Commit
·
d008a64
1
Parent(s):
7806e29
End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +277 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_loss": 1.2441085577011108,
|
| 4 |
+
"eval_runtime": 12.8952,
|
| 5 |
+
"eval_samples": 360,
|
| 6 |
+
"eval_samples_per_second": 27.917,
|
| 7 |
+
"eval_steps_per_second": 3.49,
|
| 8 |
+
"eval_wer": 0.5350842806742454,
|
| 9 |
+
"train_loss": 0.7377449184876901,
|
| 10 |
+
"train_runtime": 5722.2661,
|
| 11 |
+
"train_samples": 856,
|
| 12 |
+
"train_samples_per_second": 14.959,
|
| 13 |
+
"train_steps_per_second": 0.944
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_loss": 1.2441085577011108,
|
| 4 |
+
"eval_runtime": 12.8952,
|
| 5 |
+
"eval_samples": 360,
|
| 6 |
+
"eval_samples_per_second": 27.917,
|
| 7 |
+
"eval_steps_per_second": 3.49,
|
| 8 |
+
"eval_wer": 0.5350842806742454
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"train_loss": 0.7377449184876901,
|
| 4 |
+
"train_runtime": 5722.2661,
|
| 5 |
+
"train_samples": 856,
|
| 6 |
+
"train_samples_per_second": 14.959,
|
| 7 |
+
"train_steps_per_second": 0.944
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,277 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 100.0,
|
| 5 |
+
"global_step": 5400,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.7,
|
| 12 |
+
"learning_rate": 5.909999999999999e-05,
|
| 13 |
+
"loss": 11.4212,
|
| 14 |
+
"step": 200
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 7.41,
|
| 18 |
+
"learning_rate": 0.0001191,
|
| 19 |
+
"loss": 3.3931,
|
| 20 |
+
"step": 400
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 9.26,
|
| 24 |
+
"eval_loss": 2.6012766361236572,
|
| 25 |
+
"eval_runtime": 13.0763,
|
| 26 |
+
"eval_samples_per_second": 27.531,
|
| 27 |
+
"eval_steps_per_second": 3.441,
|
| 28 |
+
"eval_wer": 0.991375931007448,
|
| 29 |
+
"step": 500
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 11.11,
|
| 33 |
+
"learning_rate": 0.0001791,
|
| 34 |
+
"loss": 2.2716,
|
| 35 |
+
"step": 600
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 14.81,
|
| 39 |
+
"learning_rate": 0.00023909999999999998,
|
| 40 |
+
"loss": 0.5568,
|
| 41 |
+
"step": 800
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 18.52,
|
| 45 |
+
"learning_rate": 0.0002991,
|
| 46 |
+
"loss": 0.3457,
|
| 47 |
+
"step": 1000
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 18.52,
|
| 51 |
+
"eval_loss": 1.0315884351730347,
|
| 52 |
+
"eval_runtime": 12.6921,
|
| 53 |
+
"eval_samples_per_second": 28.364,
|
| 54 |
+
"eval_steps_per_second": 3.546,
|
| 55 |
+
"eval_wer": 0.6452371618972952,
|
| 56 |
+
"step": 1000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 22.22,
|
| 60 |
+
"learning_rate": 0.00032255581395348835,
|
| 61 |
+
"loss": 0.2708,
|
| 62 |
+
"step": 1200
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 25.93,
|
| 66 |
+
"learning_rate": 0.00030720697674418605,
|
| 67 |
+
"loss": 0.2118,
|
| 68 |
+
"step": 1400
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 27.78,
|
| 72 |
+
"eval_loss": 1.1182368993759155,
|
| 73 |
+
"eval_runtime": 12.9665,
|
| 74 |
+
"eval_samples_per_second": 27.764,
|
| 75 |
+
"eval_steps_per_second": 3.47,
|
| 76 |
+
"eval_wer": 0.6354370834966679,
|
| 77 |
+
"step": 1500
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 29.63,
|
| 81 |
+
"learning_rate": 0.00029185813953488374,
|
| 82 |
+
"loss": 0.1765,
|
| 83 |
+
"step": 1600
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 33.33,
|
| 87 |
+
"learning_rate": 0.0002765093023255814,
|
| 88 |
+
"loss": 0.1404,
|
| 89 |
+
"step": 1800
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 37.04,
|
| 93 |
+
"learning_rate": 0.00026116046511627907,
|
| 94 |
+
"loss": 0.1245,
|
| 95 |
+
"step": 2000
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 37.04,
|
| 99 |
+
"eval_loss": 1.2044304609298706,
|
| 100 |
+
"eval_runtime": 12.7752,
|
| 101 |
+
"eval_samples_per_second": 28.18,
|
| 102 |
+
"eval_steps_per_second": 3.522,
|
| 103 |
+
"eval_wer": 0.6295570364562917,
|
| 104 |
+
"step": 2000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 40.74,
|
| 108 |
+
"learning_rate": 0.00024581162790697676,
|
| 109 |
+
"loss": 0.1166,
|
| 110 |
+
"step": 2200
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 44.44,
|
| 114 |
+
"learning_rate": 0.00023046279069767443,
|
| 115 |
+
"loss": 0.1068,
|
| 116 |
+
"step": 2400
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 46.3,
|
| 120 |
+
"eval_loss": 1.1400295495986938,
|
| 121 |
+
"eval_runtime": 12.7392,
|
| 122 |
+
"eval_samples_per_second": 28.259,
|
| 123 |
+
"eval_steps_per_second": 3.532,
|
| 124 |
