xls-r-1b-cv_8-fr / trainer_state.json
Plim's picture
modify training files
6dc53fd
raw
history blame
30.6 kB
{
"best_metric": Infinity,
"best_model_checkpoint": "./checkpoint-13000",
"epoch": 5.999928330824912,
"global_step": 20928,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 3.7125e-06,
"loss": 7.648,
"step": 100
},
{
"epoch": 0.06,
"learning_rate": 7.4625e-06,
"loss": 2.9839,
"step": 200
},
{
"epoch": 0.09,
"learning_rate": 1.1212499999999998e-05,
"loss": 2.7908,
"step": 300
},
{
"epoch": 0.11,
"learning_rate": 1.49625e-05,
"loss": 1.6324,
"step": 400
},
{
"epoch": 0.14,
"learning_rate": 1.8712499999999997e-05,
"loss": 1.2129,
"step": 500
},
{
"epoch": 0.17,
"learning_rate": 2.2462499999999997e-05,
"loss": 1.1063,
"step": 600
},
{
"epoch": 0.2,
"learning_rate": 2.6212499999999997e-05,
"loss": 1.0385,
"step": 700
},
{
"epoch": 0.23,
"learning_rate": 2.9925e-05,
"loss": 1.0169,
"step": 800
},
{
"epoch": 0.26,
"learning_rate": 3.3675e-05,
"loss": 0.993,
"step": 900
},
{
"epoch": 0.29,
"learning_rate": 3.7424999999999995e-05,
"loss": 0.9827,
"step": 1000
},
{
"epoch": 0.29,
"eval_loss": Infinity,
"eval_runtime": 971.9761,
"eval_samples_per_second": 16.483,
"eval_steps_per_second": 1.031,
"eval_wer": 0.2936861897030796,
"step": 1000
},
{
"epoch": 0.32,
"learning_rate": 4.1175e-05,
"loss": 0.9828,
"step": 1100
},
{
"epoch": 0.34,
"learning_rate": 4.4924999999999994e-05,
"loss": 0.9736,
"step": 1200
},
{
"epoch": 0.37,
"learning_rate": 4.8675e-05,
"loss": 0.9782,
"step": 1300
},
{
"epoch": 0.4,
"learning_rate": 5.2424999999999994e-05,
"loss": 0.9913,
"step": 1400
},
{
"epoch": 0.43,
"learning_rate": 5.6175e-05,
"loss": 0.9866,
"step": 1500
},
{
"epoch": 0.46,
"learning_rate": 5.9925e-05,
"loss": 1.0025,
"step": 1600
},
{
"epoch": 0.49,
"learning_rate": 6.367499999999999e-05,
"loss": 0.9933,
"step": 1700
},
{
"epoch": 0.52,
"learning_rate": 6.7425e-05,
"loss": 1.005,
"step": 1800
},
{
"epoch": 0.54,
"learning_rate": 7.1175e-05,
"loss": 1.0231,
"step": 1900
},
{
"epoch": 0.57,
"learning_rate": 7.492499999999999e-05,
"loss": 1.0203,
"step": 2000
},
{
"epoch": 0.57,
"eval_loss": Infinity,
"eval_runtime": 974.2963,
"eval_samples_per_second": 16.444,
"eval_steps_per_second": 1.028,
"eval_wer": 0.2710914889657518,
"step": 2000
},
{
"epoch": 0.6,
"learning_rate": 7.438504016064257e-05,
"loss": 1.0416,
"step": 2100
},
{
"epoch": 0.63,
"learning_rate": 7.375753012048193e-05,
"loss": 1.0432,
"step": 2200
},
{
"epoch": 0.66,
"learning_rate": 7.313002008032128e-05,
"loss": 1.0324,
"step": 2300
},
{
"epoch": 0.69,
"learning_rate": 7.250251004016064e-05,
"loss": 1.0323,
"step": 2400
},
{
"epoch": 0.72,
"learning_rate": 7.18812751004016e-05,
"loss": 1.0347,
"step": 2500
},
{
"epoch": 0.75,
"learning_rate": 7.125376506024097e-05,
"loss": 1.0236,
"step": 2600
},
{
"epoch": 0.77,
"learning_rate": 7.062625502008031e-05,
