gom-stt-v3 / trainer_state.json
thak123's picture
Upload 9 files
551659a
{
"best_metric": 0.09221997112035751,
"best_model_checkpoint": "./openai/whisper-small-gom-LDC-v1.3-repeating-not fixed/checkpoint-11000",
"epoch": 2.846790890269151,
"global_step": 11000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 3.36e-07,
"loss": 4.6764,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 7.2e-07,
"loss": 3.7914,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 1.12e-06,
"loss": 2.6046,
"step": 75
},
{
"epoch": 0.03,
"learning_rate": 1.5199999999999998e-06,
"loss": 1.6768,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 1.92e-06,
"loss": 1.3718,
"step": 125
},
{
"epoch": 0.04,
"learning_rate": 2.32e-06,
"loss": 1.2488,
"step": 150
},
{
"epoch": 0.05,
"learning_rate": 2.7200000000000002e-06,
"loss": 1.253,
"step": 175
},
{
"epoch": 0.05,
"learning_rate": 3.1199999999999998e-06,
"loss": 1.1124,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 3.5199999999999998e-06,
"loss": 1.0475,
"step": 225
},
{
"epoch": 0.06,
"learning_rate": 3.92e-06,
"loss": 0.9747,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 4.32e-06,
"loss": 0.8999,
"step": 275
},
{
"epoch": 0.08,
"learning_rate": 4.72e-06,
"loss": 0.8971,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 5.12e-06,
"loss": 0.749,
"step": 325
},
{
"epoch": 0.09,
"learning_rate": 5.52e-06,
"loss": 0.6336,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 5.92e-06,
"loss": 0.4841,
"step": 375
},
{
"epoch": 0.1,
"learning_rate": 6.32e-06,
"loss": 0.4429,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 6.719999999999999e-06,
"loss": 0.4465,
"step": 425
},
{
"epoch": 0.12,
"learning_rate": 7.12e-06,
"loss": 0.4767,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 7.519999999999999e-06,
"loss": 0.4209,
"step": 475
},
{
"epoch": 0.13,
"learning_rate": 7.92e-06,
"loss": 0.4128,
"step": 500
},
{
"epoch": 0.13,
"eval_loss": 0.40222010016441345,
"eval_runtime": 4775.0561,
"eval_samples_per_second": 2.284,
"eval_steps_per_second": 0.286,
"eval_wer": 73.57266936884771,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 7.98557518932564e-06,
"loss": 0.3573,
"step": 525
},
{
"epoch": 0.14,
"learning_rate": 7.967544175982689e-06,
"loss": 0.411,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 7.94951316263974e-06,
"loss": 0.3807,
"step": 575
},
{
"epoch": 0.16,
"learning_rate": 7.93148214929679e-06,
"loss": 0.3108,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 7.91345113595384e-06,
"loss": 0.3786,
"step": 625
},
{
"epoch": 0.17,
"learning_rate": 7.895420122610891e-06,
"loss": 0.3971,
"step": 650
},
{
"epoch": 0.17,
"learning_rate": 7.87738910926794e-06,
"loss": 0.361,
"step": 675
},
{
"epoch": 0.18,
"learning_rate": 7.85935809592499e-06,
"loss": 0.3319,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 7.84132708258204e-06,
"loss": 0.3615,
"step": 725
},
{
"epoch": 0.19,
"learning_rate": 7.823296069239091e-06,
"loss": 0.3228,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 7.805265055896142e-06,
"loss": 0.3479,
"step": 775
},
{
"epoch": 0.21,
"learning_rate": 7.787234042553192e-06,
"loss": 0.3281,
"step": 800
},
{
"epoch": 0.21,
"learning_rate": 7.769203029210241e-06,
"loss": 0.2795,
"step": 825
},
{
"epoch": 0.22,
"learning_rate": 7.751172015867292e-06,
"loss": 0.2753,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 7.73314100252434e-06,
"loss": 0.3094,
"step": 875
},
{
"epoch": 0.23,
"learning_rate": 7.715109989181391e-06,
"loss": 0.2997,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 7.697078975838442e-06,
"loss": 0.3097,
"step": 925
},
{
"epoch": 0.25,
"learning_rate": 7.679047962495492e-06,
"loss": 0.2546,
"step": 950
},
{
"epoch": 0.25,
"learning_rate": 7.661016949152543e-06,
"loss": 0.32,
"step": 975
},
{
"epoch": 0.26,
"learning_rate": 7.642985935809593e-06,
"loss": 0.2796,
"step": 1000
},
{
"epoch": 0.26,
"eval_loss": 0.2927670180797577,
"eval_runtime": 4502.0818,
"eval_samples_per_second": 2.423,
"eval_steps_per_second": 0.303,
"eval_wer": 51.43794634240494,
"step": 1000
},
{
"epoch": 0.27,
"learning_rate": 7.624954922466642e-06,
"loss": 0.2974,
"step": 1025
},
{
"epoch": 0.27,
"learning_rate": 7.606923909123692e-06,
"loss": 0.2887,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 7.588892895780742e-06,
"loss": 0.2648,
"step": 1075
},
{
"epoch": 0.28,
"learning_rate": 7.570861882437793e-06,
"loss": 0.2784,
"step": 1100
},
{
"epoch": 0.29,
"learning_rate": 7.552830869094843e-06,
"loss": 0.241,
"step": 1125
},
{
"epoch": 0.3,
"learning_rate": 7.534799855751893e-06,
"loss": 0.2438,
"step": 1150
},
{
"epoch": 0.3,
"learning_rate": 7.516768842408943e-06,
"loss": 0.3066,
"step": 1175
},
{
"epoch": 0.31,
"learning_rate": 7.498737829065994e-06,
"loss": 0.2939,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 7.4807068157230435e-06,
"loss": 0.29,
"step": 1225
},
{
"epoch": 0.32,
"learning_rate": 7.462675802380093e-06,
"loss": 0.2659,
"step": 1250
},
{
"epoch": 0.33,
"learning_rate": 7.444644789037144e-06,
"loss": 0.2268,
"step": 1275
},
{
"epoch": 0.34,
"learning_rate": 7.426613775694193e-06,
"loss": 0.2845,
"step": 1300
},
{
"epoch": 0.34,
"learning_rate": 7.408582762351244e-06,
"loss": 0.2655,
"step": 1325
},
{
"epoch": 0.35,
"learning_rate": 7.3905517490082935e-06,
"loss": 0.2648,
"step": 1350
},
{
"epoch": 0.36,
"learning_rate": 7.372520735665344e-06,
"loss": 0.2757,
"step": 1375
},
{
"epoch": 0.36,
"learning_rate": 7.3544897223223945e-06,
"loss": 0.254,
