Upload trainer_state.json
Browse files- trainer_state.json +312 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,312 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 7.909466234149218,
|
| 5 |
+
"global_step": 6700,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.4,
|
| 12 |
+
"eval_cer": 0.06875412712023102,
|
| 13 |
+
"eval_loss": 0.2844763696193695,
|
| 14 |
+
"eval_runtime": 65.9715,
|
| 15 |
+
"eval_samples_per_second": 7.579,
|
| 16 |
+
"eval_steps_per_second": 0.955,
|
| 17 |
+
"eval_wer": 0.25628282600101654,
|
| 18 |
+
"step": 335
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.47,
|
| 22 |
+
"learning_rate": 0.0002394,
|
| 23 |
+
"loss": 3.0984,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.79,
|
| 28 |
+
"eval_cer": 0.04085233040615791,
|
| 29 |
+
"eval_loss": 0.15484967827796936,
|
| 30 |
+
"eval_runtime": 65.7396,
|
| 31 |
+
"eval_samples_per_second": 7.606,
|
| 32 |
+
"eval_steps_per_second": 0.958,
|
| 33 |
+
"eval_wer": 0.1411306263059807,
|
| 34 |
+
"step": 670
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.94,
|
| 38 |
+
"learning_rate": 0.0002857074569789675,
|
| 39 |
+
"loss": 0.2022,
|
| 40 |
+
"step": 800
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 1.19,
|
| 44 |
+
"eval_cer": 0.03730423898366892,
|
| 45 |
+
"eval_loss": 0.14234833419322968,
|
| 46 |
+
"eval_runtime": 65.8678,
|
| 47 |
+
"eval_samples_per_second": 7.591,
|
| 48 |
+
"eval_steps_per_second": 0.956,
|
| 49 |
+
"eval_wer": 0.11916191336759474,
|
| 50 |
+
"step": 1005
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.42,
|
| 54 |
+
"learning_rate": 0.0002665869980879541,
|
| 55 |
+
"loss": 0.1458,
|
| 56 |
+
"step": 1200
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.58,
|
| 60 |
+
"eval_cer": 0.03469244946433676,
|
| 61 |
+
"eval_loss": 0.12395954132080078,
|
| 62 |
+
"eval_runtime": 66.8459,
|
| 63 |
+
"eval_samples_per_second": 7.48,
|
| 64 |
+
"eval_steps_per_second": 0.942,
|
| 65 |
+
"eval_wer": 0.11165075958660417,
|
| 66 |
+
"step": 1340
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.89,
|
| 70 |
+
"learning_rate": 0.0002474665391969407,
|
| 71 |
+
"loss": 0.1354,
|
| 72 |
+
"step": 1600
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.98,
|
| 76 |
+
"eval_cer": 0.03423908222701871,
|
| 77 |
+
"eval_loss": 0.12080004811286926,
|
| 78 |
+
"eval_runtime": 65.8681,
|
| 79 |
+
"eval_samples_per_second": 7.591,
|
| 80 |
+
"eval_steps_per_second": 0.956,
|
| 81 |
+
"eval_wer": 0.10814931947817247,
|
| 82 |
+
"step": 1675
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 2.36,
|
| 86 |
+
"learning_rate": 0.00022834608030592734,
|
| 87 |
+
"loss": 0.1169,
|
| 88 |
+
"step": 2000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 2.37,
|
| 92 |
+
"eval_cer": 0.03338162679991721,
|
| 93 |
+
"eval_loss": 0.11872279644012451,
|
| 94 |
+
"eval_runtime": 66.8663,
|
| 95 |
+
"eval_samples_per_second": 7.478,
|
| 96 |
+
"eval_steps_per_second": 0.942,
|
| 97 |
+
"eval_wer": 0.10735867171175241,
|
| 98 |
+
"step": 2010
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.77,
|
| 102 |
+
"eval_cer": 0.03327321289534116,
|
| 103 |
+
"eval_loss": 0.12243428081274033,
|
| 104 |
+
"eval_runtime": 67.2919,
|
| 105 |
+
"eval_samples_per_second": 7.43,
|
| 106 |
+
"eval_steps_per_second": 0.936,
|
| 107 |
