{
  "best_metric": 0.4691739691294871,
  "best_model_checkpoint": "./whisper-small-ar_tsize_0.8/checkpoint-5000",
  "epoch": 1.6103059581320451,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 74.5753402709961,
      "learning_rate": 5.000000000000001e-07,
      "loss": 3.3084,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": 23.875898361206055,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.626,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 22.36936378479004,
      "learning_rate": 1.5e-06,
      "loss": 1.9536,
      "step": 75
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.3158016204834,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.315,
      "step": 100
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.385392189025879,
      "learning_rate": 2.5e-06,
      "loss": 1.1058,
      "step": 125
    },
    {
      "epoch": 0.05,
      "grad_norm": 14.316064834594727,
      "learning_rate": 3e-06,
      "loss": 1.0312,
      "step": 150
    },
    {
      "epoch": 0.06,
      "grad_norm": 16.03868293762207,
      "learning_rate": 3.5e-06,
      "loss": 0.9277,
      "step": 175
    },
    {
      "epoch": 0.06,
      "grad_norm": 16.684885025024414,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.867,
      "step": 200
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.64684772491455,
      "learning_rate": 4.5e-06,
      "loss": 0.8164,
      "step": 225
    },
    {
      "epoch": 0.08,
      "grad_norm": 16.835796356201172,
      "learning_rate": 5e-06,
      "loss": 0.7126,
      "step": 250
    },
    {
      "epoch": 0.09,
      "grad_norm": 17.927705764770508,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.6566,
      "step": 275
    },
    {
      "epoch": 0.1,
      "grad_norm": 20.162878036499023,
      "learning_rate": 6e-06,
      "loss": 0.5585,
      "step": 300
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.939558982849121,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.4888,
      "step": 325
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.801262855529785,
      "learning_rate": 7e-06,
      "loss": 0.4609,
      "step": 350
    },
    {
      "epoch": 0.12,
      "grad_norm": 14.835463523864746,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4432,
      "step": 375
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.00851821899414,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.4361,
      "step": 400
    },
    {
      "epoch": 0.14,
      "grad_norm": 15.16860294342041,
      "learning_rate": 8.5e-06,
      "loss": 0.4177,
      "step": 425
    },
    {
      "epoch": 0.14,
      "grad_norm": 15.027597427368164,
      "learning_rate": 9e-06,
      "loss": 0.4447,
      "step": 450
    },
    {
      "epoch": 0.15,
      "grad_norm": 15.810965538024902,
      "learning_rate": 9.5e-06,
      "loss": 0.4198,
      "step": 475
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.27136993408203,
      "learning_rate": 1e-05,
      "loss": 0.4265,
      "step": 500
    },
    {
      "epoch": 0.17,
      "grad_norm": 12.16751766204834,
      "learning_rate": 9.944444444444445e-06,
      "loss": 0.4191,
      "step": 525
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.931772232055664,
      "learning_rate": 9.88888888888889e-06,
      "loss": 0.3924,
      "step": 550
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.04465389251709,
      "learning_rate": 9.833333333333333e-06,
      "loss": 0.4176,
      "step": 575
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.589746475219727,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.4358,
      "step": 600
    },
    {
      "epoch": 0.2,
      "grad_norm": 14.16213607788086,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.3828,
      "step": 625
    },
    {
      "epoch": 0.21,
      "grad_norm": 15.552674293518066,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.4139,
      "step": 650
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.745460510253906,
      "learning_rate": 9.611111111111112e-06,
      "loss": 0.3915,
      "step": 675
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.602766990661621,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.365,
      "step": 700
    },
    {
      "epoch": 0.23,
      "grad_norm": 14.100479125976562,
      "learning_rate": 9.5e-06,
      "loss": 0.4218,
      "step": 725
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.610438346862793,
      "learning_rate": 9.444444444444445e-06,
      "loss": 0.3358,
      "step": 750
    },
    {
      "epoch": 0.25,
      "grad_norm": 11.454590797424316,
      "learning_rate": 9.38888888888889e-06,
      "loss": 0.4012,
      "step": 775
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.808696746826172,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.3754,
      "step": 800
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.447662353515625,
      "learning_rate": 9.277777777777778e-06,
      "loss": 0.3407,
      "step": 825
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.822239875793457,
      "learning_rate": 9.222222222222224e-06,
      "loss": 0.3557,
      "step": 850
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.579741477966309,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.3557,
      "step": 875
    },
    {
      "epoch": 0.29,
      "grad_norm": 13.126590728759766,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.3588,
      "step": 900
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.920930862426758,
      "learning_rate": 9.055555555555556e-06,
      "loss": 0.3761,
      "step": 925
    },
    {
      "epoch": 0.31,
      "grad_norm": 16.436307907104492,
      "learning_rate": 9e-06,
      "loss": 0.3613,
      "step": 950
    },
    {
      "epoch": 0.31,
      "grad_norm": 16.763071060180664,
      "learning_rate": 8.944444444444446e-06,
      "loss": 0.3718,
      "step": 975
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.7616376876831055,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.2768,
      "step": 1000
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.3424592614173889,
      "eval_runtime": 2548.0078,
      "eval_samples_per_second": 3.046,
      "eval_steps_per_second": 0.381,
      "eval_wer": 0.5745963257862194,
