{
  "best_metric": 0.07845821231603622,
  "best_model_checkpoint": "./checkpoints/ckpt_8_12_r32_debug_fp16_lr5e-05_batch8/checkpoint-3000",
  "epoch": 4.980286366466072,
  "eval_steps": 125,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.747252747252747e-06,
      "loss": 0.6187,
      "step": 5
    },
    {
      "epoch": 0.02,
      "learning_rate": 5.494505494505494e-06,
      "loss": 0.6047,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 8.241758241758243e-06,
      "loss": 0.5203,
      "step": 15
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.0989010989010989e-05,
      "loss": 0.4791,
      "step": 20
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.3736263736263738e-05,
      "loss": 0.4216,
      "step": 25
    },
    {
      "epoch": 0.05,
      "learning_rate": 1.6483516483516486e-05,
      "loss": 0.3502,
      "step": 30
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.923076923076923e-05,
      "loss": 0.29,
      "step": 35
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.1978021978021977e-05,
      "loss": 0.2761,
      "step": 40
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.4725274725274727e-05,
      "loss": 0.2568,
      "step": 45
    },
    {
      "epoch": 0.08,
      "learning_rate": 2.7472527472527476e-05,
      "loss": 0.2493,
      "step": 50
    },
    {
      "epoch": 0.09,
      "learning_rate": 3.021978021978022e-05,
      "loss": 0.2458,
      "step": 55
    },
    {
      "epoch": 0.1,
      "learning_rate": 3.296703296703297e-05,
      "loss": 0.2542,
      "step": 60
    },
    {
      "epoch": 0.11,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.2133,
      "step": 65
    },
    {
      "epoch": 0.12,
      "learning_rate": 3.846153846153846e-05,
      "loss": 0.2177,
      "step": 70
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.120879120879121e-05,
      "loss": 0.2397,
      "step": 75
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.3956043956043955e-05,
      "loss": 0.2287,
      "step": 80
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.670329670329671e-05,
      "loss": 0.2279,
      "step": 85
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.945054945054945e-05,
      "loss": 0.2375,
      "step": 90
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.999976833475735e-05,
      "loss": 0.2392,
      "step": 95
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.999882720206755e-05,
      "loss": 0.2276,
      "step": 100
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.999716215008542e-05,
      "loss": 0.2269,
      "step": 105
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.999477322702776e-05,
      "loss": 0.2174,
      "step": 110
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.99916605020733e-05,
      "loss": 0.2315,
      "step": 115
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.9987824065360743e-05,
      "loss": 0.2137,
      "step": 120
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.998326402798612e-05,
      "loss": 0.2415,
      "step": 125
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.19619159400463104,
      "eval_runtime": 6.7054,
      "eval_samples_per_second": 14.913,
      "eval_steps_per_second": 14.913,
      "step": 125
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.997798052199959e-05,
      "loss": 0.2074,
      "step": 130
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.99719737004016e-05,
      "loss": 0.2289,
      "step": 135
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.996524373713848e-05,
      "loss": 0.2396,
      "step": 140
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.995779082709739e-05,
      "loss": 0.2294,
      "step": 145
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.994961518610064e-05,
      "loss": 0.229,
      "step": 150
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.994071705089953e-05,
      "loss": 0.2269,
      "step": 155
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.993109667916742e-05,
      "loss": 0.2172,
      "step": 160
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.992075434949227e-05,
      "loss": 0.2143,
      "step": 165
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.990969036136862e-05,
      "loss": 0.211,
      "step": 170
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.989790503518888e-05,
      "loss": 0.2261,
      "step": 175
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.9885398712234054e-05,
      "loss": 0.2182,
      "step": 180
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.987217175466387e-05,
      "loss": 0.2304,
      "step": 185
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.9858224545506296e-05,
      "loss": 0.2293,
      "step": 190
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.9843557488646416e-05,
      "loss": 0.2252,
      "step": 195
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.9828171008814786e-05,
      "loss": 0.2294,
      "step": 200
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.9812065551575077e-05,
      "loss": 0.2137,
      "step": 205
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.979524158331123e-05,
      "loss": 0.1993,
      "step": 210
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.97776995912139e-05,
      "loss": 0.2004,
      "step": 215
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.975944008326639e-05,
      "loss": 0.2284,
      "step": 220
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.974046358822988e-05,
      "loss": 0.2117,
      "step": 225
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.972077065562821e-05,
      "loss": 0.2229,
      "step": 230
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.9700361855731883e-05,
      "loss": 0.2124,
      "step": 235
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.967923777954158e-05,
      "loss": 0.22,
      "step": 240
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.965739903877104e-05,
      "loss": 0.2345,
      "step": 245
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.963484626582937e-05,
      "loss": 0.2077,
      "step": 250
    },
    {
      "epoch": 0.42,
      "eval_loss": 0.18103449046611786,
      "eval_runtime": 6.668,
      "eval_samples_per_second": 14.997,
      "eval_steps_per_second": 14.997,
      "step": 250
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.96115801138027e-05,
      "loss": 0.2069,
      "step": 255
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.95876012564353e-05,
      "loss": 0.226,
      "step": 260
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.956291038811003e-05,
      "loss": 0.2113,
      "step": 265
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.9537508223828266e-05,
      "loss": 0.2033,
      "step": 270
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.9511395499189206e-05,
      "loss": 0.2263,
      "step": 275
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.9484572970368514e-05,
      "loss": 0.2187,
      "step": 280
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.945704141409648e-05,
      "loss": 0.2176,
      "step": 285
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.942880162763549e-05,
      "loss": 0.2297,
      "step": 290
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.939985442875695e-05,
      "loss": 0.2101,
      "step": 295
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.937020065571763e-05,
      "loss": 0.2089,
      "step": 300
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.9339841167235336e-05,
      "loss": 0.1906,
      "step": 305
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.93087768424641e-05,
      "loss": 0.222,
      "step": 310
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.927700858096866e-05,
      "loss": 0.2094,
      "step": 315
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.924453730269849e-05,
      "loss": 0.238,
      "step": 320
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.9211363947961083e-05,
      "loss": 0.2104,
      "step": 325
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.917748947739478e-05,
      "loss": 0.2025,
      "step": 330
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.9142914871940885e-05,
      "loss": 0.2037,
      "step": 335
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.9107641132815325e-05,
      "loss": 0.2179,
      "step": 340
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.907166928147965e-05,
      "loss": 0.2265,
      "step": 345
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.9035000359611396e-05,
      "loss": 0.2204,
      "step": 350
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.899763542907399e-05,
      "loss": 0.2114,
      "step": 355
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.895957557188593e-05,
      "loss": 0.2205,
      "step": 360
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.892082189018954e-05,
      "loss": 0.2267,
      "step": 365
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.8881375506218976e-05,
      "loss": 0.2338,
      "step": 370
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.8841237562267745e-05,
      "loss": 0.196,
      "step": 375
    },
    {
      "epoch": 0.62,
      "eval_loss": 0.16536256670951843,
      "eval_runtime": 6.6441,
      "eval_samples_per_second": 15.051,
      "eval_steps_per_second": 15.051,
      "step": 375
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.880040922065566e-05,
      "loss": 0.2032,
      "step": 380
