End of training
Browse files- README.md +1 -1
- all_results.json +16 -0
- eval--1hr test set_results.json +10 -0
- train_results.json +9 -0
- trainer_state.json +382 -0
README.md
CHANGED
|
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model was trained from scratch on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss:
|
| 20 |
- Wer: 1.0
|
| 21 |
- Cer: 1.0
|
| 22 |
|
|
|
|
| 16 |
|
| 17 |
This model was trained from scratch on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 1144.6145
|
| 20 |
- Wer: 1.0
|
| 21 |
- Cer: 1.0
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_cer": 1.0,
|
| 4 |
+
"eval_loss": 1144.614501953125,
|
| 5 |
+
"eval_runtime": 23.5534,
|
| 6 |
+
"eval_samples": 890,
|
| 7 |
+
"eval_samples_per_second": 37.786,
|
| 8 |
+
"eval_steps_per_second": 4.755,
|
| 9 |
+
"eval_wer": 1.0,
|
| 10 |
+
"total_flos": 3.1981440469898e+19,
|
| 11 |
+
"train_loss": 4410.907295081967,
|
| 12 |
+
"train_runtime": 25843.4041,
|
| 13 |
+
"train_samples": 24395,
|
| 14 |
+
"train_samples_per_second": 18.879,
|
| 15 |
+
"train_steps_per_second": 1.18
|
| 16 |
+
}
|
eval--1hr test set_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_cer": 1.0,
|
| 4 |
+
"eval_loss": 1144.614501953125,
|
| 5 |
+
"eval_runtime": 23.5534,
|
| 6 |
+
"eval_samples": 890,
|
| 7 |
+
"eval_samples_per_second": 37.786,
|
| 8 |
+
"eval_steps_per_second": 4.755,
|
| 9 |
+
"eval_wer": 1.0
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 3.1981440469898e+19,
|
| 4 |
+
"train_loss": 4410.907295081967,
|
| 5 |
+
"train_runtime": 25843.4041,
|
| 6 |
+
"train_samples": 24395,
|
| 7 |
+
"train_samples_per_second": 18.879,
|
| 8 |
+
"train_steps_per_second": 1.18
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,382 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.0,
|
| 3 |
+
"best_model_checkpoint": "/scratch/elec/puhe/p/palp3/sami_ASR/base_model_output/base-sami-cont-pt-22k/mixlabels/rerun/ep20/outputs/checkpoint-1525",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 30500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"grad_norm": 599.1070556640625,
|
| 14 |
+
"learning_rate": 9.908196721311476e-05,
|
| 15 |
+
"loss": 5425.9882,
|
| 16 |
+
"step": 1525
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_cer": 1.0,
|
| 21 |
+
"eval_loss": 1144.5670166015625,
|
| 22 |
+
"eval_runtime": 24.5176,
|
| 23 |
+
"eval_samples_per_second": 36.3,
|
| 24 |
+
"eval_steps_per_second": 4.568,
|
| 25 |
+
"eval_wer": 1.0,
|
| 26 |
+
"step": 1525
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.0,
|
| 30 |
+
"grad_norm": 729.23193359375,
|
| 31 |
+
"learning_rate": 0.00019908196721311476,
|
| 32 |
+
"loss": 4416.5207,
|
| 33 |
+
"step": 3050
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 2.0,
|
| 37 |
+
"eval_cer": 1.0,
|
| 38 |
+
"eval_loss": 1142.759521484375,
|
| 39 |
+
"eval_runtime": 26.3683,
|
| 40 |
+
"eval_samples_per_second": 33.753,
|
| 41 |
+
"eval_steps_per_second": 4.248,
|
| 42 |
+
"eval_wer": 1.0,
|
| 43 |
+
"step": 3050
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 3.0,
|
| 47 |
+
"grad_norm": 611.4544677734375,
|
| 48 |
+
"learning_rate": 0.00029908196721311475,
|
| 49 |
+
"loss": 4474.501,
|
| 50 |
+
"step": 4575
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.0,
|
| 54 |
+
"eval_cer": 1.0,
|
| 55 |
+
"eval_loss": 1144.2608642578125,
|
| 56 |
+
"eval_runtime": 25.6743,
