{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.882352941176471,
  "eval_steps": 10,
  "global_step": 550,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "learning_rate": 9.820788530465951e-05,
      "loss": 2.039,
      "step": 10
    },
    {
      "epoch": 0.11,
      "eval_accuracy": 0.15425531566143036,
      "eval_loss": 1.9874085187911987,
      "eval_runtime": 18.0154,
      "eval_samples_per_second": 10.436,
      "eval_steps_per_second": 2.609,
      "step": 10
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.6415770609319e-05,
      "loss": 1.8674,
      "step": 20
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.18617020547389984,
      "eval_loss": 1.915345549583435,
      "eval_runtime": 19.3568,
      "eval_samples_per_second": 9.712,
      "eval_steps_per_second": 2.428,
      "step": 20
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.46236559139785e-05,
      "loss": 1.9047,
      "step": 30
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.21808511018753052,
      "eval_loss": 1.8614839315414429,
      "eval_runtime": 16.987,
      "eval_samples_per_second": 11.067,
      "eval_steps_per_second": 2.767,
      "step": 30
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.2831541218638e-05,
      "loss": 1.8554,
      "step": 40
    },
    {
      "epoch": 0.43,
      "eval_accuracy": 0.3297872245311737,
      "eval_loss": 1.7530335187911987,
      "eval_runtime": 17.9287,
      "eval_samples_per_second": 10.486,
      "eval_steps_per_second": 2.621,
      "step": 40
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.10394265232975e-05,
      "loss": 1.7546,
      "step": 50
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.34574466943740845,
      "eval_loss": 1.693456768989563,
      "eval_runtime": 17.7507,
      "eval_samples_per_second": 10.591,
      "eval_steps_per_second": 2.648,
      "step": 50
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.924731182795699e-05,
      "loss": 1.7583,
      "step": 60
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.31382977962493896,
      "eval_loss": 1.7744777202606201,
      "eval_runtime": 17.5758,
      "eval_samples_per_second": 10.697,
      "eval_steps_per_second": 2.674,
      "step": 60
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.74551971326165e-05,
      "loss": 1.6846,
      "step": 70
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.47340425848960876,
      "eval_loss": 1.383662462234497,
      "eval_runtime": 17.9239,
      "eval_samples_per_second": 10.489,
      "eval_steps_per_second": 2.622,
      "step": 70
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.566308243727598e-05,
      "loss": 1.335,
      "step": 80
    },
    {
      "epoch": 0.86,
      "eval_accuracy": 0.4095744788646698,
      "eval_loss": 1.3189808130264282,
      "eval_runtime": 17.9015,
      "eval_samples_per_second": 10.502,
      "eval_steps_per_second": 2.625,
      "step": 80
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.387096774193549e-05,
      "loss": 1.3937,
      "step": 90
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.5,
      "eval_loss": 1.2728934288024902,
      "eval_runtime": 16.8503,
      "eval_samples_per_second": 11.157,
      "eval_steps_per_second": 2.789,
      "step": 90
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.207885304659499e-05,
      "loss": 1.3213,
      "step": 100
    },
    {
      "epoch": 1.07,
      "eval_accuracy": 0.5319148898124695,
      "eval_loss": 1.348612666130066,
      "eval_runtime": 17.9939,
      "eval_samples_per_second": 10.448,
      "eval_steps_per_second": 2.612,
      "step": 100
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.028673835125448e-05,
      "loss": 1.2577,
      "step": 110
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.43617022037506104,
      "eval_loss": 1.4858437776565552,
      "eval_runtime": 16.7284,
      "eval_samples_per_second": 11.238,
      "eval_steps_per_second": 2.81,
      "step": 110
