{
  "best_metric": 0.9372,
  "best_model_checkpoint": "output/checkpoint-22500",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 25000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 3.3333333333333337e-06,
      "loss": 0.8125,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 6.6666666666666675e-06,
      "loss": 0.7599,
      "step": 100
    },
    {
      "epoch": 0.06,
      "learning_rate": 1e-05,
      "loss": 0.7309,
      "step": 150
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.3333333333333335e-05,
      "loss": 0.7104,
      "step": 200
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.6939,
      "step": 250
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 0.6824,
      "step": 300
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.686,
      "step": 350
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.666666666666667e-05,
      "loss": 0.6693,
      "step": 400
    },
    {
      "epoch": 0.18,
      "learning_rate": 3e-05,
      "loss": 0.6436,
      "step": 450
    },
    {
      "epoch": 0.2,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.6101,
      "step": 500
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.493,
      "step": 550
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.993333333333333e-05,
      "loss": 0.4239,
      "step": 600
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.326666666666667e-05,
      "loss": 0.3391,
      "step": 650
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.66e-05,
      "loss": 0.3332,
      "step": 700
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.986666666666666e-05,
      "loss": 0.3966,
      "step": 750
    },
    {
      "epoch": 0.32,
      "learning_rate": 5.32e-05,
      "loss": 0.3137,
      "step": 800
    },
    {
      "epoch": 0.34,
      "learning_rate": 5.653333333333333e-05,
      "loss": 0.3072,
      "step": 850
    },
    {
      "epoch": 0.36,
      "learning_rate": 5.986666666666667e-05,
      "loss": 0.272,
      "step": 900
    },
    {
      "epoch": 0.38,
      "learning_rate": 6.32e-05,
      "loss": 0.2715,
      "step": 950
    },
    {
      "epoch": 0.4,
      "learning_rate": 6.653333333333333e-05,
      "loss": 0.2892,
      "step": 1000
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.986666666666667e-05,
      "loss": 0.28,
      "step": 1050
    },
    {
      "epoch": 0.44,
      "learning_rate": 7.32e-05,
      "loss": 0.3102,
      "step": 1100
    },
    {
      "epoch": 0.46,
      "learning_rate": 7.653333333333333e-05,
      "loss": 0.2663,
      "step": 1150
    },
    {
      "epoch": 0.48,
      "learning_rate": 7.986666666666667e-05,
      "loss": 0.2778,
      "step": 1200
    },
    {
      "epoch": 0.5,
      "learning_rate": 8.32e-05,
      "loss": 0.3414,
      "step": 1250
    },
    {
      "epoch": 0.52,
      "learning_rate": 8.653333333333334e-05,
      "loss": 0.2608,
      "step": 1300
    },
    {
      "epoch": 0.54,
      "learning_rate": 8.986666666666666e-05,
      "loss": 0.2847,
      "step": 1350
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.32e-05,
      "loss": 0.2801,
      "step": 1400
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.653333333333334e-05,
      "loss": 0.2777,
      "step": 1450
    },
    {
      "epoch": 0.6,
      "learning_rate": 9.986666666666668e-05,
      "loss": 0.2819,
      "step": 1500
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0001032,
      "loss": 0.2631,
      "step": 1550
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00010653333333333333,
      "loss": 0.255,
      "step": 1600
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00010986666666666668,
      "loss": 0.3253,
      "step": 1650
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0001132,
      "loss": 0.2632,
      "step": 1700
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00011653333333333334,
      "loss": 0.2808,
      "step": 1750
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00011986666666666666,
      "loss": 0.2487,
      "step": 1800
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0001232,
      "loss": 0.1958,
      "step": 1850
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00012653333333333334,
      "loss": 0.2882,
      "step": 1900
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00012986666666666666,
      "loss": 0.2877,
      "step": 1950
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0001332,
      "loss": 0.2412,
      "step": 2000
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00013653333333333334,
      "loss": 0.2608,
      "step": 2050
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00013986666666666666,
      "loss": 0.2296,
      "step": 2100
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00014319999999999998,
      "loss": 0.2311,
      "step": 2150
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00014653333333333334,
      "loss": 0.2603,
      "step": 2200
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00014986666666666669,
      "loss": 0.2923,
      "step": 2250
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0001532,
      "loss": 0.2415,
      "step": 2300
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015646666666666668,
      "loss": 0.2671,
      "step": 2350
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0001598,
      "loss": 0.2421,
      "step": 2400
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00016313333333333333,
      "loss": 0.2449,
      "step": 2450
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00016646666666666668,
      "loss": 0.2399,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9037,
      "eval_loss": 0.2539164125919342,
      "eval_runtime": 132.4594,
      "eval_samples_per_second": 75.495,
      "eval_steps_per_second": 4.718,
      "step": 2500
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0001698,
      "loss": 0.2453,
      "step": 2550
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00017313333333333333,
      "loss": 0.2485,
      "step": 2600
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00017646666666666666,
      "loss": 0.2321,
      "step": 2650
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00017979999999999998,
      "loss": 0.239,
      "step": 2700
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00018313333333333336,
      "loss": 0.2389,
      "step": 2750
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00018646666666666668,
      "loss": 0.2216,
      "step": 2800
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0001898,
      "loss": 0.3027,
      "step": 2850
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00019313333333333333,
      "loss": 0.2683,
      "step": 2900
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00019646666666666668,
      "loss": 0.2378,
      "step": 2950
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0001998,
      "loss": 0.2598,
      "step": 3000
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00020313333333333333,
      "loss": 0.2363,
      "step": 3050
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00020646666666666665,
      "loss": 0.2437,
      "step": 3100
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002098,
      "loss": 0.244,
      "step": 3150
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00021313333333333335,
      "loss": 0.2264,
      "step": 3200
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00021646666666666668,
      "loss": 0.188,
      "step": 3250
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002198,
      "loss": 0.2507,
      "step": 3300
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00022313333333333333,
      "loss": 0.2265,
      "step": 3350