|
| 57 |
+
"eval_samples_per_second": 34.665,
|
| 58 |
+
"eval_steps_per_second": 4.362,
|
| 59 |
+
"eval_wer": 1.0,
|
| 60 |
+
"step": 4575
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 4.0,
|
| 64 |
+
"grad_norm": 435.5348815917969,
|
| 65 |
+
"learning_rate": 0.00039901639344262297,
|
| 66 |
+
"loss": 4499.2403,
|
| 67 |
+
"step": 6100
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 4.0,
|
| 71 |
+
"eval_cer": 1.0,
|
| 72 |
+
"eval_loss": 1163.198486328125,
|
| 73 |
+
"eval_runtime": 24.0072,
|
| 74 |
+
"eval_samples_per_second": 37.072,
|
| 75 |
+
"eval_steps_per_second": 4.665,
|
| 76 |
+
"eval_wer": 1.0,
|
| 77 |
+
"step": 6100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.0,
|
| 81 |
+
"grad_norm": 500.39129638671875,
|
| 82 |
+
"learning_rate": 0.0004989508196721312,
|
| 83 |
+
"loss": 4492.3184,
|
| 84 |
+
"step": 7625
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 5.0,
|
| 88 |
+
"eval_cer": 1.0,
|
| 89 |
+
"eval_loss": 1145.3702392578125,
|
| 90 |
+
"eval_runtime": 23.8651,
|
| 91 |
+
"eval_samples_per_second": 37.293,
|
| 92 |
+
"eval_steps_per_second": 4.693,
|
| 93 |
+
"eval_wer": 1.0,
|
| 94 |
+
"step": 7625
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 6.0,
|
| 98 |
+
"grad_norm": 480.6201171875,
|
| 99 |
+
"learning_rate": 0.000467016393442623,
|
| 100 |
+
"loss": 4439.5584,
|
| 101 |
+
"step": 9150
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 6.0,
|
| 105 |
+
"eval_cer": 1.0,
|
| 106 |
+
"eval_loss": 1145.4093017578125,
|
| 107 |
+
"eval_runtime": 26.4383,
|
| 108 |
+
"eval_samples_per_second": 33.663,
|
| 109 |
+
"eval_steps_per_second": 4.236,
|
| 110 |
+
"eval_wer": 1.0,
|
| 111 |
+
"step": 9150
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 7.0,
|
| 115 |
+
"grad_norm": 496.0490417480469,
|
| 116 |
+
"learning_rate": 0.00043370491803278687,
|
| 117 |
+
"loss": 4400.9305,
|
| 118 |
+
"step": 10675
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 7.0,
|
| 122 |
+
"eval_cer": 1.0,
|
| 123 |
+
"eval_loss": 1144.0640869140625,
|
| 124 |
+
"eval_runtime": 25.1926,
|
| 125 |
+
"eval_samples_per_second": 35.328,
|
| 126 |
+
"eval_steps_per_second": 4.446,
|
| 127 |
+
"eval_wer": 1.0,
|
| 128 |
+
"step": 10675
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"grad_norm": 500.4902038574219,
|
| 133 |
+
"learning_rate": 0.0004003715846994536,
|
| 134 |
+
"loss": 4388.3134,
|
| 135 |
+
"step": 12200
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 8.0,
|
| 139 |
+
"eval_cer": 1.0,
|
| 140 |
+
"eval_loss": 1142.6368408203125,
|
| 141 |
+
"eval_runtime": 25.9206,
|
| 142 |
+
"eval_samples_per_second": 34.336,
|
| 143 |
+
"eval_steps_per_second": 4.321,
|
| 144 |
+
"eval_wer": 1.0,
|
| 145 |
+
"step": 12200
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 9.0,
|
| 149 |
+
"grad_norm": 519.6370239257812,
|
| 150 |
+
"learning_rate": 0.00036706010928961745,
|
| 151 |
+
"loss": 4337.4364,
|
| 152 |
+
"step": 13725
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 9.0,
|
| 156 |
+
"eval_cer": 1.0,
|
| 157 |
+
"eval_loss": 1141.1510009765625,
|
| 158 |
+
"eval_runtime": 25.2158,
|
| 159 |
+
"eval_samples_per_second": 35.295,
|
| 160 |
+
"eval_steps_per_second": 4.442,
|
| 161 |
+
"eval_wer": 1.0,
|
| 162 |
+
"step": 13725
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 10.0,
|
| 166 |
+
"grad_norm": 622.8638305664062,