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.849462365591398e-05,
      "loss": 1.3039,
      "step": 120
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.5372340679168701,
      "eval_loss": 1.1272777318954468,
      "eval_runtime": 18.0338,
      "eval_samples_per_second": 10.425,
      "eval_steps_per_second": 2.606,
      "step": 120
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.670250896057349e-05,
      "loss": 1.2018,
      "step": 130
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.6010638475418091,
      "eval_loss": 1.07073974609375,
      "eval_runtime": 18.3403,
      "eval_samples_per_second": 10.251,
      "eval_steps_per_second": 2.563,
      "step": 130
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.491039426523297e-05,
      "loss": 1.0523,
      "step": 140
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.6223404407501221,
      "eval_loss": 0.9273392558097839,
      "eval_runtime": 17.4553,
      "eval_samples_per_second": 10.77,
      "eval_steps_per_second": 2.693,
      "step": 140
    },
    {
      "epoch": 1.6,
      "learning_rate": 7.311827956989248e-05,
      "loss": 0.8089,
      "step": 150
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.6436170339584351,
      "eval_loss": 0.958331823348999,
      "eval_runtime": 18.1909,
      "eval_samples_per_second": 10.335,
      "eval_steps_per_second": 2.584,
      "step": 150
    },
    {
      "epoch": 1.71,
      "learning_rate": 7.132616487455197e-05,
      "loss": 0.86,
      "step": 160
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.728723406791687,
      "eval_loss": 0.7838733792304993,
      "eval_runtime": 17.986,
      "eval_samples_per_second": 10.453,
      "eval_steps_per_second": 2.613,
      "step": 160
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.953405017921147e-05,
      "loss": 0.7991,
      "step": 170
    },
    {
      "epoch": 1.82,
      "eval_accuracy": 0.7393617033958435,
      "eval_loss": 0.7742288708686829,
      "eval_runtime": 17.5516,
      "eval_samples_per_second": 10.711,
      "eval_steps_per_second": 2.678,
      "step": 170
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.774193548387096e-05,
      "loss": 0.6772,
      "step": 180
    },
    {
      "epoch": 1.93,
      "eval_accuracy": 0.728723406791687,
      "eval_loss": 0.7831820845603943,
      "eval_runtime": 17.1247,
      "eval_samples_per_second": 10.978,
      "eval_steps_per_second": 2.745,
      "step": 180
    },
    {
      "epoch": 2.03,
      "learning_rate": 6.594982078853047e-05,
      "loss": 0.5716,
      "step": 190
    },
    {
      "epoch": 2.03,
      "eval_accuracy": 0.6808510422706604,
      "eval_loss": 0.8607780933380127,
      "eval_runtime": 17.9663,
      "eval_samples_per_second": 10.464,
      "eval_steps_per_second": 2.616,
      "step": 190
    },
    {
      "epoch": 2.14,
      "learning_rate": 6.415770609318996e-05,
      "loss": 0.7228,
      "step": 200
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.7127659320831299,
      "eval_loss": 0.9357430338859558,
      "eval_runtime": 16.801,
      "eval_samples_per_second": 11.19,
      "eval_steps_per_second": 2.797,
      "step": 200
    },
    {
      "epoch": 2.25,
      "learning_rate": 6.236559139784946e-05,
      "loss": 0.5451,
      "step": 210
    },
    {
      "epoch": 2.25,
      "eval_accuracy": 0.6489361524581909,
      "eval_loss": 1.018636703491211,
      "eval_runtime": 17.1948,
      "eval_samples_per_second": 10.934,
      "eval_steps_per_second": 2.733,
      "step": 210
    },
    {
      "epoch": 2.35,
      "learning_rate": 6.057347670250897e-05,
      "loss": 0.6429,
      "step": 220
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.6968085169792175,
      "eval_loss": 0.917243242263794,
      "eval_runtime": 17.5045,
      "eval_samples_per_second": 10.74,
      "eval_steps_per_second": 2.685,
      "step": 220
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.878136200716846e-05,
      "loss": 0.8224,