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.875889166369517e-05,
      "loss": 0.2016,
      "step": 385
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.8716686093657066e-05,
      "loss": 0.2104,
      "step": 390
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.8673793732735753e-05,
      "loss": 0.2049,
      "step": 395
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.8630215823013805e-05,
      "loss": 0.2081,
      "step": 400
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.8585953626426e-05,
      "loss": 0.2049,
      "step": 405
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.854100842472278e-05,
      "loss": 0.2159,
      "step": 410
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.849538151943316e-05,
      "loss": 0.2305,
      "step": 415
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.8449074231826994e-05,
      "loss": 0.2217,
      "step": 420
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.840208790287676e-05,
      "loss": 0.2061,
      "step": 425
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.835442389321867e-05,
      "loss": 0.2202,
      "step": 430
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.8306083583113326e-05,
      "loss": 0.2032,
      "step": 435
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.825706837240572e-05,
      "loss": 0.2089,
      "step": 440
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.820737968048471e-05,
      "loss": 0.2101,
      "step": 445
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.815701894624191e-05,
      "loss": 0.2121,
      "step": 450
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.810598762803e-05,
      "loss": 0.2205,
      "step": 455
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.805428720362054e-05,
      "loss": 0.206,
      "step": 460
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.800191917016116e-05,
      "loss": 0.2061,
      "step": 465
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.794888504413219e-05,
      "loss": 0.2162,
      "step": 470
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.7895186361302734e-05,
      "loss": 0.1978,
      "step": 475
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.784082467668627e-05,
      "loss": 0.2248,
      "step": 480
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.7785801564495514e-05,
      "loss": 0.2083,
      "step": 485
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.773011861809693e-05,
      "loss": 0.2178,
      "step": 490
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.767377744996451e-05,
      "loss": 0.2003,
      "step": 495
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.761677969163315e-05,
      "loss": 0.2059,
      "step": 500
    },
    {
      "epoch": 0.83,
      "eval_loss": 0.1576012670993805,
      "eval_runtime": 6.6622,
      "eval_samples_per_second": 15.01,
      "eval_steps_per_second": 15.01,
      "step": 500
    },
    {
      "epoch": 0.84,
      "learning_rate": 4.755912699365135e-05,
      "loss": 0.2144,
      "step": 505
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.750082102553344e-05,
      "loss": 0.219,
      "step": 510
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.7441863475711236e-05,
      "loss": 0.1945,
      "step": 515
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.7382256051485144e-05,
      "loss": 0.2298,
      "step": 520
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.73220004789747e-05,
      "loss": 0.2052,
      "step": 525
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.726109850306864e-05,
      "loss": 0.197,
      "step": 530
    },
    {
      "epoch": 0.89,
      "learning_rate": 4.7199551887374295e-05,
      "loss": 0.2122,
      "step": 535
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.713736241416658e-05,
      "loss": 0.2088,
      "step": 540
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.707453188433635e-05,
      "loss": 0.2086,
      "step": 545
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.701106211733827e-05,
      "loss": 0.2022,
      "step": 550
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.6946954951138125e-05,
      "loss": 0.2158,
      "step": 555
    },
    {
      "epoch": 0.93,
      "learning_rate": 4.688221224215955e-05,
      "loss": 0.2191,
      "step": 560
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.681683586523037e-05,
      "loss": 0.2204,
      "step": 565
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.675082771352821e-05,
      "loss": 0.2356,
      "step": 570
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.66841896985257e-05,
      "loss": 0.2064,
      "step": 575
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.661692374993519e-05,
      "loss": 0.2086,
      "step": 580
    },
    {
      "epoch": 0.97,
      "learning_rate": 4.6549031815652744e-05,
      "loss": 0.196,
      "step": 585
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.648051586170185e-05,
      "loss": 0.2034,
      "step": 590
    },
    {
      "epoch": 0.99,
      "learning_rate": 4.641137787217641e-05,
      "loss": 0.2129,
      "step": 595
    },
    {
      "epoch": 1.0,
      "learning_rate": 4.634161984918333e-05,
      "loss": 0.2064,
      "step": 600
    },
    {
      "epoch": 1.0,
      "learning_rate": 4.6271243812784514e-05,
      "loss": 0.2097,
      "step": 605
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.62002518009384e-05,
      "loss": 0.1844,
      "step": 610
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.612864586944089e-05,
      "loss": 0.2026,
      "step": 615
    },
    {
      "epoch": 1.03,
      "learning_rate": 4.605642809186589e-05,
      "loss": 0.1695,
      "step": 620
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.598360055950519e-05,
      "loss": 0.1998,
      "step": 625
    },
    {
      "epoch": 1.04,
      "eval_loss": 0.15137994289398193,
      "eval_runtime": 6.6769,
      "eval_samples_per_second": 14.977,
      "eval_steps_per_second": 14.977,
      "step": 625
    },
    {
      "epoch": 1.05,
      "learning_rate": 4.591016538130796e-05,
      "loss": 0.1899,
      "step": 630
    },
    {
      "epoch": 1.05,
      "learning_rate": 4.583612468381964e-05,
      "loss": 0.1934,
      "step": 635
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.576148061112039e-05,
      "loss": 0.1849,
      "step": 640
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.5686235324762995e-05,
      "loss": 0.1976,
      "step": 645
    },
    {
      "epoch": 1.08,
      "learning_rate": 4.561039100371023e-05,
      "loss": 0.1927,
      "step": 650
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.553394984427184e-05,
      "loss": 0.2024,
      "step": 655
    },
    {
      "epoch": 1.1,
      "learning_rate": 4.545691406004085e-05,
      "loss": 0.1813,
      "step": 660
    },
    {
      "epoch": 1.1,
      "learning_rate": 4.5379285881829555e-05,
      "loss": 0.1706,
      "step": 665
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.5301067557604813e-05,
      "loss": 0.1894,
      "step": 670
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.522226135242306e-05,
      "loss": 0.1863,
      "step": 675
    },
    {
      "epoch": 1.13,
      "learning_rate": 4.5142869548364644e-05,
      "loss": 0.1961,
      "step": 680
    },
    {
      "epoch": 1.14,
      "learning_rate": 4.5062894444467754e-05,
      "loss": 0.1811,
      "step": 685
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.498233835666186e-05,
      "loss": 0.1864,
      "step": 690
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.490120361770064e-05,
      "loss": 0.1973,
      "step": 695
    },
    {
      "epoch": 1.16,
      "learning_rate": 4.481949257709442e-05,
      "loss": 0.1939,
      "step": 700
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.473720760104217e-05,
      "loss": 0.1906,
      "step": 705
    },
    {
      "epoch": 1.18,
      "learning_rate": 4.465435107236292e-05,
      "loss": 0.1808,
      "step": 710
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.457092539042682e-05,
      "loss": 0.192,
      "step": 715
    },
    {
      "epoch": 1.2,
      "learning_rate": 4.448693297108563e-05,
      "loss": 0.1935,
      "step": 720
    },
    {
      "epoch": 1.2,
      "learning_rate": 4.440237624660275e-05,
      "loss": 0.1744,
      "step": 725
    },
    {
      "epoch": 1.21,
      "learning_rate": 4.4317257665582836e-05,
      "loss": 0.1909,
      "step": 730
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.4231579692900816e-05,
      "loss": 0.2009,
      "step": 735
    },
    {
      "epoch": 1.23,
      "learning_rate": 4.414534480963059e-05,
      "loss": 0.2023,
      "step": 740
    },
    {
      "epoch": 1.24,
      "learning_rate": 4.405855551297313e-05,
      "loss": 0.1932,
      "step": 745
    },
    {
      "epoch": 1.25,
      "learning_rate": 4.397121431618419e-05,
      "loss": 0.1798,
      "step": 750
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.14108647406101227,
      "eval_runtime": 6.6767,
      "eval_samples_per_second": 14.977,
      "eval_steps_per_second": 14.977,
      "step": 750
    },
    {
      "epoch": 1.25,
      "learning_rate": 4.38833237485015e-05,
      "loss": 0.2114,
      "step": 755
    },
    {
      "epoch": 1.26,
      "learning_rate": 4.3794886355071575e-05,
      "loss": 0.1978,
      "step": 760
    },
    {
      "epoch": 1.27,
      "learning_rate": 4.370590469687596e-05,