"loss": 1.0245,
"step": 2700
},
{
"epoch": 0.8,
"learning_rate": 6.999874497991966e-05,
"loss": 1.0117,
"step": 2800
},
{
"epoch": 0.83,
"learning_rate": 6.937123493975902e-05,
"loss": 1.0093,
"step": 2900
},
{
"epoch": 0.86,
"learning_rate": 6.874372489959839e-05,
"loss": 1.0048,
"step": 3000
},
{
"epoch": 0.86,
"eval_loss": Infinity,
"eval_runtime": 976.2,
"eval_samples_per_second": 16.412,
"eval_steps_per_second": 1.026,
"eval_wer": 0.26197275288289384,
"step": 3000
},
{
"epoch": 0.89,
"learning_rate": 6.811621485943775e-05,
"loss": 1.0049,
"step": 3100
},
{
"epoch": 0.92,
"learning_rate": 6.74887048192771e-05,
"loss": 1.0064,
"step": 3200
},
{
"epoch": 0.95,
"learning_rate": 6.686119477911646e-05,
"loss": 1.0018,
"step": 3300
},
{
"epoch": 0.97,
"learning_rate": 6.623368473895582e-05,
"loss": 1.0046,
"step": 3400
},
{
"epoch": 1.0,
"learning_rate": 6.560617469879518e-05,
"loss": 1.0052,
"step": 3500
},
{
"epoch": 1.03,
"learning_rate": 6.497866465863453e-05,
"loss": 1.0034,
"step": 3600
},
{
"epoch": 1.06,
"learning_rate": 6.435115461847389e-05,
"loss": 0.9911,
"step": 3700
},
{
"epoch": 1.09,
"learning_rate": 6.372364457831325e-05,
"loss": 0.9923,
"step": 3800
},
{
"epoch": 1.12,
"learning_rate": 6.310240963855422e-05,
"loss": 0.9886,
"step": 3900
},
{
"epoch": 1.15,
"learning_rate": 6.247489959839357e-05,
"loss": 0.9858,
"step": 4000
},
{
"epoch": 1.15,
"eval_loss": Infinity,
"eval_runtime": 984.2539,
"eval_samples_per_second": 16.277,
"eval_steps_per_second": 1.018,
"eval_wer": 0.2522379727015459,
"step": 4000
},
{
"epoch": 1.18,
"learning_rate": 6.184738955823293e-05,
"loss": 0.991,
"step": 4100
},
{
"epoch": 1.2,
"learning_rate": 6.121987951807229e-05,
"loss": 0.9756,
"step": 4200
},
{
"epoch": 1.23,
"learning_rate": 6.059236947791164e-05,
"loss": 0.9871,
"step": 4300
},
{
"epoch": 1.26,
"learning_rate": 5.9964859437751e-05,
"loss": 0.9762,
"step": 4400
},
{
"epoch": 1.29,
"learning_rate": 5.933734939759035e-05,
"loss": 0.9901,
"step": 4500
},
{
"epoch": 1.32,
"learning_rate": 5.870983935742971e-05,
"loss": 0.9765,
"step": 4600
},
{
"epoch": 1.35,
"learning_rate": 5.808232931726907e-05,
"loss": 0.9784,
"step": 4700
},
{
"epoch": 1.38,
"learning_rate": 5.7454819277108424e-05,
"loss": 0.9736,
"step": 4800
},
{
"epoch": 1.4,
"learning_rate": 5.6827309236947786e-05,
"loss": 0.9694,
"step": 4900
},
{
"epoch": 1.43,
"learning_rate": 5.619979919678714e-05,
"loss": 0.9709,
"step": 5000
},
{
"epoch": 1.43,
"eval_loss": Infinity,
"eval_runtime": 987.1835,
"eval_samples_per_second": 16.229,
"eval_steps_per_second": 1.015,
"eval_wer": 0.23654168244210147,
"step": 5000
},
{
"epoch": 1.46,
"learning_rate": 5.55722891566265e-05,
"loss": 0.9664,
"step": 5100
},
{
"epoch": 1.49,
"learning_rate": 5.494477911646586e-05,
"loss": 0.9648,
"step": 5200
},
{
"epoch": 1.52,
"learning_rate": 5.431726907630522e-05,