      "step": 230
    },
    {
      "epoch": 2.46,
      "eval_accuracy": 0.6329787373542786,
      "eval_loss": 1.0460094213485718,
      "eval_runtime": 16.9483,
      "eval_samples_per_second": 11.093,
      "eval_steps_per_second": 2.773,
      "step": 230
    },
    {
      "epoch": 2.57,
      "learning_rate": 5.6989247311827965e-05,
      "loss": 0.6061,
      "step": 240
    },
    {
      "epoch": 2.57,
      "eval_accuracy": 0.7446808218955994,
      "eval_loss": 0.8167480230331421,
      "eval_runtime": 16.9079,
      "eval_samples_per_second": 11.119,
      "eval_steps_per_second": 2.78,
      "step": 240
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.519713261648746e-05,
      "loss": 0.3594,
      "step": 250
    },
    {
      "epoch": 2.67,
      "eval_accuracy": 0.7234042286872864,
      "eval_loss": 0.8162137866020203,
      "eval_runtime": 16.7777,
      "eval_samples_per_second": 11.205,
      "eval_steps_per_second": 2.801,
      "step": 250
    },
    {
      "epoch": 2.78,
      "learning_rate": 5.340501792114696e-05,
      "loss": 0.4122,
      "step": 260
    },
    {
      "epoch": 2.78,
      "eval_accuracy": 0.7819148898124695,
      "eval_loss": 0.7674133777618408,
      "eval_runtime": 16.8556,
      "eval_samples_per_second": 11.154,
      "eval_steps_per_second": 2.788,
      "step": 260
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.161290322580645e-05,
      "loss": 0.2874,
      "step": 270
    },
    {
      "epoch": 2.89,
      "eval_accuracy": 0.7978723645210266,
      "eval_loss": 0.6874329447746277,
      "eval_runtime": 18.1768,
      "eval_samples_per_second": 10.343,
      "eval_steps_per_second": 2.586,
      "step": 270
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.982078853046595e-05,
      "loss": 0.4881,
      "step": 280
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.7553191781044006,
      "eval_loss": 0.7479570508003235,
      "eval_runtime": 17.7772,
      "eval_samples_per_second": 10.575,
      "eval_steps_per_second": 2.644,
      "step": 280
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.802867383512545e-05,
      "loss": 0.5232,
      "step": 290
    },
    {
      "epoch": 3.1,
      "eval_accuracy": 0.7659574747085571,
      "eval_loss": 0.7912107706069946,
      "eval_runtime": 17.9423,
      "eval_samples_per_second": 10.478,
      "eval_steps_per_second": 2.62,
      "step": 290
    },
    {
      "epoch": 3.21,
      "learning_rate": 4.6236559139784944e-05,
      "loss": 0.1914,
      "step": 300
    },
    {
      "epoch": 3.21,
      "eval_accuracy": 0.7978723645210266,
      "eval_loss": 0.7023528218269348,
      "eval_runtime": 18.317,
      "eval_samples_per_second": 10.264,
      "eval_steps_per_second": 2.566,
      "step": 300
    },
    {
      "epoch": 3.32,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.2573,
      "step": 310
    },
    {
      "epoch": 3.32,
      "eval_accuracy": 0.7659574747085571,
      "eval_loss": 0.7678730487823486,
      "eval_runtime": 16.9343,
      "eval_samples_per_second": 11.102,
      "eval_steps_per_second": 2.775,
      "step": 310
    },
    {
      "epoch": 3.42,
      "learning_rate": 4.265232974910394e-05,
      "loss": 0.1959,
      "step": 320
    },
    {
      "epoch": 3.42,
      "eval_accuracy": 0.7819148898124695,
      "eval_loss": 0.7615554332733154,
      "eval_runtime": 17.2171,
      "eval_samples_per_second": 10.919,
      "eval_steps_per_second": 2.73,
      "step": 320
    },
    {
      "epoch": 3.53,
      "learning_rate": 4.0860215053763444e-05,
      "loss": 0.2317,
      "step": 330
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.8191489577293396,
      "eval_loss": 0.6963978409767151,
      "eval_runtime": 18.8086,
      "eval_samples_per_second": 9.995,
      "eval_steps_per_second": 2.499,
      "step": 330
    },
    {
      "epoch": 3.64,
      "learning_rate": 3.906810035842295e-05,
      "loss": 0.2674,
      "step": 340
    },
    {
      "epoch": 3.64,
