Model save
Browse files- README.md +3 -1
- adapter_model.safetensors +1 -1
- all_results.json +7 -7
- runs/Jan04_21-56-52_f82d07273834/events.out.tfevents.1736056683.f82d07273834.2210081.0 +2 -2
- train_results.json +7 -7
- trainer_state.json +723 -1794
README.md
CHANGED
|
@@ -40,8 +40,10 @@ The following hyperparameters were used during training:
|
|
| 40 |
- eval_batch_size: 8
|
| 41 |
- seed: 42
|
| 42 |
- distributed_type: multi-GPU
|
|
|
|
| 43 |
- gradient_accumulation_steps: 2
|
| 44 |
-
- total_train_batch_size:
|
|
|
|
| 45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
- lr_scheduler_type: cosine
|
| 47 |
- lr_scheduler_warmup_ratio: 0.1
|
|
|
|
| 40 |
- eval_batch_size: 8
|
| 41 |
- seed: 42
|
| 42 |
- distributed_type: multi-GPU
|
| 43 |
+
- num_devices: 2
|
| 44 |
- gradient_accumulation_steps: 2
|
| 45 |
+
- total_train_batch_size: 16
|
| 46 |
+
- total_eval_batch_size: 16
|
| 47 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 48 |
- lr_scheduler_type: cosine
|
| 49 |
- lr_scheduler_warmup_ratio: 0.1
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167832240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b36e01b194812bf82947e25cd6062aa7e7be470f0181e6ae76551238e04fd525
|
| 3 |
size 167832240
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 8.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 0.9994372537985369,
|
| 3 |
+
"total_flos": 8.639715708407644e+17,
|
| 4 |
+
"train_loss": 0.2980908542066007,
|
| 5 |
+
"train_runtime": 5659.1413,
|
| 6 |
+
"train_samples": 14215,
|
| 7 |
+
"train_samples_per_second": 2.512,
|
| 8 |
+
"train_steps_per_second": 0.157
|
| 9 |
}
|
runs/Jan04_21-56-52_f82d07273834/events.out.tfevents.1736056683.f82d07273834.2210081.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1515f12d9b566ff30d9c2de7655566e316d35edce4acd55b964b672e809b893a
|
| 3 |
+
size 44421
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 8.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 0.9994372537985369,
|
| 3 |
+
"total_flos": 8.639715708407644e+17,
|
| 4 |
+
"train_loss": 0.2980908542066007,
|
| 5 |
+
"train_runtime": 5659.1413,
|
| 6 |
+
"train_samples": 14215,
|
| 7 |
+
"train_samples_per_second": 2.512,
|
| 8 |
+
"train_steps_per_second": 0.157
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,2342 +1,1271 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 1.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm": 1.
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm": 1.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm": 1.
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate": 0.
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
-
"learning_rate": 0.
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm": 1.
|
| 147 |
-
"learning_rate": 0.
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate": 0.
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate": 0.
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate": 0.
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm":
|
| 175 |
-
"learning_rate": 0.
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 115
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm":
|
| 182 |
-
"learning_rate": 0.
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 120
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm":
|
| 189 |
-
"learning_rate": 0.
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 125
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch": 0.
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate": 0.
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 130
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 0.
|
| 202 |
-
"grad_norm":
|
| 203 |
-
"learning_rate": 0.
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 135
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate": 0.
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 140
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"grad_norm":
|
| 217 |
-
"learning_rate": 0.
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 145
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
-
"learning_rate": 0.
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 150
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch": 0.
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate": 0.
|
| 232 |
-
"loss": 0.
|
| 233 |
"step": 155
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch": 0.
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate": 0.
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 0.
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate": 0.
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 165
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch": 0.
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
-
"learning_rate": 0.
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 170
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
-
"learning_rate": 0.
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 175
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
-
"learning_rate": 0.
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 180
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch": 0.
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate": 0.
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 185
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
-
"learning_rate": 0.
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 190
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 0.
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate": 0.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 195
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 0.
|
| 293 |
-
"grad_norm":
|
| 294 |
-
"learning_rate": 0.
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 200
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate": 0.
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 205
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 0.
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
-
"learning_rate": 0.
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 210
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"epoch": 0.
|
| 314 |
-
"grad_norm":
|
| 315 |
-
"learning_rate": 0.
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 215
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch": 0.
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate": 0.
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 220
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch": 0.
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
-
"learning_rate": 0.
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 225
|
| 332 |
},
|
| 333 |
{
|
| 334 |
-
"epoch": 0.
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate": 0.
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 230
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
-
"learning_rate": 0.
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 235
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 0.
|
| 349 |
-
"grad_norm":
|
| 350 |
-
"learning_rate": 0.
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 240
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"epoch": 0.
|
| 356 |
-
"grad_norm":
|
| 357 |
-
"learning_rate": 0.
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 245
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"grad_norm":
|
| 364 |
-
"learning_rate": 0.
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 250
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch": 0.
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
-
"learning_rate": 0.
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 255
|
| 374 |
},
|
| 375 |
{
|
| 376 |
-
"epoch": 0.
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 0.
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 260
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch": 0.
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate": 0.
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 265
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 0.
|
| 391 |
-
"grad_norm":
|
| 392 |
-
"learning_rate": 0.
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 270
|
| 395 |
},
|
| 396 |
{
|
| 397 |
-
"epoch": 0.
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
-
"learning_rate": 0.
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 275
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch": 0.
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate": 0.
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 280
|
| 409 |
},
|
| 410 |
{
|
| 411 |
-
"epoch": 0.
|
| 412 |
-
"grad_norm":
|
| 413 |
-
"learning_rate": 0.
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 285
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch": 0.
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
-
"learning_rate": 0.
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 290
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch": 0.
|
| 426 |
-
"grad_norm":
|
| 427 |
-
"learning_rate": 0.
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 295
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 0.
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate": 0.
|
| 435 |
-
"loss": 0.
|
| 436 |
"step": 300
|
| 437 |
},
|
| 438 |
{
|
| 439 |
-
"epoch": 0.
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate": 0.
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 305
|
| 444 |
},
|
| 445 |
{
|
| 446 |
-
"epoch": 0.
|
| 447 |
-
"grad_norm":
|
| 448 |
-
"learning_rate": 0.
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 310
|
| 451 |
},
|
| 452 |
{
|
| 453 |
-
"epoch": 0.
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
-
"learning_rate": 0.
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 315
|
| 458 |
},
|
| 459 |
{
|
| 460 |
-
"epoch": 0.
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
-
"learning_rate": 0.
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 320
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
-
"grad_norm":
|
| 469 |
-
"learning_rate": 0.
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 325
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch": 0.
|
| 475 |
-
"grad_norm":
|
| 476 |
-
"learning_rate": 0.
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 330
|
| 479 |
},
|
| 480 |
{
|
| 481 |
-
"epoch": 0.
|
| 482 |
-
"grad_norm":
|
| 483 |
-
"learning_rate": 0.
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 335
|
| 486 |
},
|
| 487 |
{
|
| 488 |
-
"epoch": 0.
|
| 489 |
-
"grad_norm":
|
| 490 |
-
"learning_rate": 0.
|
| 491 |
-
"loss": 0.
|
| 492 |
"step": 340
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"epoch": 0.
|
| 496 |
-
"grad_norm":
|
| 497 |
-
"learning_rate": 0.
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 345
|
| 500 |
},
|
| 501 |
{
|
| 502 |
-
"epoch": 0.
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
-
"learning_rate": 0.
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 350
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch": 0.
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
-
"learning_rate": 0.
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 355
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch": 0.
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
-
"learning_rate": 0.
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 360
|
| 521 |
},
|
| 522 |
{
|
| 523 |
-
"epoch": 0.
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
-
"learning_rate": 0.
|
| 526 |
-
"loss": 0.
|
| 527 |
"step": 365
|
| 528 |
},
|
| 529 |
{
|
| 530 |
-
"epoch": 0.
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
-
"learning_rate": 0.
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 370
|
| 535 |
},
|
| 536 |
{
|
| 537 |
-
"epoch": 0.
|
| 538 |
-
"grad_norm":
|
| 539 |
-
"learning_rate": 0.
|
| 540 |
-
"loss": 0.
|
| 541 |
"step": 375
|
| 542 |
},
|
| 543 |
{
|
| 544 |
-
"epoch": 0.
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
-
"learning_rate": 0.
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 380
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch": 0.
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
-
"learning_rate": 0.
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 385
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch": 0.
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
-
"learning_rate": 0.
|
| 561 |
-
"loss": 0.
|
| 562 |
"step": 390
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch": 0.
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
-
"learning_rate": 0.
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 395
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch": 0.
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
-
"learning_rate": 0.
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 400
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch": 0.
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
-
"learning_rate": 0.
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 405
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch": 0.
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
-
"learning_rate": 0.
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 410
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch": 0.
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
-
"learning_rate": 0.
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 415
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch": 0.
|
| 601 |
-
"grad_norm":
|
| 602 |
-
"learning_rate": 0.
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 420
|
| 605 |
},
|
| 606 |
{
|
| 607 |
-
"epoch": 0.
|
| 608 |
-
"grad_norm":
|
| 609 |
-
"learning_rate": 0.
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 425
|
| 612 |
},
|
| 613 |
{
|
| 614 |
-
"epoch": 0.
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
-
"learning_rate": 0.
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 430
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch": 0.
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
-
"learning_rate": 0.
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 435
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch": 0.
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
-
"learning_rate": 0.
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 440
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch": 0.
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
-
"learning_rate": 0.
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 445
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch": 0.
|
| 643 |
-
"grad_norm":
|
| 644 |
-
"learning_rate": 0.
|
| 645 |
-
"loss": 0.
|
| 646 |
"step": 450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
-
"epoch": 0.
|
| 650 |
-
"grad_norm":
|
| 651 |
-
"learning_rate": 0.
|
| 652 |
-
"loss": 0.
|
| 653 |
"step": 455
|
| 654 |
},
|
| 655 |
{
|
| 656 |
-
"epoch": 0.
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
-
"learning_rate": 0.
|
| 659 |
-
"loss": 0.
|
| 660 |
"step": 460
|
| 661 |
},
|
| 662 |
{
|
| 663 |
-
"epoch": 0.
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
-
"learning_rate": 0.
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 465
|
| 668 |
},
|
| 669 |
{
|
| 670 |
-
"epoch": 0.
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
-
"learning_rate": 0.
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 470
|
| 675 |
},
|
| 676 |
{
|
| 677 |
-
"epoch": 0.
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
-
"learning_rate": 0.
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 475
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch": 0.
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
-
"learning_rate": 0.
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 480
|
| 689 |
},
|
| 690 |
{
|
| 691 |
-
"epoch": 0.
|
| 692 |
-
"grad_norm":
|
| 693 |
-
"learning_rate": 0.
|
| 694 |
-
"loss": 0.
|
| 695 |
"step": 485
|
| 696 |
},
|
| 697 |
{
|
| 698 |
-
"epoch": 0.
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 490
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"epoch": 0.