"step": 1400
},
{
"epoch": 0.37,
"learning_rate": 7.336458708979444e-06,
"loss": 0.2316,
"step": 1425
},
{
"epoch": 0.38,
"learning_rate": 7.318427695636495e-06,
"loss": 0.2831,
"step": 1450
},
{
"epoch": 0.38,
"learning_rate": 7.300396682293545e-06,
"loss": 0.2383,
"step": 1475
},
{
"epoch": 0.39,
"learning_rate": 7.282365668950595e-06,
"loss": 0.2821,
"step": 1500
},
{
"epoch": 0.39,
"eval_loss": 0.24086758494377136,
"eval_runtime": 4425.709,
"eval_samples_per_second": 2.464,
"eval_steps_per_second": 0.308,
"eval_wer": 45.50472881683073,
"step": 1500
},
{
"epoch": 0.39,
"learning_rate": 7.2643346556076454e-06,
"loss": 0.2544,
"step": 1525
},
{
"epoch": 0.4,
"learning_rate": 7.246303642264694e-06,
"loss": 0.2726,
"step": 1550
},
{
"epoch": 0.41,
"learning_rate": 7.228272628921745e-06,
"loss": 0.2381,
"step": 1575
},
{
"epoch": 0.41,
"learning_rate": 7.210241615578795e-06,
"loss": 0.2367,
"step": 1600
},
{
"epoch": 0.42,
"learning_rate": 7.192210602235845e-06,
"loss": 0.2094,
"step": 1625
},
{
"epoch": 0.43,
"learning_rate": 7.1741795888928955e-06,
"loss": 0.2214,
"step": 1650
},
{
"epoch": 0.43,
"learning_rate": 7.156148575549946e-06,
"loss": 0.2174,
"step": 1675
},
{
"epoch": 0.44,
"learning_rate": 7.138117562206996e-06,
"loss": 0.2455,
"step": 1700
},
{
"epoch": 0.45,
"learning_rate": 7.120086548864046e-06,
"loss": 0.2173,
"step": 1725
},
{
"epoch": 0.45,
"learning_rate": 7.102055535521096e-06,
"loss": 0.2356,
"step": 1750
},
{
"epoch": 0.46,
"learning_rate": 7.084024522178146e-06,
"loss": 0.2346,
"step": 1775
},
{
"epoch": 0.47,
"learning_rate": 7.065993508835197e-06,
"loss": 0.1862,
"step": 1800
},
{
"epoch": 0.47,
"learning_rate": 7.047962495492246e-06,
"loss": 0.2338,
"step": 1825
},
{
"epoch": 0.48,
"learning_rate": 7.029931482149296e-06,
"loss": 0.2051,
"step": 1850
},
{
"epoch": 0.49,
"learning_rate": 7.011900468806347e-06,
"loss": 0.2204,
"step": 1875
},
{
"epoch": 0.49,
"learning_rate": 6.993869455463396e-06,
"loss": 0.2079,
"step": 1900
},
{
"epoch": 0.5,
"learning_rate": 6.975838442120447e-06,
"loss": 0.2269,
"step": 1925
},
{
"epoch": 0.5,
"learning_rate": 6.957807428777497e-06,
"loss": 0.2508,
"step": 1950
},
{
"epoch": 0.51,
"learning_rate": 6.939776415434547e-06,
"loss": 0.2134,
"step": 1975
},
{
"epoch": 0.52,
"learning_rate": 6.921745402091598e-06,
"loss": 0.2378,
"step": 2000
},
{
"epoch": 0.52,
"eval_loss": 0.20588794350624084,
"eval_runtime": 4387.4072,
"eval_samples_per_second": 2.486,
"eval_steps_per_second": 0.311,
"eval_wer": 37.6066396448562,
"step": 2000
},
{
"epoch": 0.52,
"learning_rate": 6.903714388748647e-06,
"loss": 0.2252,
"step": 2025
},
{
"epoch": 0.53,
"learning_rate": 6.885683375405698e-06,
"loss": 0.2178,
"step": 2050
},
{
"epoch": 0.54,
"learning_rate": 6.867652362062748e-06,
"loss": 0.1849,
"step": 2075
},
{
"epoch": 0.54,
"learning_rate": 6.849621348719798e-06,
"loss": 0.1764,
"step": 2100
},
{
"epoch": 0.55,
"learning_rate": 6.831590335376848e-06,
"loss": 0.1591,
"step": 2125
},
{
"epoch": 0.56,
"learning_rate": 6.813559322033897e-06,
"loss": 0.2159,
"step": 2150
},
{
"epoch": 0.56,
"learning_rate": 6.795528308690948e-06,
"loss": 0.1972,
"step": 2175
},
{
"epoch": 0.57,
"learning_rate": 6.777497295347998e-06,
"loss": 0.1974,
"step": 2200
},
{
"epoch": 0.58,
"learning_rate": 6.759466282005048e-06,
"loss": 0.1883,
"step": 2225
},
{
"epoch": 0.58,
"learning_rate": 6.7414352686620985e-06,
"loss": 0.1917,
"step": 2250
},
{
"epoch": 0.59,
"learning_rate": 6.723404255319149e-06,
"loss": 0.1953,
"step": 2275
},
{
"epoch": 0.6,
"learning_rate": 6.705373241976199e-06,
"loss": 0.1929,
"step": 2300
},
{
"epoch": 0.6,
"learning_rate": 6.687342228633249e-06,
"loss": 0.2116,
"step": 2325
},
{
"epoch": 0.61,
"learning_rate": 6.669311215290299e-06,
"loss": 0.2196,
"step": 2350
},
{
"epoch": 0.61,
"learning_rate": 6.651280201947349e-06,
"loss": 0.1934,
"step": 2375
},
{
"epoch": 0.62,
"learning_rate": 6.6332491886044e-06,
"loss": 0.1799,
"step": 2400
},
{
"epoch": 0.63,
"learning_rate": 6.615218175261449e-06,
"loss": 0.1948,
"step": 2425
},
{
"epoch": 0.63,
"learning_rate": 6.597187161918499e-06,
"loss": 0.2599,
"step": 2450
},
{
"epoch": 0.64,
"learning_rate": 6.57915614857555e-06,
"loss": 0.1955,
"step": 2475
},
{
"epoch": 0.65,
"learning_rate": 6.5611251352325995e-06,
"loss": 0.1903,
"step": 2500
},
{
"epoch": 0.65,
"eval_loss": 0.1840863972902298,
"eval_runtime": 4427.3622,
"eval_samples_per_second": 2.464,
"eval_steps_per_second": 0.308,
"eval_wer": 34.27909669947887,
"step": 2500
},
{
"epoch": 0.65,
"learning_rate": 6.54309412188965e-06,
"loss": 0.1986,
"step": 2525
},
{
"epoch": 0.66,
"learning_rate": 6.5250631085467e-06,
"loss": 0.185,
"step": 2550
},
{
"epoch": 0.67,
"learning_rate": 6.50703209520375e-06,
"loss": 0.2186,
"step": 2575
},
{
"epoch": 0.67,
"learning_rate": 6.489001081860801e-06,
"loss": 0.2152,
"step": 2600
},
{
"epoch": 0.68,
"learning_rate": 6.47097006851785e-06,
"loss": 0.1774,
"step": 2625
},
{
"epoch": 0.69,
"learning_rate": 6.452939055174901e-06,
"loss": 0.1798,
"step": 2650
},
{
"epoch": 0.69,
"learning_rate": 6.434908041831951e-06,
"loss": 0.1746,
"step": 2675
},
{
"epoch": 0.7,
"learning_rate": 6.416877028489e-06,
"loss": 0.2079,
"step": 2700
},
{
"epoch": 0.71,
"learning_rate": 6.398846015146051e-06,
"loss": 0.187,
"step": 2725
},
{
"epoch": 0.71,
"learning_rate": 6.3808150018031e-06,
"loss": 0.1926,
"step": 2750
},
{
"epoch": 0.72,
"learning_rate": 6.362783988460151e-06,
"loss": 0.1701,