      "eval_accuracy": 0.813829779624939,
      "eval_loss": 0.6807590126991272,
      "eval_runtime": 18.085,
      "eval_samples_per_second": 10.395,
      "eval_steps_per_second": 2.599,
      "step": 340
    },
    {
      "epoch": 3.74,
      "learning_rate": 3.727598566308244e-05,
      "loss": 0.1163,
      "step": 350
    },
    {
      "epoch": 3.74,
      "eval_accuracy": 0.8297872543334961,
      "eval_loss": 0.6285024881362915,
      "eval_runtime": 18.2735,
      "eval_samples_per_second": 10.288,
      "eval_steps_per_second": 2.572,
      "step": 350
    },
    {
      "epoch": 3.85,
      "learning_rate": 3.548387096774194e-05,
      "loss": 0.1601,
      "step": 360
    },
    {
      "epoch": 3.85,
      "eval_accuracy": 0.813829779624939,
      "eval_loss": 0.68441241979599,
      "eval_runtime": 18.2649,
      "eval_samples_per_second": 10.293,
      "eval_steps_per_second": 2.573,
      "step": 360
    },
    {
      "epoch": 3.96,
      "learning_rate": 3.369175627240144e-05,
      "loss": 0.2855,
      "step": 370
    },
    {
      "epoch": 3.96,
      "eval_accuracy": 0.792553186416626,
      "eval_loss": 0.7759273052215576,
      "eval_runtime": 16.9436,
      "eval_samples_per_second": 11.096,
      "eval_steps_per_second": 2.774,
      "step": 370
    },
    {
      "epoch": 4.06,
      "learning_rate": 3.1899641577060935e-05,
      "loss": 0.2195,
      "step": 380
    },
    {
      "epoch": 4.06,
      "eval_accuracy": 0.835106372833252,
      "eval_loss": 0.6165522336959839,
      "eval_runtime": 16.8992,
      "eval_samples_per_second": 11.125,
      "eval_steps_per_second": 2.781,
      "step": 380
    },
    {
      "epoch": 4.17,
      "learning_rate": 3.010752688172043e-05,
      "loss": 0.2313,
      "step": 390
    },
    {
      "epoch": 4.17,
      "eval_accuracy": 0.8457446694374084,
      "eval_loss": 0.5867934823036194,
      "eval_runtime": 16.9793,
      "eval_samples_per_second": 11.072,
      "eval_steps_per_second": 2.768,
      "step": 390
    },
    {
      "epoch": 4.28,
      "learning_rate": 2.831541218637993e-05,
      "loss": 0.0568,
      "step": 400
    },
    {
      "epoch": 4.28,
      "eval_accuracy": 0.8510638475418091,
      "eval_loss": 0.5872617959976196,
      "eval_runtime": 17.0099,
      "eval_samples_per_second": 11.052,
      "eval_steps_per_second": 2.763,
      "step": 400
    },
    {
      "epoch": 4.39,
      "learning_rate": 2.652329749103943e-05,
      "loss": 0.1141,
      "step": 410
    },
    {
      "epoch": 4.39,
      "eval_accuracy": 0.8563829660415649,
      "eval_loss": 0.5599851012229919,
      "eval_runtime": 17.0,
      "eval_samples_per_second": 11.059,
      "eval_steps_per_second": 2.765,
      "step": 410
    },
    {
      "epoch": 4.49,
      "learning_rate": 2.4731182795698928e-05,
      "loss": 0.1607,
      "step": 420
    },
    {
      "epoch": 4.49,
      "eval_accuracy": 0.8297872543334961,
      "eval_loss": 0.6353866457939148,
      "eval_runtime": 17.4021,
      "eval_samples_per_second": 10.803,
      "eval_steps_per_second": 2.701,
      "step": 420
    },
    {
      "epoch": 4.6,
      "learning_rate": 2.2939068100358423e-05,
      "loss": 0.0994,
      "step": 430
    },
    {
      "epoch": 4.6,
      "eval_accuracy": 0.8670212626457214,
      "eval_loss": 0.5707226991653442,
      "eval_runtime": 19.3599,
      "eval_samples_per_second": 9.711,
      "eval_steps_per_second": 2.428,
      "step": 430
    },
    {
      "epoch": 4.71,
      "learning_rate": 2.1146953405017922e-05,
      "loss": 0.125,
      "step": 440
    },
    {
      "epoch": 4.71,
      "eval_accuracy": 0.8510638475418091,
      "eval_loss": 0.6016649007797241,
      "eval_runtime": 18.1587,
      "eval_samples_per_second": 10.353,
      "eval_steps_per_second": 2.588,
      "step": 440
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.935483870967742e-05,
      "loss": 0.1598,
      "step": 450
    },
    {
      "epoch": 4.81,
      "eval_accuracy": 0.8670212626457214,