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
-
"learning_rate":
|
| 708 |
-
"loss": 0.
|
| 709 |
"step": 495
|
| 710 |
},
|
| 711 |
{
|
| 712 |
-
"epoch": 0.
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
-
"learning_rate":
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 500
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch": 0.
|
| 720 |
-
"grad_norm": 0.
|
| 721 |
-
"learning_rate":
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 505
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch": 0.
|
| 727 |
-
"grad_norm":
|
| 728 |
-
"learning_rate":
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 510
|
| 731 |
},
|
| 732 |
{
|
| 733 |
-
"epoch": 0.
|
| 734 |
-
"grad_norm": 0.
|
| 735 |
-
"learning_rate":
|
| 736 |
-
"loss": 0.
|
| 737 |
"step": 515
|
| 738 |
},
|
| 739 |
{
|
| 740 |
-
"epoch": 0.
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
-
"learning_rate":
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 520
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch": 0.
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
-
"learning_rate":
|
| 750 |
-
"loss": 0.
|
| 751 |
"step": 525
|
| 752 |
},
|
| 753 |
{
|
| 754 |
-
"epoch": 0.
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
-
"learning_rate":
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 530
|
| 759 |
},
|
| 760 |
{
|
| 761 |
-
"epoch": 0.
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 535
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch": 0.
|
| 769 |
-
"grad_norm": 0.
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 540
|
| 773 |
},
|
| 774 |
{
|
| 775 |
-
"epoch": 0.
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
-
"learning_rate":
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 545
|
| 780 |
},
|
| 781 |
{
|
| 782 |
-
"epoch": 0.
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
-
"learning_rate":
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 550
|
| 787 |
},
|
| 788 |
{
|
| 789 |
-
"epoch": 0.
|
| 790 |
-
"grad_norm": 0.
|
| 791 |
-
"learning_rate":
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 555
|
| 794 |
},
|
| 795 |
{
|
| 796 |
-
"epoch": 0.
|
| 797 |
-
"grad_norm":
|
| 798 |
-
"learning_rate":
|
| 799 |
-
"loss": 0.
|
| 800 |
"step": 560
|
| 801 |
},
|
| 802 |
{
|
| 803 |
-
"epoch": 0.
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
-
"learning_rate":
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 565
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch": 0.
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
-
"learning_rate":
|
| 813 |
-
"loss": 0.
|
| 814 |
"step": 570
|
| 815 |
},
|
| 816 |
{
|
| 817 |
-
"epoch": 0.
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
-
"learning_rate":
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 575
|
| 822 |
},
|
| 823 |
{
|
| 824 |
-
"epoch": 0.
|
| 825 |
-
"grad_norm": 0.
|
| 826 |
-
"learning_rate":
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 580
|
| 829 |
},
|
| 830 |
{
|
| 831 |
-
"epoch": 0.
|
| 832 |
-
"grad_norm":
|
| 833 |
-
"learning_rate":
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 585
|
| 836 |
},
|
| 837 |
{
|
| 838 |
-
"epoch": 0.
|
| 839 |
-
"grad_norm": 0.
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 590
|
| 843 |
},
|
| 844 |
{
|
| 845 |
-
"epoch": 0.
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
-
"learning_rate":
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 595
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch": 0.
|
| 853 |
-
"grad_norm": 0.
|
| 854 |
-
"learning_rate":
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 600
|
| 857 |
},
|
| 858 |
{
|
| 859 |
-
"epoch": 0.
|
| 860 |
-
"grad_norm":
|
| 861 |
-
"learning_rate":
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 605
|
| 864 |
},
|
| 865 |
{
|
| 866 |
-
"epoch": 0.
|
| 867 |
-
"grad_norm":
|
| 868 |
-
"learning_rate":
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 610
|
| 871 |
},
|
| 872 |
{
|
| 873 |
-
"epoch": 0.
|
| 874 |
-
"grad_norm": 0.
|
| 875 |
-
"learning_rate":
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 615
|
| 878 |
},
|
| 879 |
{
|
| 880 |
-
"epoch": 0.
|
| 881 |
-
"grad_norm": 0.
|
| 882 |
-
"learning_rate":
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 620
|
| 885 |
},
|
| 886 |
{
|
| 887 |
-
"epoch": 0.
|
| 888 |
-
"grad_norm": 0.
|
| 889 |
-
"learning_rate":
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 625
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch": 0.
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
-
"learning_rate":
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 630
|
| 899 |
},
|
| 900 |
{
|
| 901 |
-
"epoch": 0.
|
| 902 |
-
"grad_norm": 0.
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 635
|
| 906 |
},
|
| 907 |
{
|
| 908 |
-
"epoch": 0.
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 640
|
| 913 |
},
|
| 914 |
{
|
| 915 |
-
"epoch": 0.
|
| 916 |
-
"grad_norm": 0.
|
| 917 |
-
"learning_rate":
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 645
|
| 920 |
},
|
| 921 |
{
|
| 922 |
-
"epoch": 0.
|
| 923 |
-
"grad_norm": 0.
|
| 924 |
-
"learning_rate":
|
| 925 |
-
"loss": 0.
|
| 926 |
"step": 650
|
| 927 |
},
|
| 928 |
{
|
| 929 |
-
"epoch": 0.
|
| 930 |
-
"grad_norm": 0.
|
| 931 |
-
"learning_rate":
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 655
|
| 934 |
},
|
| 935 |
{
|
| 936 |
-
"epoch": 0.
|
| 937 |
-
"grad_norm": 0.
|
| 938 |
-
"learning_rate":
|
| 939 |
-
"loss": 0.
|
| 940 |
"step": 660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
-
"epoch": 0.
|
| 944 |
-
"grad_norm":
|
| 945 |
-
"learning_rate":
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 665
|
| 948 |
},
|
| 949 |
{
|
| 950 |
-
"epoch": 0.
|
| 951 |
-
"grad_norm": 0.
|
| 952 |
-
"learning_rate":
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 670
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"epoch": 0.
|
| 958 |
-
"grad_norm": 0.
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 675
|
| 962 |
},
|
| 963 |
{
|
| 964 |
-
"epoch": 0.
|
| 965 |
-
"grad_norm":
|
| 966 |
-
"learning_rate":
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 680
|
| 969 |
},
|
| 970 |
{
|
| 971 |
-
"epoch": 0.
|
| 972 |
-
"grad_norm": 0.
|
| 973 |
-
"learning_rate":
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 685
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch": 0.
|
| 979 |
-
"grad_norm": 0.
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 690
|
| 983 |
},
|
| 984 |
{
|
| 985 |
-
"epoch": 0.
|
| 986 |
-
"grad_norm": 0.
|
| 987 |
-
"learning_rate":
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 695
|
| 990 |
},
|
| 991 |
{
|
| 992 |
-
"epoch": 0.
|
| 993 |
-
"grad_norm": 0.
|
| 994 |
-
"learning_rate":
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 700
|
| 997 |
},
|
| 998 |
{
|
| 999 |
-
"epoch": 0.
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
-
"learning_rate":
|
| 1002 |
-
"loss": 0.
|
| 1003 |
"step": 705
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
-
"epoch": 0.
|
| 1007 |
-
"grad_norm": 1.
|
| 1008 |
-
"learning_rate":
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 710
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
-
"epoch": 0.
|
| 1014 |
-
"grad_norm":
|
| 1015 |
-
"learning_rate":
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 715
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
-
"epoch": 0.
|
| 1021 |
-
"grad_norm": 0.
|
| 1022 |
-
"learning_rate":
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 720
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
-
"epoch": 0.
|
| 1028 |
-
"grad_norm": 0.
|
| 1029 |
-
"learning_rate":
|
| 1030 |
-
"loss": 0.
|
| 1031 |
"step": 725
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
-
"epoch": 0.
|
| 1035 |
-
"grad_norm": 0.
|
| 1036 |
-
"learning_rate":
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 730
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
-
"epoch": 0.
|
| 1042 |
-
"grad_norm": 0.
|
| 1043 |
-
"learning_rate":
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 735
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
-
"epoch": 0.
|
| 1049 |
-
"grad_norm": 0.
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 740
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
-
"epoch": 0.
|
| 1056 |
-
"grad_norm": 0.
|
| 1057 |
-
"learning_rate":
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 745
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch": 0.
|
| 1063 |
-
"grad_norm": 0.
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 750
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
-
"epoch": 0.
|
| 1070 |
-
"grad_norm":
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 755
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch": 0.
|
| 1077 |
-
"grad_norm": 0.
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 760
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
-
"epoch": 0.
|
| 1084 |
-
"grad_norm": 0.
|
| 1085 |
-
"learning_rate":
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 765
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
-
"epoch": 0.
|
| 1091 |
-
"grad_norm": 0.
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 770
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch": 0.
|
| 1098 |
-
"grad_norm":
|
| 1099 |
-
"learning_rate":
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 775
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch": 0.
|
| 1105 |
-
"grad_norm": 0.
|
| 1106 |
-
"learning_rate":
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 780
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
-
"epoch": 0.
|
| 1112 |
-
"grad_norm":
|
| 1113 |
-
"learning_rate":
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 785
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
-
"epoch": 0.
|
| 1119 |
-
"grad_norm": 0.
|
| 1120 |
-
"learning_rate":
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 790
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
-
"epoch": 0.
|
| 1126 |
-
"grad_norm": 0.
|
| 1127 |
-
"learning_rate":
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 795
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
-
"epoch": 0.
|
| 1133 |
-
"grad_norm": 0.
|
| 1134 |
-
"learning_rate":
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 800
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"epoch": 0.
|
| 1140 |
-
"grad_norm": 0.
|
| 1141 |
-
"learning_rate":
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 805
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch": 0.
|
| 1147 |
-
"grad_norm": 0.
|
| 1148 |
-
"learning_rate":
|
| 1149 |
-
"loss": 0.
|
| 1150 |
"step": 810
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
-
"epoch": 0.
|
| 1154 |
-
"grad_norm":
|
| 1155 |
-
"learning_rate":
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 815
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
-
"epoch": 0.
|
| 1161 |
-
"grad_norm": 0.
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 820
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
-
"epoch": 0.
|
| 1168 |
-
"grad_norm": 0.
|
| 1169 |
-
"learning_rate":
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 825
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
-
"epoch": 0.
|
| 1175 |
-
"grad_norm": 0.
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 830
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
-
"epoch": 0.
|
| 1182 |
-
"grad_norm": 0.