"loss": 0.9649,
"step": 5300
},
{
"epoch": 1.55,
"learning_rate": 5.368975903614457e-05,
"loss": 0.9524,
"step": 5400
},
{
"epoch": 1.58,
"learning_rate": 5.3062248995983935e-05,
"loss": 0.9512,
"step": 5500
},
{
"epoch": 1.61,
"learning_rate": 5.243473895582329e-05,
"loss": 0.9447,
"step": 5600
},
{
"epoch": 1.63,
"learning_rate": 5.180722891566265e-05,
"loss": 0.9503,
"step": 5700
},
{
"epoch": 1.66,
"learning_rate": 5.1179718875502006e-05,
"loss": 0.9494,
"step": 5800
},
{
"epoch": 1.69,
"learning_rate": 5.0558483935742966e-05,
"loss": 0.9518,
"step": 5900
},
{
"epoch": 1.72,
"learning_rate": 4.993097389558232e-05,
"loss": 0.9347,
"step": 6000
},
{
"epoch": 1.72,
"eval_loss": Infinity,
"eval_runtime": 976.0993,
"eval_samples_per_second": 16.413,
"eval_steps_per_second": 1.027,
"eval_wer": 0.23323044541271745,
"step": 6000
},
{
"epoch": 1.75,
"learning_rate": 4.930346385542168e-05,
"loss": 0.9401,
"step": 6100
},
{
"epoch": 1.78,
"learning_rate": 4.867595381526104e-05,
"loss": 0.9313,
"step": 6200
},
{
"epoch": 1.81,
"learning_rate": 4.80484437751004e-05,
"loss": 0.9346,
"step": 6300
},
{
"epoch": 1.83,
"learning_rate": 4.7420933734939753e-05,
"loss": 0.9282,
"step": 6400
},
{
"epoch": 1.86,
"learning_rate": 4.679342369477911e-05,
"loss": 0.929,
"step": 6500
},
{
"epoch": 1.89,
"learning_rate": 4.616591365461847e-05,
"loss": 0.928,
"step": 6600
},
{
"epoch": 1.92,
"learning_rate": 4.5538403614457825e-05,
"loss": 0.9283,
"step": 6700
},
{
"epoch": 1.95,
"learning_rate": 4.4910893574297186e-05,
"loss": 0.9231,
"step": 6800
},
{
"epoch": 1.98,
"learning_rate": 4.428338353413654e-05,
"loss": 0.9162,
"step": 6900
},
{
"epoch": 2.01,
"learning_rate": 4.36558734939759e-05,
"loss": 0.9256,
"step": 7000
},
{
"epoch": 2.01,
"eval_loss": Infinity,
"eval_runtime": 987.0925,
"eval_samples_per_second": 16.23,
"eval_steps_per_second": 1.015,
"eval_wer": 0.22612026977597813,
"step": 7000
},
{
"epoch": 2.04,
"learning_rate": 4.302836345381526e-05,
"loss": 0.914,
"step": 7100
},
{
"epoch": 2.06,
"learning_rate": 4.240085341365462e-05,
"loss": 0.9153,
"step": 7200
},
{
"epoch": 2.09,
"learning_rate": 4.1773343373493974e-05,
"loss": 0.9042,
"step": 7300
},
{
"epoch": 2.12,
"learning_rate": 4.1145833333333335e-05,
"loss": 0.8996,
"step": 7400
},
{
"epoch": 2.15,
"learning_rate": 4.051832329317268e-05,
"loss": 0.9048,
"step": 7500
},
{
"epoch": 2.18,
"learning_rate": 3.989708835341365e-05,
"loss": 0.9114,
"step": 7600
},
{
"epoch": 2.21,
"learning_rate": 3.9269578313253005e-05,
"loss": 0.8999,
"step": 7700
},
{
"epoch": 2.24,
"learning_rate": 3.8642068273092367e-05,
"loss": 0.8994,
"step": 7800
},
{
"epoch": 2.26,
"learning_rate": 3.801455823293172e-05,
"loss": 0.8992,
"step": 7900
},
{
"epoch": 2.29,
"learning_rate": 3.738704819277108e-05,
"loss": 0.8936,
"step": 8000
},
{
"epoch": 2.29,
"eval_loss": Infinity,