+
"eval_wer": 0.10199356186818773,
|
| 108 |
+
"step": 2345
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.83,
|
| 112 |
+
"learning_rate": 0.00020922562141491394,
|
| 113 |
+
"loss": 0.1107,
|
| 114 |
+
"step": 2400
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 3.16,
|
| 118 |
+
"eval_cer": 0.0313808974700137,
|
| 119 |
+
"eval_loss": 0.11623063683509827,
|
| 120 |
+
"eval_runtime": 66.9398,
|
| 121 |
+
"eval_samples_per_second": 7.469,
|
| 122 |
+
"eval_steps_per_second": 0.941,
|
| 123 |
+
"eval_wer": 0.09724967526966737,
|
| 124 |
+
"step": 2680
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 3.31,
|
| 128 |
+
"learning_rate": 0.00019010516252390057,
|
| 129 |
+
"loss": 0.0976,
|
| 130 |
+
"step": 2800
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 3.56,
|
| 134 |
+
"eval_cer": 0.03366744527561771,
|
| 135 |
+
"eval_loss": 0.11647358536720276,
|
| 136 |
+
"eval_runtime": 67.2006,
|
| 137 |
+
"eval_samples_per_second": 7.44,
|
| 138 |
+
"eval_steps_per_second": 0.937,
|
| 139 |
+
"eval_wer": 0.10052521601626475,
|
| 140 |
+
"step": 3015
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.78,
|
| 144 |
+
"learning_rate": 0.0001709847036328872,
|
| 145 |
+
"loss": 0.0914,
|
| 146 |
+
"step": 3200
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.95,
|
| 150 |
+
"eval_cer": 0.03200181346894927,
|
| 151 |
+
"eval_loss": 0.1139749139547348,
|
| 152 |
+
"eval_runtime": 67.5106,
|
| 153 |
+
"eval_samples_per_second": 7.406,
|
| 154 |
+
"eval_steps_per_second": 0.933,
|
| 155 |
+
"eval_wer": 0.09871802112159034,
|
| 156 |
+
"step": 3350
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 4.25,
|
| 160 |
+
"learning_rate": 0.00015186424474187378,
|
| 161 |
+
"loss": 0.0852,
|
| 162 |
+
"step": 3600
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 4.35,
|
| 166 |
+
"eval_cer": 0.03156815785064506,
|
| 167 |
+
"eval_loss": 0.12490493804216385,
|
| 168 |
+
"eval_runtime": 67.225,
|
| 169 |
+
"eval_samples_per_second": 7.438,
|
| 170 |
+
"eval_steps_per_second": 0.937,
|
| 171 |
+
"eval_wer": 0.09561190489636867,
|
| 172 |
+
"step": 3685
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 4.72,
|
| 176 |
+
"learning_rate": 0.0001327437858508604,
|
| 177 |
+
"loss": 0.0777,
|
| 178 |
+
"step": 4000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 4.75,
|
| 182 |
+
"eval_cer": 0.03226792032563595,
|
| 183 |
+
"eval_loss": 0.11327209323644638,
|
| 184 |
+
"eval_runtime": 67.468,
|
| 185 |
+
"eval_samples_per_second": 7.411,
|
| 186 |
+
"eval_steps_per_second": 0.934,
|
| 187 |
+
"eval_wer": 0.09532953069407579,
|
| 188 |
+
"step": 4020
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 5.14,
|
| 192 |
+
"eval_cer": 0.03157801366015198,
|
| 193 |
+
"eval_loss": 0.1231861561536789,
|
| 194 |
+
"eval_runtime": 67.6103,
|
| 195 |
+
"eval_samples_per_second": 7.395,
|
| 196 |
+
"eval_steps_per_second": 0.932,
|
| 197 |
+
"eval_wer": 0.09521658101315864,
|
| 198 |
+
"step": 4355
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 5.19,
|
| 202 |
+
"learning_rate": 0.00011362332695984703,
|
| 203 |
+
"loss": 0.0714,
|
| 204 |
+
"step": 4400
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 5.54,
|
| 208 |
+
"eval_cer": 0.032208785468594466,
|
| 209 |
+
"eval_loss": 0.13348866999149323,
|
| 210 |
+