+
"eval_wer": 0.582124656997256,
|
| 125 |
+
"step": 2500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 48.15,
|
| 129 |
+
"learning_rate": 0.0002151139534883721,
|
| 130 |
+
"loss": 0.0917,
|
| 131 |
+
"step": 2600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 51.85,
|
| 135 |
+
"learning_rate": 0.00019976511627906976,
|
| 136 |
+
"loss": 0.0842,
|
| 137 |
+
"step": 2800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 55.56,
|
| 141 |
+
"learning_rate": 0.00018441627906976745,
|
| 142 |
+
"loss": 0.0806,
|
| 143 |
+
"step": 3000
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 55.56,
|
| 147 |
+
"eval_loss": 1.2010185718536377,
|
| 148 |
+
"eval_runtime": 12.8794,
|
| 149 |
+
"eval_samples_per_second": 27.952,
|
| 150 |
+
"eval_steps_per_second": 3.494,
|
| 151 |
+
"eval_wer": 0.5785966287730302,
|
| 152 |
+
"step": 3000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 59.26,
|
| 156 |
+
"learning_rate": 0.00016906744186046509,
|
| 157 |
+
"loss": 0.0719,
|
| 158 |
+
"step": 3200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 62.96,
|
| 162 |
+
"learning_rate": 0.00015371860465116278,
|
| 163 |
+
"loss": 0.0679,
|
| 164 |
+
"step": 3400
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 64.81,
|
| 168 |
+
"eval_loss": 1.1644504070281982,
|
| 169 |
+
"eval_runtime": 12.9878,
|
| 170 |
+
"eval_samples_per_second": 27.718,
|
| 171 |
+
"eval_steps_per_second": 3.465,
|
| 172 |
+
"eval_wer": 0.546452371618973,
|
| 173 |
+
"step": 3500
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 66.67,
|
| 177 |
+
"learning_rate": 0.00013836976744186047,
|
| 178 |
+
"loss": 0.0582,
|
| 179 |
+
"step": 3600
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 70.37,
|
| 183 |
+
"learning_rate": 0.00012302093023255814,
|
| 184 |
+
"loss": 0.0566,
|
| 185 |
+
"step": 3800
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 74.07,
|
| 189 |
+
"learning_rate": 0.00010767209302325581,
|
| 190 |
+
"loss": 0.0506,
|
| 191 |
+
"step": 4000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 74.07,
|
| 195 |
+
"eval_loss": 1.2748003005981445,
|
| 196 |
+
"eval_runtime": 12.9412,
|
| 197 |
+
"eval_samples_per_second": 27.818,
|
| 198 |
+
"eval_steps_per_second": 3.477,
|
| 199 |
+
"eval_wer": 0.5507644061152489,
|
| 200 |
+
"step": 4000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 77.78,
|
| 204 |
+
"learning_rate": 9.232325581395348e-05,
|
| 205 |
+
"loss": 0.0443,
|
| 206 |
+
"step": 4200
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 81.48,
|
| 210 |
+
"learning_rate": 7.697441860465116e-05,
|
| 211 |
+
"loss": 0.0397,
|
| 212 |
+
"step": 4400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 83.33,
|
| 216 |
+
"eval_loss": 1.1869040727615356,
|
| 217 |
+
"eval_runtime": 12.9633,
|
| 218 |
+
"eval_samples_per_second": 27.771,
|
| 219 |
+
"eval_steps_per_second": 3.471,
|
| 220 |
+
"eval_wer": 0.5393963151705213,
|
| 221 |
+
"step": 4500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 85.19,
|
| 225 |
+
"learning_rate": 6.162558139534884e-05,
|
| 226 |
+
"loss": 0.0353,
|
| 227 |
+
"step": 4600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 88.89,
|
| 231 |
+
"learning_rate": 4.6276744186046515e-05,
|
| 232 |
+
"loss": 0.0282,
|
| 233 |
+
"step": 4800
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 92.59,
|
| 237 |
+
"learning_rate": 3.092790697674419e-05,
|
| 238 |
+
"loss": 0.0276,
|
| 239 |
+
"step": 5000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 92.59,
|
| 243 |
+
"eval_loss": 1.2239253520965576,
|
| 244 |
+
"eval_runtime": 12.8775,
|
| 245 |
+
"eval_samples_per_second": 27.956,
|
| 246 |
+
"eval_steps_per_second": 3.494,
|
| 247 |
+
"eval_wer": 0.5390043120344963,
|
| 248 |
+
"step": 5000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 96.3,
|
| 252 |
+
"learning_rate": 1.5579069767441862e-05,
|
| 253 |
+
"loss": 0.0247,
|
| 254 |
+
"step": 5200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 100.0,
|
| 258 |
+
"learning_rate": 2.302325581395349e-07,
|
| 259 |
+
"loss": 0.0218,
|
| 260 |
+
"step": 5400
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 100.0,
|
| 264 |
+
"step": 5400,
|
| 265 |
+
"total_flos": 1.3559072525710848e+19,
|
| 266 |
+
"train_loss": 0.7377449184876901,
|
| 267 |
+
"train_runtime": 5722.2661,
|
| 268 |
+
"train_samples_per_second": 14.959,
|
| 269 |
+
"train_steps_per_second": 0.944
|
| 270 |
+
}
|
| 271 |
+
],
|
| 272 |
+
"max_steps": 5400,
|
| 273 |
+
"num_train_epochs": 100,
|
| 274 |
+
"total_flos": 1.3559072525710848e+19,
|
| 275 |
+
"trial_name": null,
|
| 276 |
+
"trial_params": null
|
| 277 |
+
}
|