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 835
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch": 0.
|
| 1189 |
-
"grad_norm": 0.
|
| 1190 |
-
"learning_rate":
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 840
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
-
"epoch": 0.
|
| 1196 |
-
"grad_norm": 0.
|
| 1197 |
-
"learning_rate":
|
| 1198 |
-
"loss": 0.
|
| 1199 |
"step": 845
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
-
"epoch": 0.
|
| 1203 |
-
"grad_norm": 0.
|
| 1204 |
-
"learning_rate":
|
| 1205 |
-
"loss": 0.
|
| 1206 |
"step": 850
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
-
"epoch": 0.
|
| 1210 |
-
"grad_norm": 0.
|
| 1211 |
-
"learning_rate":
|
| 1212 |
-
"loss": 0.
|
| 1213 |
"step": 855
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
-
"epoch": 0.
|
| 1217 |
-
"grad_norm": 0.
|
| 1218 |
-
"learning_rate":
|
| 1219 |
-
"loss": 0.
|
| 1220 |
"step": 860
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
-
"epoch": 0.
|
| 1224 |
-
"grad_norm": 0.
|
| 1225 |
-
"learning_rate":
|
| 1226 |
-
"loss": 0.
|
| 1227 |
"step": 865
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
-
"epoch": 0.
|
| 1231 |
-
"grad_norm": 0.
|
| 1232 |
-
"learning_rate":
|
| 1233 |
-
"loss": 0.
|
| 1234 |
"step": 870
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
-
"epoch": 0.
|
| 1238 |
-
"grad_norm": 0.
|
| 1239 |
-
"learning_rate":
|
| 1240 |
-
"loss": 0.
|
| 1241 |
"step": 875
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
-
"epoch": 0.
|
| 1245 |
-
"grad_norm": 0.
|
| 1246 |
-
"learning_rate":
|
| 1247 |
-
"loss": 0.
|
| 1248 |
"step": 880
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
-
"epoch": 0.
|
| 1252 |
-
"grad_norm":
|
| 1253 |
-
"learning_rate":
|
| 1254 |
-
"loss": 0.
|
| 1255 |
"step": 885
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
-
"epoch": 0.
|
| 1259 |
-
"
|
| 1260 |
-
"
|
| 1261 |
-
"
|
| 1262 |
-
"
|
| 1263 |
-
|
| 1264 |
-
|
| 1265 |
-
"epoch": 0.5417675544794189,
|
| 1266 |
-
"grad_norm": 0.6645005345344543,
|
| 1267 |
-
"learning_rate": 0.00010295934561290592,
|
| 1268 |
-
"loss": 0.4596,
|
| 1269 |
-
"step": 895
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 0.5447941888619855,
|
| 1273 |
-
"grad_norm": 0.47522783279418945,
|
| 1274 |
-
"learning_rate": 0.00010190259945335478,
|
| 1275 |
-
"loss": 0.2282,
|
| 1276 |
-
"step": 900
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 0.5478208232445521,
|
| 1280 |
-
"grad_norm": 0.6378582119941711,
|
| 1281 |
-
"learning_rate": 0.00010084564070251776,
|
| 1282 |
-
"loss": 0.253,
|
| 1283 |
-
"step": 905
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 0.5508474576271186,
|
| 1287 |
-
"grad_norm": 0.2484198808670044,
|
| 1288 |
-
"learning_rate": 9.978858746208965e-05,
|
| 1289 |
-
"loss": 0.1392,
|
| 1290 |
-
"step": 910
|
| 1291 |
-
},
|
| 1292 |
-
{
|
| 1293 |
-
"epoch": 0.5538740920096852,
|
| 1294 |
-
"grad_norm": 0.6086086630821228,
|
| 1295 |
-
"learning_rate": 9.873155784432321e-05,
|
| 1296 |
-
"loss": 0.2196,
|
| 1297 |
-
"step": 915
|
| 1298 |
-
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 0.5569007263922519,
|
| 1301 |
-
"grad_norm": 0.5516247749328613,
|
| 1302 |
-
"learning_rate": 9.767466995883183e-05,
|
| 1303 |
-
"loss": 0.2337,
|
| 1304 |
-
"step": 920
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 0.5599273607748184,
|
| 1308 |
-
"grad_norm": 0.5388537049293518,
|
| 1309 |
-
"learning_rate": 9.6618041899392e-05,
|
| 1310 |
-
"loss": 0.3432,
|
| 1311 |
-
"step": 925
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 0.562953995157385,
|
| 1315 |
-
"grad_norm": 0.65997314453125,
|
| 1316 |
-
"learning_rate": 9.556179173074797e-05,
|
| 1317 |
-
"loss": 0.3198,
|
| 1318 |
-
"step": 930
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 0.5659806295399515,
|
| 1322 |
-
"grad_norm": 0.9989176392555237,
|
| 1323 |
-
"learning_rate": 9.450603747541952e-05,
|
| 1324 |
-
"loss": 0.1796,
|
| 1325 |
-
"step": 935
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 0.5690072639225182,
|
| 1329 |
-
"grad_norm": 0.39720070362091064,
|
| 1330 |
-
"learning_rate": 9.34508971005144e-05,
|
| 1331 |
-
"loss": 0.2138,
|
| 1332 |
-
"step": 940
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 0.5720338983050848,
|
| 1336 |
-
"grad_norm": 0.7215743660926819,
|
| 1337 |
-
"learning_rate": 9.239648850454709e-05,
|
| 1338 |
-
"loss": 0.365,
|
| 1339 |
-
"step": 945
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 0.5750605326876513,
|
| 1343 |
-
"grad_norm": 1.1030629873275757,
|
| 1344 |
-
"learning_rate": 9.134292950426503e-05,
|
| 1345 |
-
"loss": 0.3136,
|
| 1346 |
-
"step": 950
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 0.5780871670702179,
|
| 1350 |
-
"grad_norm": 1.4015649557113647,
|
| 1351 |
-
"learning_rate": 9.029033782148416e-05,
|
| 1352 |
-
"loss": 0.3191,
|
| 1353 |
-
"step": 955
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 0.5811138014527845,
|
| 1357 |
-
"grad_norm": 0.432902991771698,
|
| 1358 |
-
"learning_rate": 8.923883106993499e-05,
|
| 1359 |
-
"loss": 0.1939,
|
| 1360 |
-
"step": 960
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 0.5841404358353511,
|
| 1364 |
-
"grad_norm": 0.6869246959686279,
|
| 1365 |
-
"learning_rate": 8.818852674212075e-05,
|
| 1366 |
-
"loss": 0.2841,
|
| 1367 |
-
"step": 965
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 0.5871670702179177,
|
| 1371 |
-
"grad_norm": 0.7746998071670532,
|
| 1372 |
-
"learning_rate": 8.71395421961892e-05,
|
| 1373 |
-
"loss": 0.5166,
|
| 1374 |
-
"step": 970
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 0.5901937046004843,
|
| 1378 |
-
"grad_norm": 0.44155651330947876,
|
| 1379 |
-
"learning_rate": 8.609199464281916e-05,
|
| 1380 |
-
"loss": 0.2632,
|
| 1381 |
-
"step": 975
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 0.5932203389830508,
|
| 1385 |
-
"grad_norm": 0.5189603567123413,
|
| 1386 |
-
"learning_rate": 8.504600113212387e-05,
|
| 1387 |
-
"loss": 0.4156,
|
| 1388 |
-
"step": 980
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 0.5962469733656174,
|
| 1392 |
-
"grad_norm": 0.6747870445251465,
|
| 1393 |
-
"learning_rate": 8.400167854057202e-05,
|
| 1394 |
-
"loss": 0.3773,
|
| 1395 |
-
"step": 985
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 0.5992736077481841,
|
| 1399 |
-
"grad_norm": 0.9039033651351929,
|
| 1400 |
-
"learning_rate": 8.295914355792839e-05,
|
| 1401 |
-
"loss": 0.2724,
|
| 1402 |
-
"step": 990
|
| 1403 |
-
},
|
| 1404 |
-
{
|
| 1405 |
-
"epoch": 0.6023002421307506,
|
| 1406 |
-
"grad_norm": 0.5657184720039368,
|
| 1407 |
-
"learning_rate": 8.191851267421523e-05,
|
| 1408 |
-
"loss": 0.2085,
|
| 1409 |
-
"step": 995
|
| 1410 |
-
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 0.6053268765133172,
|
| 1413 |
-
"grad_norm": 0.4272811710834503,
|
| 1414 |
-
"learning_rate": 8.087990216669585e-05,
|
| 1415 |
-
"loss": 0.2214,
|
| 1416 |
-
"step": 1000
|
| 1417 |
-
},
|
| 1418 |
-
{
|
| 1419 |
-
"epoch": 0.6083535108958837,
|
| 1420 |
-
"grad_norm": 0.4147346317768097,
|
| 1421 |
-
"learning_rate": 7.984342808688235e-05,
|
| 1422 |
-
"loss": 0.2039,
|
| 1423 |
-
"step": 1005
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 0.6113801452784504,
|
| 1427 |
-
"grad_norm": 0.6329300403594971,
|
| 1428 |
-
"learning_rate": 7.88092062475681e-05,
|
| 1429 |
-
"loss": 0.3535,
|
| 1430 |
-
"step": 1010
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 0.614406779661017,
|
| 1434 |
-
"grad_norm": 0.9579744935035706,
|
| 1435 |
-
"learning_rate": 7.777735220988728e-05,
|
| 1436 |
-
"loss": 0.3241,
|
| 1437 |
-
"step": 1015
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 0.6174334140435835,
|
| 1441 |
-
"grad_norm": 0.822500467300415,
|
| 1442 |
-
"learning_rate": 7.674798127040242e-05,
|
| 1443 |
-
"loss": 0.3485,
|
| 1444 |
-
"step": 1020
|
| 1445 |
-
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 0.6204600484261501,
|
| 1448 |
-
"grad_norm": 0.30872082710266113,
|
| 1449 |
-
"learning_rate": 7.572120844822136e-05,
|
| 1450 |
-
"loss": 0.194,
|
| 1451 |
-
"step": 1025
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 0.6234866828087167,
|
| 1455 |
-
"grad_norm": 0.27451083064079285,
|
| 1456 |
-
"learning_rate": 7.469714847214538e-05,
|
| 1457 |
-
"loss": 0.19,
|
| 1458 |
-
"step": 1030
|
| 1459 |
-
},
|
| 1460 |
-
{
|
| 1461 |
-
"epoch": 0.6265133171912833,
|
| 1462 |
-
"grad_norm": 0.6881311535835266,
|
| 1463 |
-