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00022646666666666668,
      "loss": 0.2265,
      "step": 3400
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002298,
      "loss": 0.226,
      "step": 3450
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00023313333333333333,
      "loss": 0.2502,
      "step": 3500
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00023646666666666665,
      "loss": 0.2436,
      "step": 3550
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00023980000000000003,
      "loss": 0.2472,
      "step": 3600
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00024313333333333335,
      "loss": 0.2596,
      "step": 3650
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002464666666666667,
      "loss": 0.2381,
      "step": 3700
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002498,
      "loss": 0.231,
      "step": 3750
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002531333333333333,
      "loss": 0.2489,
      "step": 3800
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002564666666666667,
      "loss": 0.1849,
      "step": 3850
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00025979999999999997,
      "loss": 0.2742,
      "step": 3900
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00026313333333333335,
      "loss": 0.2268,
      "step": 3950
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002664666666666667,
      "loss": 0.287,
      "step": 4000
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002698,
      "loss": 0.208,
      "step": 4050
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002731333333333333,
      "loss": 0.2446,
      "step": 4100
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002764666666666667,
      "loss": 0.2359,
      "step": 4150
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002798,
      "loss": 0.2917,
      "step": 4200
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00028313333333333335,
      "loss": 0.2051,
      "step": 4250
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00028646666666666667,
      "loss": 0.2119,
      "step": 4300
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002898,
      "loss": 0.1946,
      "step": 4350
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002931333333333334,
      "loss": 0.2159,
      "step": 4400
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00029646666666666664,
      "loss": 0.2793,
      "step": 4450
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002998,
      "loss": 0.2675,
      "step": 4500
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0003031333333333333,
      "loss": 0.1927,
      "step": 4550
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00030646666666666667,
      "loss": 0.177,
      "step": 4600
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00030980000000000005,
      "loss": 0.2681,
      "step": 4650
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00031306666666666667,
      "loss": 0.2123,
      "step": 4700
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00031640000000000005,
      "loss": 0.2232,
      "step": 4750
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0003197333333333333,
      "loss": 0.222,
      "step": 4800
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0003230666666666667,
      "loss": 0.2126,
      "step": 4850
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0003264,
      "loss": 0.2268,
      "step": 4900
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00032973333333333334,
      "loss": 0.2239,
      "step": 4950
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00033306666666666667,
      "loss": 0.2454,
      "step": 5000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.27525001764297485,
      "eval_runtime": 132.3865,
      "eval_samples_per_second": 75.536,
      "eval_steps_per_second": 4.721,
      "step": 5000
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0003364,
      "loss": 0.2001,
      "step": 5050
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0003397333333333333,
      "loss": 0.1839,
      "step": 5100
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0003430666666666667,
      "loss": 0.2544,
      "step": 5150
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0003464,
      "loss": 0.1747,
      "step": 5200
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.00034973333333333334,
      "loss": 0.1857,
      "step": 5250
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.00035306666666666667,
      "loss": 0.2486,
      "step": 5300
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0003564,
      "loss": 0.1884,
      "step": 5350
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.00035973333333333337,
      "loss": 0.2118,
      "step": 5400
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.00036306666666666664,
      "loss": 0.2187,
      "step": 5450
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0003664,
      "loss": 0.2032,
      "step": 5500
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.00036973333333333334,
      "loss": 0.1931,
      "step": 5550
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.00037306666666666666,
      "loss": 0.2272,
      "step": 5600
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.00037640000000000004,
      "loss": 0.2196,
      "step": 5650
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0003797333333333333,
      "loss": 0.1778,
      "step": 5700
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0003830666666666667,
      "loss": 0.1637,
      "step": 5750
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.0003864,
      "loss": 0.1595,
      "step": 5800
    },
    {
      "epoch": 2.34,
      "learning_rate": 0.00038973333333333334,
      "loss": 0.2069,
      "step": 5850
    },
    {
      "epoch": 2.36,
      "learning_rate": 0.00039306666666666666,
      "loss": 0.2338,
      "step": 5900
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.0003964,
      "loss": 0.2159,
      "step": 5950
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.00039973333333333336,
      "loss": 0.2325,
      "step": 6000
    },
    {
      "epoch": 2.42,
      "learning_rate": 0.0004030666666666667,
      "loss": 0.2506,
      "step": 6050
    },
    {
      "epoch": 2.44,
      "learning_rate": 0.0004064,
      "loss": 0.2129,
      "step": 6100
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.00040973333333333334,
      "loss": 0.2239,
      "step": 6150
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0004130666666666667,
      "loss": 0.231,
      "step": 6200
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.0004164,
      "loss": 0.2157,
      "step": 6250
    },
    {
      "epoch": 2.52,
      "learning_rate": 0.00041973333333333336,
      "loss": 0.2642,
      "step": 6300
    },
    {
      "epoch": 2.54,
      "learning_rate": 0.00042306666666666663,
      "loss": 0.2308,
      "step": 6350
    },
    {
      "epoch": 2.56,
      "learning_rate": 0.0004264,
      "loss": 0.2229,
      "step": 6400
    },
    {
      "epoch": 2.58,
      "learning_rate": 0.0004297333333333334,
      "loss": 0.2231,
      "step": 6450
    },
    {
      "epoch": 2.6,
      "learning_rate": 0.00043306666666666666,
      "loss": 0.1866,
      "step": 6500
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.00043640000000000004,
      "loss": 0.1999,
      "step": 6550
    },
    {
      "epoch": 2.64,
      "learning_rate": 0.0004397333333333333,
      "loss": 0.2566,
      "step": 6600
    },
    {
      "epoch": 2.66,
      "learning_rate": 0.0004430666666666667,
      "loss": 0.2168,
      "step": 6650
    },
    {
      "epoch": 2.68,
      "learning_rate": 0.0004464,
      "loss": 0.2551,