      "loss": 0.1735,
      "step": 765
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.361638135065711e-05,
      "loss": 0.1852,
      "step": 770
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.352631890884372e-05,
      "loss": 0.1813,
      "step": 775
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.343571997947571e-05,
      "loss": 0.1968,
      "step": 780
    },
    {
      "epoch": 1.3,
      "learning_rate": 4.3344587186128676e-05,
      "loss": 0.1944,
      "step": 785
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.325292316783791e-05,
      "loss": 0.1929,
      "step": 790
    },
    {
      "epoch": 1.32,
      "learning_rate": 4.3160730579021975e-05,
      "loss": 0.1823,
      "step": 795
    },
    {
      "epoch": 1.33,
      "learning_rate": 4.306801208940587e-05,
      "loss": 0.1968,
      "step": 800
    },
    {
      "epoch": 1.34,
      "learning_rate": 4.297477038394368e-05,
      "loss": 0.1888,
      "step": 805
    },
    {
      "epoch": 1.34,
      "learning_rate": 4.2881008162740854e-05,
      "loss": 0.196,
      "step": 810
    },
    {
      "epoch": 1.35,
      "learning_rate": 4.278672814097601e-05,
      "loss": 0.1971,
      "step": 815
    },
    {
      "epoch": 1.36,
      "learning_rate": 4.269193304882226e-05,
      "loss": 0.196,
      "step": 820
    },
    {
      "epoch": 1.37,
      "learning_rate": 4.2596625631368255e-05,
      "loss": 0.1922,
      "step": 825
    },
    {
      "epoch": 1.38,
      "learning_rate": 4.250080864853858e-05,
      "loss": 0.1955,
      "step": 830
    },
    {
      "epoch": 1.39,
      "learning_rate": 4.2404484875013895e-05,
      "loss": 0.1892,
      "step": 835
    },
    {
      "epoch": 1.39,
      "learning_rate": 4.2307657100150584e-05,
      "loss": 0.1903,
      "step": 840
    },
    {
      "epoch": 1.4,
      "learning_rate": 4.2210328127899943e-05,
      "loss": 0.1953,
      "step": 845
    },
    {
      "epoch": 1.41,
      "learning_rate": 4.211250077672704e-05,
      "loss": 0.1916,
      "step": 850
    },
    {
      "epoch": 1.42,
      "learning_rate": 4.201417787952901e-05,
      "loss": 0.1951,
      "step": 855
    },
    {
      "epoch": 1.43,
      "learning_rate": 4.191536228355315e-05,
      "loss": 0.181,
      "step": 860
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.181605685031432e-05,
      "loss": 0.1976,
      "step": 865
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.17162644555122e-05,
      "loss": 0.1915,
      "step": 870
    },
    {
      "epoch": 1.45,
      "learning_rate": 4.1615987988947944e-05,
      "loss": 0.2061,
      "step": 875
    },
    {
      "epoch": 1.45,
      "eval_loss": 0.13502272963523865,
      "eval_runtime": 6.5739,
      "eval_samples_per_second": 15.212,
      "eval_steps_per_second": 15.212,
      "step": 875
    },
    {
      "epoch": 1.46,
      "learning_rate": 4.1515230354440524e-05,
      "loss": 0.2027,
      "step": 880
    },
    {
      "epoch": 1.47,
      "learning_rate": 4.1413994469742637e-05,
      "loss": 0.1916,
      "step": 885
    },
    {
      "epoch": 1.48,
      "learning_rate": 4.131228326645621e-05,
      "loss": 0.1925,
      "step": 890
    },
    {
      "epoch": 1.49,
      "learning_rate": 4.1210099689947513e-05,
      "loss": 0.1941,
      "step": 895
    },
    {
      "epoch": 1.49,
      "learning_rate": 4.110744669926184e-05,
      "loss": 0.1746,
      "step": 900
    },
    {
      "epoch": 1.5,
      "learning_rate": 4.100432726703788e-05,
      "loss": 0.1876,
      "step": 905
    },
    {
      "epoch": 1.51,
      "learning_rate": 4.090074437942155e-05,
      "loss": 0.204,
      "step": 910
    },
    {
      "epoch": 1.52,
      "learning_rate": 4.0796701035979626e-05,
      "loss": 0.1852,
      "step": 915
    },
    {
      "epoch": 1.53,
      "learning_rate": 4.069220024961275e-05,
      "loss": 0.1925,
      "step": 920
    },
    {
      "epoch": 1.54,
      "learning_rate": 4.058724504646834e-05,
      "loss": 0.192,
      "step": 925
    },
    {
      "epoch": 1.54,
      "learning_rate": 4.0481838465852806e-05,
      "loss": 0.2025,
      "step": 930
    },
    {
      "epoch": 1.55,
      "learning_rate": 4.037598356014363e-05,
      "loss": 0.1921,
      "step": 935
    },
    {
      "epoch": 1.56,
      "learning_rate": 4.026968339470096e-05,
      "loss": 0.1824,
      "step": 940
    },
    {
      "epoch": 1.57,
      "learning_rate": 4.0162941047778825e-05,
      "loss": 0.1861,
      "step": 945
    },
    {
      "epoch": 1.58,
      "learning_rate": 4.0055759610435996e-05,
      "loss": 0.1928,
      "step": 950
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.9948142186446486e-05,
      "loss": 0.1821,
      "step": 955
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.9840091892209673e-05,
      "loss": 0.1918,
      "step": 960
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.973161185666005e-05,
      "loss": 0.1779,
      "step": 965
    },
    {
      "epoch": 1.61,
      "learning_rate": 3.9622705221176596e-05,
      "loss": 0.1854,
      "step": 970
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.951337513949184e-05,
      "loss": 0.1732,
      "step": 975
    },
    {
      "epoch": 1.63,
      "learning_rate": 3.940362477760052e-05,
      "loss": 0.1893,
      "step": 980
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.92934573136679e-05,
      "loss": 0.1969,
      "step": 985
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.9182875937937755e-05,
      "loss": 0.1884,
      "step": 990
    },
    {
      "epoch": 1.65,
      "learning_rate": 3.907188385263995e-05,
      "loss": 0.1837,
      "step": 995
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.8960484271897736e-05,
      "loss": 0.1887,
      "step": 1000
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.13266006112098694,
      "eval_runtime": 6.6053,
      "eval_samples_per_second": 15.139,
      "eval_steps_per_second": 15.139,
      "step": 1000
    },
    {
      "epoch": 1.67,
      "learning_rate": 3.8848680421634676e-05,
      "loss": 0.1837,
      "step": 1005
    },
    {
      "epoch": 1.68,
      "learning_rate": 3.873647553948121e-05,
      "loss": 0.199,
      "step": 1010
    },
    {
      "epoch": 1.68,
      "learning_rate": 3.862387287468094e-05,
      "loss": 0.1933,
      "step": 1015
    },
    {
      "epoch": 1.69,
      "learning_rate": 3.851087568799648e-05,
      "loss": 0.1761,
      "step": 1020
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.8397487251615085e-05,
      "loss": 0.1848,
      "step": 1025
    },
    {
      "epoch": 1.71,
      "learning_rate": 3.8283710849053834e-05,
      "loss": 0.1883,
      "step": 1030
    },
    {
      "epoch": 1.72,
      "learning_rate": 3.816954977506462e-05,
      "loss": 0.2009,
      "step": 1035
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.8055007335538665e-05,
      "loss": 0.1863,
      "step": 1040
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.7940086847410874e-05,
      "loss": 0.1866,
      "step": 1045
    },
    {
      "epoch": 1.74,
      "learning_rate": 3.782479163856367e-05,
      "loss": 0.2085,
      "step": 1050
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.770912504773076e-05,
      "loss": 0.1874,
      "step": 1055
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.759309042440028e-05,
      "loss": 0.1885,
      "step": 1060
    },
    {
      "epoch": 1.77,
      "learning_rate": 3.747669112871801e-05,
      "loss": 0.195,
      "step": 1065
    },
    {
      "epoch": 1.78,
      "learning_rate": 3.735993053138988e-05,
      "loss": 0.1969,
      "step": 1070
    },
    {
      "epoch": 1.78,
      "learning_rate": 3.724281201358447e-05,
      "loss": 0.193,
      "step": 1075
    },
    {
      "epoch": 1.79,
      "learning_rate": 3.7125338966835064e-05,
      "loss": 0.179,
      "step": 1080
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.700751479294146e-05,
      "loss": 0.2007,
      "step": 1085
    },
    {
      "epoch": 1.81,
      "learning_rate": 3.68893429038714e-05,
      "loss": 0.1979,
      "step": 1090
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.677082672166185e-05,
      "loss": 0.1982,
      "step": 1095
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.665196967831981e-05,
      "loss": 0.1936,
      "step": 1100
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.6532775215723044e-05,
      "loss": 0.1896,
      "step": 1105
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.641324678552031e-05,
      "loss": 0.1927,
      "step": 1110
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.629338784903145e-05,
      "loss": 0.1961,
      "step": 1115
    },
    {
      "epoch": 1.86,
      "learning_rate": 3.6173201877147134e-05,
      "loss": 0.2042,
      "step": 1120
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.6052692350228404e-05,
      "loss": 0.1966,
      "step": 1125
    },
    {
      "epoch": 1.87,
      "eval_loss": 0.12441341578960419,
      "eval_runtime": 6.5924,
      "eval_samples_per_second": 15.169,
      "eval_steps_per_second": 15.169,
      "step": 1125
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.5931862758005803e-05,
      "loss": 0.1934,
      "step": 1130
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.581071659947842e-05,
      "loss": 0.1776,
      "step": 1135
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.5689257382812454e-05,
      "loss": 0.1814,
      "step": 1140
