Commit
•
dfd8113
1
Parent(s):
8780e49
End of training
Browse files- trainer_state.json +368 -8
trainer_state.json
CHANGED
|
@@ -1,15 +1,375 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint":
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
-
"log_history": [
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
"trial_name": null,
|
| 14 |
"trial_params": null
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.6447107791900635,
|
| 3 |
+
"best_model_checkpoint": "detr-r101-cd45rb-8ah-12l\\checkpoint-115150",
|
| 4 |
+
"epoch": 25.0,
|
| 5 |
+
"global_step": 115150,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"learning_rate": 9.00238818931828e-06,
|
| 13 |
+
"loss": 3.0401,
|
| 14 |
+
"step": 4606
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_loss": 2.163705587387085,
|
| 19 |
+
"eval_runtime": 222.5701,
|
| 20 |
+
"eval_samples_per_second": 8.002,
|
| 21 |
+
"eval_steps_per_second": 1.002,
|
| 22 |
+
"step": 4606
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 2.0,
|
| 26 |
+
"learning_rate": 8.002822405557968e-06,
|
| 27 |
+
"loss": 2.7151,
|
| 28 |
+
"step": 9212
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 2.0,
|
| 32 |
+
"eval_loss": 2.06657338142395,
|
| 33 |
+
"eval_runtime": 222.3447,
|
| 34 |
+
"eval_samples_per_second": 8.01,
|
| 35 |
+
"eval_steps_per_second": 1.003,
|
| 36 |
+
"step": 9212
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 3.0,
|
| 40 |
+
"learning_rate": 7.003039513677812e-06,
|
| 41 |
+
"loss": 2.6021,
|
| 42 |
+
"step": 13818
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 3.0,
|
| 46 |
+
"eval_loss": 1.986824870109558,
|
| 47 |
+
"eval_runtime": 226.29,
|
| 48 |
+
"eval_samples_per_second": 7.87,
|
| 49 |
+
"eval_steps_per_second": 0.985,
|
| 50 |
+
"step": 13818
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 4.0,
|
| 54 |
+
"learning_rate": 6.003473729917499e-06,
|
| 55 |
+
"loss": 2.5221,
|
| 56 |
+
"step": 18424
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 4.0,
|
| 60 |
+
"eval_loss": 1.8957558870315552,
|
| 61 |
+
"eval_runtime": 219.599,
|
| 62 |
+
"eval_samples_per_second": 8.11,
|
| 63 |
+
"eval_steps_per_second": 1.015,
|
| 64 |
+
"step": 18424
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 5.0,
|
| 68 |
+
"learning_rate": 5.003907946157187e-06,
|
| 69 |
+
"loss": 2.4541,
|
| 70 |
+
"step": 23030
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 5.0,
|
| 74 |
+
"eval_loss": 1.8809998035430908,
|
| 75 |
+
"eval_runtime": 222.9944,
|
| 76 |
+
"eval_samples_per_second": 7.987,
|
| 77 |
+
"eval_steps_per_second": 1.0,
|
| 78 |
+
"step": 23030
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 6.0,
|
| 82 |
+
"learning_rate": 4.0041250542770305e-06,
|
| 83 |
+
"loss": 2.4155,
|
| 84 |
+
"step": 27636
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 6.0,
|
| 88 |
+
"eval_loss": 1.8369228839874268,
|
| 89 |
+
"eval_runtime": 220.1885,
|
| 90 |
+
"eval_samples_per_second": 8.089,
|
| 91 |
+
"eval_steps_per_second": 1.013,
|
| 92 |
+
"step": 27636
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 7.0,
|
| 96 |
+
"learning_rate": 3.0047763786365614e-06,
|
| 97 |
+
"loss": 2.3531,
|
| 98 |
+
"step": 32242
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 7.0,
|
| 102 |
+
"eval_loss": 1.8039908409118652,
|
| 103 |
+
"eval_runtime": 225.4193,
|
| 104 |
+
"eval_samples_per_second": 7.901,
|
| 105 |
+
"eval_steps_per_second": 0.989,
|
| 106 |
+
"step": 32242
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 8.0,
|
| 110 |
+
"learning_rate": 2.004993486756405e-06,
|
| 111 |
+
"loss": 2.31,
|
| 112 |
+
"step": 36848
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 8.0,
|
| 116 |
+
"eval_loss": 1.797875165939331,
|
| 117 |
+
"eval_runtime": 219.4975,
|
| 118 |
+
"eval_samples_per_second": 8.114,
|
| 119 |
+
"eval_steps_per_second": 1.016,
|
| 120 |
+
"step": 36848