"learning_rate": 7.36759157678497e-05,
|
| 1464 |
-
"loss": 0.3104,
|
| 1465 |
-
"step": 1035
|
| 1466 |
-
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 0.6295399515738499,
|
| 1469 |
-
"grad_norm": 0.4620918333530426,
|
| 1470 |
-
"learning_rate": 7.265762444509792e-05,
|
| 1471 |
-
"loss": 0.2077,
|
| 1472 |
-
"step": 1040
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.6325665859564165,
|
| 1476 |
-
"grad_norm": 0.40100374817848206,
|
| 1477 |
-
"learning_rate": 7.164238828499167e-05,
|
| 1478 |
-
"loss": 0.2379,
|
| 1479 |
-
"step": 1045
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.635593220338983,
|
| 1483 |
-
"grad_norm": 0.4899684190750122,
|
| 1484 |
-
"learning_rate": 7.063032072725704e-05,
|
| 1485 |
-
"loss": 0.2746,
|
| 1486 |
-
"step": 1050
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 0.6386198547215496,
|
| 1490 |
-
"grad_norm": 0.6375607252120972,
|
| 1491 |
-
"learning_rate": 6.962153485756903e-05,
|
| 1492 |
-
"loss": 0.2895,
|
| 1493 |
-
"step": 1055
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 0.6416464891041163,
|
| 1497 |
-
"grad_norm": 0.7599363327026367,
|
| 1498 |
-
"learning_rate": 6.861614339491585e-05,
|
| 1499 |
-
"loss": 0.4866,
|
| 1500 |
-
"step": 1060
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 0.6446731234866828,
|
| 1504 |
-
"grad_norm": 0.5043110251426697,
|
| 1505 |
-
"learning_rate": 6.761425867900377e-05,
|
| 1506 |
-
"loss": 0.4566,
|
| 1507 |
-
"step": 1065
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 0.6476997578692494,
|
| 1511 |
-
"grad_norm": 0.7498160004615784,
|
| 1512 |
-
"learning_rate": 6.661599265770488e-05,
|
| 1513 |
-
"loss": 0.5193,
|
| 1514 |
-
"step": 1070
|
| 1515 |
-
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 0.6507263922518159,
|
| 1518 |
-
"grad_norm": 0.920767605304718,
|
| 1519 |
-
"learning_rate": 6.562145687454808e-05,
|
| 1520 |
-
"loss": 0.3063,
|
| 1521 |
-
"step": 1075
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 0.6537530266343826,
|
| 1525 |
-
"grad_norm": 0.41485604643821716,
|
| 1526 |
-
"learning_rate": 6.463076245625564e-05,
|
| 1527 |
-
"loss": 0.274,
|
| 1528 |
-
"step": 1080
|
| 1529 |
-
},
|
| 1530 |
-
{
|
| 1531 |
-
"epoch": 0.6567796610169492,
|
| 1532 |
-
"grad_norm": 0.5181906819343567,
|
| 1533 |
-
"learning_rate": 6.364402010032616e-05,
|
| 1534 |
-
"loss": 0.3616,
|
| 1535 |
-
"step": 1085
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 0.6598062953995157,
|
| 1539 |
-
"grad_norm": 0.7301517724990845,
|
| 1540 |
-
"learning_rate": 6.266134006266554e-05,
|
| 1541 |
-
"loss": 0.4633,
|
| 1542 |
-
"step": 1090
|
| 1543 |
-
},
|
| 1544 |
-
{
|
| 1545 |
-
"epoch": 0.6628329297820823,
|
| 1546 |
-
"grad_norm": 0.26296061277389526,
|
| 1547 |
-
"learning_rate": 6.168283214526734e-05,
|
| 1548 |
-
"loss": 0.1601,
|
| 1549 |
-
"step": 1095
|
| 1550 |
-
},
|
| 1551 |
-
{
|
| 1552 |
-
"epoch": 0.6658595641646489,
|
| 1553 |
-
"grad_norm": 0.48812544345855713,
|
| 1554 |
-
"learning_rate": 6.0708605683943676e-05,
|
| 1555 |
-
"loss": 0.3234,
|
| 1556 |
-
"step": 1100
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 0.6688861985472155,
|
| 1560 |
-
"grad_norm": 0.5620171427726746,
|
| 1561 |
-
"learning_rate": 5.9738769536108484e-05,
|
| 1562 |
-
"loss": 0.2721,
|
| 1563 |
-
"step": 1105
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 0.6719128329297821,
|
| 1567 |
-
"grad_norm": 0.35485607385635376,
|
| 1568 |
-
"learning_rate": 5.877343206861397e-05,
|
| 1569 |
-
"loss": 0.1854,
|
| 1570 |
-
"step": 1110
|
| 1571 |
-
},
|
| 1572 |
-
{
|
| 1573 |
-
"epoch": 0.6749394673123487,
|
| 1574 |
-
"grad_norm": 0.37520357966423035,
|
| 1575 |
-
"learning_rate": 5.7812701145642036e-05,
|
| 1576 |
-
"loss": 0.1946,
|
| 1577 |
-
"step": 1115
|
| 1578 |
-
},
|
| 1579 |
-
{
|
| 1580 |
-
"epoch": 0.6779661016949152,
|
| 1581 |
-
"grad_norm": 0.3933568000793457,
|
| 1582 |
-
"learning_rate": 5.6856684116651904e-05,
|
| 1583 |
-
"loss": 0.2679,
|
| 1584 |
-
"step": 1120
|
| 1585 |
-
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 0.6809927360774818,
|
| 1588 |
-
"grad_norm": 0.7496293783187866,
|
| 1589 |
-
"learning_rate": 5.5905487804385014e-05,
|
| 1590 |
-
"loss": 0.2944,
|
| 1591 |
-
"step": 1125
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 0.6840193704600485,
|
| 1595 |
-
"grad_norm": 0.4908842444419861,
|
| 1596 |
-
"learning_rate": 5.495921849292911e-05,
|
| 1597 |
-
"loss": 0.2457,
|
| 1598 |
-
"step": 1130
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 0.687046004842615,
|
| 1602 |
-
"grad_norm": 0.6135571599006653,
|
| 1603 |
-
"learning_rate": 5.401798191584222e-05,
|
| 1604 |
-
"loss": 0.334,
|
| 1605 |
-
"step": 1135
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 0.6900726392251816,
|
| 1609 |
-
"grad_norm": 0.8274165987968445,
|
| 1610 |
-
"learning_rate": 5.308188324433835e-05,
|
| 1611 |
-
"loss": 0.3017,
|
| 1612 |
-
"step": 1140
|
| 1613 |
-
},
|
| 1614 |
-
{
|
| 1615 |
-
"epoch": 0.6930992736077481,
|
| 1616 |
-
"grad_norm": 0.37864798307418823,
|
| 1617 |
-
"learning_rate": 5.215102707553594e-05,
|
| 1618 |
-
"loss": 0.184,
|
| 1619 |
-
"step": 1145
|
| 1620 |
-
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 0.6961259079903148,
|
| 1623 |
-
"grad_norm": 0.7884442806243896,
|
| 1624 |
-
"learning_rate": 5.122551742077035e-05,
|
| 1625 |
-
"loss": 0.4359,
|
| 1626 |
-
"step": 1150
|
| 1627 |
-
},
|
| 1628 |
-
{
|
| 1629 |
-
"epoch": 0.6991525423728814,
|
| 1630 |
-
"grad_norm": 0.8110864758491516,
|
| 1631 |
-
"learning_rate": 5.030545769397218e-05,
|
| 1632 |
-
"loss": 0.2803,
|
| 1633 |
-
"step": 1155
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 0.7021791767554479,
|
| 1637 |
-
"grad_norm": 0.4235096871852875,
|
| 1638 |
-
"learning_rate": 4.939095070011176e-05,
|
| 1639 |
-
"loss": 0.4761,
|
| 1640 |
-
"step": 1160
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 0.7052058111380145,
|
| 1644 |
-
"grad_norm": 1.767540454864502,
|
| 1645 |
-
"learning_rate": 4.8482098623712205e-05,
|
| 1646 |
-
"loss": 0.2602,
|
| 1647 |
-
"step": 1165
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 0.7082324455205811,
|
| 1651 |
-
"grad_norm": 0.7618401050567627,
|
| 1652 |
-
"learning_rate": 4.7579003017431526e-05,
|
| 1653 |
-
"loss": 0.3445,
|
| 1654 |
-
"step": 1170
|
| 1655 |
-
},
|
| 1656 |
-
{
|
| 1657 |
-
"epoch": 0.7112590799031477,
|
| 1658 |
-
"grad_norm": 0.6124890446662903,
|
| 1659 |
-
"learning_rate": 4.668176479071542e-05,
|
| 1660 |
-
"loss": 0.4762,
|
| 1661 |
-
"step": 1175
|
| 1662 |
-
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 0.7142857142857143,
|
| 1665 |
-
"grad_norm": 0.48818454146385193,
|
| 1666 |
-
"learning_rate": 4.579048419852197e-05,
|
| 1667 |
-
"loss": 0.2078,
|
| 1668 |
-
"step": 1180
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 0.7173123486682809,
|
| 1672 |
-
"grad_norm": 0.5148090720176697,
|
| 1673 |
-
"learning_rate": 4.490526083011934e-05,
|
| 1674 |
-
"loss": 0.2045,
|
| 1675 |
-
"step": 1185
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 0.7203389830508474,
|
| 1679 |
-
"grad_norm": 0.45713886618614197,
|
| 1680 |
-
"learning_rate": 4.402619359795803e-05,
|
| 1681 |
-
"loss": 0.111,
|
| 1682 |
-
"step": 1190
|
| 1683 |
-
},
|
| 1684 |
-
{
|
| 1685 |
-
"epoch": 0.7233656174334141,
|
| 1686 |
-
"grad_norm": 0.3432838022708893,
|
| 1687 |
-
"learning_rate": 4.31533807266186e-05,
|
| 1688 |
-
"loss": 0.2598,
|
| 1689 |
-
"step": 1195
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 0.7263922518159807,
|
| 1693 |
-
"grad_norm": 0.6573700308799744,
|
| 1694 |
-
"learning_rate": 4.228691974183631e-05,
|
| 1695 |
-
"loss": 0.2569,
|
| 1696 |
-
"step": 1200
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 0.7294188861985472,
|
| 1700 |
-
"grad_norm": 0.9488723278045654,
|
| 1701 |
-
"learning_rate": 4.1426907459604015e-05,
|
| 1702 |
-
"loss": 0.2357,
|
| 1703 |
-
"step": 1205
|
| 1704 |
-
},
|
| 1705 |
-
{
|
| 1706 |
-
"epoch": 0.7324455205811138,
|
| 1707 |
-
"grad_norm": 0.6784693598747253,
|
| 1708 |
-
"learning_rate": 4.057343997535402e-05,
|
| 1709 |
-
"loss": 0.3265,
|
| 1710 |
-
"step": 1210
|
| 1711 |
-
},
|
| 1712 |
-
{
|
| 1713 |
-
"epoch": 0.7354721549636803,
|
| 1714 |
-
"grad_norm": 0.6079845428466797,
|
| 1715 |
-
"learning_rate": 3.9726612653220695e-05,
|
| 1716 |
-
"loss": 0.2475,
|
| 1717 |
-
"step": 1215
|
| 1718 |
-
},
|
| 1719 |
-
{
|
| 1720 |
-
"epoch": 0.738498789346247,
|
| 1721 |
-