    },
    {
      "epoch": 1.9,
      "learning_rate": 3.556748862523973e-05,
      "loss": 0.1909,
      "step": 1145
    },
    {
      "epoch": 1.91,
      "learning_rate": 3.544541385295578e-05,
      "loss": 0.1925,
      "step": 1150
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.5323036601017756e-05,
      "loss": 0.1847,
      "step": 1155
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.520036041324203e-05,
      "loss": 0.1779,
      "step": 1160
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.507738884210164e-05,
      "loss": 0.1907,
      "step": 1165
    },
    {
      "epoch": 1.94,
      "learning_rate": 3.4954125448623336e-05,
      "loss": 0.1807,
      "step": 1170
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.483057380228452e-05,
      "loss": 0.1777,
      "step": 1175
    },
    {
      "epoch": 1.96,
      "learning_rate": 3.470673748090984e-05,
      "loss": 0.194,
      "step": 1180
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.4582620070567614e-05,
      "loss": 0.1821,
      "step": 1185
    },
    {
      "epoch": 1.98,
      "learning_rate": 3.445822516546598e-05,
      "loss": 0.1915,
      "step": 1190
    },
    {
      "epoch": 1.98,
      "learning_rate": 3.4333556367848776e-05,
      "loss": 0.1688,
      "step": 1195
    },
    {
      "epoch": 1.99,
      "learning_rate": 3.420861728789129e-05,
      "loss": 0.1949,
      "step": 1200
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.408341154359564e-05,
      "loss": 0.1974,
      "step": 1205
    },
    {
      "epoch": 2.01,
      "learning_rate": 3.395794276068609e-05,
      "loss": 0.1609,
      "step": 1210
    },
    {
      "epoch": 2.02,
      "learning_rate": 3.3832214572503975e-05,
      "loss": 0.1667,
      "step": 1215
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.370623061990255e-05,
      "loss": 0.1726,
      "step": 1220
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.357999455114148e-05,
      "loss": 0.1659,
      "step": 1225
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.3453510021781325e-05,
      "loss": 0.1694,
      "step": 1230
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.3326780694577506e-05,
      "loss": 0.1668,
      "step": 1235
    },
    {
      "epoch": 2.06,
      "learning_rate": 3.319981023937442e-05,
      "loss": 0.1715,
      "step": 1240
    },
    {
      "epoch": 2.07,
      "learning_rate": 3.3072602332999e-05,
      "loss": 0.1698,
      "step": 1245
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.294516065915438e-05,
      "loss": 0.174,
      "step": 1250
    },
    {
      "epoch": 2.08,
      "eval_loss": 0.11452513933181763,
      "eval_runtime": 6.5986,
      "eval_samples_per_second": 15.155,
      "eval_steps_per_second": 15.155,
      "step": 1250
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.2817488908313134e-05,
      "loss": 0.1546,
      "step": 1255
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.268959077761044e-05,
      "loss": 0.1739,
      "step": 1260
    },
    {
      "epoch": 2.1,
      "learning_rate": 3.256146997073702e-05,
      "loss": 0.1765,
      "step": 1265
    },
    {
      "epoch": 2.11,
      "learning_rate": 3.243313019783187e-05,
      "loss": 0.174,
      "step": 1270
    },
    {
      "epoch": 2.12,
      "learning_rate": 3.230457517537485e-05,
      "loss": 0.1673,
      "step": 1275
    },
    {
      "epoch": 2.12,
      "learning_rate": 3.217580862607904e-05,
      "loss": 0.1682,
      "step": 1280
    },
    {
      "epoch": 2.13,
      "learning_rate": 3.204683427878291e-05,
      "loss": 0.1641,
      "step": 1285
    },
    {
      "epoch": 2.14,
      "learning_rate": 3.191765586834243e-05,
      "loss": 0.1763,
      "step": 1290
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.178827713552281e-05,
      "loss": 0.1628,
      "step": 1295
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.165870182689024e-05,
      "loss": 0.1648,
      "step": 1300
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.152893369470337e-05,
      "loss": 0.1773,
      "step": 1305
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.139897649680466e-05,
      "loss": 0.1672,
      "step": 1310
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.126883399651158e-05,
      "loss": 0.1633,
      "step": 1315
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.1138509962507585e-05,
      "loss": 0.1543,
      "step": 1320
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.100800816873301e-05,
      "loss": 0.1667,
      "step": 1325
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.08773323942758e-05,
      "loss": 0.1568,
      "step": 1330
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.074648642326204e-05,
      "loss": 0.1622,
      "step": 1335
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.0615474044746386e-05,
      "loss": 0.1609,
      "step": 1340
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.0484299052602355e-05,
      "loss": 0.1647,
      "step": 1345
    },
    {
      "epoch": 2.24,
      "learning_rate": 3.0352965245412445e-05,
      "loss": 0.1775,
      "step": 1350
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.022147642635813e-05,
      "loss": 0.1762,
      "step": 1355
    },
    {
      "epoch": 2.26,
      "learning_rate": 3.0089836403109762e-05,
      "loss": 0.1603,
      "step": 1360
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.9958048987716265e-05,
      "loss": 0.1581,
      "step": 1365
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.9826117996494762e-05,
      "loss": 0.1643,
      "step": 1370
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.9694047249920092e-05,
      "loss": 0.1722,
      "step": 1375
    },
    {
      "epoch": 2.28,
      "eval_loss": 0.10874178260564804,
      "eval_runtime": 6.6047,
      "eval_samples_per_second": 15.141,
      "eval_steps_per_second": 15.141,
      "step": 1375
    },
    {
      "epoch": 2.29,
      "learning_rate": 2.95618405725141e-05,
      "loss": 0.1739,
      "step": 1380
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.9429501792734987e-05,
      "loss": 0.1703,
      "step": 1385
    },
    {
      "epoch": 2.31,
      "learning_rate": 2.9297034742866337e-05,
      "loss": 0.1615,
      "step": 1390
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.9164443258906245e-05,
      "loss": 0.1658,
      "step": 1395
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.9031731180456158e-05,
      "loss": 0.159,
      "step": 1400
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.889890235060972e-05,
      "loss": 0.1565,
      "step": 1405
    },
    {
      "epoch": 2.34,
      "learning_rate": 2.876596061584147e-05,
      "loss": 0.163,
      "step": 1410
    },
    {
      "epoch": 2.35,
      "learning_rate": 2.8632909825895477e-05,
      "loss": 0.1737,
      "step": 1415
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.8499753833673824e-05,
      "loss": 0.1633,
      "step": 1420
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.8366496495125096e-05,
      "loss": 0.1697,
      "step": 1425
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.8233141669132607e-05,
      "loss": 0.1669,
      "step": 1430
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.8099693217402806e-05,
      "loss": 0.1601,
      "step": 1435
    },
    {
      "epoch": 2.39,
      "learning_rate": 2.7966155004353306e-05,
      "loss": 0.1641,
      "step": 1440
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.783253089700108e-05,
      "loss": 0.1929,
      "step": 1445
    },
    {
      "epoch": 2.41,
      "learning_rate": 2.7698824764850408e-05,
      "loss": 0.1745,
      "step": 1450
    },
    {
      "epoch": 2.42,
      "learning_rate": 2.756504047978088e-05,
      "loss": 0.165,
      "step": 1455
    },
    {
      "epoch": 2.42,
      "learning_rate": 2.743118191593523e-05,
      "loss": 0.1598,
      "step": 1460
    },
    {
      "epoch": 2.43,
      "learning_rate": 2.7297252949607193e-05,
      "loss": 0.1579,
      "step": 1465
    },
    {
      "epoch": 2.44,
      "learning_rate": 2.716325745912918e-05,
      "loss": 0.1681,
      "step": 1470
    },
    {
      "epoch": 2.45,
      "learning_rate": 2.7029199324760078e-05,
      "loss": 0.1763,
      "step": 1475
    },
    {
      "epoch": 2.46,
      "learning_rate": 2.6895082428572776e-05,
      "loss": 0.1653,
      "step": 1480
    },
    {
      "epoch": 2.47,
      "learning_rate": 2.6760910654341816e-05,
      "loss": 0.1564,
      "step": 1485
    },
    {
      "epoch": 2.47,
      "learning_rate": 2.66266878874309e-05,
      "loss": 0.1587,
      "step": 1490
    },
    {
      "epoch": 2.48,
      "learning_rate": 2.6492418014680385e-05,
      "loss": 0.1532,
      "step": 1495
    },
    {
      "epoch": 2.49,
      "learning_rate": 2.635810492429473e-05,
      "loss": 0.1717,
      "step": 1500
    },
    {
      "epoch": 2.49,
      "eval_loss": 0.10279469192028046,
      "eval_runtime": 6.5822,
      "eval_samples_per_second": 15.193,
      "eval_steps_per_second": 15.193,
      "step": 1500
    },
    {
      "epoch": 2.5,
      "learning_rate": 2.622375250572988e-05,
      "loss": 0.1849,
      "step": 1505
    },
    {
      "epoch": 2.51,
      "learning_rate": 2.6089364649580676e-05,
      "loss": 0.1573,
      "step": 1510
    },
    {
      "epoch": 2.52,
      "learning_rate": 2.5954945247468137e-05,
      "loss": 0.1748,
      "step": 1515
    },
    {
      "epoch": 2.52,