"step": 2775
},
{
"epoch": 0.72,
"learning_rate": 6.344752975117201e-06,
"loss": 0.167,
"step": 2800
},
{
"epoch": 0.73,
"learning_rate": 6.326721961774251e-06,
"loss": 0.2032,
"step": 2825
},
{
"epoch": 0.74,
"learning_rate": 6.308690948431302e-06,
"loss": 0.14,
"step": 2850
},
{
"epoch": 0.74,
"learning_rate": 6.290659935088352e-06,
"loss": 0.1754,
"step": 2875
},
{
"epoch": 0.75,
"learning_rate": 6.272628921745402e-06,
"loss": 0.1639,
"step": 2900
},
{
"epoch": 0.76,
"learning_rate": 6.254597908402452e-06,
"loss": 0.2081,
"step": 2925
},
{
"epoch": 0.76,
"learning_rate": 6.236566895059502e-06,
"loss": 0.1387,
"step": 2950
},
{
"epoch": 0.77,
"learning_rate": 6.2185358817165525e-06,
"loss": 0.1526,
"step": 2975
},
{
"epoch": 0.78,
"learning_rate": 6.200504868373602e-06,
"loss": 0.2271,
"step": 3000
},
{
"epoch": 0.78,
"eval_loss": 0.1636439859867096,
"eval_runtime": 4491.1529,
"eval_samples_per_second": 2.429,
"eval_steps_per_second": 0.304,
"eval_wer": 35.32908704883227,
"step": 3000
},
{
"epoch": 0.78,
"learning_rate": 6.182473855030652e-06,
"loss": 0.1551,
"step": 3025
},
{
"epoch": 0.79,
"learning_rate": 6.164442841687702e-06,
"loss": 0.1418,
"step": 3050
},
{
"epoch": 0.8,
"learning_rate": 6.146411828344753e-06,
"loss": 0.1699,
"step": 3075
},
{
"epoch": 0.8,
"learning_rate": 6.1283808150018025e-06,
"loss": 0.1632,
"step": 3100
},
{
"epoch": 0.81,
"learning_rate": 6.110349801658853e-06,
"loss": 0.1649,
"step": 3125
},
{
"epoch": 0.82,
"learning_rate": 6.0923187883159035e-06,
"loss": 0.1265,
"step": 3150
},
{
"epoch": 0.82,
"learning_rate": 6.074287774972953e-06,
"loss": 0.1562,
"step": 3175
},
{
"epoch": 0.83,
"learning_rate": 6.056256761630004e-06,
"loss": 0.166,
"step": 3200
},
{
"epoch": 0.83,
"learning_rate": 6.038225748287053e-06,
"loss": 0.1531,
"step": 3225
},
{
"epoch": 0.84,
"learning_rate": 6.020194734944104e-06,
"loss": 0.1707,
"step": 3250
},
{
"epoch": 0.85,
"learning_rate": 6.0021637216011544e-06,
"loss": 0.174,
"step": 3275
},
{
"epoch": 0.85,
"learning_rate": 5.984132708258203e-06,
"loss": 0.1562,
"step": 3300
},
{
"epoch": 0.86,
"learning_rate": 5.966101694915254e-06,
"loss": 0.1068,
"step": 3325
},
{
"epoch": 0.87,
"learning_rate": 5.948070681572304e-06,
"loss": 0.1473,
"step": 3350
},
{
"epoch": 0.87,
"learning_rate": 5.930039668229354e-06,
"loss": 0.1381,
"step": 3375
},
{
"epoch": 0.88,
"learning_rate": 5.9120086548864045e-06,
"loss": 0.1566,
"step": 3400
},
{
"epoch": 0.89,
"learning_rate": 5.893977641543454e-06,
"loss": 0.1537,
"step": 3425
},
{
"epoch": 0.89,
"learning_rate": 5.875946628200505e-06,
"loss": 0.1339,
"step": 3450
},
{
"epoch": 0.9,
"learning_rate": 5.857915614857555e-06,
"loss": 0.1187,
"step": 3475
},
{
"epoch": 0.91,
"learning_rate": 5.839884601514605e-06,
"loss": 0.1794,
"step": 3500
},
{
"epoch": 0.91,
"eval_loss": 0.15343192219734192,
"eval_runtime": 4341.7028,
"eval_samples_per_second": 2.512,
"eval_steps_per_second": 0.314,
"eval_wer": 28.291835552982047,
"step": 3500
},
{
"epoch": 0.91,
"learning_rate": 5.821853588171655e-06,
"loss": 0.1772,
"step": 3525
},
{
"epoch": 0.92,
"learning_rate": 5.803822574828706e-06,
"loss": 0.146,
"step": 3550
},
{
"epoch": 0.93,
"learning_rate": 5.785791561485755e-06,
"loss": 0.1211,
"step": 3575
},
{
"epoch": 0.93,
"learning_rate": 5.767760548142805e-06,
"loss": 0.1193,
"step": 3600
},
{
"epoch": 0.94,
"learning_rate": 5.749729534799855e-06,
"loss": 0.1324,
"step": 3625
},
{
"epoch": 0.94,
"learning_rate": 5.731698521456905e-06,
"loss": 0.144,
"step": 3650
},
{
"epoch": 0.95,
"learning_rate": 5.713667508113956e-06,
"loss": 0.1294,
"step": 3675
},
{
"epoch": 0.96,
"learning_rate": 5.6956364947710056e-06,
"loss": 0.1487,
"step": 3700
},
{
"epoch": 0.96,
"learning_rate": 5.677605481428056e-06,
"loss": 0.1453,
"step": 3725
},
{
"epoch": 0.97,
"learning_rate": 5.659574468085107e-06,
"loss": 0.1538,
"step": 3750
},
{
"epoch": 0.98,
"learning_rate": 5.641543454742156e-06,
"loss": 0.1789,
"step": 3775
},
{
"epoch": 0.98,
"learning_rate": 5.623512441399207e-06,
"loss": 0.139,
"step": 3800
},
{
"epoch": 0.99,
"learning_rate": 5.6054814280562565e-06,
"loss": 0.1509,
"step": 3825
},
{
"epoch": 1.0,
"learning_rate": 5.587450414713307e-06,
"loss": 0.1444,
"step": 3850
},
{
"epoch": 1.0,
"learning_rate": 5.569419401370357e-06,
"loss": 0.1174,
"step": 3875
},
{
"epoch": 1.01,
"learning_rate": 5.551388388027406e-06,
"loss": 0.0813,
"step": 3900
},
{
"epoch": 1.02,
"learning_rate": 5.533357374684457e-06,
"loss": 0.1102,
"step": 3925
},
{
"epoch": 1.02,
"learning_rate": 5.515326361341507e-06,
"loss": 0.1029,
"step": 3950
},
{
"epoch": 1.03,
"learning_rate": 5.497295347998557e-06,
"loss": 0.1184,
"step": 3975
},
{
"epoch": 1.04,
"learning_rate": 5.4792643346556075e-06,
"loss": 0.1327,
"step": 4000
},
{
"epoch": 1.04,
"eval_loss": 0.14156095683574677,
"eval_runtime": 4290.1057,
"eval_samples_per_second": 2.542,
"eval_steps_per_second": 0.318,
"eval_wer": 24.898668210770122,
"step": 4000
},
{
"epoch": 1.04,
"learning_rate": 5.461233321312657e-06,
"loss": 0.1076,
"step": 4025
},
{
"epoch": 1.05,
"learning_rate": 5.443202307969708e-06,
"loss": 0.1316,
"step": 4050
},
{
"epoch": 1.05,
"learning_rate": 5.425171294626758e-06,
"loss": 0.1088,
"step": 4075
},
{
"epoch": 1.06,
"learning_rate": 5.407140281283808e-06,
"loss": 0.0914,
"step": 4100
},
{
"epoch": 1.07,
"learning_rate": 5.389109267940858e-06,
"loss": 0.1141,
"step": 4125
},
{
"epoch": 1.07,
"learning_rate": 5.371078254597909e-06,