      "step": 1000
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.518533706665039,
      "learning_rate": 8.833333333333334e-06,
      "loss": 0.3837,
      "step": 1025
    },
    {
      "epoch": 0.34,
      "grad_norm": 15.829094886779785,
      "learning_rate": 8.777777777777778e-06,
      "loss": 0.324,
      "step": 1050
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.936345100402832,
      "learning_rate": 8.722222222222224e-06,
      "loss": 0.3251,
      "step": 1075
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.695411682128906,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.3596,
      "step": 1100
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.362096786499023,
      "learning_rate": 8.611111111111112e-06,
      "loss": 0.3336,
      "step": 1125
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.786664009094238,
      "learning_rate": 8.555555555555556e-06,
      "loss": 0.3427,
      "step": 1150
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.013471603393555,
      "learning_rate": 8.5e-06,
      "loss": 0.3265,
      "step": 1175
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.64340877532959,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.3561,
      "step": 1200
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.622845649719238,
      "learning_rate": 8.38888888888889e-06,
      "loss": 0.3457,
      "step": 1225
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.175597190856934,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.3605,
      "step": 1250
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.138216018676758,
      "learning_rate": 8.277777777777778e-06,
      "loss": 0.3493,
      "step": 1275
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.549818992614746,
      "learning_rate": 8.222222222222222e-06,
      "loss": 0.2964,
      "step": 1300
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.058015823364258,
      "learning_rate": 8.166666666666668e-06,
      "loss": 0.3219,
      "step": 1325
    },
    {
      "epoch": 0.43,
      "grad_norm": 13.490798950195312,
      "learning_rate": 8.111111111111112e-06,
      "loss": 0.2941,
      "step": 1350
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.73831844329834,
      "learning_rate": 8.055555555555557e-06,
      "loss": 0.329,
      "step": 1375
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.883594512939453,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.3076,
      "step": 1400
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.517281532287598,
      "learning_rate": 7.944444444444445e-06,
      "loss": 0.3153,
      "step": 1425
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.685423851013184,
      "learning_rate": 7.88888888888889e-06,
      "loss": 0.2991,
      "step": 1450
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.92092514038086,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.3064,
      "step": 1475
    },
    {
      "epoch": 0.48,
      "grad_norm": 13.166866302490234,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.3195,
      "step": 1500
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.179815292358398,
      "learning_rate": 7.722222222222223e-06,
      "loss": 0.3214,
      "step": 1525
    },
    {
      "epoch": 0.5,
      "grad_norm": 12.424599647521973,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.3062,
      "step": 1550
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.103777885437012,
      "learning_rate": 7.611111111111111e-06,
      "loss": 0.2858,
      "step": 1575
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.597440719604492,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.3194,
      "step": 1600
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.370293617248535,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3285,
      "step": 1625
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.53341007232666,
      "learning_rate": 7.444444444444445e-06,
      "loss": 0.2848,
      "step": 1650
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.32474422454834,
      "learning_rate": 7.38888888888889e-06,
      "loss": 0.304,
      "step": 1675
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.368809700012207,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.3135,
      "step": 1700
    },
    {
      "epoch": 0.56,
      "grad_norm": 13.282475471496582,
      "learning_rate": 7.277777777777778e-06,
      "loss": 0.2797,
      "step": 1725
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.41550350189209,
      "learning_rate": 7.222222222222223e-06,
      "loss": 0.3019,
      "step": 1750
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.789618492126465,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.3085,
      "step": 1775
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.927020072937012,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.271,
      "step": 1800
    },
    {
      "epoch": 0.59,
      "grad_norm": 14.091415405273438,
      "learning_rate": 7.055555555555557e-06,
      "loss": 0.2985,
      "step": 1825
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.496126174926758,
      "learning_rate": 7e-06,
      "loss": 0.3029,
      "step": 1850
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.859543323516846,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.271,
      "step": 1875
    },
    {
      "epoch": 0.61,
      "grad_norm": 16.182022094726562,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.3039,
      "step": 1900
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.995250225067139,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.2867,
      "step": 1925
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.519378662109375,
      "learning_rate": 6.777777777777779e-06,
      "loss": 0.3339,
      "step": 1950
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.880196571350098,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.2873,
      "step": 1975
    },
    {
      "epoch": 0.64,
      "grad_norm": 16.335350036621094,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2977,
      "step": 2000
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.2855232059955597,
      "eval_runtime": 2086.2971,
      "eval_samples_per_second": 3.72,