      "learning_rate": 2.5820498191926816e-05,
      "loss": 0.1663,
      "step": 1520
    },
    {
      "epoch": 2.53,
      "learning_rate": 2.5686027376292037e-05,
      "loss": 0.1678,
      "step": 1525
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.5551536694587203e-05,
      "loss": 0.1633,
      "step": 1530
    },
    {
      "epoch": 2.55,
      "learning_rate": 2.541703004141096e-05,
      "loss": 0.1543,
      "step": 1535
    },
    {
      "epoch": 2.56,
      "learning_rate": 2.5282511311824497e-05,
      "loss": 0.1719,
      "step": 1540
    },
    {
      "epoch": 2.56,
      "learning_rate": 2.514798440123869e-05,
      "loss": 0.1532,
      "step": 1545
    },
    {
      "epoch": 2.57,
      "learning_rate": 2.501345320530133e-05,
      "loss": 0.1539,
      "step": 1550
    },
    {
      "epoch": 2.58,
      "learning_rate": 2.4878921619784295e-05,
      "loss": 0.1671,
      "step": 1555
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.474439354047076e-05,
      "loss": 0.1702,
      "step": 1560
    },
    {
      "epoch": 2.6,
      "learning_rate": 2.4609872863042344e-05,
      "loss": 0.1835,
      "step": 1565
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.4475363482966353e-05,
      "loss": 0.1621,
      "step": 1570
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.4340869295382924e-05,
      "loss": 0.1673,
      "step": 1575
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.4206394194992232e-05,
      "loss": 0.1648,
      "step": 1580
    },
    {
      "epoch": 2.63,
      "learning_rate": 2.4071942075941746e-05,
      "loss": 0.162,
      "step": 1585
    },
    {
      "epoch": 2.64,
      "learning_rate": 2.3937516831713438e-05,
      "loss": 0.1881,
      "step": 1590
    },
    {
      "epoch": 2.65,
      "learning_rate": 2.380312235501102e-05,
      "loss": 0.1662,
      "step": 1595
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.3668762537647254e-05,
      "loss": 0.1641,
      "step": 1600
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.3534441270431225e-05,
      "loss": 0.16,
      "step": 1605
    },
    {
      "epoch": 2.67,
      "learning_rate": 2.3400162443055656e-05,
      "loss": 0.1682,
      "step": 1610
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.3265929943984314e-05,
      "loss": 0.1821,
      "step": 1615
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.3131747660339394e-05,
      "loss": 0.1728,
      "step": 1620
    },
    {
      "epoch": 2.7,
      "learning_rate": 2.2997619477788924e-05,
      "loss": 0.1596,
      "step": 1625
    },
    {
      "epoch": 2.7,
      "eval_loss": 0.0988275557756424,
      "eval_runtime": 6.5679,
      "eval_samples_per_second": 15.226,
      "eval_steps_per_second": 15.226,
      "step": 1625
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.2863549280434282e-05,
      "loss": 0.1636,
      "step": 1630
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.2729540950697677e-05,
      "loss": 0.1544,
      "step": 1635
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.2595598369209782e-05,
      "loss": 0.1867,
      "step": 1640
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.2461725414697303e-05,
      "loss": 0.1592,
      "step": 1645
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.232792596387068e-05,
      "loss": 0.1583,
      "step": 1650
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.2194203891311842e-05,
      "loss": 0.1569,
      "step": 1655
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.2060563069361955e-05,
      "loss": 0.1698,
      "step": 1660
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.1927007368009345e-05,
      "loss": 0.1771,
      "step": 1665
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.17935406547774e-05,
      "loss": 0.1644,
      "step": 1670
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.1660166794612575e-05,
      "loss": 0.1586,
      "step": 1675
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.1526889649772478e-05,
      "loss": 0.172,
      "step": 1680
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.1393713079714e-05,
      "loss": 0.1558,
      "step": 1685
    },
    {
      "epoch": 2.81,
      "learning_rate": 2.1260640940981617e-05,
      "loss": 0.168,
      "step": 1690
    },
    {
      "epoch": 2.81,
      "learning_rate": 2.1127677087095632e-05,
      "loss": 0.1763,
      "step": 1695
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.099482536844063e-05,
      "loss": 0.181,
      "step": 1700
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.0862089632153988e-05,
      "loss": 0.1656,
      "step": 1705
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.0729473722014436e-05,
      "loss": 0.1723,
      "step": 1710
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.0596981478330748e-05,
      "loss": 0.1516,
      "step": 1715
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.0464616737830567e-05,
      "loss": 0.1763,
      "step": 1720
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.0332383333549288e-05,
      "loss": 0.1791,
      "step": 1725
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.020028509471904e-05,
      "loss": 0.1577,
      "step": 1730
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.006832584665783e-05,
      "loss": 0.1635,
      "step": 1735
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.9936509410658723e-05,
      "loss": 0.1613,
      "step": 1740
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.9804839603879238e-05,
      "loss": 0.1557,
      "step": 1745
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.967332023923078e-05,
      "loss": 0.1732,
      "step": 1750
    },
    {
      "epoch": 2.91,
      "eval_loss": 0.09610436111688614,
      "eval_runtime": 6.569,
      "eval_samples_per_second": 15.223,
      "eval_steps_per_second": 15.223,
      "step": 1750
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.9541955125268235e-05,
      "loss": 0.1803,
      "step": 1755
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.941074806607968e-05,
      "loss": 0.1544,
      "step": 1760
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.9279702861176202e-05,
      "loss": 0.1784,
      "step": 1765
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.9148823305381898e-05,
      "loss": 0.1631,
      "step": 1770
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.9018113188723998e-05,
      "loss": 0.1675,
      "step": 1775
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.888757629632307e-05,
      "loss": 0.1702,
      "step": 1780
    },
    {
      "epoch": 2.96,
      "learning_rate": 1.875721640828344e-05,
      "loss": 0.1641,
      "step": 1785
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.8627037299583703e-05,
      "loss": 0.1687,
      "step": 1790
    },
    {
      "epoch": 2.98,
      "learning_rate": 1.8497042739967445e-05,
      "loss": 0.1706,
      "step": 1795
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.836723649383404e-05,
      "loss": 0.1575,
      "step": 1800
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.8237622320129665e-05,
      "loss": 0.1608,
      "step": 1805
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.8108203972238436e-05,
      "loss": 0.1506,
      "step": 1810
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.797898519787372e-05,
      "loss": 0.134,
      "step": 1815
    },
    {
      "epoch": 3.02,
      "learning_rate": 1.784996973896959e-05,
      "loss": 0.148,
      "step": 1820
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.7721161331572523e-05,
      "loss": 0.1457,
      "step": 1825
    },
    {
      "epoch": 3.04,
      "learning_rate": 1.759256370573313e-05,
      "loss": 0.1503,
      "step": 1830
    },
    {
      "epoch": 3.05,
      "learning_rate": 1.7464180585398205e-05,
      "loss": 0.142,
      "step": 1835
    },
    {
      "epoch": 3.05,
      "learning_rate": 1.733601568830287e-05,
      "loss": 0.1492,
      "step": 1840
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.720807272586288e-05,
      "loss": 0.1662,
      "step": 1845
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.708035540306721e-05,
      "loss": 0.1461,
      "step": 1850
    },
    {
      "epoch": 3.08,
      "learning_rate": 1.695286741837071e-05,
      "loss": 0.1496,
      "step": 1855
    },
    {
      "epoch": 3.09,
      "learning_rate": 1.6825612463587044e-05,
      "loss": 0.1448,
      "step": 1860
    },
    {
      "epoch": 3.1,
      "learning_rate": 1.669859422378176e-05,
      "loss": 0.1588,
      "step": 1865
    },
    {
      "epoch": 3.1,
      "learning_rate": 1.6571816377165567e-05,
      "loss": 0.153,
      "step": 1870
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.6445282594987862e-05,
      "loss": 0.1511,
      "step": 1875
    },
    {
      "epoch": 3.11,
      "eval_loss": 0.0892980769276619,
      "eval_runtime": 6.5864,
      "eval_samples_per_second": 15.183,
      "eval_steps_per_second": 15.183,
      "step": 1875
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.631899654143038e-05,
      "loss": 0.1455,
      "step": 1880
    },
    {
      "epoch": 3.13,
      "learning_rate": 1.6192961873501096e-05,
      "loss": 0.141,
      "step": 1885
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.606718224092833e-05,
      "loss": 0.1423,
      "step": 1890
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.5941661286055068e-05,
      "loss": 0.1583,