"grad_norm": 0.6726992130279541,
|
| 1722 |
-
"learning_rate": 3.8886520115384796e-05,
|
| 1723 |
-
"loss": 0.2672,
|
| 1724 |
-
"step": 1220
|
| 1725 |
-
},
|
| 1726 |
-
{
|
| 1727 |
-
"epoch": 0.7415254237288136,
|
| 1728 |
-
"grad_norm": 0.6581181883811951,
|
| 1729 |
-
"learning_rate": 3.80532562315006e-05,
|
| 1730 |
-
"loss": 0.237,
|
| 1731 |
-
"step": 1225
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 0.7445520581113801,
|
| 1735 |
-
"grad_norm": 0.5347960591316223,
|
| 1736 |
-
"learning_rate": 3.722691410820716e-05,
|
| 1737 |
-
"loss": 0.3184,
|
| 1738 |
-
"step": 1230
|
| 1739 |
-
},
|
| 1740 |
-
{
|
| 1741 |
-
"epoch": 0.7475786924939467,
|
| 1742 |
-
"grad_norm": 0.613092839717865,
|
| 1743 |
-
"learning_rate": 3.640758607872484e-05,
|
| 1744 |
-
"loss": 0.3132,
|
| 1745 |
-
"step": 1235
|
| 1746 |
-
},
|
| 1747 |
-
{
|
| 1748 |
-
"epoch": 0.7506053268765133,
|
| 1749 |
-
"grad_norm": 0.7819620966911316,
|
| 1750 |
-
"learning_rate": 3.559536369253828e-05,
|
| 1751 |
-
"loss": 0.2976,
|
| 1752 |
-
"step": 1240
|
| 1753 |
-
},
|
| 1754 |
-
{
|
| 1755 |
-
"epoch": 0.7536319612590799,
|
| 1756 |
-
"grad_norm": 0.5769484043121338,
|
| 1757 |
-
"learning_rate": 3.4790337705166807e-05,
|
| 1758 |
-
"loss": 0.3245,
|
| 1759 |
-
"step": 1245
|
| 1760 |
-
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 0.7566585956416465,
|
| 1763 |
-
"grad_norm": 0.4908241629600525,
|
| 1764 |
-
"learning_rate": 3.3992598068023764e-05,
|
| 1765 |
-
"loss": 0.2678,
|
| 1766 |
-
"step": 1250
|
| 1767 |
-
},
|
| 1768 |
-
{
|
| 1769 |
-
"epoch": 0.7596852300242131,
|
| 1770 |
-
"grad_norm": 0.630765974521637,
|
| 1771 |
-
"learning_rate": 3.3202233918365585e-05,
|
| 1772 |
-
"loss": 0.3196,
|
| 1773 |
-
"step": 1255
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 0.7627118644067796,
|
| 1777 |
-
"grad_norm": 0.9001298546791077,
|
| 1778 |
-
"learning_rate": 3.2419333569331745e-05,
|
| 1779 |
-
"loss": 0.4613,
|
| 1780 |
-
"step": 1260
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 0.7657384987893463,
|
| 1784 |
-
"grad_norm": 0.9717426896095276,
|
| 1785 |
-
"learning_rate": 3.164398450007695e-05,
|
| 1786 |
-
"loss": 0.2501,
|
| 1787 |
-
"step": 1265
|
| 1788 |
-
},
|
| 1789 |
-
{
|
| 1790 |
-
"epoch": 0.7687651331719129,
|
| 1791 |
-
"grad_norm": 0.8477965593338013,
|
| 1792 |
-
"learning_rate": 3.087627334599647e-05,
|
| 1793 |
-
"loss": 0.3467,
|
| 1794 |
-
"step": 1270
|
| 1795 |
-
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 0.7717917675544794,
|
| 1798 |
-
"grad_norm": 0.6805775165557861,
|
| 1799 |
-
"learning_rate": 3.0116285889045724e-05,
|
| 1800 |
-
"loss": 0.405,
|
| 1801 |
-
"step": 1275
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 0.774818401937046,
|
| 1805 |
-
"grad_norm": 0.5589771270751953,
|
| 1806 |
-
"learning_rate": 2.9364107048155187e-05,
|
| 1807 |
-
"loss": 0.3572,
|
| 1808 |
-
"step": 1280
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 0.7778450363196125,
|
| 1812 |
-
"grad_norm": 0.3014810085296631,
|
| 1813 |
-
"learning_rate": 2.8619820869741875e-05,
|
| 1814 |
-
"loss": 0.2763,
|
| 1815 |
-
"step": 1285
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 0.7808716707021792,
|
| 1819 |
-
"grad_norm": 1.0074349641799927,
|
| 1820 |
-
"learning_rate": 2.7883510518318135e-05,
|
| 1821 |
-
"loss": 0.1978,
|
| 1822 |
-
"step": 1290
|
| 1823 |
-
},
|
| 1824 |
-
{
|
| 1825 |
-
"epoch": 0.7838983050847458,
|
| 1826 |
-
"grad_norm": 0.6734029054641724,
|
| 1827 |
-
"learning_rate": 2.7155258267199124e-05,
|
| 1828 |
-
"loss": 0.1619,
|
| 1829 |
-
"step": 1295
|
| 1830 |
-
},
|
| 1831 |
-
{
|
| 1832 |
-
"epoch": 0.7869249394673123,
|
| 1833 |
-
"grad_norm": 0.3331570625305176,
|
| 1834 |
-
"learning_rate": 2.6435145489309775e-05,
|
| 1835 |
-
"loss": 0.1658,
|
| 1836 |
-
"step": 1300
|
| 1837 |
-
},
|
| 1838 |
-
{
|
| 1839 |
-
"epoch": 0.7899515738498789,
|
| 1840 |
-
"grad_norm": 0.5702409744262695,
|
| 1841 |
-
"learning_rate": 2.572325264809252e-05,
|
| 1842 |
-
"loss": 0.2729,
|
| 1843 |
-
"step": 1305
|
| 1844 |
-
},
|
| 1845 |
-
{
|
| 1846 |
-
"epoch": 0.7929782082324455,
|
| 1847 |
-
"grad_norm": 0.6430597305297852,
|
| 1848 |
-
"learning_rate": 2.5019659288516327e-05,
|
| 1849 |
-
"loss": 0.3226,
|
| 1850 |
-
"step": 1310
|
| 1851 |
-
},
|
| 1852 |
-
{
|
| 1853 |
-
"epoch": 0.7960048426150121,
|
| 1854 |
-
"grad_norm": 1.3691140413284302,
|
| 1855 |
-
"learning_rate": 2.432444402818871e-05,
|
| 1856 |
-
"loss": 0.2465,
|
| 1857 |
-
"step": 1315
|
| 1858 |
-
},
|
| 1859 |
-
{
|
| 1860 |
-
"epoch": 0.7990314769975787,
|
| 1861 |
-
"grad_norm": 0.4949992895126343,
|
| 1862 |
-
"learning_rate": 2.363768454857117e-05,
|
| 1863 |
-
"loss": 0.3212,
|
| 1864 |
-
"step": 1320
|
| 1865 |
-
},
|
| 1866 |
-
{
|
| 1867 |
-
"epoch": 0.8020581113801453,
|
| 1868 |
-
"grad_norm": 0.742047905921936,
|
| 1869 |
-
"learning_rate": 2.295945758629927e-05,
|
| 1870 |
-
"loss": 0.2433,
|
| 1871 |
-
"step": 1325
|
| 1872 |
-
},
|
| 1873 |
-
{
|
| 1874 |
-
"epoch": 0.8050847457627118,
|
| 1875 |
-
"grad_norm": 0.3381989896297455,
|
| 1876 |
-
"learning_rate": 2.228983892460831e-05,
|
| 1877 |
-
"loss": 0.2479,
|
| 1878 |
-
"step": 1330
|
| 1879 |
-
},
|
| 1880 |
-
{
|
| 1881 |
-
"epoch": 0.8081113801452785,
|
| 1882 |
-
"grad_norm": 0.7100020051002502,
|
| 1883 |
-
"learning_rate": 2.1628903384865517e-05,
|
| 1884 |
-
"loss": 0.2148,
|
| 1885 |
-
"step": 1335
|
| 1886 |
-
},
|
| 1887 |
-
{
|
| 1888 |
-
"epoch": 0.8111380145278451,
|
| 1889 |
-
"grad_norm": 0.5883369445800781,
|
| 1890 |
-
"learning_rate": 2.097672481820968e-05,
|
| 1891 |
-
"loss": 0.3359,
|
| 1892 |
-
"step": 1340
|
| 1893 |
-
},
|
| 1894 |
-
{
|
| 1895 |
-
"epoch": 0.8141646489104116,
|
| 1896 |
-
"grad_norm": 0.4386845827102661,
|
| 1897 |
-
"learning_rate": 2.033337609729925e-05,
|
| 1898 |
-
"loss": 0.3101,
|
| 1899 |
-
"step": 1345
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 0.8171912832929782,
|
| 1903 |
-
"grad_norm": 0.5113346576690674,
|
| 1904 |
-
"learning_rate": 1.9698929108169717e-05,
|
| 1905 |
-
"loss": 0.3446,
|
| 1906 |
-
"step": 1350
|
| 1907 |
-
},
|
| 1908 |
-
{
|
| 1909 |
-
"epoch": 0.8202179176755447,
|
| 1910 |
-
"grad_norm": 0.4762308895587921,
|
| 1911 |
-
"learning_rate": 1.907345474220136e-05,
|
| 1912 |
-
"loss": 0.2668,
|
| 1913 |
-
"step": 1355
|
| 1914 |
-
},
|
| 1915 |
-
{
|
| 1916 |
-
"epoch": 0.8232445520581114,
|
| 1917 |
-
"grad_norm": 0.7801670432090759,
|
| 1918 |
-
"learning_rate": 1.845702288819787e-05,
|
| 1919 |
-
"loss": 0.1855,
|
| 1920 |
-
"step": 1360
|
| 1921 |
-
},
|
| 1922 |
-
{
|
| 1923 |
-
"epoch": 0.826271186440678,
|
| 1924 |
-
"grad_norm": 0.5607767701148987,
|
| 1925 |
-
"learning_rate": 1.784970242457732e-05,
|
| 1926 |
-
"loss": 0.1905,
|
| 1927 |
-
"step": 1365
|
| 1928 |
-
},
|
| 1929 |
-
{
|
| 1930 |
-
"epoch": 0.8292978208232445,
|
| 1931 |
-
"grad_norm": 0.8111298680305481,
|
| 1932 |
-
"learning_rate": 1.725156121167576e-05,
|
| 1933 |
-
"loss": 0.2512,
|
| 1934 |
-
"step": 1370
|
| 1935 |
-
},
|
| 1936 |
-
{
|
| 1937 |
-
"epoch": 0.8323244552058111,
|
| 1938 |
-
"grad_norm": 0.6997355818748474,
|
| 1939 |
-
"learning_rate": 1.666266608416479e-05,
|
| 1940 |
-
"loss": 0.3077,
|
| 1941 |
-
"step": 1375
|
| 1942 |
-
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 0.8353510895883777,
|
| 1945 |
-
"grad_norm": 0.7996625304222107,
|
| 1946 |
-
"learning_rate": 1.6083082843583554e-05,
|
| 1947 |
-
"loss": 0.2562,
|
| 1948 |
-
"step": 1380
|
| 1949 |
-
},
|
| 1950 |
-
{
|
| 1951 |
-
"epoch": 0.8383777239709443,
|
| 1952 |
-
"grad_norm": 0.37021154165267944,
|
| 1953 |
-
"learning_rate": 1.5512876250986307e-05,
|
| 1954 |
-
"loss": 0.2525,
|
| 1955 |
-
"step": 1385
|
| 1956 |
-
},
|
| 1957 |
-
{
|
| 1958 |
-
"epoch": 0.8414043583535109,
|
| 1959 |
-
"grad_norm": 0.42876264452934265,
|
| 1960 |
-
"learning_rate": 1.4952110019706157e-05,
|
| 1961 |
-
"loss": 0.2538,
|
| 1962 |
-
"step": 1390
|
| 1963 |
-
},
|
| 1964 |
-
{
|
| 1965 |
-
"epoch": 0.8444309927360775,
|
| 1966 |
-
"grad_norm": 0.44964295625686646,
|
| 1967 |
-
"learning_rate": 1.4400846808235947e-05,
|
| 1968 |
-
"loss": 0.2089,
|
| 1969 |
-
"step": 1395
|
| 1970 |
-
},
|
| 1971 |
-
{
|
| 1972 |
-
"epoch": 0.847457627118644,
|
| 1973 |
-
"grad_norm": 0.6915865540504456,
|
| 1974 |
-
"learning_rate": 1.3859148213226903e-05,
|
| 1975 |
-
"loss": 0.2569,
|
| 1976 |
-
"step": 1400
|
| 1977 |
-
},
|
| 1978 |
-
{
|
| 1979 |
-
"epoch": 0.8504842615012107,
|
| 1980 |
-