      "step": 6700
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.00044973333333333333,
      "loss": 0.1921,
      "step": 6750
    },
    {
      "epoch": 2.72,
      "learning_rate": 0.00045306666666666666,
      "loss": 0.2241,
      "step": 6800
    },
    {
      "epoch": 2.74,
      "learning_rate": 0.0004564,
      "loss": 0.2359,
      "step": 6850
    },
    {
      "epoch": 2.76,
      "learning_rate": 0.00045966666666666665,
      "loss": 0.2348,
      "step": 6900
    },
    {
      "epoch": 2.78,
      "learning_rate": 0.00046300000000000003,
      "loss": 0.2228,
      "step": 6950
    },
    {
      "epoch": 2.8,
      "learning_rate": 0.00046633333333333336,
      "loss": 0.2302,
      "step": 7000
    },
    {
      "epoch": 2.82,
      "learning_rate": 0.0004696666666666667,
      "loss": 0.2455,
      "step": 7050
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.000473,
      "loss": 0.2731,
      "step": 7100
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.00047633333333333333,
      "loss": 0.2365,
      "step": 7150
    },
    {
      "epoch": 2.88,
      "learning_rate": 0.0004796666666666667,
      "loss": 0.2554,
      "step": 7200
    },
    {
      "epoch": 2.9,
      "learning_rate": 0.000483,
      "loss": 0.2386,
      "step": 7250
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.00048633333333333335,
      "loss": 0.1938,
      "step": 7300
    },
    {
      "epoch": 2.94,
      "learning_rate": 0.0004896666666666667,
      "loss": 0.23,
      "step": 7350
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.0004930000000000001,
      "loss": 0.2151,
      "step": 7400
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0004963333333333333,
      "loss": 0.2073,
      "step": 7450
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0004996666666666667,
      "loss": 0.2251,
      "step": 7500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9167,
      "eval_loss": 0.24363669753074646,
      "eval_runtime": 132.3921,
      "eval_samples_per_second": 75.533,
      "eval_steps_per_second": 4.721,
      "step": 7500
    },
    {
      "epoch": 3.02,
      "learning_rate": 0.0004999918425141952,
      "loss": 0.1875,
      "step": 7550
    },
    {
      "epoch": 3.04,
      "learning_rate": 0.0004999636444812977,
      "loss": 0.2006,
      "step": 7600
    },
    {
      "epoch": 3.06,
      "learning_rate": 0.0004999153074629148,
      "loss": 0.2359,
      "step": 7650
    },
    {
      "epoch": 3.08,
      "learning_rate": 0.0004998468353534468,
      "loss": 0.1878,
      "step": 7700
    },
    {
      "epoch": 3.1,
      "learning_rate": 0.0004997582336695312,
      "loss": 0.2309,
      "step": 7750
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.0004996495095495983,
      "loss": 0.2101,
      "step": 7800
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.0004995206717532957,
      "loss": 0.2599,
      "step": 7850
    },
    {
      "epoch": 3.16,
      "learning_rate": 0.0004993717306607832,
      "loss": 0.2405,
      "step": 7900
    },
    {
      "epoch": 3.18,
      "learning_rate": 0.0004992026982718955,
      "loss": 0.2082,
      "step": 7950
    },
    {
      "epoch": 3.2,
      "learning_rate": 0.0004990135882051766,
      "loss": 0.2071,
      "step": 8000
    },
    {
      "epoch": 3.22,
      "learning_rate": 0.0004988044156967816,
      "loss": 0.1868,
      "step": 8050
    },
    {
      "epoch": 3.24,
      "learning_rate": 0.0004985751975992497,
      "loss": 0.2029,
      "step": 8100
    },
    {
      "epoch": 3.26,
      "learning_rate": 0.0004983259523801463,
      "loss": 0.2113,
      "step": 8150
    },
    {
      "epoch": 3.28,
      "learning_rate": 0.000498056700120575,
      "loss": 0.2172,
      "step": 8200
    },
    {
      "epoch": 3.3,
      "learning_rate": 0.0004977674625135598,
      "loss": 0.1988,
      "step": 8250
    },
    {
      "epoch": 3.32,
      "learning_rate": 0.0004974582628622975,
      "loss": 0.2154,
      "step": 8300
    },
    {
      "epoch": 3.34,
      "learning_rate": 0.0004971291260782797,
      "loss": 0.2122,
      "step": 8350
    },
    {
      "epoch": 3.36,
      "learning_rate": 0.0004967800786792866,
      "loss": 0.1891,
      "step": 8400
    },
    {
      "epoch": 3.38,
      "learning_rate": 0.0004964111487872495,
      "loss": 0.2461,
      "step": 8450
    },
    {
      "epoch": 3.4,
      "learning_rate": 0.000496022366125986,
      "loss": 0.1967,
      "step": 8500
    },
    {
      "epoch": 3.42,
      "learning_rate": 0.0004956137620188048,
      "loss": 0.2278,
      "step": 8550
    },
    {
      "epoch": 3.44,
      "learning_rate": 0.000495185369385982,
      "loss": 0.2029,
      "step": 8600
    },
    {
      "epoch": 3.46,
      "learning_rate": 0.0004947372227421084,
      "loss": 0.2094,
      "step": 8650
    },
    {
      "epoch": 3.48,
      "learning_rate": 0.0004942693581933101,
      "loss": 0.2701,
      "step": 8700
    },
    {
      "epoch": 3.5,
      "learning_rate": 0.000493781813434338,
      "loss": 0.1984,
      "step": 8750
    },
    {
      "epoch": 3.52,
      "learning_rate": 0.0004932746277455317,
      "loss": 0.2277,
      "step": 8800
    },
    {
      "epoch": 3.54,
      "learning_rate": 0.0004927478419896543,
      "loss": 0.1903,
      "step": 8850
    },
    {
      "epoch": 3.56,
      "learning_rate": 0.0004922014986086007,
      "loss": 0.1693,
      "step": 8900
    },
    {
      "epoch": 3.58,
      "learning_rate": 0.0004916356416199778,
      "loss": 0.2198,
      "step": 8950
    },
    {
      "epoch": 3.6,
      "learning_rate": 0.0004910503166135578,
      "loss": 0.1728,
      "step": 9000
    },
    {
      "epoch": 3.62,
      "learning_rate": 0.000490445570747606,
      "loss": 0.2523,
      "step": 9050
    },
    {
      "epoch": 3.64,
      "learning_rate": 0.0004898214527450807,
      "loss": 0.2319,
      "step": 9100
    },
    {
      "epoch": 3.66,
      "learning_rate": 0.0004891910707082325,
      "loss": 0.2402,
      "step": 9150
    },
    {
      "epoch": 3.68,
      "learning_rate": 0.0004885287457226171,
      "loss": 0.1891,
      "step": 9200
    },
    {
      "epoch": 3.7,
      "learning_rate": 0.00048784720303453175,
      "loss": 0.1828,
      "step": 9250
    },
    {
      "epoch": 3.72,
      "learning_rate": 0.0004871464975542746,
      "loss": 0.2131,
      "step": 9300
    },
    {
      "epoch": 3.74,
      "learning_rate": 0.0004864266857360456,
      "loss": 0.1854,
      "step": 9350
    },
    {
      "epoch": 3.76,
      "learning_rate": 0.00048568782557339713,
      "loss": 0.1927,
      "step": 9400
    },
    {
      "epoch": 3.78,
      "learning_rate": 0.0004849299765945623,
      "loss": 0.206,
      "step": 9450
    },
    {
      "epoch": 3.8,
      "learning_rate": 0.00048415319985765877,
      "loss": 0.2008,
      "step": 9500
    },
    {
      "epoch": 3.82,
      "learning_rate": 0.0004833575579457691,
      "loss": 0.1961,
      "step": 9550
    },
    {
      "epoch": 3.84,
      "learning_rate": 0.000482543114961899,
      "loss": 0.1752,
      "step": 9600
    },
    {
      "epoch": 3.86,
      "learning_rate": 0.00048170993652381247,
      "loss": 0.1751,
      "step": 9650
    },
    {
      "epoch": 3.88,
      "learning_rate": 0.00048085808975874506,
      "loss": 0.2128,
      "step": 9700
    },
    {
      "epoch": 3.9,
      "learning_rate": 0.000479987643297996,
      "loss": 0.188,
      "step": 9750
    },
    {
      "epoch": 3.92,
      "learning_rate": 0.0004790986672713982,
      "loss": 0.227,
      "step": 9800
    },
    {
      "epoch": 3.94,
      "learning_rate": 0.0004781912333016683,
      "loss": 0.1947,
