{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.758733462167385,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 5e-06,
      "loss": 5.2948,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 1e-05,
      "loss": 4.9959,
      "step": 100
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.5e-05,
      "loss": 4.6899,
      "step": 150
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 4.5063,
      "step": 200
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.5e-05,
      "loss": 4.4335,
      "step": 250
    },
    {
      "epoch": 0.18,
      "learning_rate": 3e-05,
      "loss": 4.3943,
      "step": 300
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.5e-05,
      "loss": 4.35,
      "step": 350
    },
    {
      "epoch": 0.24,
      "learning_rate": 4e-05,
      "loss": 4.2928,
      "step": 400
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.5e-05,
      "loss": 4.2246,
      "step": 450
    },
    {
      "epoch": 0.3,
      "learning_rate": 5e-05,
      "loss": 4.1448,
      "step": 500
    },
    {
      "epoch": 0.33,
      "learning_rate": 5.500000000000001e-05,
      "loss": 4.0853,
      "step": 550
    },
    {
      "epoch": 0.36,
      "learning_rate": 6e-05,
      "loss": 4.0358,
      "step": 600
    },
    {
      "epoch": 0.39,
      "learning_rate": 6.500000000000001e-05,
      "loss": 3.9763,
      "step": 650
    },
    {
      "epoch": 0.42,
      "learning_rate": 7e-05,
      "loss": 3.8816,
      "step": 700
    },
    {
      "epoch": 0.45,
      "learning_rate": 7.500000000000001e-05,
      "loss": 3.7246,
      "step": 750
    },
    {
      "epoch": 0.48,
      "learning_rate": 8e-05,
      "loss": 3.5085,
      "step": 800
    },
    {
      "epoch": 0.51,
      "learning_rate": 8.5e-05,
      "loss": 3.2879,
      "step": 850
    },
    {
      "epoch": 0.54,
      "learning_rate": 9e-05,
      "loss": 3.0898,
      "step": 900
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.5e-05,
      "loss": 2.8867,
      "step": 950
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0001,
      "loss": 2.6323,
      "step": 1000
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.932478055367996e-05,
      "loss": 2.3103,
      "step": 1050
    },
    {
      "epoch": 0.65,
      "learning_rate": 9.86495611073599e-05,
      "loss": 2.0185,
      "step": 1100
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.797434166103985e-05,
      "loss": 1.7382,
      "step": 1150
    },
    {
      "epoch": 0.71,
      "learning_rate": 9.729912221471978e-05,
      "loss": 1.5627,
      "step": 1200
    },
    {
      "epoch": 0.74,
      "learning_rate": 9.662390276839974e-05,
      "loss": 1.4679,
      "step": 1250
    },
    {
      "epoch": 0.77,
      "learning_rate": 9.594868332207968e-05,
      "loss": 1.4046,
      "step": 1300
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.527346387575963e-05,
      "loss": 1.3544,
      "step": 1350
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.459824442943957e-05,
      "loss": 1.325,
      "step": 1400
    },
    {
      "epoch": 0.86,
      "learning_rate": 9.392302498311952e-05,
      "loss": 1.2984,
      "step": 1450
    },
    {
      "epoch": 0.89,
      "learning_rate": 9.324780553679947e-05,
      "loss": 1.2735,
      "step": 1500
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.25725860904794e-05,
      "loss": 1.2617,
      "step": 1550
    },
    {
      "epoch": 0.95,
      "learning_rate": 9.189736664415936e-05,
      "loss": 1.2475,
      "step": 1600
    },
    {
      "epoch": 0.98,
      "learning_rate": 9.12221471978393e-05,
      "loss": 1.2348,
      "step": 1650
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.054692775151925e-05,
      "loss": 1.2387,
      "step": 1700
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.98717083051992e-05,
      "loss": 1.2149,
      "step": 1750
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.919648885887914e-05,
      "loss": 1.2004,
      "step": 1800
    },
    {
      "epoch": 1.1,
      "learning_rate": 8.852126941255908e-05,
      "loss": 1.1918,
      "step": 1850
    },
    {
      "epoch": 1.13,
      "learning_rate": 8.784604996623903e-05,
      "loss": 1.1825,
      "step": 1900
    },
    {
      "epoch": 1.16,
      "learning_rate": 8.717083051991897e-05,
      "loss": 1.1794,
      "step": 1950
    },
    {
      "epoch": 1.19,
      "learning_rate": 8.649561107359893e-05,
      "loss": 1.175,
      "step": 2000
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.582039162727888e-05,
      "loss": 1.1657,
      "step": 2050
    },
    {
      "epoch": 1.25,
      "learning_rate": 8.514517218095882e-05,
      "loss": 1.1621,
      "step": 2100