      "step": 1895
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.581640264373344e-05,
      "loss": 0.138,
      "step": 1900
    },
    {
      "epoch": 3.16,
      "learning_rate": 1.5691409941219526e-05,
      "loss": 0.1543,
      "step": 1905
    },
    {
      "epoch": 3.17,
      "learning_rate": 1.556668679806828e-05,
      "loss": 0.1432,
      "step": 1910
    },
    {
      "epoch": 3.18,
      "learning_rate": 1.5442236826028706e-05,
      "loss": 0.146,
      "step": 1915
    },
    {
      "epoch": 3.19,
      "learning_rate": 1.531806362893932e-05,
      "loss": 0.1438,
      "step": 1920
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.5194170802623691e-05,
      "loss": 0.1331,
      "step": 1925
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.5070561934786437e-05,
      "loss": 0.1517,
      "step": 1930
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.4947240604909232e-05,
      "loss": 0.1432,
      "step": 1935
    },
    {
      "epoch": 3.22,
      "learning_rate": 1.4824210384147207e-05,
      "loss": 0.1499,
      "step": 1940
    },
    {
      "epoch": 3.23,
      "learning_rate": 1.4701474835225515e-05,
      "loss": 0.1368,
      "step": 1945
    },
    {
      "epoch": 3.24,
      "learning_rate": 1.4579037512336138e-05,
      "loss": 0.1385,
      "step": 1950
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.445690196103503e-05,
      "loss": 0.146,
      "step": 1955
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.4335071718139378e-05,
      "loss": 0.1489,
      "step": 1960
    },
    {
      "epoch": 3.26,
      "learning_rate": 1.4213550311625216e-05,
      "loss": 0.1434,
      "step": 1965
    },
    {
      "epoch": 3.27,
      "learning_rate": 1.4092341260525248e-05,
      "loss": 0.1645,
      "step": 1970
    },
    {
      "epoch": 3.28,
      "learning_rate": 1.397144807482696e-05,
      "loss": 0.1562,
      "step": 1975
    },
    {
      "epoch": 3.29,
      "learning_rate": 1.385087425537096e-05,
      "loss": 0.1426,
      "step": 1980
    },
    {
      "epoch": 3.3,
      "learning_rate": 1.3730623293749606e-05,
      "loss": 0.1546,
      "step": 1985
    },
    {
      "epoch": 3.3,
      "learning_rate": 1.3610698672205895e-05,
      "loss": 0.1568,
      "step": 1990
    },
    {
      "epoch": 3.31,
      "learning_rate": 1.3491103863532626e-05,
      "loss": 0.1401,
      "step": 1995
    },
    {
      "epoch": 3.32,
      "learning_rate": 1.3371842330971862e-05,
      "loss": 0.1417,
      "step": 2000
    },
    {
      "epoch": 3.32,
      "eval_loss": 0.08844374865293503,
      "eval_runtime": 6.6133,
      "eval_samples_per_second": 15.121,
      "eval_steps_per_second": 15.121,
      "step": 2000
    },
    {
      "epoch": 3.33,
      "learning_rate": 1.3252917528114556e-05,
      "loss": 0.1392,
      "step": 2005
    },
    {
      "epoch": 3.34,
      "learning_rate": 1.3134332898800645e-05,
      "loss": 0.1457,
      "step": 2010
    },
    {
      "epoch": 3.35,
      "learning_rate": 1.3016091877019248e-05,
      "loss": 0.1524,
      "step": 2015
    },
    {
      "epoch": 3.35,
      "learning_rate": 1.2898197886809287e-05,
      "loss": 0.1538,
      "step": 2020
    },
    {
      "epoch": 3.36,
      "learning_rate": 1.2780654342160277e-05,
      "loss": 0.1491,
      "step": 2025
    },
    {
      "epoch": 3.37,
      "learning_rate": 1.2663464646913458e-05,
      "loss": 0.1402,
      "step": 2030
    },
    {
      "epoch": 3.38,
      "learning_rate": 1.2546632194663305e-05,
      "loss": 0.1525,
      "step": 2035
    },
    {
      "epoch": 3.39,
      "learning_rate": 1.2430160368659171e-05,
      "loss": 0.1396,
      "step": 2040
    },
    {
      "epoch": 3.39,
      "learning_rate": 1.2314052541707355e-05,
      "loss": 0.151,
      "step": 2045
    },
    {
      "epoch": 3.4,
      "learning_rate": 1.2198312076073426e-05,
      "loss": 0.1485,
      "step": 2050
    },
    {
      "epoch": 3.41,
      "learning_rate": 1.2082942323384852e-05,
      "loss": 0.1523,
      "step": 2055
    },
    {
      "epoch": 3.42,
      "learning_rate": 1.1967946624533948e-05,
      "loss": 0.141,
      "step": 2060
    },
    {
      "epoch": 3.43,
      "learning_rate": 1.1853328309581139e-05,
      "loss": 0.1451,
      "step": 2065
    },
    {
      "epoch": 3.44,
      "learning_rate": 1.1739090697658497e-05,
      "loss": 0.1636,
      "step": 2070
    },
    {
      "epoch": 3.44,
      "learning_rate": 1.162523709687367e-05,
      "loss": 0.1564,
      "step": 2075
    },
    {
      "epoch": 3.45,
      "learning_rate": 1.1511770804214044e-05,
      "loss": 0.143,
      "step": 2080
    },
    {
      "epoch": 3.46,
      "learning_rate": 1.1398695105451298e-05,
      "loss": 0.1379,
      "step": 2085
    },
    {
      "epoch": 3.47,
      "learning_rate": 1.1286013275046232e-05,
      "loss": 0.1386,
      "step": 2090
    },
    {
      "epoch": 3.48,
      "learning_rate": 1.1173728576053949e-05,
      "loss": 0.145,
      "step": 2095
    },
    {
      "epoch": 3.49,
      "learning_rate": 1.10618442600294e-05,
      "loss": 0.148,
      "step": 2100
    },
    {
      "epoch": 3.49,
      "learning_rate": 1.0950363566933172e-05,
      "loss": 0.1501,
      "step": 2105
    },
    {
      "epoch": 3.5,
      "learning_rate": 1.0839289725037668e-05,
      "loss": 0.1774,
      "step": 2110
    },
    {
      "epoch": 3.51,
      "learning_rate": 1.0728625950833665e-05,
      "loss": 0.1456,
      "step": 2115
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.0618375448937157e-05,
      "loss": 0.1518,
      "step": 2120
    },
    {
      "epoch": 3.53,
      "learning_rate": 1.050854141199652e-05,
      "loss": 0.1581,
      "step": 2125
    },
    {
      "epoch": 3.53,
      "eval_loss": 0.08569944649934769,
      "eval_runtime": 6.6352,
      "eval_samples_per_second": 15.071,
      "eval_steps_per_second": 15.071,
      "step": 2125
    },
    {
      "epoch": 3.54,
      "learning_rate": 1.0399127020600103e-05,
      "loss": 0.1498,
      "step": 2130
    },
    {
      "epoch": 3.54,
      "learning_rate": 1.0290135443184068e-05,
      "loss": 0.145,
      "step": 2135
    },
    {
      "epoch": 3.55,
      "learning_rate": 1.018156983594073e-05,
      "loss": 0.1418,
      "step": 2140
    },
    {
      "epoch": 3.56,
      "learning_rate": 1.0073433342727069e-05,
      "loss": 0.1575,
      "step": 2145
    },
    {
      "epoch": 3.57,
      "learning_rate": 9.965729094973725e-06,
      "loss": 0.1474,
      "step": 2150
    },
    {
      "epoch": 3.58,
      "learning_rate": 9.85846021159433e-06,
      "loss": 0.1456,
      "step": 2155
    },
    {
      "epoch": 3.59,
      "learning_rate": 9.751629798895173e-06,
      "loss": 0.1369,
      "step": 2160
    },
    {
      "epoch": 3.59,
      "learning_rate": 9.645240950485248e-06,
      "loss": 0.1391,
      "step": 2165
    },
    {
      "epoch": 3.6,
      "learning_rate": 9.53929674718668e-06,
      "loss": 0.1429,
      "step": 2170
    },
    {
      "epoch": 3.61,
      "learning_rate": 9.433800256945499e-06,
      "loss": 0.1555,
      "step": 2175
    },
    {
      "epoch": 3.62,
      "learning_rate": 9.328754534742793e-06,
      "loss": 0.1435,
      "step": 2180
    },
    {
      "epoch": 3.63,
      "learning_rate": 9.224162622506283e-06,
      "loss": 0.1709,
      "step": 2185
    },
    {
      "epoch": 3.64,
      "learning_rate": 9.120027549022156e-06,
      "loss": 0.1467,
      "step": 2190
    },
    {
      "epoch": 3.64,
      "learning_rate": 9.016352329847433e-06,
      "loss": 0.1572,
      "step": 2195
    },
    {
      "epoch": 3.65,
      "learning_rate": 8.9131399672226e-06,
      "loss": 0.1394,
      "step": 2200
    },
    {
      "epoch": 3.66,
      "learning_rate": 8.810393449984705e-06,
      "loss": 0.1266,
      "step": 2205
    },
    {
      "epoch": 3.67,
      "learning_rate": 8.708115753480769e-06,
      "loss": 0.152,
      "step": 2210
    },
    {
      "epoch": 3.68,
      "learning_rate": 8.606309839481627e-06,
      "loss": 0.147,
      "step": 2215
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.504978656096185e-06,
      "loss": 0.1555,
      "step": 2220
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.404125137686061e-06,
      "loss": 0.1414,
      "step": 2225
    },
    {
      "epoch": 3.7,
      "learning_rate": 8.303752204780563e-06,
      "loss": 0.1362,
      "step": 2230
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.203862763992153e-06,
      "loss": 0.142,
      "step": 2235
    },
    {
      "epoch": 3.72,
      "learning_rate": 8.104459707932237e-06,
      "loss": 0.1435,
      "step": 2240
    },
    {
      "epoch": 3.73,
      "learning_rate": 8.005545915127479e-06,
      "loss": 0.1547,
      "step": 2245
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.907124249936356e-06,
      "loss": 0.1463,
      "step": 2250
    },
    {
      "epoch": 3.74,
      "eval_loss": 0.08357314765453339,
      "eval_runtime": 6.642,
      "eval_samples_per_second": 15.056,
      "eval_steps_per_second": 15.056,
      "step": 2250
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.809197562466264e-06,
      "loss": 0.144,
      "step": 2255
    },
    {
      "epoch": 3.75,
      "learning_rate": 7.711768688490962e-06,
      "loss": 0.1512,
      "step": 2260
    },
    {
      "epoch": 3.76,
      "learning_rate": 7.614840449368469e-06,
      "loss": 0.156,
      "step": 2265
    },
    {