      "step": 9850
    },
    {
      "epoch": 3.96,
      "learning_rate": 0.00047726541449863636,
      "loss": 0.2057,
      "step": 9900
    },
    {
      "epoch": 3.98,
      "learning_rate": 0.00047632128545335527,
      "loss": 0.195,
      "step": 9950
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.0004753589222320913,
      "loss": 0.1996,
      "step": 10000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9246,
      "eval_loss": 0.2271093726158142,
      "eval_runtime": 132.5243,
      "eval_samples_per_second": 75.458,
      "eval_steps_per_second": 4.716,
      "step": 10000
    },
    {
      "epoch": 4.02,
      "learning_rate": 0.00047437840237019567,
      "loss": 0.1785,
      "step": 10050
    },
    {
      "epoch": 4.04,
      "learning_rate": 0.00047337980486585735,
      "loss": 0.2067,
      "step": 10100
    },
    {
      "epoch": 4.06,
      "learning_rate": 0.00047236321017373893,
      "loss": 0.1896,
      "step": 10150
    },
    {
      "epoch": 4.08,
      "learning_rate": 0.00047132870019849415,
      "loss": 0.1719,
      "step": 10200
    },
    {
      "epoch": 4.1,
      "learning_rate": 0.00047027635828816915,
      "loss": 0.1836,
      "step": 10250
    },
    {
      "epoch": 4.12,
      "learning_rate": 0.0004692062692274873,
      "loss": 0.1785,
      "step": 10300
    },
    {
      "epoch": 4.14,
      "learning_rate": 0.0004681185192310183,
      "loss": 0.2045,
      "step": 10350
    },
    {
      "epoch": 4.16,
      "learning_rate": 0.0004670131959362323,
      "loss": 0.175,
      "step": 10400
    },
    {
      "epoch": 4.18,
      "learning_rate": 0.00046589038839643865,
      "loss": 0.1609,
      "step": 10450
    },
    {
      "epoch": 4.2,
      "learning_rate": 0.0004647501870736115,
      "loss": 0.2075,
      "step": 10500
    },
    {
      "epoch": 4.22,
      "learning_rate": 0.0004635926838311014,
      "loss": 0.1618,
      "step": 10550
    },
    {
      "epoch": 4.24,
      "learning_rate": 0.0004624179719262341,
      "loss": 0.2275,
      "step": 10600
    },
    {
      "epoch": 4.26,
      "learning_rate": 0.0004612261460027968,
      "loss": 0.1881,
      "step": 10650
    },
    {
      "epoch": 4.28,
      "learning_rate": 0.0004600173020834132,
      "loss": 0.1863,
      "step": 10700
    },
    {
      "epoch": 4.3,
      "learning_rate": 0.00045879153756180714,
      "loss": 0.1765,
      "step": 10750
    },
    {
      "epoch": 4.32,
      "learning_rate": 0.00045754895119495576,
      "loss": 0.1964,
      "step": 10800
    },
    {
      "epoch": 4.34,
      "learning_rate": 0.00045628964309513267,
      "loss": 0.2081,
      "step": 10850
    },
    {
      "epoch": 4.36,
      "learning_rate": 0.00045501371472184257,
      "loss": 0.1695,
      "step": 10900
    },
    {
      "epoch": 4.38,
      "learning_rate": 0.0004537212688736466,
      "loss": 0.2042,
      "step": 10950
    },
    {
      "epoch": 4.4,
      "learning_rate": 0.00045241240967988015,
      "loss": 0.2003,
      "step": 11000
    },
    {
      "epoch": 4.42,
      "learning_rate": 0.0004510872425922635,
      "loss": 0.2015,
      "step": 11050
    },
    {
      "epoch": 4.44,
      "learning_rate": 0.0004497458743764057,
      "loss": 0.1409,
      "step": 11100
    },
    {
      "epoch": 4.46,
      "learning_rate": 0.00044838841310320266,
      "loss": 0.1446,
      "step": 11150
    },
    {
      "epoch": 4.48,
      "learning_rate": 0.0004470149681401305,
      "loss": 0.1834,
      "step": 11200
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.0004456256501424335,
      "loss": 0.1908,
      "step": 11250
    },
    {
      "epoch": 4.52,
      "learning_rate": 0.0004442205710442094,
      "loss": 0.1727,
      "step": 11300
    },
    {
      "epoch": 4.54,
      "learning_rate": 0.0004427998440493908,
      "loss": 0.1853,
      "step": 11350
    },
    {
      "epoch": 4.56,
      "learning_rate": 0.00044139246031239603,
      "loss": 0.2239,
      "step": 11400
    },
    {
      "epoch": 4.58,
      "learning_rate": 0.0004399410893820698,
      "loss": 0.1986,
      "step": 11450
    },
    {
      "epoch": 4.6,
      "learning_rate": 0.00043847441534296505,
      "loss": 0.1802,
      "step": 11500
    },
    {
      "epoch": 4.62,
      "learning_rate": 0.00043699255636157463,
      "loss": 0.1795,
      "step": 11550
    },
    {
      "epoch": 4.64,
      "learning_rate": 0.0004354956318278065,
      "loss": 0.1828,
      "step": 11600
    },
    {
      "epoch": 4.66,
      "learning_rate": 0.000433983762345365,
      "loss": 0.1806,
      "step": 11650
    },
    {
      "epoch": 4.68,
      "learning_rate": 0.00043245706972203383,
      "loss": 0.1666,
      "step": 11700
    },
    {
      "epoch": 4.7,
      "learning_rate": 0.00043091567695986265,
      "loss": 0.1972,
      "step": 11750
    },
    {
      "epoch": 4.72,
      "learning_rate": 0.0004293597082452566,
      "loss": 0.1707,
      "step": 11800
    },
    {
      "epoch": 4.74,
      "learning_rate": 0.0004277892889389715,
      "loss": 0.1942,
      "step": 11850
    },
    {
      "epoch": 4.76,
      "learning_rate": 0.0004262045455660132,
      "loss": 0.192,
      "step": 11900
    },
    {
      "epoch": 4.78,
      "learning_rate": 0.0004246056058054444,
      "loss": 0.2138,
      "step": 11950
    },
    {
      "epoch": 4.8,
      "learning_rate": 0.000422992598480097,
      "loss": 0.2031,
      "step": 12000
    },
    {
      "epoch": 4.82,
      "learning_rate": 0.0004213656535461942,
      "loss": 0.1608,
      "step": 12050
    },
    {
      "epoch": 4.84,
      "learning_rate": 0.0004197249020828794,
      "loss": 0.2179,
      "step": 12100
    },
    {
      "epoch": 4.86,
      "learning_rate": 0.00041807047628165576,
      "loss": 0.2141,
      "step": 12150
    },
    {
      "epoch": 4.88,
      "learning_rate": 0.00041640250943573574,
      "loss": 0.187,
      "step": 12200
    },
    {
      "epoch": 4.9,
      "learning_rate": 0.0004147211359293023,
      "loss": 0.1854,
      "step": 12250
    },
    {
      "epoch": 4.92,
      "learning_rate": 0.00041302649122668155,
      "loss": 0.1687,
      "step": 12300
    },
    {
      "epoch": 4.94,
      "learning_rate": 0.00041131871186142867,
      "loss": 0.1844,
      "step": 12350
    },
    {
      "epoch": 4.96,
      "learning_rate": 0.00040959793542532784,
      "loss": 0.2203,
      "step": 12400
    },
    {
      "epoch": 4.98,
      "learning_rate": 0.0004078643005573071,
      "loss": 0.1733,
      "step": 12450
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.00040611794693226795,
      "loss": 0.1845,
      "step": 12500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9269,
      "eval_loss": 0.21157951653003693,
      "eval_runtime": 132.481,
      "eval_samples_per_second": 75.483,
      "eval_steps_per_second": 4.718,
      "step": 12500
    },
    {
      "epoch": 5.02,
      "learning_rate": 0.00040435901524983245,
      "loss": 0.1696,
      "step": 12550
    },
    {
      "epoch": 5.04,
      "learning_rate": 0.00040258764722300727,
      "loss": 0.1566,
      "step": 12600
    },
    {
      "epoch": 5.06,
      "learning_rate": 0.00040080398556676625,
      "loss": 0.1533,
      "step": 12650
    },
    {
      "epoch": 5.08,
      "learning_rate": 0.00039900817398655223,
      "loss": 0.1357,
      "step": 12700
    },
    {
      "epoch": 5.1,
      "learning_rate": 0.0003972003571666988,
      "loss": 0.1493,
      "step": 12750
    },
    {
      "epoch": 5.12,
      "learning_rate": 0.00039538068075877376,
      "loss": 0.2111,
      "step": 12800
    },
    {
      "epoch": 5.14,
      "learning_rate": 0.0003935492913698441,