"eval_runtime": 980.3676,
"eval_samples_per_second": 16.342,
"eval_steps_per_second": 1.022,
"eval_wer": 0.2203063535964783,
"step": 8000
},
{
"epoch": 2.32,
"learning_rate": 3.675953815261044e-05,
"loss": 0.8957,
"step": 8100
},
{
"epoch": 2.35,
"learning_rate": 3.613202811244979e-05,
"loss": 0.8956,
"step": 8200
},
{
"epoch": 2.38,
"learning_rate": 3.5504518072289154e-05,
"loss": 0.8816,
"step": 8300
},
{
"epoch": 2.41,
"learning_rate": 3.487700803212851e-05,
"loss": 0.8851,
"step": 8400
},
{
"epoch": 2.44,
"learning_rate": 3.424949799196787e-05,
"loss": 0.8703,
"step": 8500
},
{
"epoch": 2.47,
"learning_rate": 3.3621987951807225e-05,
"loss": 0.8809,
"step": 8600
},
{
"epoch": 2.49,
"learning_rate": 3.299447791164658e-05,
"loss": 0.8733,
"step": 8700
},
{
"epoch": 2.52,
"learning_rate": 3.236696787148594e-05,
"loss": 0.8715,
"step": 8800
},
{
"epoch": 2.55,
"learning_rate": 3.1739457831325296e-05,
"loss": 0.8686,
"step": 8900
},
{
"epoch": 2.58,
"learning_rate": 3.111822289156626e-05,
"loss": 0.877,
"step": 9000
},
{
"epoch": 2.58,
"eval_loss": Infinity,
"eval_runtime": 976.1269,
"eval_samples_per_second": 16.413,
"eval_steps_per_second": 1.027,
"eval_wer": 0.20961542163726554,
"step": 9000
},
{
"epoch": 2.61,
"learning_rate": 3.049071285140562e-05,
"loss": 0.8666,
"step": 9100
},
{
"epoch": 2.64,
"learning_rate": 2.986320281124498e-05,
"loss": 0.8581,
"step": 9200
},
{
"epoch": 2.67,
"learning_rate": 2.9235692771084334e-05,
"loss": 0.8644,
"step": 9300
},
{
"epoch": 2.69,
"learning_rate": 2.8608182730923693e-05,
"loss": 0.856,
"step": 9400
},
{
"epoch": 2.72,
"learning_rate": 2.7980672690763047e-05,
"loss": 0.854,
"step": 9500
},
{
"epoch": 2.75,
"learning_rate": 2.7353162650602406e-05,
"loss": 0.8491,
"step": 9600
},
{
"epoch": 2.78,
"learning_rate": 2.6725652610441764e-05,
"loss": 0.8514,
"step": 9700
},
{
"epoch": 2.81,
"learning_rate": 2.6098142570281122e-05,
"loss": 0.847,
"step": 9800
},
{
"epoch": 2.84,
"learning_rate": 2.547063253012048e-05,
"loss": 0.8439,
"step": 9900
},
{
"epoch": 2.87,
"learning_rate": 2.4843122489959838e-05,
"loss": 0.8393,
"step": 10000
},
{
"epoch": 2.87,
"eval_loss": Infinity,
"eval_runtime": 976.7065,
"eval_samples_per_second": 16.403,
"eval_steps_per_second": 1.026,
"eval_wer": 0.20169027099523207,
"step": 10000
},
{
"epoch": 2.9,
"learning_rate": 2.4215612449799196e-05,
"loss": 0.8417,
"step": 10100
},
{
"epoch": 2.92,
"learning_rate": 2.3588102409638555e-05,
"loss": 0.8348,
"step": 10200
},
{
"epoch": 2.95,
"learning_rate": 2.296059236947791e-05,
"loss": 0.8356,
"step": 10300
},
{
"epoch": 2.98,
"learning_rate": 2.2333082329317267e-05,
"loss": 0.8368,
"step": 10400
},
{
"epoch": 3.01,
"learning_rate": 2.1705572289156622e-05,
"loss": 0.8378,
"step": 10500
},
{
"epoch": 3.04,
"learning_rate": 2.107806224899598e-05,
"loss": 0.8301,
"step": 10600
},
{
"epoch": 3.07,