|
| 167 |
+
"learning_rate": 0.00033372677595628415,
|
| 168 |
+
"loss": 4328.8672,
|
| 169 |
+
"step": 15250
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.0,
|
| 173 |
+
"eval_cer": 1.0,
|
| 174 |
+
"eval_loss": 1140.218017578125,
|
| 175 |
+
"eval_runtime": 24.9601,
|
| 176 |
+
"eval_samples_per_second": 35.657,
|
| 177 |
+
"eval_steps_per_second": 4.487,
|
| 178 |
+
"eval_wer": 1.0,
|
| 179 |
+
"step": 15250
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 11.0,
|
| 183 |
+
"grad_norm": 536.9533081054688,
|
| 184 |
+
"learning_rate": 0.00030041530054644814,
|
| 185 |
+
"loss": 4315.4475,
|
| 186 |
+
"step": 16775
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 11.0,
|
| 190 |
+
"eval_cer": 1.0,
|
| 191 |
+
"eval_loss": 1140.9918212890625,
|
| 192 |
+
"eval_runtime": 24.0596,
|
| 193 |
+
"eval_samples_per_second": 36.991,
|
| 194 |
+
"eval_steps_per_second": 4.655,
|
| 195 |
+
"eval_wer": 1.0,
|
| 196 |
+
"step": 16775
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 12.0,
|
| 200 |
+
"grad_norm": 479.0085754394531,
|
| 201 |
+
"learning_rate": 0.0002671256830601093,
|
| 202 |
+
"loss": 4340.9885,
|
| 203 |
+
"step": 18300
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 12.0,
|
| 207 |
+
"eval_cer": 1.0,
|
| 208 |
+
"eval_loss": 1141.5726318359375,
|
| 209 |
+
"eval_runtime": 24.2656,
|
| 210 |
+
"eval_samples_per_second": 36.677,
|
| 211 |
+
"eval_steps_per_second": 4.616,
|
| 212 |
+
"eval_wer": 1.0,
|
| 213 |
+
"step": 18300
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 13.0,
|
| 217 |
+
"grad_norm": 435.52410888671875,
|
| 218 |
+
"learning_rate": 0.00023379234972677597,
|
| 219 |
+
"loss": 4303.5944,
|
| 220 |
+
"step": 19825
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 13.0,
|
| 224 |
+
"eval_cer": 1.0,
|
| 225 |
+
"eval_loss": 1140.975830078125,
|
| 226 |
+
"eval_runtime": 25.0082,
|
| 227 |
+
"eval_samples_per_second": 35.588,
|
| 228 |
+
"eval_steps_per_second": 4.479,
|
| 229 |
+
"eval_wer": 1.0,
|
| 230 |
+
"step": 19825
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 14.0,
|
| 234 |
+
"grad_norm": 576.5276489257812,
|
| 235 |
+
"learning_rate": 0.00020045901639344265,
|
| 236 |
+
"loss": 4305.4659,
|
| 237 |
+
"step": 21350
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 14.0,
|
| 241 |
+
"eval_cer": 1.0,
|
| 242 |
+
"eval_loss": 1141.2747802734375,
|
| 243 |
+
"eval_runtime": 24.5459,
|
| 244 |
+
"eval_samples_per_second": 36.259,
|
| 245 |
+
"eval_steps_per_second": 4.563,
|
| 246 |
+
"eval_wer": 1.0,
|
| 247 |
+
"step": 21350
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 15.0,
|
| 251 |
+
"grad_norm": 485.7404479980469,
|
| 252 |
+
"learning_rate": 0.00016714754098360658,
|
| 253 |
+
"loss": 4296.8341,
|
| 254 |
+
"step": 22875
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 15.0,
|
| 258 |
+
"eval_cer": 1.0,
|
| 259 |
+
"eval_loss": 1140.5506591796875,
|
| 260 |
+
"eval_runtime": 27.5372,
|
| 261 |
+
"eval_samples_per_second": 32.32,
|
| 262 |
+
"eval_steps_per_second": 4.067,
|
| 263 |
+
"eval_wer": 1.0,
|
| 264 |
+
"step": 22875
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 16.0,
|
| 268 |
+
"grad_norm": 455.784423828125,
|
| 269 |
+
"learning_rate": 0.00013381420765027323,
|
| 270 |
+
"loss": 4293.78,
|
| 271 |
+
"step": 24400
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 16.0,
|
| 275 |
+
"eval_cer": 1.0,
