End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +301 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 27.12,
|
| 3 |
+
"eval_loss": 0.7805310487747192,
|
| 4 |
+
"eval_runtime": 81.6634,
|
| 5 |
+
"eval_samples": 2095,
|
| 6 |
+
"eval_samples_per_second": 25.654,
|
| 7 |
+
"eval_steps_per_second": 3.208,
|
| 8 |
+
"eval_wer": 0.4340145266527145,
|
| 9 |
+
"train_loss": 0.48406150817871096,
|
| 10 |
+
"train_runtime": 11109.9321,
|
| 11 |
+
"train_samples": 4711,
|
| 12 |
+
"train_samples_per_second": 11.521,
|
| 13 |
+
"train_steps_per_second": 0.72
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 27.12,
|
| 3 |
+
"eval_loss": 0.7805310487747192,
|
| 4 |
+
"eval_runtime": 81.6634,
|
| 5 |
+
"eval_samples": 2095,
|
| 6 |
+
"eval_samples_per_second": 25.654,
|
| 7 |
+
"eval_steps_per_second": 3.208,
|
| 8 |
+
"eval_wer": 0.4340145266527145
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 27.12,
|
| 3 |
+
"train_loss": 0.48406150817871096,
|
| 4 |
+
"train_runtime": 11109.9321,
|
| 5 |
+
"train_samples": 4711,
|
| 6 |
+
"train_samples_per_second": 11.521,
|
| 7 |
+
"train_steps_per_second": 0.72
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,301 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 27.11864406779661,
|
| 5 |
+
"global_step": 8000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.36,
|
| 12 |
+
"eval_loss": 1.9129917621612549,
|
| 13 |
+
"eval_runtime": 85.8103,
|
| 14 |
+
"eval_samples_per_second": 24.414,
|
| 15 |
+
"eval_steps_per_second": 3.053,
|
| 16 |
+
"eval_wer": 0.9244121629939678,
|
| 17 |
+
"step": 400
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.69,
|
| 21 |
+
"learning_rate": 0.0002982,
|
| 22 |
+
"loss": 5.0013,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.71,
|
| 27 |
+
"eval_loss": 0.7788760662078857,
|
| 28 |
+
"eval_runtime": 85.5378,
|
| 29 |
+
"eval_samples_per_second": 24.492,
|
| 30 |
+
"eval_steps_per_second": 3.063,
|
| 31 |
+
"eval_wer": 0.5943616890311462,
|
| 32 |
+
"step": 800
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.39,
|
| 36 |
+
"learning_rate": 0.00028011999999999997,
|
| 37 |
+
"loss": 0.6544,
|
| 38 |
+
"step": 1000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.07,
|
| 42 |
+
"eval_loss": 0.7297950983047485,
|
| 43 |
+
"eval_runtime": 84.8704,
|
| 44 |
+
"eval_samples_per_second": 24.685,
|
| 45 |
+
"eval_steps_per_second": 3.087,
|
| 46 |
+
"eval_wer": 0.585190200664779,
|
| 47 |
+
"step": 1200
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 5.08,
|
| 51 |
+
"learning_rate": 0.00026011999999999997,
|
| 52 |
+
"loss": 0.4021,
|
| 53 |
+
"step": 1500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 5.42,
|
| 57 |
+
"eval_loss": 0.697790265083313,
|
| 58 |
+
"eval_runtime": 84.1609,
|
| 59 |
+
"eval_samples_per_second": 24.893,
|
| 60 |
+
"eval_steps_per_second": 3.113,
|
| 61 |
+
"eval_wer": 0.5667241167056506,
|
| 62 |
+
"step": 1600
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 6.78,
|
| 66 |
+
"learning_rate": 0.00024011999999999997,
|
| 67 |
+
"loss": 0.3003,
|
| 68 |
+
"step": 2000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 6.78,
|
| 72 |
+
"eval_loss": 0.6764048337936401,
|
| 73 |
+
"eval_runtime": 85.0271,
|
| 74 |
+
"eval_samples_per_second": 24.639,
|
| 75 |
+
"eval_steps_per_second": 3.081,
|
| 76 |
+
"eval_wer": 0.5382247937953958,
|
| 77 |
+
"step": 2000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 8.14,
|
| 81 |
+