"learning_rate": 2.045055220883534e-05,
"loss": 0.8254,
"step": 10700
},
{
"epoch": 3.1,
"learning_rate": 1.9823042168674697e-05,
"loss": 0.8241,
"step": 10800
},
{
"epoch": 3.12,
"learning_rate": 1.9195532128514055e-05,
"loss": 0.8152,
"step": 10900
},
{
"epoch": 3.15,
"learning_rate": 1.8568022088353413e-05,
"loss": 0.8156,
"step": 11000
},
{
"epoch": 3.15,
"eval_loss": Infinity,
"eval_runtime": 979.3948,
"eval_samples_per_second": 16.358,
"eval_steps_per_second": 1.023,
"eval_wer": 0.1935918579504983,
"step": 11000
},
{
"epoch": 3.18,
"learning_rate": 1.794051204819277e-05,
"loss": 0.8104,
"step": 11100
},
{
"epoch": 3.21,
"learning_rate": 1.731300200803213e-05,
"loss": 0.8161,
"step": 11200
},
{
"epoch": 3.24,
"learning_rate": 1.6685491967871484e-05,
"loss": 0.8113,
"step": 11300
},
{
"epoch": 3.27,
"learning_rate": 1.6057981927710842e-05,
"loss": 0.8055,
"step": 11400
},
{
"epoch": 3.3,
"learning_rate": 1.54304718875502e-05,
"loss": 0.8106,
"step": 11500
},
{
"epoch": 3.33,
"learning_rate": 1.4802961847389557e-05,
"loss": 0.8028,
"step": 11600
},
{
"epoch": 3.35,
"learning_rate": 1.4175451807228913e-05,
"loss": 0.808,
"step": 11700
},
{
"epoch": 3.38,
"learning_rate": 1.3547941767068272e-05,
"loss": 0.806,
"step": 11800
},
{
"epoch": 3.41,
"learning_rate": 1.292043172690763e-05,
"loss": 0.7995,
"step": 11900
},
{
"epoch": 3.44,
"learning_rate": 1.2292921686746988e-05,
"loss": 0.8015,
"step": 12000
},
{
"epoch": 3.44,
"eval_loss": Infinity,
"eval_runtime": 977.0381,
"eval_samples_per_second": 16.398,
"eval_steps_per_second": 1.026,
"eval_wer": 0.18802820968601003,
"step": 12000
},
{
"epoch": 3.47,
"learning_rate": 1.1665411646586344e-05,
"loss": 0.7873,
"step": 12100
},
{
"epoch": 3.5,
"learning_rate": 1.1037901606425701e-05,
"loss": 0.7848,
"step": 12200
},
{
"epoch": 3.53,
"learning_rate": 1.0410391566265059e-05,
"loss": 0.7923,
"step": 12300
},
{
"epoch": 3.56,
"learning_rate": 9.782881526104417e-06,
"loss": 0.7897,
"step": 12400
},
{
"epoch": 3.58,
"learning_rate": 9.155371485943775e-06,
"loss": 0.7843,
"step": 12500
},
{
"epoch": 3.61,
"learning_rate": 8.527861445783132e-06,
"loss": 0.7896,
"step": 12600
},
{
"epoch": 3.64,
"learning_rate": 7.90035140562249e-06,
"loss": 0.7768,
"step": 12700
},
{
"epoch": 3.67,
"learning_rate": 7.2728413654618465e-06,
"loss": 0.775,
"step": 12800
},
{
"epoch": 3.7,
"learning_rate": 6.645331325301205e-06,
"loss": 0.7818,
"step": 12900
},
{
"epoch": 3.73,
"learning_rate": 6.017821285140562e-06,
"loss": 0.774,
"step": 13000
},
{
"epoch": 3.73,
"eval_loss": Infinity,
"eval_runtime": 979.8012,
"eval_samples_per_second": 16.351,
"eval_steps_per_second": 1.023,
"eval_wer": 0.18340787894733465,
"step": 13000
},
{
"epoch": 3.76,
"learning_rate": 3.1045276838546063e-05,
"loss": 0.805,
"step": 13100
},
{
"epoch": 3.78,
"learning_rate": 3.0649038461538456e-05,
"loss": 0.816,
"step": 13200