"grad_norm": 0.6955189108848572,
|
| 1981 |
-
"learning_rate": 1.3327074762606096e-05,
|
| 1982 |
-
"loss": 0.3002,
|
| 1983 |
-
"step": 1405
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 0.8535108958837773,
|
| 1987 |
-
"grad_norm": 0.22800378501415253,
|
| 1988 |
-
"learning_rate": 1.2804685908813118e-05,
|
| 1989 |
-
"loss": 0.2723,
|
| 1990 |
-
"step": 1410
|
| 1991 |
-
},
|
| 1992 |
-
{
|
| 1993 |
-
"epoch": 0.8565375302663438,
|
| 1994 |
-
"grad_norm": 0.4008031487464905,
|
| 1995 |
-
"learning_rate": 1.2292040022157058e-05,
|
| 1996 |
-
"loss": 0.2237,
|
| 1997 |
-
"step": 1415
|
| 1998 |
-
},
|
| 1999 |
-
{
|
| 2000 |
-
"epoch": 0.8595641646489104,
|
| 2001 |
-
"grad_norm": 0.4650994837284088,
|
| 2002 |
-
"learning_rate": 1.1789194384294378e-05,
|
| 2003 |
-
"loss": 0.2479,
|
| 2004 |
-
"step": 1420
|
| 2005 |
-
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 0.862590799031477,
|
| 2008 |
-
"grad_norm": 0.7434440851211548,
|
| 2009 |
-
"learning_rate": 1.129620518182838e-05,
|
| 2010 |
-
"loss": 0.1985,
|
| 2011 |
-
"step": 1425
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 0.8656174334140436,
|
| 2015 |
-
"grad_norm": 0.44593191146850586,
|
| 2016 |
-
"learning_rate": 1.0813127500031139e-05,
|
| 2017 |
-
"loss": 0.2782,
|
| 2018 |
-
"step": 1430
|
| 2019 |
-
},
|
| 2020 |
-
{
|
| 2021 |
-
"epoch": 0.8686440677966102,
|
| 2022 |
-
"grad_norm": 0.38460445404052734,
|
| 2023 |
-
"learning_rate": 1.0340015316688357e-05,
|
| 2024 |
-
"loss": 0.2248,
|
| 2025 |
-
"step": 1435
|
| 2026 |
-
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 0.8716707021791767,
|
| 2029 |
-
"grad_norm": 0.8492452502250671,
|
| 2030 |
-
"learning_rate": 9.876921496068059e-06,
|
| 2031 |
-
"loss": 0.4202,
|
| 2032 |
-
"step": 1440
|
| 2033 |
-
},
|
| 2034 |
-
{
|
| 2035 |
-
"epoch": 0.8746973365617433,
|
| 2036 |
-
"grad_norm": 0.6205224394798279,
|
| 2037 |
-
"learning_rate": 9.423897783013658e-06,
|
| 2038 |
-
"loss": 0.3328,
|
| 2039 |
-
"step": 1445
|
| 2040 |
-
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 0.8777239709443099,
|
| 2043 |
-
"grad_norm": 0.33269113302230835,
|
| 2044 |
-
"learning_rate": 8.980994797162157e-06,
|
| 2045 |
-
"loss": 0.2071,
|
| 2046 |
-
"step": 1450
|
| 2047 |
-
},
|
| 2048 |
-
{
|
| 2049 |
-
"epoch": 0.8807506053268765,
|
| 2050 |
-
"grad_norm": 0.7258440256118774,
|
| 2051 |
-
"learning_rate": 8.548262027288102e-06,
|
| 2052 |
-
"loss": 0.3225,
|
| 2053 |
-
"step": 1455
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 0.8837772397094431,
|
| 2057 |
-
"grad_norm": 0.3106304407119751,
|
| 2058 |
-
"learning_rate": 8.125747825773689e-06,
|
| 2059 |
-
"loss": 0.1594,
|
| 2060 |
-
"step": 1460
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 0.8868038740920097,
|
| 2064 |
-
"grad_norm": 0.7520200610160828,
|
| 2065 |
-
"learning_rate": 7.713499403206137e-06,
|
| 2066 |
-
"loss": 0.1702,
|
| 2067 |
-
"step": 1465
|
| 2068 |
-
},
|
| 2069 |
-
{
|
| 2070 |
-
"epoch": 0.8898305084745762,
|
| 2071 |
-
"grad_norm": 0.6596885919570923,
|
| 2072 |
-
"learning_rate": 7.311562823102469e-06,
|
| 2073 |
-
"loss": 0.2627,
|
| 2074 |
-
"step": 1470
|
| 2075 |
-
},
|
| 2076 |
-
{
|
| 2077 |
-
"epoch": 0.8928571428571429,
|
| 2078 |
-
"grad_norm": 0.49822545051574707,
|
| 2079 |
-
"learning_rate": 6.919982996762431e-06,
|
| 2080 |
-
"loss": 0.2594,
|
| 2081 |
-
"step": 1475
|
| 2082 |
-
},
|
| 2083 |
-
{
|
| 2084 |
-
"epoch": 0.8958837772397095,
|
| 2085 |
-
"grad_norm": 0.5468152761459351,
|
| 2086 |
-
"learning_rate": 6.538803678250338e-06,
|
| 2087 |
-
"loss": 0.2448,
|
| 2088 |
-
"step": 1480
|
| 2089 |
-
},
|
| 2090 |
-
{
|
| 2091 |
-
"epoch": 0.898910411622276,
|
| 2092 |
-
"grad_norm": 0.6610658168792725,
|
| 2093 |
-
"learning_rate": 6.168067459506066e-06,
|
| 2094 |
-
"loss": 0.1538,
|
| 2095 |
-
"step": 1485
|
| 2096 |
-
},
|
| 2097 |
-
{
|
| 2098 |
-
"epoch": 0.9019370460048426,
|
| 2099 |
-
"grad_norm": 0.7339268326759338,
|
| 2100 |
-
"learning_rate": 5.8078157655858775e-06,
|
| 2101 |
-
"loss": 0.3897,
|
| 2102 |
-
"step": 1490
|
| 2103 |
-
},
|
| 2104 |
-
{
|
| 2105 |
-
"epoch": 0.9049636803874092,
|
| 2106 |
-
"grad_norm": 0.46854448318481445,
|
| 2107 |
-
"learning_rate": 5.458088850033849e-06,
|
| 2108 |
-
"loss": 0.2082,
|
| 2109 |
-
"step": 1495
|
| 2110 |
-
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 0.9079903147699758,
|
| 2113 |
-
"grad_norm": 0.516205370426178,
|
| 2114 |
-
"learning_rate": 5.118925790383899e-06,
|
| 2115 |
-
"loss": 0.1918,
|
| 2116 |
-
"step": 1500
|
| 2117 |
-
},
|
| 2118 |
-
{
|
| 2119 |
-
"epoch": 0.9110169491525424,
|
| 2120 |
-
"grad_norm": 0.5182480812072754,
|
| 2121 |
-
"learning_rate": 4.79036448379343e-06,
|
| 2122 |
-
"loss": 0.3493,
|
| 2123 |
-
"step": 1505
|
| 2124 |
-
},
|
| 2125 |
-
{
|
| 2126 |
-
"epoch": 0.914043583535109,
|
| 2127 |
-
"grad_norm": 0.43342140316963196,
|
| 2128 |
-
"learning_rate": 4.472441642808844e-06,
|
| 2129 |
-
"loss": 0.3326,
|
| 2130 |
-
"step": 1510
|
| 2131 |
-
},
|
| 2132 |
-
{
|
| 2133 |
-
"epoch": 0.9170702179176755,
|
| 2134 |
-
"grad_norm": 0.5030437111854553,
|
| 2135 |
-
"learning_rate": 4.165192791263295e-06,
|
| 2136 |
-
"loss": 0.2271,
|
| 2137 |
-
"step": 1515
|
| 2138 |
-
},
|
| 2139 |
-
{
|
| 2140 |
-
"epoch": 0.9200968523002422,
|
| 2141 |
-
"grad_norm": 0.4448809325695038,
|
| 2142 |
-
"learning_rate": 3.868652260307437e-06,
|
| 2143 |
-
"loss": 0.1694,
|
| 2144 |
-
"step": 1520
|
| 2145 |
-
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 0.9231234866828087,
|
| 2148 |
-
"grad_norm": 0.6059044003486633,
|
| 2149 |
-
"learning_rate": 3.5828531845733202e-06,
|
| 2150 |
-
"loss": 0.2699,
|
| 2151 |
-
"step": 1525
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 0.9261501210653753,
|
| 2155 |
-
"grad_norm": 0.7686289548873901,
|
| 2156 |
-
"learning_rate": 3.3078274984719828e-06,
|
| 2157 |
-
"loss": 0.4511,
|
| 2158 |
-
"step": 1530
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 0.9291767554479419,
|
| 2162 |
-
"grad_norm": 0.7025601267814636,
|
| 2163 |
-
"learning_rate": 3.0436059326252597e-06,
|
| 2164 |
-
"loss": 0.2399,
|
| 2165 |
-
"step": 1535
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 0.9322033898305084,
|
| 2169 |
-
"grad_norm": 0.3700661063194275,
|
| 2170 |
-
"learning_rate": 2.7902180104319443e-06,
|
| 2171 |
-
"loss": 0.2198,
|
| 2172 |
-
"step": 1540
|
| 2173 |
-
},
|
| 2174 |
-
{
|
| 2175 |
-
"epoch": 0.9352300242130751,
|
| 2176 |
-
"grad_norm": 0.6774312257766724,
|
| 2177 |
-
"learning_rate": 2.5476920447690124e-06,
|
| 2178 |
-
"loss": 0.2386,
|
| 2179 |
-
"step": 1545
|
| 2180 |
-
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 0.9382566585956417,
|
| 2183 |
-
"grad_norm": 0.6344419121742249,
|
| 2184 |
-
"learning_rate": 2.3160551348279436e-06,
|
| 2185 |
-
"loss": 0.2138,
|
| 2186 |
-
"step": 1550
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 0.9412832929782082,
|
| 2190 |
-
"grad_norm": 0.4920920729637146,
|
| 2191 |
-
"learning_rate": 2.0953331630867766e-06,
|
| 2192 |
-
"loss": 0.1751,
|
| 2193 |
-
"step": 1555
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 0.9443099273607748,
|
| 2197 |
-
"grad_norm": 0.8836653828620911,
|
| 2198 |
-
"learning_rate": 1.8855507924180338e-06,
|
| 2199 |
-
"loss": 0.2963,
|
| 2200 |
-
"step": 1560
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 0.9473365617433414,
|
| 2204 |
-
"grad_norm": 0.6120699048042297,
|
| 2205 |
-
"learning_rate": 1.6867314633330023e-06,
|
| 2206 |
-
"loss": 0.2666,
|
| 2207 |
-
"step": 1565
|
| 2208 |
-
},
|
| 2209 |
-
{
|
| 2210 |
-
"epoch": 0.950363196125908,
|
| 2211 |
-
"grad_norm": 0.45984670519828796,
|
| 2212 |
-
"learning_rate": 1.4988973913625082e-06,
|
| 2213 |
-
"loss": 0.2108,
|
| 2214 |
-
"step": 1570
|
| 2215 |
-
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 0.9533898305084746,
|
| 2218 |
-
"grad_norm": 0.386865496635437,
|
| 2219 |
-
"learning_rate": 1.3220695645746684e-06,
|
| 2220 |
-
"loss": 0.2447,
|
| 2221 |
-
"step": 1575
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 0.9564164648910412,
|
| 2225 |
-
"grad_norm": 0.46179696917533875,
|
| 2226 |
-
"learning_rate": 1.1562677412296997e-06,
|
| 2227 |
-
"loss": 0.3123,
|
| 2228 |
-
"step": 1580
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 0.9594430992736077,
|
| 2232 |
-
"grad_norm": 0.4733109176158905,
|
| 2233 |
-
"learning_rate": 1.001510447572185e-06,
|
| 2234 |
-
"loss": 0.2871,
|
| 2235 |
-
"step": 1585