"loss": 0.103,
"step": 4150
},
{
"epoch": 1.08,
"learning_rate": 5.353047241254958e-06,
"loss": 0.1203,
"step": 4175
},
{
"epoch": 1.09,
"learning_rate": 5.335016227912008e-06,
"loss": 0.1036,
"step": 4200
},
{
"epoch": 1.09,
"learning_rate": 5.316985214569058e-06,
"loss": 0.149,
"step": 4225
},
{
"epoch": 1.1,
"learning_rate": 5.2989542012261084e-06,
"loss": 0.1277,
"step": 4250
},
{
"epoch": 1.11,
"learning_rate": 5.280923187883159e-06,
"loss": 0.0915,
"step": 4275
},
{
"epoch": 1.11,
"learning_rate": 5.262892174540209e-06,
"loss": 0.0906,
"step": 4300
},
{
"epoch": 1.12,
"learning_rate": 5.244861161197259e-06,
"loss": 0.1038,
"step": 4325
},
{
"epoch": 1.13,
"learning_rate": 5.22683014785431e-06,
"loss": 0.117,
"step": 4350
},
{
"epoch": 1.13,
"learning_rate": 5.208799134511359e-06,
"loss": 0.0978,
"step": 4375
},
{
"epoch": 1.14,
"learning_rate": 5.19076812116841e-06,
"loss": 0.1158,
"step": 4400
},
{
"epoch": 1.15,
"learning_rate": 5.1727371078254595e-06,
"loss": 0.1096,
"step": 4425
},
{
"epoch": 1.15,
"learning_rate": 5.154706094482509e-06,
"loss": 0.0935,
"step": 4450
},
{
"epoch": 1.16,
"learning_rate": 5.13667508113956e-06,
"loss": 0.0889,
"step": 4475
},
{
"epoch": 1.16,
"learning_rate": 5.118644067796609e-06,
"loss": 0.1224,
"step": 4500
},
{
"epoch": 1.16,
"eval_loss": 0.13514740765094757,
"eval_runtime": 4589.5901,
"eval_samples_per_second": 2.376,
"eval_steps_per_second": 0.297,
"eval_wer": 24.898668210770122,
"step": 4500
},
{
"epoch": 1.17,
"learning_rate": 5.10061305445366e-06,
"loss": 0.1034,
"step": 4525
},
{
"epoch": 1.18,
"learning_rate": 5.08258204111071e-06,
"loss": 0.1161,
"step": 4550
},
{
"epoch": 1.18,
"learning_rate": 5.06455102776776e-06,
"loss": 0.0899,
"step": 4575
},
{
"epoch": 1.19,
"learning_rate": 5.046520014424811e-06,
"loss": 0.0711,
"step": 4600
},
{
"epoch": 1.2,
"learning_rate": 5.02848900108186e-06,
"loss": 0.1138,
"step": 4625
},
{
"epoch": 1.2,
"learning_rate": 5.010457987738911e-06,
"loss": 0.0782,
"step": 4650
},
{
"epoch": 1.21,
"learning_rate": 4.992426974395961e-06,
"loss": 0.114,
"step": 4675
},
{
"epoch": 1.22,
"learning_rate": 4.974395961053011e-06,
"loss": 0.0766,
"step": 4700
},
{
"epoch": 1.22,
"learning_rate": 4.9563649477100615e-06,
"loss": 0.116,
"step": 4725
},
{
"epoch": 1.23,
"learning_rate": 4.938333934367111e-06,
"loss": 0.092,
"step": 4750
},
{
"epoch": 1.24,
"learning_rate": 4.920302921024161e-06,
"loss": 0.0969,
"step": 4775
},
{
"epoch": 1.24,
"learning_rate": 4.902271907681211e-06,
"loss": 0.1045,
"step": 4800
},
{
"epoch": 1.25,
"learning_rate": 4.884240894338262e-06,
"loss": 0.0981,
"step": 4825
},
{
"epoch": 1.26,
"learning_rate": 4.8662098809953115e-06,
"loss": 0.124,
"step": 4850
},
{
"epoch": 1.26,
"learning_rate": 4.848178867652362e-06,
"loss": 0.1049,
"step": 4875
},
{
"epoch": 1.27,
"learning_rate": 4.830147854309412e-06,
"loss": 0.0812,
"step": 4900
},
{
"epoch": 1.27,
"learning_rate": 4.812116840966462e-06,
"loss": 0.0978,
"step": 4925
},
{
"epoch": 1.28,
"learning_rate": 4.794085827623513e-06,
"loss": 0.1069,
"step": 4950
},
{
"epoch": 1.29,
"learning_rate": 4.776054814280562e-06,
"loss": 0.0977,
"step": 4975
},
{
"epoch": 1.29,
"learning_rate": 4.758023800937613e-06,
"loss": 0.0769,
"step": 5000
},
{
"epoch": 1.29,
"eval_loss": 0.12578198313713074,
"eval_runtime": 4416.0732,
"eval_samples_per_second": 2.47,
"eval_steps_per_second": 0.309,
"eval_wer": 25.57421347230264,
"step": 5000
},
{
"epoch": 1.3,
"learning_rate": 4.7399927875946634e-06,
"loss": 0.1234,
"step": 5025
},
{
"epoch": 1.31,
"learning_rate": 4.721961774251712e-06,
"loss": 0.0676,
"step": 5050
},
{
"epoch": 1.31,
"learning_rate": 4.703930760908763e-06,
"loss": 0.094,
"step": 5075
},
{
"epoch": 1.32,
"learning_rate": 4.6858997475658124e-06,
"loss": 0.0979,
"step": 5100
},
{
"epoch": 1.33,
"learning_rate": 4.667868734222863e-06,
"loss": 0.0945,
"step": 5125
},
{
"epoch": 1.33,
"learning_rate": 4.6498377208799135e-06,
"loss": 0.0886,
"step": 5150
},
{
"epoch": 1.34,
"learning_rate": 4.631806707536963e-06,
"loss": 0.1179,
"step": 5175
},
{
"epoch": 1.35,
"learning_rate": 4.613775694194014e-06,
"loss": 0.1152,
"step": 5200
},
{
"epoch": 1.35,
"learning_rate": 4.595744680851064e-06,
"loss": 0.1208,
"step": 5225
},
{
"epoch": 1.36,
"learning_rate": 4.577713667508114e-06,
"loss": 0.0786,
"step": 5250
},
{
"epoch": 1.37,
"learning_rate": 4.559682654165164e-06,
"loss": 0.1059,
"step": 5275
},
{
"epoch": 1.37,
"learning_rate": 4.541651640822214e-06,
"loss": 0.1075,
"step": 5300
},
{
"epoch": 1.38,
"learning_rate": 4.523620627479264e-06,
"loss": 0.0817,
"step": 5325
},
{
"epoch": 1.38,
"learning_rate": 4.505589614136314e-06,
"loss": 0.0903,
"step": 5350
},
{
"epoch": 1.39,
"learning_rate": 4.487558600793364e-06,
"loss": 0.0841,
"step": 5375
},
{
"epoch": 1.4,
"learning_rate": 4.469527587450414e-06,
"loss": 0.1095,
"step": 5400
},
{
"epoch": 1.4,
"learning_rate": 4.451496574107465e-06,
"loss": 0.1035,
"step": 5425
},
{
"epoch": 1.41,
"learning_rate": 4.4334655607645146e-06,
"loss": 0.066,
"step": 5450
},
{
"epoch": 1.42,
"learning_rate": 4.415434547421565e-06,
"loss": 0.1118,
"step": 5475
},
{
"epoch": 1.42,
"learning_rate": 4.397403534078615e-06,
"loss": 0.0946,
"step": 5500
},
{
"epoch": 1.42,
"eval_loss": 0.12158416956663132,
"eval_runtime": 4406.9698,
"eval_samples_per_second": 2.475,
"eval_steps_per_second": 0.31,
"eval_wer": 27.319050376375216,
"step": 5500
},
{
"epoch": 1.43,
"learning_rate": 4.379372520735665e-06,