},
{
"epoch": 3.81,
"learning_rate": 3.025280008453085e-05,
"loss": 0.8112,
"step": 13300
},
{
"epoch": 3.84,
"learning_rate": 2.9860524091293318e-05,
"loss": 0.8274,
"step": 13400
},
{
"epoch": 3.87,
"learning_rate": 2.946428571428571e-05,
"loss": 0.8249,
"step": 13500
},
{
"epoch": 3.9,
"learning_rate": 2.9068047337278103e-05,
"loss": 0.8276,
"step": 13600
},
{
"epoch": 3.93,
"learning_rate": 2.8675771344040573e-05,
"loss": 0.8227,
"step": 13700
},
{
"epoch": 3.96,
"learning_rate": 2.8279532967032966e-05,
"loss": 0.8246,
"step": 13800
},
{
"epoch": 3.99,
"learning_rate": 2.7883294590025355e-05,
"loss": 0.8198,
"step": 13900
},
{
"epoch": 4.01,
"learning_rate": 2.748705621301775e-05,
"loss": 0.8372,
"step": 14000
},
{
"epoch": 4.01,
"eval_loss": Infinity,
"eval_runtime": 973.1623,
"eval_samples_per_second": 16.463,
"eval_steps_per_second": 1.03,
"eval_wer": 0.19336725853959047,
"step": 14000
},
{
"epoch": 4.04,
"learning_rate": 2.7090817836010143e-05,
"loss": 0.8228,
"step": 14100
},
{
"epoch": 4.07,
"learning_rate": 2.6694579459002535e-05,
"loss": 0.8197,
"step": 14200
},
{
"epoch": 4.1,
"learning_rate": 2.6302303465765005e-05,
"loss": 0.8295,
"step": 14300
},
{
"epoch": 4.13,
"learning_rate": 2.5906065088757394e-05,
"loss": 0.8147,
"step": 14400
},
{
"epoch": 4.16,
"learning_rate": 2.5509826711749787e-05,
"loss": 0.8153,
"step": 14500
},
{
"epoch": 4.19,
"learning_rate": 2.511358833474218e-05,
"loss": 0.8155,
"step": 14600
},
{
"epoch": 4.21,
"learning_rate": 2.471734995773457e-05,
"loss": 0.8215,
"step": 14700
},
{
"epoch": 4.24,
"learning_rate": 2.4321111580726964e-05,
"loss": 0.8214,
"step": 14800
},
{
"epoch": 4.27,
"learning_rate": 2.3924873203719357e-05,
"loss": 0.811,
"step": 14900
},
{
"epoch": 4.3,
"learning_rate": 2.352863482671175e-05,
"loss": 0.8075,
"step": 15000
},
{
"epoch": 4.3,
"eval_loss": Infinity,
"eval_runtime": 973.424,
"eval_samples_per_second": 16.458,
"eval_steps_per_second": 1.029,
"eval_wer": 0.1923276841233885,
"step": 15000
},
{
"epoch": 4.33,
"learning_rate": 2.313239644970414e-05,
"loss": 0.8085,
"step": 15100
},
{
"epoch": 4.36,
"learning_rate": 2.273615807269653e-05,
"loss": 0.8149,
"step": 15200
},
{
"epoch": 4.39,
"learning_rate": 2.2339919695688927e-05,
"loss": 0.8103,
"step": 15300
},
{
"epoch": 4.42,
"learning_rate": 2.194368131868132e-05,
"loss": 0.8053,
"step": 15400
},
{
"epoch": 4.44,
"learning_rate": 2.154744294167371e-05,
"loss": 0.8038,
"step": 15500
},
{
"epoch": 4.47,
"learning_rate": 2.11512045646661e-05,
"loss": 0.8062,
"step": 15600
},
{
"epoch": 4.5,
"learning_rate": 2.075892857142857e-05,
"loss": 0.8074,
"step": 15700
},
{
"epoch": 4.53,
"learning_rate": 2.0362690194420963e-05,
"loss": 0.8082,
"step": 15800
},
{
"epoch": 4.56,
"learning_rate": 1.9966451817413352e-05,
"loss": 0.8093,
"step": 15900
},
{
"epoch": 4.59,
"learning_rate": 1.9570213440405745e-05,