      "eval_steps_per_second": 0.465,
      "eval_wer": 0.5296694986877808,
      "step": 2000
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.254558563232422,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 0.2668,
      "step": 2025
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.6058931350708,
      "learning_rate": 6.555555555555556e-06,
      "loss": 0.3302,
      "step": 2050
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.349474906921387,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.3228,
      "step": 2075
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.22000789642334,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.2944,
      "step": 2100
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.740598201751709,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 0.2875,
      "step": 2125
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.818449974060059,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.2918,
      "step": 2150
    },
    {
      "epoch": 0.7,
      "grad_norm": 17.033401489257812,
      "learning_rate": 6.277777777777778e-06,
      "loss": 0.2803,
      "step": 2175
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.139738082885742,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.2964,
      "step": 2200
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.335383892059326,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.2548,
      "step": 2225
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.209503173828125,
      "learning_rate": 6.111111111111112e-06,
      "loss": 0.2768,
      "step": 2250
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.060270309448242,
      "learning_rate": 6.055555555555555e-06,
      "loss": 0.3085,
      "step": 2275
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.265747547149658,
      "learning_rate": 6e-06,
      "loss": 0.2695,
      "step": 2300
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.466809272766113,
      "learning_rate": 5.944444444444445e-06,
      "loss": 0.2711,
      "step": 2325
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.156736373901367,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.2893,
      "step": 2350
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.08906364440918,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.2621,
      "step": 2375
    },
    {
      "epoch": 0.77,
      "grad_norm": 14.466202735900879,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.2988,
      "step": 2400
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.54973030090332,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.2747,
      "step": 2425
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.84890079498291,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.2845,
      "step": 2450
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.711421966552734,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.2513,
      "step": 2475
    },
    {
      "epoch": 0.81,
      "grad_norm": 13.456315994262695,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.2653,
      "step": 2500
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.786430358886719,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.2562,
      "step": 2525
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.972217559814453,
      "learning_rate": 5.444444444444445e-06,
      "loss": 0.2499,
      "step": 2550
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.539140224456787,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.2587,
      "step": 2575
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.782011032104492,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.2563,
      "step": 2600
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.317708015441895,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 0.2669,
      "step": 2625
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.227994441986084,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.2516,
      "step": 2650
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.91267728805542,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.2708,
      "step": 2675
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.71472454071045,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.2752,
      "step": 2700
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.538637161254883,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.2541,
      "step": 2725
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.09244441986084,
      "learning_rate": 5e-06,
      "loss": 0.2667,
      "step": 2750
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.292573928833008,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.3007,
      "step": 2775
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.321602821350098,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.245,
      "step": 2800
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.209299087524414,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.2337,
      "step": 2825
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.65815258026123,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.2553,
      "step": 2850
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.273401260375977,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.247,
      "step": 2875
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.994607925415039,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.2761,
      "step": 2900
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.799399375915527,
      "learning_rate": 4.611111111111112e-06,
      "loss": 0.2637,
      "step": 2925
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.812761306762695,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.2926,
      "step": 2950
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.971693992614746,
      "learning_rate": 4.5e-06,
      "loss": 0.2571,
      "step": 2975
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.44893741607666,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.2541,
      "step": 3000
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.254996120929718,
      "eval_runtime": 2083.5916,
      "eval_samples_per_second": 3.725,
      "eval_steps_per_second": 0.466,