      "loss": 0.1651,
      "step": 12850
    },
    {
      "epoch": 5.16,
      "learning_rate": 0.00039170633655066415,
      "loss": 0.1535,
      "step": 12900
    },
    {
      "epoch": 5.18,
      "learning_rate": 0.000389851964783788,
      "loss": 0.1618,
      "step": 12950
    },
    {
      "epoch": 5.2,
      "learning_rate": 0.00038798632547160653,
      "loss": 0.1672,
      "step": 13000
    },
    {
      "epoch": 5.22,
      "learning_rate": 0.00038610956892431015,
      "loss": 0.1675,
      "step": 13050
    },
    {
      "epoch": 5.24,
      "learning_rate": 0.0003842218463477791,
      "loss": 0.1482,
      "step": 13100
    },
    {
      "epoch": 5.26,
      "learning_rate": 0.0003823233098314008,
      "loss": 0.1637,
      "step": 13150
    },
    {
      "epoch": 5.28,
      "learning_rate": 0.00038041411233581636,
      "loss": 0.1497,
      "step": 13200
    },
    {
      "epoch": 5.3,
      "learning_rate": 0.0003784944076805974,
      "loss": 0.1742,
      "step": 13250
    },
    {
      "epoch": 5.32,
      "learning_rate": 0.00037656435053185226,
      "loss": 0.1357,
      "step": 13300
    },
    {
      "epoch": 5.34,
      "learning_rate": 0.0003746240963897659,
      "loss": 0.1938,
      "step": 13350
    },
    {
      "epoch": 5.36,
      "learning_rate": 0.000372673801576071,
      "loss": 0.1506,
      "step": 13400
    },
    {
      "epoch": 5.38,
      "learning_rate": 0.0003707136232214534,
      "loss": 0.1912,
      "step": 13450
    },
    {
      "epoch": 5.4,
      "learning_rate": 0.0003687437192528932,
      "loss": 0.1429,
      "step": 13500
    },
    {
      "epoch": 5.42,
      "learning_rate": 0.0003667642483809398,
      "loss": 0.1579,
      "step": 13550
    },
    {
      "epoch": 5.44,
      "learning_rate": 0.0003647753700869262,
      "loss": 0.1754,
      "step": 13600
    },
    {
      "epoch": 5.46,
      "learning_rate": 0.00036277724461011905,
      "loss": 0.1572,
      "step": 13650
    },
    {
      "epoch": 5.48,
      "learning_rate": 0.0003607700329348088,
      "loss": 0.1964,
      "step": 13700
    },
    {
      "epoch": 5.5,
      "learning_rate": 0.00035875389677734,
      "loss": 0.1813,
      "step": 13750
    },
    {
      "epoch": 5.52,
      "learning_rate": 0.0003567289985730813,
      "loss": 0.1673,
      "step": 13800
    },
    {
      "epoch": 5.54,
      "learning_rate": 0.00035469550146333947,
      "loss": 0.1418,
      "step": 13850
    },
    {
      "epoch": 5.56,
      "learning_rate": 0.0003526535692822146,
      "loss": 0.1536,
      "step": 13900
    },
    {
      "epoch": 5.58,
      "learning_rate": 0.0003506033665434007,
      "loss": 0.1923,
      "step": 13950
    },
    {
      "epoch": 5.6,
      "learning_rate": 0.0003485450584269317,
      "loss": 0.1772,
      "step": 14000
    },
    {
      "epoch": 5.62,
      "learning_rate": 0.0003464788107658724,
      "loss": 0.152,
      "step": 14050
    },
    {
      "epoch": 5.64,
      "learning_rate": 0.000344446345543944,
      "loss": 0.1638,
      "step": 14100
    },
    {
      "epoch": 5.66,
      "learning_rate": 0.00034236486931606976,
      "loss": 0.195,
      "step": 14150
    },
    {
      "epoch": 5.68,
      "learning_rate": 0.00034027595146696936,
      "loss": 0.1738,
      "step": 14200
    },
    {
      "epoch": 5.7,
      "learning_rate": 0.0003381797602958623,
      "loss": 0.1539,
      "step": 14250
    },
    {
      "epoch": 5.72,
      "learning_rate": 0.00033607646468796263,
      "loss": 0.1749,
      "step": 14300
    },
    {
      "epoch": 5.74,
      "learning_rate": 0.0003339662341008717,
      "loss": 0.1713,
      "step": 14350
    },
    {
      "epoch": 5.76,
      "learning_rate": 0.0003318492385509267,
      "loss": 0.151,
      "step": 14400
    },
    {
      "epoch": 5.78,
      "learning_rate": 0.00032972564859950137,
      "loss": 0.1818,
      "step": 14450
    },
    {
      "epoch": 5.8,
      "learning_rate": 0.0003275956353392653,
      "loss": 0.1499,
      "step": 14500
    },
    {
      "epoch": 5.82,
      "learning_rate": 0.00032545937038039904,
      "loss": 0.1601,
      "step": 14550
    },
    {
      "epoch": 5.84,
      "learning_rate": 0.00032331702583676745,
      "loss": 0.1754,
      "step": 14600
    },
    {
      "epoch": 5.86,
      "learning_rate": 0.0003211687743120539,
      "loss": 0.1466,
      "step": 14650
    },
    {
      "epoch": 5.88,
      "learning_rate": 0.0003190147888858529,
      "loss": 0.1606,
      "step": 14700
    },
    {
      "epoch": 5.9,
      "learning_rate": 0.0003168552430997262,
      "loss": 0.1654,
      "step": 14750
    },
    {
      "epoch": 5.92,
      "learning_rate": 0.0003146903109432208,
      "loss": 0.166,
      "step": 14800
    },
    {
      "epoch": 5.94,
      "learning_rate": 0.0003125201668398506,
      "loss": 0.1513,
      "step": 14850
    },
    {
      "epoch": 5.96,
      "learning_rate": 0.0003103449856330443,
      "loss": 0.1736,
      "step": 14900
    },
    {
      "epoch": 5.98,
      "learning_rate": 0.0003081649425720579,
      "loss": 0.2036,
      "step": 14950
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.0003059802132978558,
      "loss": 0.205,
      "step": 15000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9312,
      "eval_loss": 0.19459494948387146,
      "eval_runtime": 132.3124,
      "eval_samples_per_second": 75.579,
      "eval_steps_per_second": 4.724,
      "step": 15000
    },
    {
      "epoch": 6.02,
      "learning_rate": 0.00030379097382895955,
      "loss": 0.1528,
      "step": 15050
    },
    {
      "epoch": 6.04,
      "learning_rate": 0.0003015974005472664,
      "loss": 0.1183,
      "step": 15100
    },
    {
      "epoch": 6.06,
      "learning_rate": 0.0002993996701838391,
      "loss": 0.1508,
      "step": 15150
    },
    {
      "epoch": 6.08,
      "learning_rate": 0.0002971979598046662,
      "loss": 0.1454,
      "step": 15200
    },
    {
      "epoch": 6.1,
      "learning_rate": 0.00029499244679639746,
      "loss": 0.1068,
      "step": 15250
    },
    {
      "epoch": 6.12,
      "learning_rate": 0.000292783308852051,
      "loss": 0.1451,
      "step": 15300
    },
    {
      "epoch": 6.14,
      "learning_rate": 0.0002905707239566978,
      "loss": 0.1349,
      "step": 15350
    },
    {
      "epoch": 6.16,
      "learning_rate": 0.00028835487037312155,
      "loss": 0.1293,
      "step": 15400
    },
    {
      "epoch": 6.18,
      "learning_rate": 0.00028613592662745665,
      "loss": 0.1029,
      "step": 15450
    },
    {
      "epoch": 6.2,
      "learning_rate": 0.0002839140714948043,
      "loss": 0.1677,
      "step": 15500
    },
    {
      "epoch": 6.22,
      "learning_rate": 0.0002816894839848297,
      "loss": 0.137,
      "step": 15550
    },
    {
      "epoch": 6.24,
      "learning_rate": 0.000279462343327339,
      "loss": 0.1447,
      "step": 15600
    },
    {
      "epoch": 6.26,
      "learning_rate": 0.00027723282895783963,
      "loss": 0.1268,
      "step": 15650
    },
    {
      "epoch": 6.28,
      "learning_rate": 0.0002750011205030834,
      "loss": 0.1075,
      "step": 15700
    },
    {
      "epoch": 6.3,
      "learning_rate": 0.00027276739776659444,
      "loss": 0.2144,
      "step": 15750
    },
    {
      "epoch": 6.32,
      "learning_rate": 0.00027053184071418275,
      "loss": 0.1229,
      "step": 15800
    },
    {
      "epoch": 6.34,
      "learning_rate": 0.00026829462945944505,
      "loss": 0.1633,
      "step": 15850
    },
    {
      "epoch": 6.36,
      "learning_rate": 0.0002660559442492531,
      "loss": 0.1457,
      "step": 15900
    },
    {
      "epoch": 6.38,