"loss": 0.8069,
"step": 16000
},
{
"epoch": 4.59,
"eval_loss": Infinity,
"eval_runtime": 966.1996,
"eval_samples_per_second": 16.581,
"eval_steps_per_second": 1.037,
"eval_wer": 0.18768168488060938,
"step": 16000
},
{
"epoch": 4.62,
"learning_rate": 1.917397506339814e-05,
"loss": 0.8039,
"step": 16100
},
{
"epoch": 4.64,
"learning_rate": 1.8777736686390533e-05,
"loss": 0.8069,
"step": 16200
},
{
"epoch": 4.67,
"learning_rate": 1.8381498309382922e-05,
"loss": 0.8108,
"step": 16300
},
{
"epoch": 4.7,
"learning_rate": 1.7985259932375315e-05,
"loss": 0.7968,
"step": 16400
},
{
"epoch": 4.73,
"learning_rate": 1.7589021555367707e-05,
"loss": 0.8013,
"step": 16500
},
{
"epoch": 4.76,
"learning_rate": 1.71927831783601e-05,
"loss": 0.8059,
"step": 16600
},
{
"epoch": 4.79,
"learning_rate": 1.6796544801352492e-05,
"loss": 0.8076,
"step": 16700
},
{
"epoch": 4.82,
"learning_rate": 1.6400306424344885e-05,
"loss": 0.808,
"step": 16800
},
{
"epoch": 4.85,
"learning_rate": 1.6004068047337277e-05,
"loss": 0.8053,
"step": 16900
},
{
"epoch": 4.87,
"learning_rate": 1.560782967032967e-05,
"loss": 0.8064,
"step": 17000
},
{
"epoch": 4.87,
"eval_loss": Infinity,
"eval_runtime": 964.7985,
"eval_samples_per_second": 16.606,
"eval_steps_per_second": 1.039,
"eval_wer": 0.19554266426238345,
"step": 17000
},
{
"epoch": 4.9,
"learning_rate": 1.521159129332206e-05,
"loss": 0.8192,
"step": 17100
},
{
"epoch": 4.93,
"learning_rate": 1.4815352916314453e-05,
"loss": 0.8165,
"step": 17200
},
{
"epoch": 4.96,
"learning_rate": 1.4419114539306845e-05,
"loss": 0.8164,
"step": 17300
},
{
"epoch": 4.99,
"learning_rate": 1.4022876162299238e-05,
"loss": 0.8161,
"step": 17400
},
{
"epoch": 5.02,
"learning_rate": 1.362663778529163e-05,
"loss": 0.8215,
"step": 17500
},
{
"epoch": 5.05,
"learning_rate": 1.3230399408284023e-05,
"loss": 0.8304,
"step": 17600
},
{
"epoch": 5.07,
"learning_rate": 1.2834161031276414e-05,
"loss": 0.8087,
"step": 17700
},
{
"epoch": 5.1,
"learning_rate": 1.2437922654268808e-05,
"loss": 0.8133,
"step": 17800
},
{
"epoch": 5.13,
"learning_rate": 1.2041684277261199e-05,
"loss": 0.7987,
"step": 17900
},
{
"epoch": 5.16,
"learning_rate": 1.1645445900253593e-05,
"loss": 0.801,
"step": 18000
},
{
"epoch": 5.16,
"eval_loss": Infinity,
"eval_runtime": 968.3005,
"eval_samples_per_second": 16.545,
"eval_steps_per_second": 1.035,
"eval_wer": 0.18908061835426385,
"step": 18000
},
{
"epoch": 5.19,
"learning_rate": 1.1249207523245984e-05,
"loss": 0.8075,
"step": 18100
},
{
"epoch": 5.22,
"learning_rate": 1.0852969146238374e-05,
"loss": 0.8069,
"step": 18200
},
{
"epoch": 5.25,
"learning_rate": 1.0456730769230769e-05,
"loss": 0.8095,
"step": 18300
},
{
"epoch": 5.28,
"learning_rate": 1.006049239222316e-05,
"loss": 0.8069,
"step": 18400
},
{
"epoch": 5.3,
"learning_rate": 9.664254015215553e-06,
"loss": 0.8127,
"step": 18500