"loss": 0.0877,
"step": 5525
},
{
"epoch": 1.44,
"learning_rate": 4.361341507392716e-06,
"loss": 0.1039,
"step": 5550
},
{
"epoch": 1.44,
"learning_rate": 4.3433104940497655e-06,
"loss": 0.0891,
"step": 5575
},
{
"epoch": 1.45,
"learning_rate": 4.325279480706816e-06,
"loss": 0.095,
"step": 5600
},
{
"epoch": 1.46,
"learning_rate": 4.307248467363866e-06,
"loss": 0.0606,
"step": 5625
},
{
"epoch": 1.46,
"learning_rate": 4.289217454020915e-06,
"loss": 0.1231,
"step": 5650
},
{
"epoch": 1.47,
"learning_rate": 4.271186440677966e-06,
"loss": 0.0967,
"step": 5675
},
{
"epoch": 1.48,
"learning_rate": 4.2531554273350155e-06,
"loss": 0.0922,
"step": 5700
},
{
"epoch": 1.48,
"learning_rate": 4.235124413992066e-06,
"loss": 0.1196,
"step": 5725
},
{
"epoch": 1.49,
"learning_rate": 4.2170934006491165e-06,
"loss": 0.0746,
"step": 5750
},
{
"epoch": 1.49,
"learning_rate": 4.199062387306166e-06,
"loss": 0.0851,
"step": 5775
},
{
"epoch": 1.5,
"learning_rate": 4.181031373963217e-06,
"loss": 0.1046,
"step": 5800
},
{
"epoch": 1.51,
"learning_rate": 4.163000360620267e-06,
"loss": 0.0924,
"step": 5825
},
{
"epoch": 1.51,
"learning_rate": 4.144969347277317e-06,
"loss": 0.1033,
"step": 5850
},
{
"epoch": 1.52,
"learning_rate": 4.126938333934367e-06,
"loss": 0.1042,
"step": 5875
},
{
"epoch": 1.53,
"learning_rate": 4.108907320591416e-06,
"loss": 0.0839,
"step": 5900
},
{
"epoch": 1.53,
"learning_rate": 4.090876307248467e-06,
"loss": 0.0854,
"step": 5925
},
{
"epoch": 1.54,
"learning_rate": 4.072845293905517e-06,
"loss": 0.0915,
"step": 5950
},
{
"epoch": 1.55,
"learning_rate": 4.054814280562567e-06,
"loss": 0.0883,
"step": 5975
},
{
"epoch": 1.55,
"learning_rate": 4.0367832672196174e-06,
"loss": 0.0645,
"step": 6000
},
{
"epoch": 1.55,
"eval_loss": 0.11707902699708939,
"eval_runtime": 4301.5745,
"eval_samples_per_second": 2.536,
"eval_steps_per_second": 0.317,
"eval_wer": 21.86064466319243,
"step": 6000
},
{
"epoch": 1.56,
"learning_rate": 4.018752253876668e-06,
"loss": 0.0777,
"step": 6025
},
{
"epoch": 1.57,
"learning_rate": 4.000721240533718e-06,
"loss": 0.073,
"step": 6050
},
{
"epoch": 1.57,
"learning_rate": 3.982690227190768e-06,
"loss": 0.0746,
"step": 6075
},
{
"epoch": 1.58,
"learning_rate": 3.964659213847818e-06,
"loss": 0.07,
"step": 6100
},
{
"epoch": 1.59,
"learning_rate": 3.946628200504868e-06,
"loss": 0.0836,
"step": 6125
},
{
"epoch": 1.59,
"learning_rate": 3.928597187161918e-06,
"loss": 0.0907,
"step": 6150
},
{
"epoch": 1.6,
"learning_rate": 3.9105661738189685e-06,
"loss": 0.0983,
"step": 6175
},
{
"epoch": 1.6,
"learning_rate": 3.892535160476018e-06,
"loss": 0.1061,
"step": 6200
},
{
"epoch": 1.61,
"learning_rate": 3.874504147133069e-06,
"loss": 0.0824,
"step": 6225
},
{
"epoch": 1.62,
"learning_rate": 3.856473133790118e-06,
"loss": 0.0835,
"step": 6250
},
{
"epoch": 1.62,
"learning_rate": 3.838442120447169e-06,
"loss": 0.0818,
"step": 6275
},
{
"epoch": 1.63,
"learning_rate": 3.820411107104219e-06,
"loss": 0.0903,
"step": 6300
},
{
"epoch": 1.64,
"learning_rate": 3.802380093761269e-06,
"loss": 0.1191,
"step": 6325
},
{
"epoch": 1.64,
"learning_rate": 3.784349080418319e-06,
"loss": 0.1132,
"step": 6350
},
{
"epoch": 1.65,
"learning_rate": 3.7663180670753697e-06,
"loss": 0.0779,
"step": 6375
},
{
"epoch": 1.66,
"learning_rate": 3.7482870537324198e-06,
"loss": 0.0606,
"step": 6400
},
{
"epoch": 1.66,
"learning_rate": 3.7302560403894694e-06,
"loss": 0.0997,
"step": 6425
},
{
"epoch": 1.67,
"learning_rate": 3.7122250270465195e-06,
"loss": 0.1081,
"step": 6450
},
{
"epoch": 1.68,
"learning_rate": 3.69419401370357e-06,
"loss": 0.0811,
"step": 6475
},
{
"epoch": 1.68,
"learning_rate": 3.67616300036062e-06,
"loss": 0.0885,
"step": 6500
},
{
"epoch": 1.68,
"eval_loss": 0.11251728981733322,
"eval_runtime": 4322.1358,
"eval_samples_per_second": 2.524,
"eval_steps_per_second": 0.316,
"eval_wer": 21.841343370005788,
"step": 6500
},
{
"epoch": 1.69,
"learning_rate": 3.6581319870176702e-06,
"loss": 0.109,
"step": 6525
},
{
"epoch": 1.7,
"learning_rate": 3.6401009736747207e-06,
"loss": 0.1101,
"step": 6550
},
{
"epoch": 1.7,
"learning_rate": 3.6220699603317704e-06,
"loss": 0.071,
"step": 6575
},
{
"epoch": 1.71,
"learning_rate": 3.6040389469888205e-06,
"loss": 0.0807,
"step": 6600
},
{
"epoch": 1.71,
"learning_rate": 3.5860079336458706e-06,
"loss": 0.1195,
"step": 6625
},
{
"epoch": 1.72,
"learning_rate": 3.567976920302921e-06,
"loss": 0.1163,
"step": 6650
},
{
"epoch": 1.73,
"learning_rate": 3.549945906959971e-06,
"loss": 0.0897,
"step": 6675
},
{
"epoch": 1.73,
"learning_rate": 3.5319148936170213e-06,
"loss": 0.0928,
"step": 6700
},
{
"epoch": 1.74,
"learning_rate": 3.513883880274071e-06,
"loss": 0.0717,
"step": 6725
},
{
"epoch": 1.75,
"learning_rate": 3.4958528669311215e-06,
"loss": 0.0915,
"step": 6750
},
{
"epoch": 1.75,
"learning_rate": 3.4778218535881716e-06,
"loss": 0.0706,
"step": 6775
},
{
"epoch": 1.76,
"learning_rate": 3.4597908402452217e-06,
"loss": 0.087,
"step": 6800
},
{
"epoch": 1.77,
"learning_rate": 3.4417598269022718e-06,
"loss": 0.1036,
"step": 6825
},
{
"epoch": 1.77,
"learning_rate": 3.4237288135593223e-06,
"loss": 0.0733,
"step": 6850
},
{
"epoch": 1.78,
"learning_rate": 3.405697800216372e-06,
"loss": 0.097,
"step": 6875
},
{
"epoch": 1.79,
"learning_rate": 3.387666786873422e-06,
"loss": 0.1069,
"step": 6900
},
{
"epoch": 1.79,
"learning_rate": 3.369635773530472e-06,
"loss": 0.1047,
"step": 6925
},
{
"epoch": 1.8,
"learning_rate": 3.3516047601875226e-06,