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 9.0,
|
| 124 |
+
"learning_rate": 1.0054277029960922e-06,
|
| 125 |
+
"loss": 2.2841,
|
| 126 |
+
"step": 41454
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 9.0,
|
| 130 |
+
"eval_loss": 1.7520661354064941,
|
| 131 |
+
"eval_runtime": 222.6235,
|
| 132 |
+
"eval_samples_per_second": 8.0,
|
| 133 |
+
"eval_steps_per_second": 1.002,
|
| 134 |
+
"step": 41454
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 10.0,
|
| 138 |
+
"learning_rate": 5.644811115935737e-09,
|
| 139 |
+
"loss": 2.2555,
|
| 140 |
+
"step": 46060
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 10.0,
|
| 144 |
+
"eval_loss": 1.7242921590805054,
|
| 145 |
+
"eval_runtime": 220.3572,
|
| 146 |
+
"eval_samples_per_second": 8.082,
|
| 147 |
+
"eval_steps_per_second": 1.012,
|
| 148 |
+
"step": 46060
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 11.0,
|
| 152 |
+
"learning_rate": 5.602518454190186e-06,
|
| 153 |
+
"loss": 2.3388,
|
| 154 |
+
"step": 50666
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 11.0,
|
| 158 |
+
"eval_loss": 1.8519611358642578,
|
| 159 |
+
"eval_runtime": 224.9013,
|
| 160 |
+
"eval_samples_per_second": 7.919,
|
| 161 |
+
"eval_steps_per_second": 0.992,
|
| 162 |
+
"step": 50666
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 12.0,
|
| 166 |
+
"learning_rate": 5.202605297438124e-06,
|
| 167 |
+
"loss": 2.3523,
|
| 168 |
+
"step": 55272
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 12.0,
|
| 172 |
+
"eval_loss": 1.8499144315719604,
|
| 173 |
+
"eval_runtime": 224.2008,
|
| 174 |
+
"eval_samples_per_second": 7.944,
|
| 175 |
+
"eval_steps_per_second": 0.995,
|
| 176 |
+
"step": 55272
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 13.0,
|
| 180 |
+
"learning_rate": 4.802778983933999e-06,
|
| 181 |
+
"loss": 2.3515,
|
| 182 |
+
"step": 59878
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 13.0,
|
| 186 |
+
"eval_loss": 1.7634978294372559,
|
| 187 |
+
"eval_runtime": 225.8943,
|
| 188 |
+
"eval_samples_per_second": 7.884,
|
| 189 |
+
"eval_steps_per_second": 0.987,
|
| 190 |
+
"step": 59878
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 14.0,
|
| 194 |
+
"learning_rate": 4.402865827181937e-06,
|
| 195 |
+
"loss": 2.3236,
|
| 196 |
+
"step": 64484
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 14.0,
|
| 200 |
+
"eval_loss": 1.7787134647369385,
|
| 201 |
+
"eval_runtime": 223.7953,
|
| 202 |
+
"eval_samples_per_second": 7.958,
|
| 203 |
+
"eval_steps_per_second": 0.996,
|
| 204 |
+
"step": 64484
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 15.0,
|
| 208 |
+
"learning_rate": 4.002952670429874e-06,
|
| 209 |
+
"loss": 2.2676,
|
| 210 |
+
"step": 69090
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 15.0,
|
| 214 |
+
"eval_loss": 1.7517672777175903,
|
| 215 |
+
"eval_runtime": 226.211,
|
| 216 |
+
"eval_samples_per_second": 7.873,
|
| 217 |
+
"eval_steps_per_second": 0.986,
|
| 218 |
+
"step": 69090
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 16.0,
|
| 222 |
+
"learning_rate": 3.6033000434216243e-06,
|
| 223 |
+
"loss": 2.2787,
|
| 224 |
+
"step": 73696
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 16.0,
|
| 228 |
+
"eval_loss": 1.7879087924957275,
|
| 229 |
+
"eval_runtime": 223.8881,
|
| 230 |
+
"eval_samples_per_second": 7.955,
|
| 231 |
+
"eval_steps_per_second": 0.996,
|
| 232 |
+
"step": 73696
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 17.0,
|
| 236 |
+
"learning_rate": 3.2033868866695615e-06,
|
| 237 |
+
"loss": 2.2523,
|
| 238 |
+
"step": 78302
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 17.0,
|
| 242 |
+
"eval_loss": 1.7303467988967896,
|
| 243 |
+
"eval_runtime": 223.161,
|
| 244 |
+
"eval_samples_per_second": 7.981,
|
| 245 |
+
"eval_steps_per_second": 0.999,
|
| 246 |
+
"step": 78302
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 18.0,
|
| 250 |
+