      "eval_loss": 0.5569073557853699,
      "eval_runtime": 17.221,
      "eval_samples_per_second": 10.917,
      "eval_steps_per_second": 2.729,
      "step": 450
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.7562724014336916e-05,
      "loss": 0.1274,
      "step": 460
    },
    {
      "epoch": 4.92,
      "eval_accuracy": 0.8670212626457214,
      "eval_loss": 0.5385098457336426,
      "eval_runtime": 17.4593,
      "eval_samples_per_second": 10.768,
      "eval_steps_per_second": 2.692,
      "step": 460
    },
    {
      "epoch": 5.03,
      "learning_rate": 1.5770609318996415e-05,
      "loss": 0.131,
      "step": 470
    },
    {
      "epoch": 5.03,
      "eval_accuracy": 0.8563829660415649,
      "eval_loss": 0.5594031810760498,
      "eval_runtime": 17.5202,
      "eval_samples_per_second": 10.73,
      "eval_steps_per_second": 2.683,
      "step": 470
    },
    {
      "epoch": 5.13,
      "learning_rate": 1.3978494623655914e-05,
      "loss": 0.0754,
      "step": 480
    },
    {
      "epoch": 5.13,
      "eval_accuracy": 0.8563829660415649,
      "eval_loss": 0.5571564435958862,
      "eval_runtime": 18.3965,
      "eval_samples_per_second": 10.219,
      "eval_steps_per_second": 2.555,
      "step": 480
    },
    {
      "epoch": 5.24,
      "learning_rate": 1.2186379928315413e-05,
      "loss": 0.0298,
      "step": 490
    },
    {
      "epoch": 5.24,
      "eval_accuracy": 0.8776595592498779,
      "eval_loss": 0.5584980845451355,
      "eval_runtime": 18.2497,
      "eval_samples_per_second": 10.302,
      "eval_steps_per_second": 2.575,
      "step": 490
    },
    {
      "epoch": 5.35,
      "learning_rate": 1.039426523297491e-05,
      "loss": 0.012,
      "step": 500
    },
    {
      "epoch": 5.35,
      "eval_accuracy": 0.8723404407501221,
      "eval_loss": 0.5599969625473022,
      "eval_runtime": 17.3404,
      "eval_samples_per_second": 10.842,
      "eval_steps_per_second": 2.71,
      "step": 500
    },
    {
      "epoch": 5.45,
      "learning_rate": 8.602150537634409e-06,
      "loss": 0.178,
      "step": 510
    },
    {
      "epoch": 5.45,
      "eval_accuracy": 0.8723404407501221,
      "eval_loss": 0.5632631778717041,
      "eval_runtime": 17.983,
      "eval_samples_per_second": 10.454,
      "eval_steps_per_second": 2.614,
      "step": 510
    },
    {
      "epoch": 5.56,
      "learning_rate": 6.810035842293908e-06,
      "loss": 0.0365,
      "step": 520
    },
    {
      "epoch": 5.56,
      "eval_accuracy": 0.8670212626457214,
      "eval_loss": 0.5653506517410278,
      "eval_runtime": 18.484,
      "eval_samples_per_second": 10.171,
      "eval_steps_per_second": 2.543,
      "step": 520
    },
    {
      "epoch": 5.67,
      "learning_rate": 5.017921146953405e-06,
      "loss": 0.0958,
      "step": 530
    },
    {
      "epoch": 5.67,
      "eval_accuracy": 0.8723404407501221,
      "eval_loss": 0.5490068197250366,
      "eval_runtime": 16.8627,
      "eval_samples_per_second": 11.149,
      "eval_steps_per_second": 2.787,
      "step": 530
    },
    {
      "epoch": 5.78,
      "learning_rate": 3.225806451612903e-06,
      "loss": 0.0525,
      "step": 540
    },
    {
      "epoch": 5.78,
      "eval_accuracy": 0.8776595592498779,
      "eval_loss": 0.542866051197052,
      "eval_runtime": 18.0905,
      "eval_samples_per_second": 10.392,
      "eval_steps_per_second": 2.598,
      "step": 540
    },
    {
      "epoch": 5.88,
      "learning_rate": 1.4336917562724014e-06,
      "loss": 0.0144,
      "step": 550
    },
    {
      "epoch": 5.88,
      "eval_accuracy": 0.8829787373542786,
      "eval_loss": 0.5416831970214844,
      "eval_runtime": 17.9022,
      "eval_samples_per_second": 10.502,
      "eval_steps_per_second": 2.625,
      "step": 550
    }
  ],
  "logging_steps": 10,
  "max_steps": 558,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 10,
  "total_flos": 7.083294890110614e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}