"loss": 0.0897,
"step": 6950
},
{
"epoch": 1.81,
"learning_rate": 3.3335737468445727e-06,
"loss": 0.0821,
"step": 6975
},
{
"epoch": 1.81,
"learning_rate": 3.3155427335016224e-06,
"loss": 0.0825,
"step": 7000
},
{
"epoch": 1.81,
"eval_loss": 0.10709645599126816,
"eval_runtime": 4356.5437,
"eval_samples_per_second": 2.504,
"eval_steps_per_second": 0.313,
"eval_wer": 21.89924724956572,
"step": 7000
},
{
"epoch": 1.82,
"learning_rate": 3.2975117201586725e-06,
"loss": 0.0845,
"step": 7025
},
{
"epoch": 1.82,
"learning_rate": 3.279480706815723e-06,
"loss": 0.0831,
"step": 7050
},
{
"epoch": 1.83,
"learning_rate": 3.261449693472773e-06,
"loss": 0.0629,
"step": 7075
},
{
"epoch": 1.84,
"learning_rate": 3.243418680129823e-06,
"loss": 0.0844,
"step": 7100
},
{
"epoch": 1.84,
"learning_rate": 3.2253876667868733e-06,
"loss": 0.0794,
"step": 7125
},
{
"epoch": 1.85,
"learning_rate": 3.2073566534439234e-06,
"loss": 0.0828,
"step": 7150
},
{
"epoch": 1.86,
"learning_rate": 3.1893256401009735e-06,
"loss": 0.0818,
"step": 7175
},
{
"epoch": 1.86,
"learning_rate": 3.1712946267580236e-06,
"loss": 0.0926,
"step": 7200
},
{
"epoch": 1.87,
"learning_rate": 3.1532636134150737e-06,
"loss": 0.0519,
"step": 7225
},
{
"epoch": 1.88,
"learning_rate": 3.135232600072124e-06,
"loss": 0.0748,
"step": 7250
},
{
"epoch": 1.88,
"learning_rate": 3.1172015867291743e-06,
"loss": 0.1114,
"step": 7275
},
{
"epoch": 1.89,
"learning_rate": 3.099170573386224e-06,
"loss": 0.074,
"step": 7300
},
{
"epoch": 1.9,
"learning_rate": 3.081139560043274e-06,
"loss": 0.0784,
"step": 7325
},
{
"epoch": 1.9,
"learning_rate": 3.0631085467003245e-06,
"loss": 0.0695,
"step": 7350
},
{
"epoch": 1.91,
"learning_rate": 3.0450775333573746e-06,
"loss": 0.0725,
"step": 7375
},
{
"epoch": 1.92,
"learning_rate": 3.0270465200144247e-06,
"loss": 0.0757,
"step": 7400
},
{
"epoch": 1.92,
"learning_rate": 3.009015506671475e-06,
"loss": 0.0942,
"step": 7425
},
{
"epoch": 1.93,
"learning_rate": 2.990984493328525e-06,
"loss": 0.108,
"step": 7450
},
{
"epoch": 1.93,
"learning_rate": 2.972953479985575e-06,
"loss": 0.103,
"step": 7475
},
{
"epoch": 1.94,
"learning_rate": 2.954922466642625e-06,
"loss": 0.0943,
"step": 7500
},
{
"epoch": 1.94,
"eval_loss": 0.10313993692398071,
"eval_runtime": 4346.2637,
"eval_samples_per_second": 2.51,
"eval_steps_per_second": 0.314,
"eval_wer": 20.069484655471918,
"step": 7500
},
{
"epoch": 1.95,
"learning_rate": 2.936891453299675e-06,
"loss": 0.0985,
"step": 7525
},
{
"epoch": 1.95,
"learning_rate": 2.9188604399567257e-06,
"loss": 0.082,
"step": 7550
},
{
"epoch": 1.96,
"learning_rate": 2.900829426613776e-06,
"loss": 0.0577,
"step": 7575
},
{
"epoch": 1.97,
"learning_rate": 2.8827984132708255e-06,
"loss": 0.0853,
"step": 7600
},
{
"epoch": 1.97,
"learning_rate": 2.8647673999278756e-06,
"loss": 0.0809,
"step": 7625
},
{
"epoch": 1.98,
"learning_rate": 2.846736386584926e-06,
"loss": 0.0653,
"step": 7650
},
{
"epoch": 1.99,
"learning_rate": 2.828705373241976e-06,
"loss": 0.0835,
"step": 7675
},
{
"epoch": 1.99,
"learning_rate": 2.8106743598990263e-06,
"loss": 0.083,
"step": 7700
},
{
"epoch": 2.0,
"learning_rate": 2.7926433465560763e-06,
"loss": 0.0799,
"step": 7725
},
{
"epoch": 2.01,
"learning_rate": 2.7746123332131264e-06,
"loss": 0.0534,
"step": 7750
},
{
"epoch": 2.01,
"learning_rate": 2.7565813198701765e-06,
"loss": 0.0411,
"step": 7775
},
{
"epoch": 2.02,
"learning_rate": 2.7385503065272266e-06,
"loss": 0.0559,
"step": 7800
},
{
"epoch": 2.03,
"learning_rate": 2.7205192931842767e-06,
"loss": 0.0524,
"step": 7825
},
{
"epoch": 2.03,
"learning_rate": 2.7024882798413272e-06,
"loss": 0.055,
"step": 7850
},
{
"epoch": 2.04,
"learning_rate": 2.6844572664983773e-06,
"loss": 0.0403,
"step": 7875
},
{
"epoch": 2.04,
"learning_rate": 2.666426253155427e-06,
"loss": 0.0595,
"step": 7900
},
{
"epoch": 2.05,
"learning_rate": 2.648395239812477e-06,
"loss": 0.058,
"step": 7925
},
{
"epoch": 2.06,
"learning_rate": 2.6303642264695276e-06,
"loss": 0.0652,
"step": 7950
},
{
"epoch": 2.06,
"learning_rate": 2.6123332131265777e-06,
"loss": 0.0481,
"step": 7975
},
{
"epoch": 2.07,
"learning_rate": 2.5943021997836278e-06,
"loss": 0.0399,
"step": 8000
},
{
"epoch": 2.07,
"eval_loss": 0.10383369773626328,
"eval_runtime": 4337.689,
"eval_samples_per_second": 2.514,
"eval_steps_per_second": 0.314,
"eval_wer": 20.200733449141094,
"step": 8000
},
{
"epoch": 2.08,
"learning_rate": 2.5762711864406775e-06,
"loss": 0.0611,
"step": 8025
},
{
"epoch": 2.08,
"learning_rate": 2.558240173097728e-06,
"loss": 0.0318,
"step": 8050
},
{
"epoch": 2.09,
"learning_rate": 2.540209159754778e-06,
"loss": 0.0797,
"step": 8075
},
{
"epoch": 2.1,
"learning_rate": 2.522178146411828e-06,
"loss": 0.0799,
"step": 8100
},
{
"epoch": 2.1,
"learning_rate": 2.5041471330688782e-06,
"loss": 0.0541,
"step": 8125
},
{
"epoch": 2.11,
"learning_rate": 2.4861161197259288e-06,
"loss": 0.0389,
"step": 8150
},
{
"epoch": 2.12,
"learning_rate": 2.4680851063829784e-06,
"loss": 0.053,
"step": 8175
},
{
"epoch": 2.12,
"learning_rate": 2.4500540930400285e-06,
"loss": 0.0447,
"step": 8200
},
{
"epoch": 2.13,
"learning_rate": 2.4320230796970786e-06,
"loss": 0.0467,
"step": 8225
},
{
"epoch": 2.14,
"learning_rate": 2.413992066354129e-06,
"loss": 0.0463,
"step": 8250
},
{
"epoch": 2.14,
"learning_rate": 2.3959610530111792e-06,
"loss": 0.0623,
"step": 8275
},
{
"epoch": 2.15,
"learning_rate": 2.3779300396682293e-06,
"loss": 0.0446,
"step": 8300
},
{
"epoch": 2.15,