    },
    {
      "epoch": 1.28,
      "learning_rate": 8.446995273463876e-05,
      "loss": 1.1564,
      "step": 2150
    },
    {
      "epoch": 1.31,
      "learning_rate": 8.37947332883187e-05,
      "loss": 1.149,
      "step": 2200
    },
    {
      "epoch": 1.34,
      "learning_rate": 8.311951384199866e-05,
      "loss": 1.1478,
      "step": 2250
    },
    {
      "epoch": 1.37,
      "learning_rate": 8.24442943956786e-05,
      "loss": 1.1449,
      "step": 2300
    },
    {
      "epoch": 1.4,
      "learning_rate": 8.176907494935855e-05,
      "loss": 1.1393,
      "step": 2350
    },
    {
      "epoch": 1.43,
      "learning_rate": 8.109385550303849e-05,
      "loss": 1.135,
      "step": 2400
    },
    {
      "epoch": 1.46,
      "learning_rate": 8.041863605671843e-05,
      "loss": 1.1308,
      "step": 2450
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.974341661039839e-05,
      "loss": 1.1282,
      "step": 2500
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.906819716407833e-05,
      "loss": 1.1197,
      "step": 2550
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.839297771775828e-05,
      "loss": 1.1227,
      "step": 2600
    },
    {
      "epoch": 1.58,
      "learning_rate": 7.771775827143822e-05,
      "loss": 1.1192,
      "step": 2650
    },
    {
      "epoch": 1.61,
      "learning_rate": 7.704253882511818e-05,
      "loss": 1.1109,
      "step": 2700
    },
    {
      "epoch": 1.64,
      "learning_rate": 7.63673193787981e-05,
      "loss": 1.1166,
      "step": 2750
    },
    {
      "epoch": 1.67,
      "learning_rate": 7.569209993247806e-05,
      "loss": 1.1132,
      "step": 2800
    },
    {
      "epoch": 1.7,
      "learning_rate": 7.5016880486158e-05,
      "loss": 1.1043,
      "step": 2850
    },
    {
      "epoch": 1.72,
      "learning_rate": 7.434166103983795e-05,
      "loss": 1.1085,
      "step": 2900
    },
    {
      "epoch": 1.75,
      "learning_rate": 7.366644159351789e-05,
      "loss": 1.1018,
      "step": 2950
    },
    {
      "epoch": 1.78,
      "learning_rate": 7.299122214719785e-05,
      "loss": 1.1034,
      "step": 3000
    },
    {
      "epoch": 1.81,
      "learning_rate": 7.23160027008778e-05,
      "loss": 1.0954,
      "step": 3050
    },
    {
      "epoch": 1.84,
      "learning_rate": 7.164078325455773e-05,
      "loss": 1.0973,
      "step": 3100
    },
    {
      "epoch": 1.87,
      "learning_rate": 7.096556380823768e-05,
      "loss": 1.0971,
      "step": 3150
    },
    {
      "epoch": 1.9,
      "learning_rate": 7.029034436191762e-05,
      "loss": 1.0893,
      "step": 3200
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.961512491559758e-05,
      "loss": 1.0891,
      "step": 3250
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.893990546927752e-05,
      "loss": 1.0866,
      "step": 3300
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.826468602295747e-05,
      "loss": 1.0938,
      "step": 3350
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.758946657663741e-05,
      "loss": 1.1025,
      "step": 3400
    },
    {
      "epoch": 2.05,
      "learning_rate": 6.691424713031735e-05,
      "loss": 1.0858,
      "step": 3450
    },
    {
      "epoch": 2.08,
      "learning_rate": 6.623902768399731e-05,
      "loss": 1.0835,
      "step": 3500
    },
    {
      "epoch": 2.11,
      "learning_rate": 6.556380823767725e-05,
      "loss": 1.0793,
      "step": 3550
    },
    {
      "epoch": 2.14,
      "learning_rate": 6.48885887913572e-05,
      "loss": 1.0754,
      "step": 3600
    },
    {
      "epoch": 2.17,
      "learning_rate": 6.421336934503714e-05,
      "loss": 1.073,
      "step": 3650
    },
    {
      "epoch": 2.2,
      "learning_rate": 6.353814989871708e-05,
      "loss": 1.0774,
      "step": 3700
    },
    {
      "epoch": 2.23,
      "learning_rate": 6.286293045239702e-05,
      "loss": 1.0763,
      "step": 3750
    },
    {
      "epoch": 2.26,
      "learning_rate": 6.218771100607698e-05,
      "loss": 1.0693,
      "step": 3800
    },
    {
      "epoch": 2.29,
      "learning_rate": 6.151249155975692e-05,
      "loss": 1.0736,
      "step": 3850
    },
    {
      "epoch": 2.32,
      "learning_rate": 6.083727211343687e-05,
      "loss": 1.0762,
      "step": 3900
    },
    {
      "epoch": 2.35,
      "learning_rate": 6.016205266711682e-05,
      "loss": 1.0691,
      "step": 3950
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.948683322079677e-05,
      "loss": 1.0668,
      "step": 4000
    },
    {
      "epoch": 2.41,
      "learning_rate": 5.88116137744767e-05,
      "loss": 1.0638,
      "step": 4050
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.813639432815665e-05,