},
{
"epoch": 5.33,
"learning_rate": 9.268015638207944e-06,
"loss": 0.8132,
"step": 18600
},
{
"epoch": 5.36,
"learning_rate": 8.871777261200337e-06,
"loss": 0.8086,
"step": 18700
},
{
"epoch": 5.39,
"learning_rate": 8.47553888419273e-06,
"loss": 0.8018,
"step": 18800
},
{
"epoch": 5.42,
"learning_rate": 8.079300507185122e-06,
"loss": 0.8102,
"step": 18900
},
{
"epoch": 5.45,
"learning_rate": 7.683062130177514e-06,
"loss": 0.8022,
"step": 19000
},
{
"epoch": 5.45,
"eval_loss": Infinity,
"eval_runtime": 968.9288,
"eval_samples_per_second": 16.535,
"eval_steps_per_second": 1.034,
"eval_wer": 0.18952981717607953,
"step": 19000
},
{
"epoch": 5.48,
"learning_rate": 7.286823753169907e-06,
"loss": 0.7979,
"step": 19100
},
{
"epoch": 5.5,
"learning_rate": 6.890585376162298e-06,
"loss": 0.797,
"step": 19200
},
{
"epoch": 5.53,
"learning_rate": 6.494346999154691e-06,
"loss": 0.7969,
"step": 19300
},
{
"epoch": 5.56,
"learning_rate": 6.098108622147083e-06,
"loss": 0.8034,
"step": 19400
},
{
"epoch": 5.59,
"learning_rate": 5.701870245139475e-06,
"loss": 0.8025,
"step": 19500
},
{
"epoch": 5.62,
"learning_rate": 5.305631868131867e-06,
"loss": 0.7938,
"step": 19600
},
{
"epoch": 5.65,
"learning_rate": 4.90939349112426e-06,
"loss": 0.7942,
"step": 19700
},
{
"epoch": 5.68,
"learning_rate": 4.513155114116652e-06,
"loss": 0.7855,
"step": 19800
},
{
"epoch": 5.71,
"learning_rate": 4.116916737109044e-06,
"loss": 0.7798,
"step": 19900
},
{
"epoch": 5.73,
"learning_rate": 3.724640743871513e-06,
"loss": 0.792,
"step": 20000
},
{
"epoch": 5.73,
"eval_loss": Infinity,
"eval_runtime": 967.729,
"eval_samples_per_second": 16.555,
"eval_steps_per_second": 1.035,
"eval_wer": 0.1853586852592198,
"step": 20000
},
{
"epoch": 5.76,
"learning_rate": 3.3284023668639047e-06,
"loss": 0.786,
"step": 20100
},
{
"epoch": 5.79,
"learning_rate": 2.932163989856297e-06,
"loss": 0.7907,
"step": 20200
},
{
"epoch": 5.82,
"learning_rate": 2.5359256128486896e-06,
"loss": 0.7837,
"step": 20300
},
{
"epoch": 5.85,
"learning_rate": 2.1396872358410817e-06,
"loss": 0.7769,
"step": 20400
},
{
"epoch": 5.88,
"learning_rate": 1.743448858833474e-06,
"loss": 0.7821,
"step": 20500
},
{
"epoch": 5.91,
"learning_rate": 1.3511728655959423e-06,
"loss": 0.7837,
"step": 20600
},
{
"epoch": 5.93,
"learning_rate": 9.549344885883346e-07,
"loss": 0.7856,
"step": 20700
},
{
"epoch": 5.96,
"learning_rate": 5.586961115807269e-07,
"loss": 0.7895,
"step": 20800
},
{
"epoch": 5.99,
"learning_rate": 1.6245773457311917e-07,
"loss": 0.7739,
"step": 20900
},
{
"epoch": 6.0,
"step": 20928,
"total_flos": 1.2767684631907514e+21,
"train_loss": 0.30566064197718185,
"train_runtime": 111829.698,
"train_samples_per_second": 23.955,
"train_steps_per_second": 0.187
}
],
"max_steps": 20928,
"num_train_epochs": 6,
"total_flos": 1.2767684631907514e+21,
"trial_name": null,
"trial_params": null
}