"learning_rate": 2.359899026325279e-06,
"loss": 0.0572,
"step": 8325
},
{
"epoch": 2.16,
"learning_rate": 2.3418680129823295e-06,
"loss": 0.0379,
"step": 8350
},
{
"epoch": 2.17,
"learning_rate": 2.3238369996393796e-06,
"loss": 0.0517,
"step": 8375
},
{
"epoch": 2.17,
"learning_rate": 2.3058059862964297e-06,
"loss": 0.0565,
"step": 8400
},
{
"epoch": 2.18,
"learning_rate": 2.28777497295348e-06,
"loss": 0.041,
"step": 8425
},
{
"epoch": 2.19,
"learning_rate": 2.2697439596105303e-06,
"loss": 0.0444,
"step": 8450
},
{
"epoch": 2.19,
"learning_rate": 2.25171294626758e-06,
"loss": 0.055,
"step": 8475
},
{
"epoch": 2.2,
"learning_rate": 2.23368193292463e-06,
"loss": 0.0444,
"step": 8500
},
{
"epoch": 2.2,
"eval_loss": 0.100523442029953,
"eval_runtime": 4279.5394,
"eval_samples_per_second": 2.549,
"eval_steps_per_second": 0.319,
"eval_wer": 19.81856784404555,
"step": 8500
},
{
"epoch": 2.21,
"learning_rate": 2.2156509195816806e-06,
"loss": 0.0569,
"step": 8525
},
{
"epoch": 2.21,
"learning_rate": 2.1976199062387307e-06,
"loss": 0.0785,
"step": 8550
},
{
"epoch": 2.22,
"learning_rate": 2.1795888928957808e-06,
"loss": 0.0602,
"step": 8575
},
{
"epoch": 2.23,
"learning_rate": 2.161557879552831e-06,
"loss": 0.0404,
"step": 8600
},
{
"epoch": 2.23,
"learning_rate": 2.143526866209881e-06,
"loss": 0.0429,
"step": 8625
},
{
"epoch": 2.24,
"learning_rate": 2.126217093400649e-06,
"loss": 0.0822,
"step": 8650
},
{
"epoch": 2.25,
"learning_rate": 2.108186080057699e-06,
"loss": 0.0532,
"step": 8675
},
{
"epoch": 2.25,
"learning_rate": 2.0901550667147494e-06,
"loss": 0.0416,
"step": 8700
},
{
"epoch": 2.26,
"learning_rate": 2.0721240533717995e-06,
"loss": 0.0495,
"step": 8725
},
{
"epoch": 2.26,
"learning_rate": 2.0540930400288496e-06,
"loss": 0.0781,
"step": 8750
},
{
"epoch": 2.27,
"learning_rate": 2.0360620266858993e-06,
"loss": 0.0421,
"step": 8775
},
{
"epoch": 2.28,
"learning_rate": 2.01803101334295e-06,
"loss": 0.0705,
"step": 8800
},
{
"epoch": 2.28,
"learning_rate": 2e-06,
"loss": 0.0542,
"step": 8825
},
{
"epoch": 2.29,
"learning_rate": 1.98196898665705e-06,
"loss": 0.0603,
"step": 8850
},
{
"epoch": 2.3,
"learning_rate": 1.9639379733141e-06,
"loss": 0.0512,
"step": 8875
},
{
"epoch": 2.3,
"learning_rate": 1.94590695997115e-06,
"loss": 0.0513,
"step": 8900
},
{
"epoch": 2.31,
"learning_rate": 1.9278759466282003e-06,
"loss": 0.0409,
"step": 8925
},
{
"epoch": 2.32,
"learning_rate": 1.909844933285251e-06,
"loss": 0.0508,
"step": 8950
},
{
"epoch": 2.32,
"learning_rate": 1.8918139199423007e-06,
"loss": 0.0619,
"step": 8975
},
{
"epoch": 2.33,
"learning_rate": 1.873782906599351e-06,
"loss": 0.0671,
"step": 9000
},
{
"epoch": 2.33,
"eval_loss": 0.09908398240804672,
"eval_runtime": 4327.228,
"eval_samples_per_second": 2.521,
"eval_steps_per_second": 0.315,
"eval_wer": 19.08125844431577,
"step": 9000
},
{
"epoch": 2.34,
"learning_rate": 1.8557518932564009e-06,
"loss": 0.0563,
"step": 9025
},
{
"epoch": 2.34,
"learning_rate": 1.8377208799134512e-06,
"loss": 0.0469,
"step": 9050
},
{
"epoch": 2.35,
"learning_rate": 1.8196898665705013e-06,
"loss": 0.0606,
"step": 9075
},
{
"epoch": 2.36,
"learning_rate": 1.8016588532275513e-06,
"loss": 0.0674,
"step": 9100
},
{
"epoch": 2.36,
"learning_rate": 1.7836278398846014e-06,
"loss": 0.0496,
"step": 9125
},
{
"epoch": 2.37,
"learning_rate": 1.7655968265416517e-06,
"loss": 0.0729,
"step": 9150
},
{
"epoch": 2.37,
"learning_rate": 1.7475658131987016e-06,
"loss": 0.0556,
"step": 9175
},
{
"epoch": 2.38,
"learning_rate": 1.729534799855752e-06,
"loss": 0.0645,
"step": 9200
},
{
"epoch": 2.39,
"learning_rate": 1.711503786512802e-06,
"loss": 0.0291,
"step": 9225
},
{
"epoch": 2.39,
"learning_rate": 1.6934727731698521e-06,
"loss": 0.0462,
"step": 9250
},
{
"epoch": 2.4,
"learning_rate": 1.6754417598269022e-06,
"loss": 0.0908,
"step": 9275
},
{
"epoch": 2.41,
"learning_rate": 1.6574107464839525e-06,
"loss": 0.0518,
"step": 9300
},
{
"epoch": 2.41,
"learning_rate": 1.6393797331410024e-06,
"loss": 0.0424,
"step": 9325
},
{
"epoch": 2.42,
"learning_rate": 1.6213487197980527e-06,
"loss": 0.0541,
"step": 9350
},
{
"epoch": 2.43,
"learning_rate": 1.6033177064551026e-06,
"loss": 0.0497,
"step": 9375
},
{
"epoch": 2.43,
"learning_rate": 1.5852866931121529e-06,
"loss": 0.0495,
"step": 9400
},
{
"epoch": 2.44,
"learning_rate": 1.567255679769203e-06,
"loss": 0.0651,
"step": 9425
},
{
"epoch": 2.45,
"learning_rate": 1.549224666426253e-06,
"loss": 0.0301,
"step": 9450
},
{
"epoch": 2.45,
"learning_rate": 1.5311936530833032e-06,
"loss": 0.058,
"step": 9475
},
{
"epoch": 2.46,
"learning_rate": 1.5131626397403535e-06,
"loss": 0.0731,
"step": 9500
},
{
"epoch": 2.46,
"eval_loss": 0.09717500954866409,
"eval_runtime": 4319.2897,
"eval_samples_per_second": 2.525,
"eval_steps_per_second": 0.316,
"eval_wer": 19.10441999613974,
"step": 9500
},
{
"epoch": 2.47,
"learning_rate": 1.4951316263974033e-06,
"loss": 0.0789,
"step": 9525
},
{
"epoch": 2.47,
"learning_rate": 1.4771006130544536e-06,
"loss": 0.0496,
"step": 9550
},
{
"epoch": 2.48,
"learning_rate": 1.4590695997115037e-06,
"loss": 0.038,
"step": 9575
},
{
"epoch": 2.48,
"learning_rate": 1.4410385863685538e-06,
"loss": 0.0553,
"step": 9600
},
{
"epoch": 2.49,
"learning_rate": 1.423007573025604e-06,
"loss": 0.035,
"step": 9625
},
{
"epoch": 2.5,
"learning_rate": 1.4049765596826542e-06,
"loss": 0.0598,
"step": 9650
},
{
"epoch": 2.5,
"learning_rate": 1.386945546339704e-06,
"loss": 0.0353,
"step": 9675
},
{
"epoch": 2.51,