"learning_rate": 2.8035605731654363e-06,
|
| 251 |
+
"loss": 2.2357,
|
| 252 |
+
"step": 82908
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 18.0,
|
| 256 |
+
"eval_loss": 1.7361239194869995,
|
| 257 |
+
"eval_runtime": 222.8596,
|
| 258 |
+
"eval_samples_per_second": 7.992,
|
| 259 |
+
"eval_steps_per_second": 1.001,
|
| 260 |
+
"step": 82908
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 19.0,
|
| 264 |
+
"learning_rate": 2.4036474164133743e-06,
|
| 265 |
+
"loss": 2.2068,
|
| 266 |
+
"step": 87514
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 19.0,
|
| 270 |
+
"eval_loss": 1.691589117050171,
|
| 271 |
+
"eval_runtime": 223.3285,
|
| 272 |
+
"eval_samples_per_second": 7.975,
|
| 273 |
+
"eval_steps_per_second": 0.999,
|
| 274 |
+
"step": 87514
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 20.0,
|
| 278 |
+
"learning_rate": 2.0037342596613115e-06,
|
| 279 |
+
"loss": 2.1972,
|
| 280 |
+
"step": 92120
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 20.0,
|
| 284 |
+
"eval_loss": 1.6940640211105347,
|
| 285 |
+
"eval_runtime": 222.4012,
|
| 286 |
+
"eval_samples_per_second": 8.008,
|
| 287 |
+
"eval_steps_per_second": 1.003,
|
| 288 |
+
"step": 92120
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 21.0,
|
| 292 |
+
"learning_rate": 1.6039079461571863e-06,
|
| 293 |
+
"loss": 2.1856,
|
| 294 |
+
"step": 96726
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 21.0,
|
| 298 |
+
"eval_loss": 1.6824039220809937,
|
| 299 |
+
"eval_runtime": 219.7189,
|
| 300 |
+
"eval_samples_per_second": 8.106,
|
| 301 |
+
"eval_steps_per_second": 1.015,
|
| 302 |
+
"step": 96726
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 22.0,
|
| 306 |
+
"learning_rate": 1.2039947894051239e-06,
|
| 307 |
+
"loss": 2.1611,
|
| 308 |
+
"step": 101332
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 22.0,
|
| 312 |
+
"eval_loss": 1.6710505485534668,
|
| 313 |
+
"eval_runtime": 219.2608,
|
| 314 |
+
"eval_samples_per_second": 8.123,
|
| 315 |
+
"eval_steps_per_second": 1.017,
|
| 316 |
+
"step": 101332
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 23.0,
|
| 320 |
+
"learning_rate": 8.041684759009987e-07,
|
| 321 |
+
"loss": 2.1419,
|
| 322 |
+
"step": 105938
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 23.0,
|
| 326 |
+
"eval_loss": 1.6535351276397705,
|
| 327 |
+
"eval_runtime": 220.9951,
|
| 328 |
+
"eval_samples_per_second": 8.059,
|
| 329 |
+
"eval_steps_per_second": 1.009,
|
| 330 |
+
"step": 105938
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 24.0,
|
| 334 |
+
"learning_rate": 4.0425531914893625e-07,
|
| 335 |
+
"loss": 2.1412,
|
| 336 |
+
"step": 110544
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 24.0,
|
| 340 |
+
"eval_loss": 1.6602009534835815,
|
| 341 |
+
"eval_runtime": 477.2637,
|
| 342 |
+
"eval_samples_per_second": 3.732,
|
| 343 |
+
"eval_steps_per_second": 0.467,
|
| 344 |
+
"step": 110544
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 25.0,
|
| 348 |
+
"learning_rate": 4.515848892748589e-09,
|
| 349 |
+
"loss": 2.1285,
|
| 350 |
+
"step": 115150
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 25.0,
|
| 354 |
+
"eval_loss": 1.6447107791900635,
|
| 355 |
+
"eval_runtime": 581.8045,
|
| 356 |
+
"eval_samples_per_second": 3.061,
|
| 357 |
+
"eval_steps_per_second": 0.383,
|
| 358 |
+
"step": 115150
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 25.0,
|
| 362 |
+
"step": 115150,
|
| 363 |
+
"total_flos": 4.126770236983678e+20,
|
| 364 |
+
"train_loss": 0.4303329726172384,
|
| 365 |
+
"train_runtime": 23864.0904,
|
| 366 |
+
"train_samples_per_second": 19.298,
|
| 367 |
+
"train_steps_per_second": 4.825
|
| 368 |
+
}
|
| 369 |
+
],
|
| 370 |
+
"max_steps": 115150,
|
| 371 |
+
"num_train_epochs": 25,
|
| 372 |
+
"total_flos": 4.126770236983678e+20,
|
| 373 |
"trial_name": null,
|
| 374 |
"trial_params": null
|
| 375 |
}
|