      "learning_rate": 0.0002638159654492318,
      "loss": 0.1168,
      "step": 15950
    },
    {
      "epoch": 6.4,
      "learning_rate": 0.00026157487352922783,
      "loss": 0.1294,
      "step": 16000
    },
    {
      "epoch": 6.42,
      "learning_rate": 0.0002593328490487688,
      "loss": 0.1347,
      "step": 16050
    },
    {
      "epoch": 6.44,
      "learning_rate": 0.0002570900726425172,
      "loss": 0.1396,
      "step": 16100
    },
    {
      "epoch": 6.46,
      "learning_rate": 0.00025484672500571576,
      "loss": 0.1444,
      "step": 16150
    },
    {
      "epoch": 6.48,
      "learning_rate": 0.0002526029868796305,
      "loss": 0.1418,
      "step": 16200
    },
    {
      "epoch": 6.5,
      "learning_rate": 0.00025035903903698785,
      "loss": 0.1275,
      "step": 16250
    },
    {
      "epoch": 6.52,
      "learning_rate": 0.00024811506226741077,
      "loss": 0.131,
      "step": 16300
    },
    {
      "epoch": 6.54,
      "learning_rate": 0.000245916111203349,
      "loss": 0.0828,
      "step": 16350
    },
    {
      "epoch": 6.56,
      "learning_rate": 0.00024367261051914146,
      "loss": 0.1857,
      "step": 16400
    },
    {
      "epoch": 6.58,
      "learning_rate": 0.00024142961961788734,
      "loss": 0.1173,
      "step": 16450
    },
    {
      "epoch": 6.6,
      "learning_rate": 0.000239187319212111,
      "loss": 0.1757,
      "step": 16500
    },
    {
      "epoch": 6.62,
      "learning_rate": 0.0002369458899587052,
      "loss": 0.1108,
      "step": 16550
    },
    {
      "epoch": 6.64,
      "learning_rate": 0.00023470551244437603,
      "loss": 0.1178,
      "step": 16600
    },
    {
      "epoch": 6.66,
      "learning_rate": 0.00023246636717109357,
      "loss": 0.1566,
      "step": 16650
    },
    {
      "epoch": 6.68,
      "learning_rate": 0.0002302286345415488,
      "loss": 0.1369,
      "step": 16700
    },
    {
      "epoch": 6.7,
      "learning_rate": 0.00022799249484461954,
      "loss": 0.1529,
      "step": 16750
    },
    {
      "epoch": 6.72,
      "learning_rate": 0.0002257581282408449,
      "loss": 0.1445,
      "step": 16800
    },
    {
      "epoch": 6.74,
      "learning_rate": 0.00022352571474790954,
      "loss": 0.1802,
      "step": 16850
    },
    {
      "epoch": 6.76,
      "learning_rate": 0.000221295434226141,
      "loss": 0.1218,
      "step": 16900
    },
    {
      "epoch": 6.78,
      "learning_rate": 0.00021906746636401824,
      "loss": 0.1158,
      "step": 16950
    },
    {
      "epoch": 6.8,
      "learning_rate": 0.00021684199066369448,
      "loss": 0.1802,
      "step": 17000
    },
    {
      "epoch": 6.82,
      "learning_rate": 0.0002146191864265354,
      "loss": 0.1179,
      "step": 17050
    },
    {
      "epoch": 6.84,
      "learning_rate": 0.00021239923273867317,
      "loss": 0.1403,
      "step": 17100
    },
    {
      "epoch": 6.86,
      "learning_rate": 0.00021018230845657804,
      "loss": 0.1483,
      "step": 17150
    },
    {
      "epoch": 6.88,
      "learning_rate": 0.0002079685921926476,
      "loss": 0.1175,
      "step": 17200
    },
    {
      "epoch": 6.9,
      "learning_rate": 0.00020575826230081757,
      "loss": 0.1689,
      "step": 17250
    },
    {
      "epoch": 6.92,
      "learning_rate": 0.00020355149686219123,
      "loss": 0.1084,
      "step": 17300
    },
    {
      "epoch": 6.94,
      "learning_rate": 0.00020134847367069225,
      "loss": 0.1598,
      "step": 17350
    },
    {
      "epoch": 6.96,
      "learning_rate": 0.0001991493702187403,
      "loss": 0.1352,
      "step": 17400
    },
    {
      "epoch": 6.98,
      "learning_rate": 0.0001969543636829508,
      "loss": 0.1491,
      "step": 17450
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.00019476363090986011,
      "loss": 0.1352,
      "step": 17500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9328,
      "eval_loss": 0.22326812148094177,
      "eval_runtime": 132.2934,
      "eval_samples_per_second": 75.59,
      "eval_steps_per_second": 4.724,
      "step": 17500
    },
    {
      "epoch": 7.02,
      "learning_rate": 0.0001925773484016777,
      "loss": 0.0964,
      "step": 17550
    },
    {
      "epoch": 7.04,
      "learning_rate": 0.00019039569230206555,
      "loss": 0.1568,
      "step": 17600
    },
    {
      "epoch": 7.06,
      "learning_rate": 0.00018821883838194664,
      "loss": 0.0789,
      "step": 17650
    },
    {
      "epoch": 7.08,
      "learning_rate": 0.00018604696202534368,
      "loss": 0.0736,
      "step": 17700
    },
    {
      "epoch": 7.1,
      "learning_rate": 0.00018388023821524889,
      "loss": 0.1445,
      "step": 17750
    },
    {
      "epoch": 7.12,
      "learning_rate": 0.0001817188415195257,
      "loss": 0.0839,
      "step": 17800
    },
    {
      "epoch": 7.14,
      "learning_rate": 0.00017956294607684436,
      "loss": 0.161,
      "step": 17850
    },
    {
      "epoch": 7.16,
      "learning_rate": 0.0001774127255826522,
      "loss": 0.1418,
      "step": 17900
    },
    {
      "epoch": 7.18,
      "learning_rate": 0.0001752683532751792,
      "loss": 0.1222,
      "step": 17950
    },
    {
      "epoch": 7.2,
      "learning_rate": 0.0001731300019214805,
      "loss": 0.1034,
      "step": 18000
    },
    {
      "epoch": 7.22,
      "learning_rate": 0.00017099784380351728,
      "loss": 0.114,
      "step": 18050
    },
    {
      "epoch": 7.24,
      "learning_rate": 0.0001688720507042762,
      "loss": 0.0874,
      "step": 18100
    },
    {
      "epoch": 7.26,
      "learning_rate": 0.00016675279389392905,
      "loss": 0.1183,
      "step": 18150
    },
    {
      "epoch": 7.28,
      "learning_rate": 0.00016464024411603435,
      "loss": 0.1099,
      "step": 18200
    },
    {
      "epoch": 7.3,
      "learning_rate": 0.0001625345715737808,
      "loss": 0.1183,
      "step": 18250
    },
    {
      "epoch": 7.32,
      "learning_rate": 0.00016043594591627392,
      "loss": 0.1456,
      "step": 18300
    },
    {
      "epoch": 7.34,
      "learning_rate": 0.0001583445362248687,
      "loss": 0.1185,
      "step": 18350
    },
    {
      "epoch": 7.36,
      "learning_rate": 0.00015630211804928086,
      "loss": 0.0823,
      "step": 18400
    },
    {
      "epoch": 7.38,
      "learning_rate": 0.00015422549250609542,
      "loss": 0.1193,
      "step": 18450
    },
    {
      "epoch": 7.4,
      "learning_rate": 0.0001521565832906994,
      "loss": 0.0982,
      "step": 18500
    },
    {
      "epoch": 7.42,
      "learning_rate": 0.00015009555709026364,
      "loss": 0.1538,
      "step": 18550
    },
    {
      "epoch": 7.44,
      "learning_rate": 0.00014804257995684273,
      "loss": 0.1161,
      "step": 18600
    },
    {
      "epoch": 7.46,
      "learning_rate": 0.00014599781729399667,
      "loss": 0.0928,
      "step": 18650
    },
    {
      "epoch": 7.48,
      "learning_rate": 0.0001439614338434649,
      "loss": 0.1397,
      "step": 18700
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.00014193359367189345,
      "loss": 0.1524,
      "step": 18750
    },
    {
      "epoch": 7.52,
      "learning_rate": 0.0001399144601576161,
      "loss": 0.1006,
      "step": 18800
    },
    {
      "epoch": 7.54,
      "learning_rate": 0.000137904195977492,
      "loss": 0.1066,
      "step": 18850
    },
    {
      "epoch": 7.56,
      "learning_rate": 0.00013590296309379834,
      "loss": 0.1155,
      "step": 18900
    },
    {
      "epoch": 7.58,
      "learning_rate": 0.00013391092274118256,
      "loss": 0.09,
      "step": 18950
    },
    {
      "epoch": 7.6,
      "learning_rate": 0.00013192823541367078,
      "loss": 0.0766,
      "step": 19000