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 0.9624697336561744,
|
| 2239 |
-
"grad_norm": 1.2949782609939575,
|
| 2240 |
-
"learning_rate": 8.578149757610177e-07,
|
| 2241 |
-
"loss": 0.3129,
|
| 2242 |
-
"step": 1590
|
| 2243 |
-
},
|
| 2244 |
-
{
|
| 2245 |
-
"epoch": 0.965496368038741,
|
| 2246 |
-
"grad_norm": 0.5484541654586792,
|
| 2247 |
-
"learning_rate": 7.251973819372371e-07,
|
| 2248 |
-
"loss": 0.2189,
|
| 2249 |
-
"step": 1595
|
| 2250 |
-
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 0.9685230024213075,
|
| 2253 |
-
"grad_norm": 1.1343319416046143,
|
| 2254 |
-
"learning_rate": 6.036724844299069e-07,
|
| 2255 |
-
"loss": 0.3559,
|
| 2256 |
-
"step": 1600
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 0.9715496368038741,
|
| 2260 |
-
"grad_norm": 0.5619347095489502,
|
| 2261 |
-
"learning_rate": 4.932538621004068e-07,
|
| 2262 |
-
"loss": 0.2438,
|
| 2263 |
-
"step": 1605
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 0.9745762711864406,
|
| 2267 |
-
"grad_norm": 0.7649571895599365,
|
| 2268 |
-
"learning_rate": 3.939538528251463e-07,
|
| 2269 |
-
"loss": 0.3436,
|
| 2270 |
-
"step": 1610
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 0.9776029055690073,
|
| 2274 |
-
"grad_norm": 0.7493587732315063,
|
| 2275 |
-
"learning_rate": 3.0578355211697826e-07,
|
| 2276 |
-
"loss": 0.2304,
|
| 2277 |
-
"step": 1615
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 0.9806295399515739,
|
| 2281 |
-
"grad_norm": 0.5108829140663147,
|
| 2282 |
-
"learning_rate": 2.287528118853688e-07,
|
| 2283 |
-
"loss": 0.4229,
|
| 2284 |
-
"step": 1620
|
| 2285 |
-
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 0.9836561743341404,
|
| 2288 |
-
"grad_norm": 0.8058522343635559,
|
| 2289 |
-
"learning_rate": 1.6287023933564405e-07,
|
| 2290 |
-
"loss": 0.2617,
|
| 2291 |
-
"step": 1625
|
| 2292 |
-
},
|
| 2293 |
-
{
|
| 2294 |
-
"epoch": 0.986682808716707,
|
| 2295 |
-
"grad_norm": 0.3378280699253082,
|
| 2296 |
-
"learning_rate": 1.0814319600718215e-07,
|
| 2297 |
-
"loss": 0.2096,
|
| 2298 |
-
"step": 1630
|
| 2299 |
-
},
|
| 2300 |
-
{
|
| 2301 |
-
"epoch": 0.9897094430992736,
|
| 2302 |
-
"grad_norm": 0.8221763372421265,
|
| 2303 |
-
"learning_rate": 6.457779695090427e-08,
|
| 2304 |
-
"loss": 0.326,
|
| 2305 |
-
"step": 1635
|
| 2306 |
-
},
|
| 2307 |
-
{
|
| 2308 |
-
"epoch": 0.9927360774818402,
|
| 2309 |
-
"grad_norm": 0.685775101184845,
|
| 2310 |
-
"learning_rate": 3.2178910045965736e-08,
|
| 2311 |
-
"loss": 0.326,
|
| 2312 |
-
"step": 1640
|
| 2313 |
-
},
|
| 2314 |
-
{
|
| 2315 |
-
"epoch": 0.9957627118644068,
|
| 2316 |
-
"grad_norm": 0.35869747400283813,
|
| 2317 |
-
"learning_rate": 1.0950155455802246e-08,
|
| 2318 |
-
"loss": 0.2954,
|
| 2319 |
-
"step": 1645
|
| 2320 |
-
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 0.9987893462469734,
|
| 2323 |
-
"grad_norm": 0.6434110403060913,
|
| 2324 |
-
"learning_rate": 8.939052237089129e-10,
|
| 2325 |
-
"loss": 0.3243,
|
| 2326 |
-
"step": 1650
|
| 2327 |
-
},
|
| 2328 |
-
{
|
| 2329 |
-
"epoch": 1.0,
|
| 2330 |
-
"step": 1652,
|
| 2331 |
-
"total_flos": 8.124741191309722e+17,
|
| 2332 |
-
"train_loss": 0.33491338237434554,
|
| 2333 |
-
"train_runtime": 8909.3997,
|
| 2334 |
-
"train_samples_per_second": 1.483,
|
| 2335 |
-
"train_steps_per_second": 0.185
|
| 2336 |
}
|
| 2337 |
],
|
| 2338 |
"logging_steps": 5,
|
| 2339 |
-
"max_steps":
|
| 2340 |
"num_input_tokens_seen": 0,
|
| 2341 |
"num_train_epochs": 1,
|
| 2342 |
"save_steps": 50,
|
|
@@ -2352,7 +1281,7 @@
|
|
| 2352 |
"attributes": {}
|
| 2353 |
}
|
| 2354 |
},
|
| 2355 |
-
"total_flos": 8.
|
| 2356 |
"train_batch_size": 4,
|
| 2357 |
"trial_name": null,
|
| 2358 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9994372537985369,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 888,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.0011254924029262803,
|
| 13 |
+
"grad_norm": 0.8479363918304443,
|
| 14 |
+
"learning_rate": 2.247191011235955e-06,
|
| 15 |
+
"loss": 1.0084,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.005627462014631401,
|
| 20 |
+
"grad_norm": 1.0625563859939575,
|
| 21 |
+
"learning_rate": 1.1235955056179776e-05,
|
| 22 |
+
"loss": 1.1513,
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.011254924029262802,
|
| 27 |
+
"grad_norm": 1.9530349969863892,
|
| 28 |
+
"learning_rate": 2.2471910112359552e-05,
|
| 29 |
+
"loss": 1.3647,
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.016882386043894203,
|
| 34 |
+
"grad_norm": 1.0856108665466309,
|
| 35 |
+
"learning_rate": 3.370786516853933e-05,
|
| 36 |
+
"loss": 1.1026,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.022509848058525603,
|
| 41 |
+
"grad_norm": 1.3664439916610718,
|
| 42 |
+
"learning_rate": 4.4943820224719104e-05,
|
| 43 |
+
"loss": 0.836,
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.028137310073157007,
|
| 48 |
+
"grad_norm": 0.8295888304710388,
|
| 49 |
+
"learning_rate": 5.6179775280898885e-05,
|
| 50 |
+
"loss": 0.6505,
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.03376477208778841,
|
| 55 |
+
"grad_norm": 0.8303141593933105,
|
| 56 |
+
"learning_rate": 6.741573033707866e-05,
|
| 57 |
+
"loss": 0.6232,
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.03939223410241981,
|
| 62 |
+
"grad_norm": 0.6008478999137878,
|
| 63 |
+
"learning_rate": 7.865168539325843e-05,
|
| 64 |
+
"loss": 0.428,
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.04501969611705121,
|
| 69 |
+
"grad_norm": 0.731120228767395,
|
| 70 |
+
"learning_rate": 8.988764044943821e-05,
|
| 71 |
+
"loss": 0.4453,
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.050647158131682614,
|
| 76 |
+
"grad_norm": 0.503491997718811,
|
| 77 |
+
"learning_rate": 0.00010112359550561799,
|
| 78 |
+
"loss": 0.3312,
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.056274620146314014,
|
| 83 |
+
"grad_norm": 0.736613392829895,
|
| 84 |
+
"learning_rate": 0.00011235955056179777,
|
| 85 |
+
"loss": 0.3825,
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.061902082160945414,
|
| 90 |
+
"grad_norm": 1.1015475988388062,
|
| 91 |
+
"learning_rate": 0.00012359550561797752,
|
| 92 |
+
"loss": 0.2907,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.06752954417557681,
|
| 97 |
+
"grad_norm": 0.7216058373451233,
|
| 98 |
+
"learning_rate": 0.00013483146067415732,
|
| 99 |
+
"loss": 0.3717,
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.07315700619020822,
|
| 104 |
+
"grad_norm": 1.0304406881332397,
|
| 105 |
+
"learning_rate": 0.0001460674157303371,
|
| 106 |
+
"loss": 0.4053,
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.07878446820483961,
|
| 111 |
+
"grad_norm": 0.717244029045105,
|
| 112 |
+
"learning_rate": 0.00015730337078651685,
|
| 113 |
+
"loss": 0.3471,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.08441193021947102,
|
| 118 |
+
"grad_norm": 0.8437204360961914,
|
| 119 |
+
"learning_rate": 0.00016853932584269662,
|
| 120 |
+
"loss": 0.3412,
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.09003939223410241,
|
| 125 |
+
"grad_norm": 0.5658055543899536,
|
| 126 |
+
"learning_rate": 0.00017977528089887642,
|
| 127 |
+
"loss": 0.4183,
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.09566685424873382,
|
| 132 |
+
"grad_norm": 0.6256535053253174,
|
| 133 |
+
"learning_rate": 0.00019101123595505618,
|
| 134 |
+
"loss": 0.3304,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.10129431626336523,
|
| 139 |
+
"grad_norm": 0.41338804364204407,
|
| 140 |
+
"learning_rate": 0.00019999922700687455,
|
| 141 |
+
"loss": 0.324,
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.10692177827799662,
|
| 146 |
+
"grad_norm": 1.2900114059448242,
|
| 147 |
+
"learning_rate": 0.000199972173502251,
|
| 148 |
+
"loss": 0.3403,
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.11254924029262803,
|
| 153 |
+
"grad_norm": 0.5315810441970825,
|
| 154 |
+
"learning_rate": 0.00019990648229089103,
|
| 155 |
+
"loss": 0.3061,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.11817670230725942,
|
| 160 |
+
"grad_norm": 0.6736317276954651,
|
| 161 |
+
"learning_rate": 0.00019980217876143698,
|
| 162 |
+
"loss": 0.307,
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.12380416432189083,
|
| 167 |
+
"grad_norm": 0.5415677428245544,
|
| 168 |
+
"learning_rate": 0.000199659303225598,
|
| 169 |
+
"loss": 0.3489,
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.12943162633652222,
|
| 174 |
+
"grad_norm": 0.6013039946556091,