      "eval_wer": 0.5009118811440773,
      "step": 3000
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.639594078063965,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.2846,
      "step": 3025
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.723342895507812,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.2692,
      "step": 3050
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.255222320556641,
      "learning_rate": 4.277777777777778e-06,
      "loss": 0.2406,
      "step": 3075
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.54223346710205,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.2596,
      "step": 3100
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.074874401092529,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.2201,
      "step": 3125
    },
    {
      "epoch": 1.01,
      "grad_norm": 5.8523640632629395,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.1695,
      "step": 3150
    },
    {
      "epoch": 1.02,
      "grad_norm": 6.729598522186279,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.1396,
      "step": 3175
    },
    {
      "epoch": 1.03,
      "grad_norm": 7.361201286315918,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.1583,
      "step": 3200
    },
    {
      "epoch": 1.04,
      "grad_norm": 5.176375389099121,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.1543,
      "step": 3225
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.241930961608887,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.1555,
      "step": 3250
    },
    {
      "epoch": 1.05,
      "grad_norm": 6.182642936706543,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.1441,
      "step": 3275
    },
    {
      "epoch": 1.06,
      "grad_norm": 13.024508476257324,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.1503,
      "step": 3300
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.090968608856201,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.1714,
      "step": 3325
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.179136276245117,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.1663,
      "step": 3350
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.39560604095459,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.1643,
      "step": 3375
    },
    {
      "epoch": 1.1,
      "grad_norm": 6.854971408843994,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.1767,
      "step": 3400
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.17493200302124,
      "learning_rate": 3.5e-06,
      "loss": 0.1408,
      "step": 3425
    },
    {
      "epoch": 1.11,
      "grad_norm": 5.211622714996338,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.1592,
      "step": 3450
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.639506816864014,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.1428,
      "step": 3475
    },
    {
      "epoch": 1.13,
      "grad_norm": 7.832518100738525,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.1564,
      "step": 3500
    },
    {
      "epoch": 1.14,
      "grad_norm": 5.339199542999268,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.1722,
      "step": 3525
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.612499713897705,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.1667,
      "step": 3550
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.03056812286377,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.1636,
      "step": 3575
    },
    {
      "epoch": 1.16,
      "grad_norm": 6.799817085266113,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.1454,
      "step": 3600
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.477158069610596,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.1604,
      "step": 3625
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.446293354034424,
      "learning_rate": 3e-06,
      "loss": 0.1545,
      "step": 3650
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.985503673553467,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.1508,
      "step": 3675
    },
    {
      "epoch": 1.19,
      "grad_norm": 6.837751388549805,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.1587,
      "step": 3700
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.51887845993042,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.1485,
      "step": 3725
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.242449760437012,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.1656,
      "step": 3750
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.832524299621582,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.1594,
      "step": 3775
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.77171516418457,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.1495,
      "step": 3800
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.725485801696777,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.1473,
      "step": 3825
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.141379356384277,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.1421,
      "step": 3850
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.970383167266846,
      "learning_rate": 2.5e-06,
      "loss": 0.1423,
      "step": 3875
    },
    {
      "epoch": 1.26,
      "grad_norm": 6.767889499664307,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.1741,
      "step": 3900
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.690521240234375,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.1614,
      "step": 3925
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.897395133972168,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.153,
      "step": 3950
    },
    {
      "epoch": 1.28,
      "grad_norm": 7.075612545013428,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.1563,
      "step": 3975
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.673244953155518,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.1512,
      "step": 4000
    },
    {
      "epoch": 1.29,
      "eval_loss": 0.2438921481370926,
      "eval_runtime": 2435.1055,
      "eval_samples_per_second": 3.187,
      "eval_steps_per_second": 0.399,
      "eval_wer": 0.47462301499043635,
      "step": 4000
    },
    {
      "epoch": 1.3,
