch": 1.936322869955157,
      "grad_norm": 0.318359375,
      "learning_rate": 3.0402675944814496e-07,
      "loss": 0.359324312210083,
      "mean_token_accuracy": 0.9096325069665909,
      "num_tokens": 5742787.0,
      "step": 540
    },
    {
      "entropy": 0.3250444334000349,
      "epoch": 1.9721973094170404,
      "grad_norm": 0.30078125,
      "learning_rate": 6.827023325626237e-08,
      "loss": 0.3184437990188599,
      "mean_token_accuracy": 0.9155865594744682,
      "num_tokens": 5844462.0,
      "step": 550
    },
    {
      "epoch": 2.0,
      "eval_entropy": 0.3912128670111535,
      "eval_loss": 0.5274220705032349,
      "eval_mean_token_accuracy": 0.8794930301330708,
      "eval_num_tokens": 5924816.0,
      "eval_runtime": 277.7162,
      "eval_samples_per_second": 5.434,
      "eval_steps_per_second": 1.361,
      "step": 558
    }
  ],
  "logging_steps": 10,
  "max_steps": 558,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.960190639222784e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}