      "loss": 1.0671,
      "step": 4100
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.7461174881836596e-05,
      "loss": 1.0595,
      "step": 4150
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.6785955435516544e-05,
      "loss": 1.0606,
      "step": 4200
    },
    {
      "epoch": 2.53,
      "learning_rate": 5.61107359891965e-05,
      "loss": 1.0621,
      "step": 4250
    },
    {
      "epoch": 2.56,
      "learning_rate": 5.5435516542876445e-05,
      "loss": 1.0633,
      "step": 4300
    },
    {
      "epoch": 2.59,
      "learning_rate": 5.476029709655638e-05,
      "loss": 1.056,
      "step": 4350
    },
    {
      "epoch": 2.62,
      "learning_rate": 5.4085077650236326e-05,
      "loss": 1.0598,
      "step": 4400
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.3409858203916274e-05,
      "loss": 1.0532,
      "step": 4450
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.273463875759622e-05,
      "loss": 1.0536,
      "step": 4500
    },
    {
      "epoch": 2.71,
      "learning_rate": 5.205941931127617e-05,
      "loss": 1.0522,
      "step": 4550
    },
    {
      "epoch": 2.74,
      "learning_rate": 5.1384199864956116e-05,
      "loss": 1.0478,
      "step": 4600
    },
    {
      "epoch": 2.77,
      "learning_rate": 5.0708980418636057e-05,
      "loss": 1.0514,
      "step": 4650
    },
    {
      "epoch": 2.8,
      "learning_rate": 5.0033760972316004e-05,
      "loss": 1.0526,
      "step": 4700
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.935854152599595e-05,
      "loss": 1.0476,
      "step": 4750
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.86833220796759e-05,
      "loss": 1.0459,
      "step": 4800
    },
    {
      "epoch": 2.88,
      "learning_rate": 4.8008102633355846e-05,
      "loss": 1.0463,
      "step": 4850
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.733288318703579e-05,
      "loss": 1.0469,
      "step": 4900
    },
    {
      "epoch": 2.94,
      "learning_rate": 4.6657663740715734e-05,
      "loss": 1.0444,
      "step": 4950
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.598244429439568e-05,
      "loss": 1.0461,
      "step": 5000
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.530722484807562e-05,
      "loss": 1.0613,
      "step": 5050
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.463200540175557e-05,
      "loss": 1.0418,
      "step": 5100
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.395678595543552e-05,
      "loss": 1.0415,
      "step": 5150
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.3281566509115464e-05,
      "loss": 1.0401,
      "step": 5200
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.260634706279541e-05,
      "loss": 1.0404,
      "step": 5250
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.193112761647536e-05,
      "loss": 1.0372,
      "step": 5300
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.125590817015531e-05,
      "loss": 1.0423,
      "step": 5350
    },
    {
      "epoch": 3.21,
      "learning_rate": 4.058068872383525e-05,
      "loss": 1.0343,
      "step": 5400
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.9905469277515195e-05,
      "loss": 1.0374,
      "step": 5450
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.923024983119514e-05,
      "loss": 1.0355,
      "step": 5500
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.855503038487508e-05,
      "loss": 1.0312,
      "step": 5550
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.787981093855503e-05,
      "loss": 1.0374,
      "step": 5600
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.720459149223498e-05,
      "loss": 1.034,
      "step": 5650
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.6529372045914925e-05,
      "loss": 1.0327,
      "step": 5700
    },
    {
      "epoch": 3.42,
      "learning_rate": 3.585415259959487e-05,
      "loss": 1.0317,
      "step": 5750
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.517893315327482e-05,
      "loss": 1.0317,
      "step": 5800
    },
    {
      "epoch": 3.48,
      "learning_rate": 3.450371370695476e-05,
      "loss": 1.028,
      "step": 5850
    },
    {
      "epoch": 3.51,
      "learning_rate": 3.382849426063471e-05,
      "loss": 1.027,
      "step": 5900
    },
    {
      "epoch": 3.54,
      "learning_rate": 3.3153274814314655e-05,
      "loss": 1.0253,
      "step": 5950
    },
    {
      "epoch": 3.57,
      "learning_rate": 3.2478055367994596e-05,
      "loss": 1.0291,
      "step": 6000
    },
    {
      "epoch": 3.6,
      "learning_rate": 3.180283592167454e-05,
      "loss": 1.0266,
      "step": 6050
    },
    {
      "epoch": 3.63,
      "learning_rate": 3.112761647535449e-05,