    },
    {
      "epoch": 7.62,
      "learning_rate": 0.000129955060851738,
      "loss": 0.1123,
      "step": 19050
    },
    {
      "epoch": 7.64,
      "learning_rate": 0.00012799155802943818,
      "loss": 0.0839,
      "step": 19100
    },
    {
      "epoch": 7.66,
      "learning_rate": 0.00012603788514159553,
      "loss": 0.1349,
      "step": 19150
    },
    {
      "epoch": 7.68,
      "learning_rate": 0.0001240941995910598,
      "loss": 0.1229,
      "step": 19200
    },
    {
      "epoch": 7.7,
      "learning_rate": 0.0001221606579760241,
      "loss": 0.1318,
      "step": 19250
    },
    {
      "epoch": 7.72,
      "learning_rate": 0.00012023741607740837,
      "loss": 0.124,
      "step": 19300
    },
    {
      "epoch": 7.74,
      "learning_rate": 0.00011832462884630888,
      "loss": 0.1481,
      "step": 19350
    },
    {
      "epoch": 7.76,
      "learning_rate": 0.00011642245039151345,
      "loss": 0.1024,
      "step": 19400
    },
    {
      "epoch": 7.78,
      "learning_rate": 0.0001145310339670857,
      "loss": 0.0985,
      "step": 19450
    },
    {
      "epoch": 7.8,
      "learning_rate": 0.00011265053196001793,
      "loss": 0.1173,
      "step": 19500
    },
    {
      "epoch": 7.82,
      "learning_rate": 0.0001107810958779531,
      "loss": 0.1126,
      "step": 19550
    },
    {
      "epoch": 7.84,
      "learning_rate": 0.00010892287633697847,
      "loss": 0.0899,
      "step": 19600
    },
    {
      "epoch": 7.86,
      "learning_rate": 0.00010707602304949115,
      "loss": 0.1293,
      "step": 19650
    },
    {
      "epoch": 7.88,
      "learning_rate": 0.0001052406848121357,
      "loss": 0.094,
      "step": 19700
    },
    {
      "epoch": 7.9,
      "learning_rate": 0.00010341700949381593,
      "loss": 0.1207,
      "step": 19750
    },
    {
      "epoch": 7.92,
      "learning_rate": 0.0001016051440237816,
      "loss": 0.116,
      "step": 19800
    },
    {
      "epoch": 7.94,
      "learning_rate": 9.980523437979085e-05,
      "loss": 0.145,
      "step": 19850
    },
    {
      "epoch": 7.96,
      "learning_rate": 9.801742557634872e-05,
      "loss": 0.0873,
      "step": 19900
    },
    {
      "epoch": 7.98,
      "learning_rate": 9.624186165302378e-05,
      "loss": 0.1202,
      "step": 19950
    },
    {
      "epoch": 8.0,
      "learning_rate": 9.447868566284355e-05,
      "loss": 0.1306,
      "step": 20000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.936,
      "eval_loss": 0.22573316097259521,
      "eval_runtime": 132.6149,
      "eval_samples_per_second": 75.406,
      "eval_steps_per_second": 4.713,
      "step": 20000
    },
    {
      "epoch": 8.02,
      "learning_rate": 9.272803966076882e-05,
      "loss": 0.0627,
      "step": 20050
    },
    {
      "epoch": 8.04,
      "learning_rate": 9.099006469224803e-05,
      "loss": 0.0855,
      "step": 20100
    },
    {
      "epoch": 8.06,
      "learning_rate": 8.926490078185472e-05,
      "loss": 0.0945,
      "step": 20150
    },
    {
      "epoch": 8.08,
      "learning_rate": 8.755268692200535e-05,
      "loss": 0.0888,
      "step": 20200
    },
    {
      "epoch": 8.1,
      "learning_rate": 8.585356106176093e-05,
      "loss": 0.1055,
      "step": 20250
    },
    {
      "epoch": 8.12,
      "learning_rate": 8.416766009571306e-05,
      "loss": 0.1047,
      "step": 20300
    },
    {
      "epoch": 8.14,
      "learning_rate": 8.249511985295474e-05,
      "loss": 0.0911,
      "step": 20350
    },
    {
      "epoch": 8.16,
      "learning_rate": 8.083607508613663e-05,
      "loss": 0.0679,
      "step": 20400
    },
    {
      "epoch": 8.18,
      "learning_rate": 7.919065946061043e-05,
      "loss": 0.1069,
      "step": 20450
    },
    {
      "epoch": 8.2,
      "learning_rate": 7.755900554366008e-05,
      "loss": 0.1176,
      "step": 20500
    },
    {
      "epoch": 8.22,
      "learning_rate": 7.594124479382086e-05,
      "loss": 0.0975,
      "step": 20550
    },
    {
      "epoch": 8.24,
      "learning_rate": 7.433750755028773e-05,
      "loss": 0.0936,
      "step": 20600
    },
    {
      "epoch": 8.26,
      "learning_rate": 7.277957518611719e-05,
      "loss": 0.087,
      "step": 20650
    },
    {
      "epoch": 8.28,
      "learning_rate": 7.12039845796604e-05,
      "loss": 0.0672,
      "step": 20700
    },
    {
      "epoch": 8.3,
      "learning_rate": 6.964279914947188e-05,
      "loss": 0.1241,
      "step": 20750
    },
    {
      "epoch": 8.32,
      "learning_rate": 6.809614467660574e-05,
      "loss": 0.0675,
      "step": 20800
    },
    {
      "epoch": 8.34,
      "learning_rate": 6.656414577139106e-05,
      "loss": 0.0568,
      "step": 20850
    },
    {
      "epoch": 8.36,
      "learning_rate": 6.504692586339217e-05,
      "loss": 0.0861,
      "step": 20900
    },
    {
      "epoch": 8.38,
      "learning_rate": 6.354460719146405e-05,
      "loss": 0.0744,
      "step": 20950
    },
    {
      "epoch": 8.4,
      "learning_rate": 6.205731079390395e-05,
      "loss": 0.1317,
      "step": 21000
    },
    {
      "epoch": 8.42,
      "learning_rate": 6.0585156498699975e-05,
      "loss": 0.106,
      "step": 21050
    },
    {
      "epoch": 8.44,
      "learning_rate": 5.912826291387613e-05,
      "loss": 0.0581,
      "step": 21100
    },
    {
      "epoch": 8.46,
      "learning_rate": 5.7686747417936864e-05,
      "loss": 0.0843,
      "step": 21150
    },
    {
      "epoch": 8.48,
      "learning_rate": 5.626072615041011e-05,
      "loss": 0.0788,
      "step": 21200
    },
    {
      "epoch": 8.5,
      "learning_rate": 5.485031400249002e-05,
      "loss": 0.0708,
      "step": 21250
    },
    {
      "epoch": 8.52,
      "learning_rate": 5.345562460778042e-05,
      "loss": 0.0954,
      "step": 21300
    },
    {
      "epoch": 8.54,
      "learning_rate": 5.2076770333139956e-05,
      "loss": 0.1097,
      "step": 21350
    },
    {
      "epoch": 8.56,
      "learning_rate": 5.0713862269628516e-05,
      "loss": 0.0538,
      "step": 21400
    },
    {
      "epoch": 8.58,
      "learning_rate": 4.936701022355733e-05,
      "loss": 0.0764,
      "step": 21450
    },
    {
      "epoch": 8.6,
      "learning_rate": 4.803632270764161e-05,
      "loss": 0.0731,
      "step": 21500
    },
    {
      "epoch": 8.62,
      "learning_rate": 4.672190693225853e-05,
      "loss": 0.1174,
      "step": 21550
    },
    {
      "epoch": 8.64,
      "learning_rate": 4.542386879680904e-05,
      "loss": 0.0934,
      "step": 21600
    },
    {
      "epoch": 8.66,
      "learning_rate": 4.414231288118598e-05,
      "loss": 0.0703,
      "step": 21650
    },
    {
      "epoch": 8.68,
      "learning_rate": 4.287734243734848e-05,
      "loss": 0.1005,
      "step": 21700
    },
    {
      "epoch": 8.7,
      "learning_rate": 4.162905938100289e-05,
      "loss": 0.1247,
      "step": 21750
    },
    {
      "epoch": 8.72,
      "learning_rate": 4.0397564283392005e-05,
      "loss": 0.0823,
      "step": 21800
    },
    {
      "epoch": 8.74,
      "learning_rate": 3.918295636319191e-05,
      "loss": 0.0866,
      "step": 21850
    },
    {
      "epoch": 8.76,
      "learning_rate": 3.798533347851843e-05,
      "loss": 0.0831,
      "step": 21900
    },
    {
      "epoch": 8.78,
      "learning_rate": 3.680479211904295e-05,
      "loss": 0.096,
      "step": 21950
    },
    {
      "epoch": 8.8,
      "learning_rate": 3.5641427398218243e-05,
      "loss": 0.0906,
      "step": 22000
    },
    {
      "epoch": 8.82,
      "learning_rate": 3.449533304561545e-05,
      "loss": 0.0891,
      "step": 22050
    },
    {
      "epoch": 8.84,
      "learning_rate": 3.3366601399372866e-05,