"learning_rate": 1.3689145329967544e-06,
"loss": 0.0822,
"step": 9700
},
{
"epoch": 2.52,
"learning_rate": 1.3508835196538045e-06,
"loss": 0.048,
"step": 9725
},
{
"epoch": 2.52,
"learning_rate": 1.3328525063108546e-06,
"loss": 0.042,
"step": 9750
},
{
"epoch": 2.53,
"learning_rate": 1.3148214929679047e-06,
"loss": 0.0331,
"step": 9775
},
{
"epoch": 2.54,
"learning_rate": 1.296790479624955e-06,
"loss": 0.0451,
"step": 9800
},
{
"epoch": 2.54,
"learning_rate": 1.2787594662820049e-06,
"loss": 0.0425,
"step": 9825
},
{
"epoch": 2.55,
"learning_rate": 1.2607284529390552e-06,
"loss": 0.0346,
"step": 9850
},
{
"epoch": 2.56,
"learning_rate": 1.2426974395961053e-06,
"loss": 0.0477,
"step": 9875
},
{
"epoch": 2.56,
"learning_rate": 1.2246664262531554e-06,
"loss": 0.0501,
"step": 9900
},
{
"epoch": 2.57,
"learning_rate": 1.2066354129102054e-06,
"loss": 0.0465,
"step": 9925
},
{
"epoch": 2.58,
"learning_rate": 1.1886043995672558e-06,
"loss": 0.0438,
"step": 9950
},
{
"epoch": 2.58,
"learning_rate": 1.1705733862243056e-06,
"loss": 0.0532,
"step": 9975
},
{
"epoch": 2.59,
"learning_rate": 1.152542372881356e-06,
"loss": 0.0577,
"step": 10000
},
{
"epoch": 2.59,
"eval_loss": 0.09593097865581512,
"eval_runtime": 4299.1151,
"eval_samples_per_second": 2.537,
"eval_steps_per_second": 0.317,
"eval_wer": 18.208839992279483,
"step": 10000
},
{
"epoch": 2.59,
"learning_rate": 1.134511359538406e-06,
"loss": 0.0685,
"step": 10025
},
{
"epoch": 2.6,
"learning_rate": 1.1164803461954561e-06,
"loss": 0.0671,
"step": 10050
},
{
"epoch": 2.61,
"learning_rate": 1.0984493328525062e-06,
"loss": 0.0492,
"step": 10075
},
{
"epoch": 2.61,
"learning_rate": 1.0804183195095565e-06,
"loss": 0.053,
"step": 10100
},
{
"epoch": 2.62,
"learning_rate": 1.0623873061666064e-06,
"loss": 0.0603,
"step": 10125
},
{
"epoch": 2.63,
"learning_rate": 1.0443562928236567e-06,
"loss": 0.0647,
"step": 10150
},
{
"epoch": 2.63,
"learning_rate": 1.0263252794807066e-06,
"loss": 0.0435,
"step": 10175
},
{
"epoch": 2.64,
"learning_rate": 1.0082942661377569e-06,
"loss": 0.0503,
"step": 10200
},
{
"epoch": 2.65,
"learning_rate": 9.90263252794807e-07,
"loss": 0.0397,
"step": 10225
},
{
"epoch": 2.65,
"learning_rate": 9.72232239451857e-07,
"loss": 0.0388,
"step": 10250
},
{
"epoch": 2.66,
"learning_rate": 9.542012261089074e-07,
"loss": 0.0432,
"step": 10275
},
{
"epoch": 2.67,
"learning_rate": 9.361702127659575e-07,
"loss": 0.0406,
"step": 10300
},
{
"epoch": 2.67,
"learning_rate": 9.181391994230076e-07,
"loss": 0.0319,
"step": 10325
},
{
"epoch": 2.68,
"learning_rate": 9.001081860800577e-07,
"loss": 0.0571,
"step": 10350
},
{
"epoch": 2.69,
"learning_rate": 8.820771727371079e-07,
"loss": 0.0561,
"step": 10375
},
{
"epoch": 2.69,
"learning_rate": 8.640461593941579e-07,
"loss": 0.0437,
"step": 10400
},
{
"epoch": 2.7,
"learning_rate": 8.46015146051208e-07,
"loss": 0.0562,
"step": 10425
},
{
"epoch": 2.7,
"learning_rate": 8.279841327082581e-07,
"loss": 0.042,
"step": 10450
},
{
"epoch": 2.71,
"learning_rate": 8.099531193653083e-07,
"loss": 0.0394,
"step": 10475
},
{
"epoch": 2.72,
"learning_rate": 7.919221060223584e-07,
"loss": 0.049,
"step": 10500
},
{
"epoch": 2.72,
"eval_loss": 0.09447470307350159,
"eval_runtime": 4280.6879,
"eval_samples_per_second": 2.548,
"eval_steps_per_second": 0.319,
"eval_wer": 18.166377147268868,
"step": 10500
},
{
"epoch": 2.72,
"learning_rate": 7.738910926794085e-07,
"loss": 0.0412,
"step": 10525
},
{
"epoch": 2.73,
"learning_rate": 7.558600793364587e-07,
"loss": 0.0415,
"step": 10550
},
{
"epoch": 2.74,
"learning_rate": 7.378290659935088e-07,
"loss": 0.0632,
"step": 10575
},
{
"epoch": 2.74,
"learning_rate": 7.197980526505589e-07,
"loss": 0.0557,
"step": 10600
},
{
"epoch": 2.75,
"learning_rate": 7.017670393076091e-07,
"loss": 0.0368,
"step": 10625
},
{
"epoch": 2.76,
"learning_rate": 6.844572664983772e-07,
"loss": 0.049,
"step": 10650
},
{
"epoch": 2.76,
"learning_rate": 6.664262531554273e-07,
"loss": 0.0547,
"step": 10675
},
{
"epoch": 2.77,
"learning_rate": 6.483952398124775e-07,
"loss": 0.0371,
"step": 10700
},
{
"epoch": 2.78,
"learning_rate": 6.303642264695276e-07,
"loss": 0.0324,
"step": 10725
},
{
"epoch": 2.78,
"learning_rate": 6.123332131265777e-07,
"loss": 0.051,
"step": 10750
},
{
"epoch": 2.79,
"learning_rate": 5.943021997836279e-07,
"loss": 0.0459,
"step": 10775
},
{
"epoch": 2.8,
"learning_rate": 5.76271186440678e-07,
"loss": 0.0548,
"step": 10800
},
{
"epoch": 2.8,
"learning_rate": 5.582401730977281e-07,
"loss": 0.0593,
"step": 10825
},
{
"epoch": 2.81,
"learning_rate": 5.402091597547783e-07,
"loss": 0.0325,
"step": 10850
},
{
"epoch": 2.81,
"learning_rate": 5.221781464118284e-07,
"loss": 0.0506,
"step": 10875
},
{
"epoch": 2.82,
"learning_rate": 5.041471330688784e-07,
"loss": 0.0338,
"step": 10900
},
{
"epoch": 2.83,
"learning_rate": 4.861161197259285e-07,
"loss": 0.0493,
"step": 10925
},
{
"epoch": 2.83,
"learning_rate": 4.6808510638297873e-07,
"loss": 0.0608,
"step": 10950
},
{
"epoch": 2.84,
"learning_rate": 4.500540930400288e-07,
"loss": 0.0433,
"step": 10975
},
{
"epoch": 2.85,
"learning_rate": 4.3202307969707897e-07,
"loss": 0.0468,
"step": 11000
},
{
"epoch": 2.85,
"eval_loss": 0.09221997112035751,
"eval_runtime": 4294.3534,
"eval_samples_per_second": 2.54,
"eval_steps_per_second": 0.318,
"eval_wer": 18.591005597375023,
"step": 11000
}
],
"max_steps": 11592,
"num_train_epochs": 3,
"total_flos": 5.078872177311744e+19,
"trial_name": null,
"trial_params": null
}