      "epoch": 3.77,
      "learning_rate": 7.518415651959349e-06,
      "loss": 0.1491,
      "step": 2270
    },
    {
      "epoch": 3.78,
      "learning_rate": 7.422497088545435e-06,
      "loss": 0.1508,
      "step": 2275
    },
    {
      "epoch": 3.79,
      "learning_rate": 7.32708753674897e-06,
      "loss": 0.1321,
      "step": 2280
    },
    {
      "epoch": 3.79,
      "learning_rate": 7.2321897594521875e-06,
      "loss": 0.1382,
      "step": 2285
    },
    {
      "epoch": 3.8,
      "learning_rate": 7.137806504717276e-06,
      "loss": 0.1567,
      "step": 2290
    },
    {
      "epoch": 3.81,
      "learning_rate": 7.0439405057068e-06,
      "loss": 0.1428,
      "step": 2295
    },
    {
      "epoch": 3.82,
      "learning_rate": 6.950594480604572e-06,
      "loss": 0.149,
      "step": 2300
    },
    {
      "epoch": 3.83,
      "learning_rate": 6.8577711325369584e-06,
      "loss": 0.1476,
      "step": 2305
    },
    {
      "epoch": 3.83,
      "learning_rate": 6.765473149494544e-06,
      "loss": 0.1463,
      "step": 2310
    },
    {
      "epoch": 3.84,
      "learning_rate": 6.673703204254347e-06,
      "loss": 0.1459,
      "step": 2315
    },
    {
      "epoch": 3.85,
      "learning_rate": 6.5824639543023674e-06,
      "loss": 0.1573,
      "step": 2320
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.491758041756701e-06,
      "loss": 0.1415,
      "step": 2325
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.401588093290964e-06,
      "loss": 0.1464,
      "step": 2330
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.311956720058263e-06,
      "loss": 0.1644,
      "step": 2335
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.222866517615569e-06,
      "loss": 0.145,
      "step": 2340
    },
    {
      "epoch": 3.89,
      "learning_rate": 6.1343200658485645e-06,
      "loss": 0.1371,
      "step": 2345
    },
    {
      "epoch": 3.9,
      "learning_rate": 6.046319928896926e-06,
      "loss": 0.1406,
      "step": 2350
    },
    {
      "epoch": 3.91,
      "learning_rate": 5.958868655080077e-06,
      "loss": 0.1483,
      "step": 2355
    },
    {
      "epoch": 3.92,
      "learning_rate": 5.871968776823383e-06,
      "loss": 0.1397,
      "step": 2360
    },
    {
      "epoch": 3.93,
      "learning_rate": 5.785622810584834e-06,
      "loss": 0.1467,
      "step": 2365
    },
    {
      "epoch": 3.93,
      "learning_rate": 5.699833256782161e-06,
      "loss": 0.1545,
      "step": 2370
    },
    {
      "epoch": 3.94,
      "learning_rate": 5.614602599720431e-06,
      "loss": 0.1487,
      "step": 2375
    },
    {
      "epoch": 3.94,
      "eval_loss": 0.08115073293447495,
      "eval_runtime": 6.6449,
      "eval_samples_per_second": 15.049,
      "eval_steps_per_second": 15.049,
      "step": 2375
    },
    {
      "epoch": 3.95,
      "learning_rate": 5.529933307520102e-06,
      "loss": 0.156,
      "step": 2380
    },
    {
      "epoch": 3.96,
      "learning_rate": 5.445827832045558e-06,
      "loss": 0.1516,
      "step": 2385
    },
    {
      "epoch": 3.97,
      "learning_rate": 5.362288608834115e-06,
      "loss": 0.1514,
      "step": 2390
    },
    {
      "epoch": 3.98,
      "learning_rate": 5.2793180570254755e-06,
      "loss": 0.1449,
      "step": 2395
    },
    {
      "epoch": 3.98,
      "learning_rate": 5.196918579291668e-06,
      "loss": 0.1507,
      "step": 2400
    },
    {
      "epoch": 3.99,
      "learning_rate": 5.11509256176749e-06,
      "loss": 0.1579,
      "step": 2405
    },
    {
      "epoch": 4.0,
      "learning_rate": 5.0338423739814285e-06,
      "loss": 0.1376,
      "step": 2410
    },
    {
      "epoch": 4.01,
      "learning_rate": 4.9531703687869855e-06,
      "loss": 0.1462,
      "step": 2415
    },
    {
      "epoch": 4.02,
      "learning_rate": 4.873078882294599e-06,
      "loss": 0.1319,
      "step": 2420
    },
    {
      "epoch": 4.03,
      "learning_rate": 4.79357023380394e-06,
      "loss": 0.1369,
      "step": 2425
    },
    {
      "epoch": 4.03,
      "learning_rate": 4.714646725736821e-06,
      "loss": 0.1475,
      "step": 2430
    },
    {
      "epoch": 4.04,
      "learning_rate": 4.636310643570454e-06,
      "loss": 0.1369,
      "step": 2435
    },
    {
      "epoch": 4.05,
      "learning_rate": 4.5585642557713085e-06,
      "loss": 0.1303,
      "step": 2440
    },
    {
      "epoch": 4.06,
      "learning_rate": 4.481409813729398e-06,
      "loss": 0.1395,
      "step": 2445
    },
    {
      "epoch": 4.07,
      "learning_rate": 4.404849551693102e-06,
      "loss": 0.1342,
      "step": 2450
    },
    {
      "epoch": 4.08,
      "learning_rate": 4.328885686704456e-06,
      "loss": 0.1372,
      "step": 2455
    },
    {
      "epoch": 4.08,
      "learning_rate": 4.2535204185349455e-06,
      "loss": 0.1436,
      "step": 2460
    },
    {
      "epoch": 4.09,
      "learning_rate": 4.178755929621822e-06,
      "loss": 0.129,
      "step": 2465
    },
    {
      "epoch": 4.1,
      "learning_rate": 4.104594385004884e-06,
      "loss": 0.1378,
      "step": 2470
    },
    {
      "epoch": 4.11,
      "learning_rate": 4.031037932263798e-06,
      "loss": 0.1472,
      "step": 2475
    },
    {
      "epoch": 4.12,
      "learning_rate": 3.958088701455895e-06,
      "loss": 0.1155,
      "step": 2480
    },
    {
      "epoch": 4.13,
      "learning_rate": 3.88574880505449e-06,
      "loss": 0.1397,
      "step": 2485
    },
    {
      "epoch": 4.13,
      "learning_rate": 3.814020337887736e-06,
      "loss": 0.1313,
      "step": 2490
    },
    {
      "epoch": 4.14,
      "learning_rate": 3.7429053770779127e-06,
      "loss": 0.1342,
      "step": 2495
    },
    {
      "epoch": 4.15,
      "learning_rate": 3.672405981981322e-06,
      "loss": 0.1453,
      "step": 2500
    },
    {
      "epoch": 4.15,
      "eval_loss": 0.07911655306816101,
      "eval_runtime": 6.6443,
      "eval_samples_per_second": 15.051,
      "eval_steps_per_second": 15.051,
      "step": 2500
    },
    {
      "epoch": 4.16,
      "learning_rate": 3.602524194128601e-06,
      "loss": 0.1416,
      "step": 2505
    },
    {
      "epoch": 4.17,
      "learning_rate": 3.5332620371656826e-06,
      "loss": 0.1426,
      "step": 2510
    },
    {
      "epoch": 4.18,
      "learning_rate": 3.464621516795116e-06,
      "loss": 0.148,
      "step": 2515
    },
    {
      "epoch": 4.18,
      "learning_rate": 3.396604620718025e-06,
      "loss": 0.1457,
      "step": 2520
    },
    {
      "epoch": 4.19,
      "learning_rate": 3.3292133185765407e-06,
      "loss": 0.1431,
      "step": 2525
    },
    {
      "epoch": 4.2,
      "learning_rate": 3.262449561896766e-06,
      "loss": 0.1398,
      "step": 2530
    },
    {
      "epoch": 4.21,
      "learning_rate": 3.19631528403225e-06,
      "loss": 0.1401,
      "step": 2535
    },
    {
      "epoch": 4.22,
      "learning_rate": 3.1308124001080296e-06,
      "loss": 0.1478,
      "step": 2540
    },
    {
      "epoch": 4.22,
      "learning_rate": 3.0659428069651388e-06,
      "loss": 0.1401,
      "step": 2545
    },
    {
      "epoch": 4.23,
      "learning_rate": 3.0017083831057004e-06,
      "loss": 0.1376,
      "step": 2550
    },
    {
      "epoch": 4.24,
      "learning_rate": 2.9381109886385205e-06,
      "loss": 0.1448,
      "step": 2555
    },
    {
      "epoch": 4.25,
      "learning_rate": 2.8751524652252335e-06,
      "loss": 0.1431,
      "step": 2560
    },
    {
      "epoch": 4.26,
      "learning_rate": 2.812834636026948e-06,
      "loss": 0.1484,
      "step": 2565
    },
    {
      "epoch": 4.27,
      "learning_rate": 2.751159305651477e-06,
      "loss": 0.1597,
      "step": 2570
    },
    {
      "epoch": 4.27,
      "learning_rate": 2.6901282601010687e-06,
      "loss": 0.1329,
      "step": 2575
    },
    {
      "epoch": 4.28,
      "learning_rate": 2.6297432667206912e-06,
      "loss": 0.1423,
      "step": 2580
    },
    {
      "epoch": 4.29,
      "learning_rate": 2.5700060741468274e-06,
      "loss": 0.1391,
      "step": 2585
    },
    {
      "epoch": 4.3,
      "learning_rate": 2.5109184122568795e-06,
      "loss": 0.1423,
      "step": 2590
    },
    {
      "epoch": 4.31,
      "learning_rate": 2.4524819921190654e-06,
      "loss": 0.1452,
      "step": 2595
    },
    {
      "epoch": 4.32,
      "learning_rate": 2.394698505942833e-06,
      "loss": 0.1364,
      "step": 2600
    },
    {
      "epoch": 4.32,
      "learning_rate": 2.337569627029909e-06,
      "loss": 0.1422,
      "step": 2605
    },
    {
      "epoch": 4.33,
      "learning_rate": 2.2810970097257913e-06,
      "loss": 0.1406,
      "step": 2610
    },
    {
      "epoch": 4.34,
      "learning_rate": 2.2252822893718912e-06,
      "loss": 0.1397,
      "step": 2615
    },
    {
      "epoch": 4.35,
      "learning_rate": 2.1701270822581472e-06,
      "loss": 0.1281,
      "step": 2620
    },
    {
      "epoch": 4.36,
      "learning_rate": 2.115632985576224e-06,
      "loss": 0.1457,
      "step": 2625
    },
    {
      "epoch": 4.36,
      "eval_loss": 0.07916607707738876,
      "eval_runtime": 6.6471,
      "eval_samples_per_second": 15.044,
      "eval_steps_per_second": 15.044,
      "step": 2625
    },
    {
      "epoch": 4.37,
      "learning_rate": 2.061801577373268e-06,
      "loss": 0.1481,
      "step": 2630
    },
    {
      "epoch": 4.37,
      "learning_rate": 2.0086344165062056e-06,
      "loss": 0.1433,
      "step": 2635