      "loss": 1.0311,
      "step": 6100
    },
    {
      "epoch": 3.66,
      "learning_rate": 3.0452397029034435e-05,
      "loss": 1.0281,
      "step": 6150
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.9777177582714382e-05,
      "loss": 1.0238,
      "step": 6200
    },
    {
      "epoch": 3.72,
      "learning_rate": 2.910195813639433e-05,
      "loss": 1.0248,
      "step": 6250
    },
    {
      "epoch": 3.75,
      "learning_rate": 2.8426738690074277e-05,
      "loss": 1.0235,
      "step": 6300
    },
    {
      "epoch": 3.78,
      "learning_rate": 2.775151924375422e-05,
      "loss": 1.0252,
      "step": 6350
    },
    {
      "epoch": 3.81,
      "learning_rate": 2.7076299797434168e-05,
      "loss": 1.0221,
      "step": 6400
    },
    {
      "epoch": 3.84,
      "learning_rate": 2.6401080351114116e-05,
      "loss": 1.0204,
      "step": 6450
    },
    {
      "epoch": 3.87,
      "learning_rate": 2.572586090479406e-05,
      "loss": 1.0205,
      "step": 6500
    },
    {
      "epoch": 3.9,
      "learning_rate": 2.5050641458474007e-05,
      "loss": 1.0207,
      "step": 6550
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.437542201215395e-05,
      "loss": 1.0165,
      "step": 6600
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.37002025658339e-05,
      "loss": 1.0182,
      "step": 6650
    },
    {
      "epoch": 3.99,
      "learning_rate": 2.3024983119513842e-05,
      "loss": 1.0165,
      "step": 6700
    },
    {
      "epoch": 4.02,
      "learning_rate": 2.234976367319379e-05,
      "loss": 1.0316,
      "step": 6750
    },
    {
      "epoch": 4.05,
      "learning_rate": 2.1674544226873737e-05,
      "loss": 1.0146,
      "step": 6800
    },
    {
      "epoch": 4.07,
      "learning_rate": 2.099932478055368e-05,
      "loss": 1.015,
      "step": 6850
    },
    {
      "epoch": 4.1,
      "learning_rate": 2.0324105334233625e-05,
      "loss": 1.0156,
      "step": 6900
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.9648885887913573e-05,
      "loss": 1.016,
      "step": 6950
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.8973666441593517e-05,
      "loss": 1.0149,
      "step": 7000
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.8298446995273467e-05,
      "loss": 1.0127,
      "step": 7050
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.762322754895341e-05,
      "loss": 1.0085,
      "step": 7100
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.6948008102633355e-05,
      "loss": 1.0151,
      "step": 7150
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.6272788656313303e-05,
      "loss": 1.0136,
      "step": 7200
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.5597569209993247e-05,
      "loss": 1.0077,
      "step": 7250
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.4922349763673194e-05,
      "loss": 1.0103,
      "step": 7300
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.4247130317353142e-05,
      "loss": 1.0155,
      "step": 7350
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.3571910871033086e-05,
      "loss": 1.0098,
      "step": 7400
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.2896691424713031e-05,
      "loss": 1.0093,
      "step": 7450
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.2221471978392979e-05,
      "loss": 1.0112,
      "step": 7500
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.1546252532072925e-05,
      "loss": 1.0075,
      "step": 7550
    },
    {
      "epoch": 4.52,
      "learning_rate": 1.087103308575287e-05,
      "loss": 1.0086,
      "step": 7600
    },
    {
      "epoch": 4.55,
      "learning_rate": 1.0195813639432816e-05,
      "loss": 1.0072,
      "step": 7650
    },
    {
      "epoch": 4.58,
      "learning_rate": 9.520594193112763e-06,
      "loss": 1.0057,
      "step": 7700
    },
    {
      "epoch": 4.61,
      "learning_rate": 8.845374746792707e-06,
      "loss": 1.0101,
      "step": 7750
    },
    {
      "epoch": 4.64,
      "learning_rate": 8.170155300472653e-06,
      "loss": 1.006,
      "step": 7800
    },
    {
      "epoch": 4.67,
      "learning_rate": 7.4949358541526005e-06,
      "loss": 1.0079,
      "step": 7850
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.819716407832546e-06,
      "loss": 1.0066,
      "step": 7900
    },
    {
      "epoch": 4.73,
      "learning_rate": 6.144496961512492e-06,
      "loss": 1.0113,
      "step": 7950
    },
    {
      "epoch": 4.76,
      "learning_rate": 5.4692775151924376e-06,
      "loss": 1.0065,
      "step": 8000
    }
  ],
  "max_steps": 8405,
  "num_train_epochs": 5,
  "total_flos": 5.140843815167534e+17,
  "trial_name": null,
  "trial_params": null
}