      "grad_norm": 12.350273132324219,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.1631,
      "step": 4025
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.359116554260254,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.149,
      "step": 4050
    },
    {
      "epoch": 1.31,
      "grad_norm": 6.066259860992432,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.1642,
      "step": 4075
    },
    {
      "epoch": 1.32,
      "grad_norm": 5.781357765197754,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.1451,
      "step": 4100
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.008292198181152,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.1498,
      "step": 4125
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.501078605651855,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.1614,
      "step": 4150
    },
    {
      "epoch": 1.34,
      "grad_norm": 5.542455673217773,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.1538,
      "step": 4175
    },
    {
      "epoch": 1.35,
      "grad_norm": 6.979973316192627,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.151,
      "step": 4200
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.080177307128906,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.1462,
      "step": 4225
    },
    {
      "epoch": 1.37,
      "grad_norm": 7.01210355758667,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.1578,
      "step": 4250
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.484158515930176,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.1468,
      "step": 4275
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.0456132888793945,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.1509,
      "step": 4300
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.766803741455078,
      "learning_rate": 1.5e-06,
      "loss": 0.1495,
      "step": 4325
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.22208023071289,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.1621,
      "step": 4350
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.47338581085205,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.1585,
      "step": 4375
    },
    {
      "epoch": 1.42,
      "grad_norm": 7.198250770568848,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.1726,
      "step": 4400
    },
    {
      "epoch": 1.43,
      "grad_norm": 5.910817623138428,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.1451,
      "step": 4425
    },
    {
      "epoch": 1.43,
      "grad_norm": 5.221580982208252,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.1469,
      "step": 4450
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.902259826660156,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.1522,
      "step": 4475
    },
    {
      "epoch": 1.45,
      "grad_norm": 7.130344390869141,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.1477,
      "step": 4500
    },
    {
      "epoch": 1.46,
      "grad_norm": 4.117496490478516,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.1308,
      "step": 4525
    },
    {
      "epoch": 1.47,
      "grad_norm": 4.9565935134887695,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.143,
      "step": 4550
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.52632999420166,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.1488,
      "step": 4575
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.335168838500977,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.1467,
      "step": 4600
    },
    {
      "epoch": 1.49,
      "grad_norm": 4.118008613586426,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.1602,
      "step": 4625
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.684025764465332,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.1767,
      "step": 4650
    },
    {
      "epoch": 1.51,
      "grad_norm": 7.431414604187012,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.1622,
      "step": 4675
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.669175148010254,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.1587,
      "step": 4700
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.074918270111084,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.1389,
      "step": 4725
    },
    {
      "epoch": 1.53,
      "grad_norm": 4.829873561859131,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.1273,
      "step": 4750
    },
    {
      "epoch": 1.54,
      "grad_norm": 6.028909683227539,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.1522,
      "step": 4775
    },
    {
      "epoch": 1.55,
      "grad_norm": 6.762262344360352,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.1484,
      "step": 4800
    },
    {
      "epoch": 1.55,
      "grad_norm": 6.13216495513916,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.1423,
      "step": 4825
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.262957572937012,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.1463,
      "step": 4850
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.5338327884674072,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.1241,
      "step": 4875
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.249900817871094,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.1419,
      "step": 4900
    },
    {
      "epoch": 1.59,
      "grad_norm": 6.7409539222717285,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.1308,
      "step": 4925
    },
    {
      "epoch": 1.59,
      "grad_norm": 5.887229919433594,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.1581,
      "step": 4950
    },
    {
      "epoch": 1.6,
      "grad_norm": 6.0119099617004395,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.1407,
      "step": 4975
    },
    {
      "epoch": 1.61,
      "grad_norm": 5.794888019561768,
      "learning_rate": 0.0,
      "loss": 0.1415,
      "step": 5000
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.23594366014003754,
      "eval_runtime": 2284.612,
      "eval_samples_per_second": 3.397,
      "eval_steps_per_second": 0.425,
      "eval_wer": 0.4691739691294871,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "total_flos": 1.15419730857984e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}