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.956133042596603e-06,
      "loss": 0.1302,
      "step": 2640
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.904298975986077e-06,
      "loss": 0.1364,
      "step": 2645
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.853133717692279e-06,
      "loss": 0.146,
      "step": 2650
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.8026387493654206e-06,
      "loss": 0.1277,
      "step": 2655
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.752815533245364e-06,
      "loss": 0.1367,
      "step": 2660
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.7036655121192874e-06,
      "loss": 0.1422,
      "step": 2665
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.6551901092799027e-06,
      "loss": 0.1442,
      "step": 2670
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.6073907284842315e-06,
      "loss": 0.135,
      "step": 2675
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.5602687539129746e-06,
      "loss": 0.1372,
      "step": 2680
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.5138255501304015e-06,
      "loss": 0.1296,
      "step": 2685
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.4680624620448419e-06,
      "loss": 0.1421,
      "step": 2690
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.4229808148697733e-06,
      "loss": 0.1319,
      "step": 2695
    },
    {
      "epoch": 4.48,
      "learning_rate": 1.3785819140853973e-06,
      "loss": 0.1338,
      "step": 2700
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.3348670454008589e-06,
      "loss": 0.1387,
      "step": 2705
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.2918374747170225e-06,
      "loss": 0.1414,
      "step": 2710
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.249494448089794e-06,
      "loss": 0.1473,
      "step": 2715
    },
    {
      "epoch": 4.52,
      "learning_rate": 1.207839191694049e-06,
      "loss": 0.1453,
      "step": 2720
    },
    {
      "epoch": 4.52,
      "learning_rate": 1.1668729117881249e-06,
      "loss": 0.13,
      "step": 2725
    },
    {
      "epoch": 4.53,
      "learning_rate": 1.1265967946788914e-06,
      "loss": 0.1335,
      "step": 2730
    },
    {
      "epoch": 4.54,
      "learning_rate": 1.0870120066873878e-06,
      "loss": 0.1368,
      "step": 2735
    },
    {
      "epoch": 4.55,
      "learning_rate": 1.0481196941150572e-06,
      "loss": 0.139,
      "step": 2740
    },
    {
      "epoch": 4.56,
      "learning_rate": 1.0099209832105521e-06,
      "loss": 0.1417,
      "step": 2745
    },
    {
      "epoch": 4.57,
      "learning_rate": 9.724169801371168e-07,
      "loss": 0.1429,
      "step": 2750
    },
    {
      "epoch": 4.57,
      "eval_loss": 0.07882240414619446,
      "eval_runtime": 6.6515,
      "eval_samples_per_second": 15.034,
      "eval_steps_per_second": 15.034,
      "step": 2750
    },
    {
      "epoch": 4.57,
      "learning_rate": 9.356087709405465e-07,
      "loss": 0.1244,
      "step": 2755
    },
    {
      "epoch": 4.58,
      "learning_rate": 8.994974215177642e-07,
      "loss": 0.1291,
      "step": 2760
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.640839775859221e-07,
      "loss": 0.1393,
      "step": 2765
    },
    {
      "epoch": 4.6,
      "learning_rate": 8.293694646521411e-07,
      "loss": 0.1441,
      "step": 2770
    },
    {
      "epoch": 4.61,
      "learning_rate": 7.953548879838013e-07,
      "loss": 0.1438,
      "step": 2775
    },
    {
      "epoch": 4.62,
      "learning_rate": 7.620412325794468e-07,
      "loss": 0.1533,
      "step": 2780
    },
    {
      "epoch": 4.62,
      "learning_rate": 7.294294631402493e-07,
      "loss": 0.1251,
      "step": 2785
    },
    {
      "epoch": 4.63,
      "learning_rate": 6.975205240420696e-07,
      "loss": 0.1355,
      "step": 2790
    },
    {
      "epoch": 4.64,
      "learning_rate": 6.663153393081184e-07,
      "loss": 0.1264,
      "step": 2795
    },
    {
      "epoch": 4.65,
      "learning_rate": 6.358148125822e-07,
      "loss": 0.1417,
      "step": 2800
    },
    {
      "epoch": 4.66,
      "learning_rate": 6.060198271025358e-07,
      "loss": 0.1311,
      "step": 2805
    },
    {
      "epoch": 4.66,
      "learning_rate": 5.769312456761989e-07,
      "loss": 0.1433,
      "step": 2810
    },
    {
      "epoch": 4.67,
      "learning_rate": 5.485499106541086e-07,
      "loss": 0.1417,
      "step": 2815
    },
    {
      "epoch": 4.68,
      "learning_rate": 5.208766439066676e-07,
      "loss": 0.1364,
      "step": 2820
    },
    {
      "epoch": 4.69,
      "learning_rate": 4.939122467999297e-07,
      "loss": 0.136,
      "step": 2825
    },
    {
      "epoch": 4.7,
      "learning_rate": 4.6765750017242206e-07,
      "loss": 0.1257,
      "step": 2830
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.421131643125104e-07,
      "loss": 0.1329,
      "step": 2835
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.172799789363996e-07,
      "loss": 0.1283,
      "step": 2840
    },
    {
      "epoch": 4.72,
      "learning_rate": 3.931586631667039e-07,
      "loss": 0.144,
      "step": 2845
    },
    {
      "epoch": 4.73,
      "learning_rate": 3.697499155116302e-07,
      "loss": 0.1349,
      "step": 2850
    },
    {
      "epoch": 4.74,
      "learning_rate": 3.4705441384473293e-07,
      "loss": 0.1394,
      "step": 2855
    },
    {
      "epoch": 4.75,
      "learning_rate": 3.25072815385305e-07,
      "loss": 0.1388,
      "step": 2860
    },
    {
      "epoch": 4.76,
      "learning_rate": 3.0380575667934285e-07,
      "loss": 0.1254,
      "step": 2865
    },
    {
      "epoch": 4.76,
      "learning_rate": 2.8325385358109467e-07,
      "loss": 0.1394,
      "step": 2870
    },
    {
      "epoch": 4.77,
      "learning_rate": 2.6341770123524676e-07,
      "loss": 0.1376,
      "step": 2875
    },
    {
      "epoch": 4.77,
      "eval_loss": 0.07855534553527832,
      "eval_runtime": 6.664,
      "eval_samples_per_second": 15.006,
      "eval_steps_per_second": 15.006,
      "step": 2875
    },
    {
      "epoch": 4.78,
      "learning_rate": 2.442978740596791e-07,
      "loss": 0.1286,
      "step": 2880
    },
    {
      "epoch": 4.79,
      "learning_rate": 2.258949257288423e-07,
      "loss": 0.1273,
      "step": 2885
    },
    {
      "epoch": 4.8,
      "learning_rate": 2.0820938915770416e-07,
      "loss": 0.1382,
      "step": 2890
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.9124177648633667e-07,
      "loss": 0.1438,
      "step": 2895
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.7499257906507526e-07,
      "loss": 0.1409,
      "step": 2900
    },
    {
      "epoch": 4.82,
      "learning_rate": 1.5946226744029402e-07,
      "loss": 0.1347,
      "step": 2905
    },
    {
      "epoch": 4.83,
      "learning_rate": 1.4465129134078048e-07,
      "loss": 0.1368,
      "step": 2910
    },
    {
      "epoch": 4.84,
      "learning_rate": 1.3056007966470717e-07,
      "loss": 0.1292,
      "step": 2915
    },
    {
      "epoch": 4.85,
      "learning_rate": 1.1718904046721658e-07,
      "loss": 0.1307,
      "step": 2920
    },
    {
      "epoch": 4.86,
      "learning_rate": 1.045385609486027e-07,
      "loss": 0.1525,
      "step": 2925
    },
    {
      "epoch": 4.86,
      "learning_rate": 9.260900744309797e-08,
      "loss": 0.1287,
      "step": 2930
    },
    {
      "epoch": 4.87,
      "learning_rate": 8.140072540826771e-08,
      "loss": 0.1312,
      "step": 2935
    },
    {
      "epoch": 4.88,
      "learning_rate": 7.091403941499598e-08,
      "loss": 0.1335,
      "step": 2940
    },
    {
      "epoch": 4.89,
      "learning_rate": 6.114925313810693e-08,
      "loss": 0.1395,
      "step": 2945
    },
    {
      "epoch": 4.9,
      "learning_rate": 5.210664934754972e-08,
      "loss": 0.1405,
      "step": 2950
    },
    {
      "epoch": 4.91,
      "learning_rate": 4.378648990022438e-08,
      "loss": 0.1343,
      "step": 2955
    },
    {
      "epoch": 4.91,
      "learning_rate": 3.618901573238798e-08,
      "loss": 0.1409,
      "step": 2960
    },
    {
      "epoch": 4.92,
      "learning_rate": 2.931444685269069e-08,
      "loss": 0.1327,
      "step": 2965
    },
    {
      "epoch": 4.93,
      "learning_rate": 2.3162982335789286e-08,
      "loss": 0.1367,
      "step": 2970
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.7734800316596135e-08,
      "loss": 0.1259,
      "step": 2975
    },
    {
      "epoch": 4.95,
      "learning_rate": 1.3030057985108369e-08,
      "loss": 0.1325,
      "step": 2980
    },
    {
      "epoch": 4.96,
      "learning_rate": 9.048891581875385e-09,
      "loss": 0.1275,
      "step": 2985
    },
    {
      "epoch": 4.96,
      "learning_rate": 5.791416394024252e-09,
      "loss": 0.1253,
      "step": 2990
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.2577267519484643e-09,
      "loss": 0.1408,
      "step": 2995
    },
    {
      "epoch": 4.98,
      "learning_rate": 1.4478960265629227e-09,
      "loss": 0.1278,
      "step": 3000
    },
    {
      "epoch": 4.98,
      "eval_loss": 0.07845821231603622,
      "eval_runtime": 6.6657,
      "eval_samples_per_second": 15.002,
      "eval_steps_per_second": 15.002,
      "step": 3000
    }
  ],
  "logging_steps": 5,
  "max_steps": 3010,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 125,
  "total_flos": 1.0266248056637276e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}