"eval_loss": 0.72489994764328,
|
| 82 |
+
"eval_runtime": 83.797,
|
| 83 |
+
"eval_samples_per_second": 25.001,
|
| 84 |
+
"eval_steps_per_second": 3.127,
|
| 85 |
+
"eval_wer": 0.5462883171242152,
|
| 86 |
+
"step": 2400
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 8.47,
|
| 90 |
+
"learning_rate": 0.00022011999999999997,
|
| 91 |
+
"loss": 0.2345,
|
| 92 |
+
"step": 2500
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 9.49,
|
| 96 |
+
"eval_loss": 0.7279900908470154,
|
| 97 |
+
"eval_runtime": 81.6862,
|
| 98 |
+
"eval_samples_per_second": 25.647,
|
| 99 |
+
"eval_steps_per_second": 3.207,
|
| 100 |
+
"eval_wer": 0.5124338298658131,
|
| 101 |
+
"step": 2800
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 10.17,
|
| 105 |
+
"learning_rate": 0.00020012,
|
| 106 |
+
"loss": 0.1993,
|
| 107 |
+
"step": 3000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 10.85,
|
| 111 |
+
"eval_loss": 0.7288674712181091,
|
| 112 |
+
"eval_runtime": 81.4638,
|
| 113 |
+
"eval_samples_per_second": 25.717,
|
| 114 |
+
"eval_steps_per_second": 3.216,
|
| 115 |
+
"eval_wer": 0.4690385325618614,
|
| 116 |
+
"step": 3200
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 11.86,
|
| 120 |
+
"learning_rate": 0.00018012,
|
| 121 |
+
"loss": 0.1617,
|
| 122 |
+
"step": 3500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 12.2,
|
| 126 |
+
"eval_loss": 0.7430842518806458,
|
| 127 |
+
"eval_runtime": 81.3254,
|
| 128 |
+
"eval_samples_per_second": 25.761,
|
| 129 |
+
"eval_steps_per_second": 3.222,
|
| 130 |
+
"eval_wer": 0.4732857318724609,
|
| 131 |
+
"step": 3600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 13.56,
|
| 135 |
+
"learning_rate": 0.00016011999999999998,
|
| 136 |
+
"loss": 0.1432,
|
| 137 |
+
"step": 4000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 13.56,
|
| 141 |
+
"eval_loss": 0.7448425889015198,
|
| 142 |
+
"eval_runtime": 83.7392,
|
| 143 |
+
"eval_samples_per_second": 25.018,
|
| 144 |
+
"eval_steps_per_second": 3.129,
|
| 145 |
+
"eval_wer": 0.4732857318724609,
|
| 146 |
+
"step": 4000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 14.92,
|
| 150 |
+
"eval_loss": 0.7745729088783264,
|
| 151 |
+
"eval_runtime": 83.3791,
|
| 152 |
+
"eval_samples_per_second": 25.126,
|
| 153 |
+
"eval_steps_per_second": 3.142,
|
| 154 |
+
"eval_wer": 0.4484796257540318,
|
| 155 |
+
"step": 4400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 15.25,
|
| 159 |
+
"learning_rate": 0.00014012,
|
| 160 |
+
"loss": 0.1172,
|
| 161 |
+
"step": 4500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 16.27,
|
| 165 |
+
"eval_loss": 0.7588675022125244,
|
| 166 |
+
"eval_runtime": 83.1746,
|
| 167 |
+
"eval_samples_per_second": 25.188,
|
| 168 |
+
"eval_steps_per_second": 3.15,
|
| 169 |
+
"eval_wer": 0.4742090360704173,
|
| 170 |
+
"step": 4800
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 16.95,
|
| 174 |
+
"learning_rate": 0.00012011999999999998,
|
| 175 |
+
"loss": 0.1035,
|
| 176 |
+
"step": 5000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 17.63,
|
| 180 |
+
"eval_loss": 0.7539412975311279,
|
| 181 |
+
"eval_runtime": 83.8808,
|
| 182 |
+
"eval_samples_per_second": 24.976,
|
| 183 |
+
"eval_steps_per_second": 3.123,
|
| 184 |
+
"eval_wer": 0.4353071525298535,
|
| 185 |
+
"step": 5200
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 18.64,
|
| 189 |
+
"learning_rate": 0.00010011999999999998,
|
| 190 |
+
"loss": 0.0956,
|
| 191 |
+