|
| 276 |
+
"eval_loss": 1141.3170166015625,
|
| 277 |
+
"eval_runtime": 23.8888,
|
| 278 |
+
"eval_samples_per_second": 37.256,
|
| 279 |
+
"eval_steps_per_second": 4.688,
|
| 280 |
+
"eval_wer": 1.0,
|
| 281 |
+
"step": 24400
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.0,
|
| 285 |
+
"grad_norm": 530.2776489257812,
|
| 286 |
+
"learning_rate": 0.00010050273224043716,
|
| 287 |
+
"loss": 4291.8813,
|
| 288 |
+
"step": 25925
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 17.0,
|
| 292 |
+
"eval_cer": 1.0,
|
| 293 |
+
"eval_loss": 1140.3837890625,
|
| 294 |
+
"eval_runtime": 25.3819,
|
| 295 |
+
"eval_samples_per_second": 35.064,
|
| 296 |
+
"eval_steps_per_second": 4.413,
|
| 297 |
+
"eval_wer": 1.0,
|
| 298 |
+
"step": 25925
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 18.0,
|
| 302 |
+
"grad_norm": 510.512451171875,
|
| 303 |
+
"learning_rate": 6.716939890710383e-05,
|
| 304 |
+
"loss": 4291.3587,
|
| 305 |
+
"step": 27450
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 18.0,
|
| 309 |
+
"eval_cer": 1.0,
|
| 310 |
+
"eval_loss": 1141.1961669921875,
|
| 311 |
+
"eval_runtime": 24.3571,
|
| 312 |
+
"eval_samples_per_second": 36.54,
|
| 313 |
+
"eval_steps_per_second": 4.598,
|
| 314 |
+
"eval_wer": 1.0,
|
| 315 |
+
"step": 27450
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 19.0,
|
| 319 |
+
"grad_norm": 453.8408508300781,
|
| 320 |
+
"learning_rate": 3.383606557377049e-05,
|
| 321 |
+
"loss": 4288.1016,
|
| 322 |
+
"step": 28975
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 19.0,
|
| 326 |
+
"eval_cer": 1.0,
|
| 327 |
+
"eval_loss": 1140.9141845703125,
|
| 328 |
+
"eval_runtime": 23.8605,
|
| 329 |
+
"eval_samples_per_second": 37.3,
|
| 330 |
+
"eval_steps_per_second": 4.694,
|
| 331 |
+
"eval_wer": 1.0,
|
| 332 |
+
"step": 28975
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 20.0,
|
| 336 |
+
"grad_norm": 636.82763671875,
|
| 337 |
+
"learning_rate": 5.245901639344263e-07,
|
| 338 |
+
"loss": 4287.0193,
|
| 339 |
+
"step": 30500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 20.0,
|
| 343 |
+
"eval_cer": 1.0,
|
| 344 |
+
"eval_loss": 1141.5001220703125,
|
| 345 |
+
"eval_runtime": 26.8787,
|
| 346 |
+
"eval_samples_per_second": 33.112,
|
| 347 |
+
"eval_steps_per_second": 4.167,
|
| 348 |
+
"eval_wer": 1.0,
|
| 349 |
+
"step": 30500
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 20.0,
|
| 353 |
+
"step": 30500,
|
| 354 |
+
"total_flos": 3.1981440469898e+19,
|
| 355 |
+
"train_loss": 4410.907295081967,
|
| 356 |
+
"train_runtime": 25843.4041,
|
| 357 |
+
"train_samples_per_second": 18.879,
|
| 358 |
+
"train_steps_per_second": 1.18
|
| 359 |
+
}
|
| 360 |
+
],
|
| 361 |
+
"logging_steps": 500,
|
| 362 |
+
"max_steps": 30500,
|
| 363 |
+
"num_input_tokens_seen": 0,
|
| 364 |
+
"num_train_epochs": 20,
|
| 365 |
+
"save_steps": 500,
|
| 366 |
+
"stateful_callbacks": {
|
| 367 |
+
"TrainerControl": {
|
| 368 |
+
"args": {
|
| 369 |
+
"should_epoch_stop": false,
|
| 370 |
+
"should_evaluate": false,
|
| 371 |
+
"should_log": false,
|
| 372 |
+
"should_save": true,
|
| 373 |
+
"should_training_stop": true
|
| 374 |
+
},
|
| 375 |
+
"attributes": {}
|
| 376 |
+
}
|
| 377 |
+
},
|
| 378 |
+
"total_flos": 3.1981440469898e+19,
|
| 379 |
+
"train_batch_size": 16,
|
| 380 |
+
"trial_name": null,
|
| 381 |
+
"trial_params": null
|
| 382 |
+
}
|