      "loss": 0.0919,
      "step": 22100
    },
    {
      "epoch": 8.86,
      "learning_rate": 3.225532339875592e-05,
      "loss": 0.1024,
      "step": 22150
    },
    {
      "epoch": 8.88,
      "learning_rate": 3.116158857683088e-05,
      "loss": 0.0791,
      "step": 22200
    },
    {
      "epoch": 8.9,
      "learning_rate": 3.0085485053250984e-05,
      "loss": 0.0707,
      "step": 22250
    },
    {
      "epoch": 8.92,
      "learning_rate": 2.902709952715732e-05,
      "loss": 0.0693,
      "step": 22300
    },
    {
      "epoch": 8.94,
      "learning_rate": 2.798651727019319e-05,
      "loss": 0.0846,
      "step": 22350
    },
    {
      "epoch": 8.96,
      "learning_rate": 2.696382211963422e-05,
      "loss": 0.1264,
      "step": 22400
    },
    {
      "epoch": 8.98,
      "learning_rate": 2.595909647163386e-05,
      "loss": 0.0944,
      "step": 22450
    },
    {
      "epoch": 9.0,
      "learning_rate": 2.497242127458474e-05,
      "loss": 0.0849,
      "step": 22500
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9372,
      "eval_loss": 0.2581612169742584,
      "eval_runtime": 132.2599,
      "eval_samples_per_second": 75.609,
      "eval_steps_per_second": 4.726,
      "step": 22500
    },
    {
      "epoch": 9.02,
      "learning_rate": 2.4003876022596904e-05,
      "loss": 0.0662,
      "step": 22550
    },
    {
      "epoch": 9.04,
      "learning_rate": 2.305353874909327e-05,
      "loss": 0.0709,
      "step": 22600
    },
    {
      "epoch": 9.06,
      "learning_rate": 2.2121486020522636e-05,
      "loss": 0.0425,
      "step": 22650
    },
    {
      "epoch": 9.08,
      "learning_rate": 2.1207792930190456e-05,
      "loss": 0.0385,
      "step": 22700
    },
    {
      "epoch": 9.1,
      "learning_rate": 2.031253309220954e-05,
      "loss": 0.0708,
      "step": 22750
    },
    {
      "epoch": 9.12,
      "learning_rate": 1.943577863556861e-05,
      "loss": 0.0898,
      "step": 22800
    },
    {
      "epoch": 9.14,
      "learning_rate": 1.8577600198321028e-05,
      "loss": 0.0911,
      "step": 22850
    },
    {
      "epoch": 9.16,
      "learning_rate": 1.7754674424872773e-05,
      "loss": 0.0721,
      "step": 22900
    },
    {
      "epoch": 9.18,
      "learning_rate": 1.69334790393732e-05,
      "loss": 0.0756,
      "step": 22950
    },
    {
      "epoch": 9.2,
      "learning_rate": 1.6131061277685822e-05,
      "loss": 0.0807,
      "step": 23000
    },
    {
      "epoch": 9.22,
      "learning_rate": 1.5347485788730464e-05,
      "loss": 0.0915,
      "step": 23050
    },
    {
      "epoch": 9.24,
      "learning_rate": 1.4582815703349406e-05,
      "loss": 0.0759,
      "step": 23100
    },
    {
      "epoch": 9.26,
      "learning_rate": 1.38371126292208e-05,
      "loss": 0.0553,
      "step": 23150
    },
    {
      "epoch": 9.28,
      "learning_rate": 1.3110436645895063e-05,
      "loss": 0.0574,
      "step": 23200
    },
    {
      "epoch": 9.3,
      "learning_rate": 1.2402846299954445e-05,
      "loss": 0.0973,
      "step": 23250
    },
    {
      "epoch": 9.32,
      "learning_rate": 1.1714398600296133e-05,
      "loss": 0.0692,
      "step": 23300
    },
    {
      "epoch": 9.34,
      "learning_rate": 1.1045149013539257e-05,
      "loss": 0.0708,
      "step": 23350
    },
    {
      "epoch": 9.36,
      "learning_rate": 1.0395151459555607e-05,
      "loss": 0.1098,
      "step": 23400
    },
    {
      "epoch": 9.38,
      "learning_rate": 9.76445830712605e-06,
      "loss": 0.0761,
      "step": 23450
    },
    {
      "epoch": 9.4,
      "learning_rate": 9.153120369721047e-06,
      "loss": 0.0803,
      "step": 23500
    },
    {
      "epoch": 9.42,
      "learning_rate": 8.561186901406532e-06,
      "loss": 0.0692,
      "step": 23550
    },
    {
      "epoch": 9.44,
      "learning_rate": 7.9887055928759e-06,
      "loss": 0.0811,
      "step": 23600
    },
    {
      "epoch": 9.46,
      "learning_rate": 7.4357225676076056e-06,
      "loss": 0.0989,
      "step": 23650
    },
    {
      "epoch": 9.48,
      "learning_rate": 6.902282378149049e-06,
      "loss": 0.0747,
      "step": 23700
    },
    {
      "epoch": 9.5,
      "learning_rate": 6.388428002527036e-06,
      "loss": 0.0433,
      "step": 23750
    },
    {
      "epoch": 9.52,
      "learning_rate": 5.894200840785402e-06,
      "loss": 0.0421,
      "step": 23800
    },
    {
      "epoch": 9.54,
      "learning_rate": 5.419640711649188e-06,
      "loss": 0.1055,
      "step": 23850
    },
    {
      "epoch": 9.56,
      "learning_rate": 4.964785849316644e-06,
      "loss": 0.0832,
      "step": 23900
    },
    {
      "epoch": 9.58,
      "learning_rate": 4.529672900378923e-06,
      "loss": 0.0741,
      "step": 23950
    },
    {
      "epoch": 9.6,
      "learning_rate": 4.114336920867434e-06,
      "loss": 0.067,
      "step": 24000
    },
    {
      "epoch": 9.62,
      "learning_rate": 3.7188113734293592e-06,
      "loss": 0.101,
      "step": 24050
    },
    {
      "epoch": 9.64,
      "learning_rate": 3.3431281246318613e-06,
      "loss": 0.0772,
      "step": 24100
    },
    {
      "epoch": 9.66,
      "learning_rate": 2.9873174423945824e-06,
      "loss": 0.0935,
      "step": 24150
    },
    {
      "epoch": 9.68,
      "learning_rate": 2.6514079935509583e-06,
      "loss": 0.0859,
      "step": 24200
    },
    {
      "epoch": 9.7,
      "learning_rate": 2.335426841538618e-06,
      "loss": 0.0631,
      "step": 24250
    },
    {
      "epoch": 9.72,
      "learning_rate": 2.0393994442191033e-06,
      "loss": 0.0765,
      "step": 24300
    },
    {
      "epoch": 9.74,
      "learning_rate": 1.7633496518265347e-06,
      "loss": 0.0937,
      "step": 24350
    },
    {
      "epoch": 9.76,
      "learning_rate": 1.507299705046289e-06,
      "loss": 0.0827,
      "step": 24400
    },
    {
      "epoch": 9.78,
      "learning_rate": 1.2712702332229875e-06,
      "loss": 0.0815,
      "step": 24450
    },
    {
      "epoch": 9.8,
      "learning_rate": 1.0552802526983819e-06,
      "loss": 0.0385,
      "step": 24500
    },
    {
      "epoch": 9.82,
      "learning_rate": 8.593471652794949e-07,
      "loss": 0.0743,
      "step": 24550
    },
    {
      "epoch": 9.84,
      "learning_rate": 6.834867568362991e-07,
      "loss": 0.0868,
      "step": 24600
    },
    {
      "epoch": 9.86,
      "learning_rate": 5.277131960301218e-07,
      "loss": 0.0709,
      "step": 24650
    },
    {
      "epoch": 9.88,
      "learning_rate": 3.920390331719481e-07,
      "loss": 0.0814,
      "step": 24700
    },
    {
      "epoch": 9.9,
      "learning_rate": 2.7647519921136786e-07,
      "loss": 0.0834,
      "step": 24750
    },
    {
      "epoch": 9.92,
      "learning_rate": 1.810310048558639e-07,
      "loss": 0.0773,
      "step": 24800
    },
    {
      "epoch": 9.94,
      "learning_rate": 1.0571413982068956e-07,
      "loss": 0.0787,
      "step": 24850
    },
    {
      "epoch": 9.96,
      "learning_rate": 5.1437002834192616e-08,
      "loss": 0.0584,
      "step": 24900
    },
    {
      "epoch": 9.98,
      "learning_rate": 1.5988588703552686e-08,
      "loss": 0.0847,
      "step": 24950
    },
    {
      "epoch": 10.0,
      "learning_rate": 6.808009741787746e-10,
      "loss": 0.0609,
      "step": 25000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9364,
      "eval_loss": 0.2748859226703644,
      "eval_runtime": 132.1385,
      "eval_samples_per_second": 75.678,
      "eval_steps_per_second": 4.73,
      "step": 25000
    }
  ],
  "logging_steps": 50,
  "max_steps": 25000,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 5.833970619727296e+16,
  "trial_name": null,
  "trial_params": null
}