"eval_runtime": 67.4714,
|
| 211 |
+
"eval_samples_per_second": 7.411,
|
| 212 |
+
"eval_steps_per_second": 0.934,
|
| 213 |
+
"eval_wer": 0.09555543005591009,
|
| 214 |
+
"step": 4690
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 5.67,
|
| 218 |
+
"learning_rate": 9.450286806883364e-05,
|
| 219 |
+
"loss": 0.0631,
|
| 220 |
+
"step": 4800
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 5.93,
|
| 224 |
+
"eval_cer": 0.03205109251648384,
|
| 225 |
+
"eval_loss": 0.12228710949420929,
|
| 226 |
+
"eval_runtime": 67.5997,
|
| 227 |
+
"eval_samples_per_second": 7.396,
|
| 228 |
+
"eval_steps_per_second": 0.932,
|
| 229 |
+
"eval_wer": 0.09521658101315864,
|
| 230 |
+
"step": 5025
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 6.14,
|
| 234 |
+
"learning_rate": 7.543021032504778e-05,
|
| 235 |
+
"loss": 0.0621,
|
| 236 |
+
"step": 5200
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.33,
|
| 240 |
+
"eval_cer": 0.032080659945004586,
|
| 241 |
+
"eval_loss": 0.1337049901485443,
|
| 242 |
+
"eval_runtime": 67.7409,
|
| 243 |
+
"eval_samples_per_second": 7.381,
|
| 244 |
+
"eval_steps_per_second": 0.93,
|
| 245 |
+
"eval_wer": 0.09504715649178291,
|
| 246 |
+
"step": 5360
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 6.61,
|
| 250 |
+
"learning_rate": 5.6309751434034414e-05,
|
| 251 |
+
"loss": 0.0538,
|
| 252 |
+
"step": 5600
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 6.72,
|
| 256 |
+
"eval_cer": 0.03255373880133645,
|
| 257 |
+
"eval_loss": 0.13896578550338745,
|
| 258 |
+
"eval_runtime": 67.8406,
|
| 259 |
+
"eval_samples_per_second": 7.37,
|
| 260 |
+
"eval_steps_per_second": 0.929,
|
| 261 |
+
"eval_wer": 0.09431298356582142,
|
| 262 |
+
"step": 5695
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 7.08,
|
| 266 |
+
"learning_rate": 3.718929254302103e-05,
|
| 267 |
+
"loss": 0.0513,
|
| 268 |
+
"step": 6000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 7.12,
|
| 272 |
+
"eval_cer": 0.03282970146753004,
|
| 273 |
+
"eval_loss": 0.1389976292848587,
|
| 274 |
+
"eval_runtime": 68.1201,
|
| 275 |
+
"eval_samples_per_second": 7.34,
|
| 276 |
+
"eval_steps_per_second": 0.925,
|
| 277 |
+
"eval_wer": 0.09549895521545151,
|
| 278 |
+
"step": 6030
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.51,
|
| 282 |
+
"eval_cer": 0.03255373880133645,
|
| 283 |
+
"eval_loss": 0.1410442739725113,
|
| 284 |
+
"eval_runtime": 67.6146,
|
| 285 |
+
"eval_samples_per_second": 7.395,
|
| 286 |
+
"eval_steps_per_second": 0.932,
|
| 287 |
+
"eval_wer": 0.09470830744903146,
|
| 288 |
+
"step": 6365
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 7.56,
|
| 292 |
+
"learning_rate": 1.8068833652007646e-05,
|
| 293 |
+
"loss": 0.0442,
|
| 294 |
+
"step": 6400
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 7.91,
|
| 298 |
+
"eval_cer": 0.03214965061155298,
|
| 299 |
+
"eval_loss": 0.1414366066455841,
|
| 300 |
+
"eval_runtime": 68.0581,
|
| 301 |
+
"eval_samples_per_second": 7.347,
|
| 302 |
+
"eval_steps_per_second": 0.926,
|
| 303 |
+
"eval_wer": 0.09453888292765573,
|
| 304 |
+
"step": 6700
|
| 305 |
+
}
|
| 306 |
+
],
|
| 307 |
+
"max_steps": 6776,
|
| 308 |
+
"num_train_epochs": 8,
|
| 309 |
+
"total_flos": 2.1224029808181802e+20,
|
| 310 |
+
"trial_name": null,
|
| 311 |
+
"trial_params": null
|
| 312 |
+
}
|