"step": 5500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 18.98,
|
| 195 |
+
"eval_loss": 0.7648215293884277,
|
| 196 |
+
"eval_runtime": 83.2596,
|
| 197 |
+
"eval_samples_per_second": 25.162,
|
| 198 |
+
"eval_steps_per_second": 3.147,
|
| 199 |
+
"eval_wer": 0.44946448356518526,
|
| 200 |
+
"step": 5600
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 20.34,
|
| 204 |
+
"learning_rate": 8.012e-05,
|
| 205 |
+
"loss": 0.0845,
|
| 206 |
+
"step": 6000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 20.34,
|
| 210 |
+
"eval_loss": 0.7876783013343811,
|
| 211 |
+
"eval_runtime": 84.6623,
|
| 212 |
+
"eval_samples_per_second": 24.745,
|
| 213 |
+
"eval_steps_per_second": 3.095,
|
| 214 |
+
"eval_wer": 0.4718699987689277,
|
| 215 |
+
"step": 6000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 21.69,
|
| 219 |
+
"eval_loss": 0.7883597016334534,
|
| 220 |
+
"eval_runtime": 84.2577,
|
| 221 |
+
"eval_samples_per_second": 24.864,
|
| 222 |
+
"eval_steps_per_second": 3.11,
|
| 223 |
+
"eval_wer": 0.44343222947187,
|
| 224 |
+
"step": 6400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 22.03,
|
| 228 |
+
"learning_rate": 6.0119999999999994e-05,
|
| 229 |
+
"loss": 0.0761,
|
| 230 |
+
"step": 6500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 23.05,
|
| 234 |
+
"eval_loss": 0.7796189188957214,
|
| 235 |
+
"eval_runtime": 81.0587,
|
| 236 |
+
"eval_samples_per_second": 25.845,
|
| 237 |
+
"eval_steps_per_second": 3.232,
|
| 238 |
+
"eval_wer": 0.4385694940292995,
|
| 239 |
+
"step": 6800
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 23.73,
|
| 243 |
+
"learning_rate": 4.012e-05,
|
| 244 |
+
"loss": 0.0634,
|
| 245 |
+
"step": 7000
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 24.41,
|
| 249 |
+
"eval_loss": 0.7729123830795288,
|
| 250 |
+
"eval_runtime": 81.4812,
|
| 251 |
+
"eval_samples_per_second": 25.711,
|
| 252 |
+
"eval_steps_per_second": 3.215,
|
| 253 |
+
"eval_wer": 0.4306290779268743,
|
| 254 |
+
"step": 7200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 25.42,
|
| 258 |
+
"learning_rate": 2.0159999999999997e-05,
|
| 259 |
+
"loss": 0.0571,
|
| 260 |
+
"step": 7500
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 25.76,
|
| 264 |
+
"eval_loss": 0.7826283574104309,
|
| 265 |
+
"eval_runtime": 81.9341,
|
| 266 |
+
"eval_samples_per_second": 25.569,
|
| 267 |
+
"eval_steps_per_second": 3.198,
|
| 268 |
+
"eval_wer": 0.4298288809553121,
|
| 269 |
+
"step": 7600
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 27.12,
|
| 273 |
+
"learning_rate": 1.6e-07,
|
| 274 |
+
"loss": 0.0508,
|
| 275 |
+
"step": 8000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 27.12,
|
| 279 |
+
"eval_loss": 0.7805310487747192,
|
| 280 |
+
"eval_runtime": 81.3643,
|
| 281 |
+
"eval_samples_per_second": 25.748,
|
| 282 |
+
"eval_steps_per_second": 3.22,
|
| 283 |
+
"eval_wer": 0.4340145266527145,
|
| 284 |
+
"step": 8000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 27.12,
|
| 288 |
+
"step": 8000,
|
| 289 |
+
"total_flos": 1.5928886672052732e+19,
|
| 290 |
+
"train_loss": 0.48406150817871096,
|
| 291 |
+
"train_runtime": 11109.9321,
|
| 292 |
+
"train_samples_per_second": 11.521,
|
| 293 |
+
"train_steps_per_second": 0.72
|
| 294 |
+
}
|
| 295 |
+
],
|
| 296 |
+
"max_steps": 8000,
|
| 297 |
+
"num_train_epochs": 28,
|
| 298 |
+
"total_flos": 1.5928886672052732e+19,
|
| 299 |
+
"trial_name": null,
|
| 300 |
+
"trial_params": null
|
| 301 |
+
}
|