|
| 175 |
+
"learning_rate": 0.0001994779109025702,
|
| 176 |
+
"loss": 0.3426,
|
| 177 |
"step": 115
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.13505908835115363,
|
| 181 |
+
"grad_norm": 0.6924653053283691,
|
| 182 |
+
"learning_rate": 0.00019925807189769533,
|
| 183 |
+
"loss": 0.3847,
|
| 184 |
"step": 120
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.14068655036578503,
|
| 188 |
+
"grad_norm": 2.3059215545654297,
|
| 189 |
+
"learning_rate": 0.00019899987117536587,
|
| 190 |
+
"loss": 0.3521,
|
| 191 |
"step": 125
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.14631401238041644,
|
| 195 |
+
"grad_norm": 0.7815282940864563,
|
| 196 |
+
"learning_rate": 0.00019870340852618803,
|
| 197 |
+
"loss": 0.2937,
|
| 198 |
"step": 130
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 0.15194147439504782,
|
| 202 |
+
"grad_norm": 0.7751880288124084,
|
| 203 |
+
"learning_rate": 0.00019836879852841387,
|
| 204 |
+
"loss": 0.3299,
|
| 205 |
"step": 135
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.15756893640967923,
|
| 209 |
+
"grad_norm": 0.6737527251243591,
|
| 210 |
+
"learning_rate": 0.0001979961705036587,
|
| 211 |
+
"loss": 0.307,
|
| 212 |
"step": 140
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.16319639842431063,
|
| 216 |
+
"grad_norm": 0.4875199794769287,
|
| 217 |
+
"learning_rate": 0.00019758566846692029,
|
| 218 |
+
"loss": 0.3564,
|
| 219 |
"step": 145
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.16882386043894204,
|
| 223 |
+
"grad_norm": 0.48927029967308044,
|
| 224 |
+
"learning_rate": 0.00019713745107091923,
|
| 225 |
+
"loss": 0.3709,
|
| 226 |
"step": 150
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 0.17445132245357345,
|
| 230 |
+
"grad_norm": 0.8061405420303345,
|
| 231 |
+
"learning_rate": 0.00019665169154478213,
|
| 232 |
+
"loss": 0.2479,
|
| 233 |
"step": 155
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 0.18007878446820483,
|
| 237 |
+
"grad_norm": 0.5379148721694946,
|
| 238 |
+
"learning_rate": 0.00019612857762709124,
|
| 239 |
+
"loss": 0.2649,
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 0.18570624648283623,
|
| 244 |
+
"grad_norm": 0.659520149230957,
|
| 245 |
+
"learning_rate": 0.0001955683114933263,
|
| 246 |
+
"loss": 0.3598,
|
| 247 |
"step": 165
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 0.19133370849746764,
|
| 251 |
+
"grad_norm": 0.6669427752494812,
|
| 252 |
+
"learning_rate": 0.00019497110967772692,
|
| 253 |
+
"loss": 0.2604,
|
| 254 |
"step": 170
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.19696117051209905,
|
| 258 |
+
"grad_norm": 0.48957839608192444,
|
| 259 |
+
"learning_rate": 0.00019433720298960537,
|
| 260 |
+
"loss": 0.2664,
|
| 261 |
"step": 175
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.20258863252673046,
|
| 265 |
+
"grad_norm": 0.4717267155647278,
|
| 266 |
+
"learning_rate": 0.0001936668364241424,
|
| 267 |
+
"loss": 0.2949,
|
| 268 |
"step": 180
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 0.20821609454136183,
|
| 272 |
+
"grad_norm": 0.5135904550552368,
|
| 273 |
+
"learning_rate": 0.00019296026906770027,
|
| 274 |
+
"loss": 0.3402,
|
| 275 |
"step": 185
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 0.21384355655599324,
|
| 279 |
+
"grad_norm": 0.36506161093711853,
|
| 280 |
+
"learning_rate": 0.00019221777399768998,
|
| 281 |
+
"loss": 0.2935,
|
| 282 |
"step": 190
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 0.21947101857062465,
|
| 286 |
+
"grad_norm": 0.6421459317207336,
|
| 287 |
+
"learning_rate": 0.00019143963817703087,
|
| 288 |
+
"loss": 0.2803,
|
| 289 |
"step": 195
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 0.22509848058525606,
|
| 293 |
+
"grad_norm": 0.6440847516059875,
|
| 294 |
+
"learning_rate": 0.0001906261623432441,
|
| 295 |
+
"loss": 0.2791,
|
| 296 |
"step": 200
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.23072594259988746,
|
| 300 |
+
"grad_norm": 0.4226614832878113,
|
| 301 |
+
"learning_rate": 0.00018977766089222208,
|
| 302 |
+
"loss": 0.2942,
|
| 303 |
"step": 205
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.23635340461451884,
|
| 307 |
+
"grad_norm": 0.3801310360431671,
|
| 308 |
+
"learning_rate": 0.00018889446175671926,
|
| 309 |
+
"loss": 0.3874,
|
| 310 |
"step": 210
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 0.24198086662915025,
|
| 314 |
+
"grad_norm": 0.626059353351593,
|
| 315 |
+
"learning_rate": 0.00018797690627961132,
|
| 316 |
+
"loss": 0.4083,
|
| 317 |
"step": 215
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 0.24760832864378166,
|
| 321 |
+
"grad_norm": 0.35303783416748047,
|
| 322 |
+
"learning_rate": 0.0001870253490819713,
|
| 323 |
+
"loss": 0.2474,
|
| 324 |
"step": 220
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 0.25323579065841306,
|
| 328 |
+
"grad_norm": 0.5032233595848083,
|
| 329 |
+
"learning_rate": 0.00018604015792601396,
|
| 330 |
+
"loss": 0.236,
|
| 331 |
"step": 225
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 0.25886325267304444,
|
| 335 |
+
"grad_norm": 0.7859108448028564,
|
| 336 |
+
"learning_rate": 0.00018502171357296144,
|
| 337 |
+
"loss": 0.2065,
|
| 338 |
"step": 230
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.2644907146876759,
|
| 342 |
+
"grad_norm": 0.47997885942459106,
|
| 343 |
+
"learning_rate": 0.00018397040963588488,
|
| 344 |
+
"loss": 0.2757,
|
| 345 |
"step": 235
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.27011817670230726,
|
| 349 |
+
"grad_norm": 0.3313188850879669,
|
| 350 |
+
"learning_rate": 0.00018288665242757903,
|
| 351 |
+
"loss": 0.2234,
|
| 352 |
"step": 240
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 0.27574563871693863,
|
| 356 |
+
"grad_norm": 0.42563286423683167,
|
| 357 |
+
"learning_rate": 0.0001817708608035286,
|
| 358 |
+
"loss": 0.2735,
|
| 359 |
"step": 245
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 0.28137310073157007,
|
| 363 |
+
"grad_norm": 0.7245692610740662,
|
| 364 |
+
"learning_rate": 0.00018062346600002699,
|
| 365 |
+
"loss": 0.247,
|
| 366 |
"step": 250
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 0.28700056274620145,
|
| 370 |
+
"grad_norm": 0.41776588559150696,
|
| 371 |
+
"learning_rate": 0.00017944491146751026,
|
| 372 |
+
"loss": 0.3202,
|
| 373 |
"step": 255
|
| 374 |
},
|
| 375 |
{
|
| 376 |
+
"epoch": 0.2926280247608329,
|
| 377 |
+
"grad_norm": 0.46748408675193787,
|
| 378 |
+
"learning_rate": 0.0001782356526991702,
|
| 379 |
+
"loss": 0.2366,
|
| 380 |
"step": 260
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.29825548677546426,
|
| 384 |
+
"grad_norm": 0.3817427456378937,
|
| 385 |
+
"learning_rate": 0.00017699615705491325,
|
| 386 |
+
"loss": 0.2053,
|
| 387 |
"step": 265
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 0.30388294879009564,
|
| 391 |
+
"grad_norm": 0.3615594208240509,
|
| 392 |
+
"learning_rate": 0.00017572690358073326,
|
| 393 |
+
"loss": 0.1965,
|
| 394 |
"step": 270
|
| 395 |
},
|
| 396 |
{
|
| 397 |
+
"epoch": 0.3095104108047271,
|
| 398 |
+
"grad_norm": 0.48413321375846863,
|
| 399 |
+
"learning_rate": 0.00017442838282356727,
|
| 400 |
+
"loss": 0.2121,
|
| 401 |
"step": 275
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 0.31513787281935846,
|
| 405 |
+
"grad_norm": 0.7311267256736755,
|
| 406 |
+
"learning_rate": 0.00017310109664170703,
|
| 407 |
+
"loss": 0.3424,
|
| 408 |
"step": 280
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 0.3207653348339899,
|
| 412 |
+
"grad_norm": 0.4602237641811371,
|
| 413 |
+
"learning_rate": 0.00017174555801083814,
|
| 414 |
+
"loss": 0.2099,
|
| 415 |
"step": 285
|
| 416 |
},
|
| 417 |
{
|
| 418 |
+
"epoch": 0.32639279684862127,
|
| 419 |
+
"grad_norm": 0.5406469702720642,
|
| 420 |
+
"learning_rate": 0.00017036229082578307,
|
| 421 |
+
"loss": 0.3583,
|
| 422 |
"step": 290
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.33202025886325265,
|
| 426 |
+
"grad_norm": 0.26748359203338623,
|
| 427 |
+
"learning_rate": 0.00016895182969802386,
|
| 428 |
+
"loss": 0.2119,
|
| 429 |
"step": 295
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 0.3376477208778841,
|
| 433 |
+
"grad_norm": 0.40785279870033264,
|
| 434 |
+
"learning_rate": 0.00016751471974908288,
|
| 435 |
+
"loss": 0.2793,
|
| 436 |
"step": 300
|
| 437 |
},
|
| 438 |
{
|
| 439 |
+
"epoch": 0.34327518289251546,
|
| 440 |
+
"grad_norm": 0.5709397196769714,
|
| 441 |
+
"learning_rate": 0.00016605151639984187,
|
| 442 |
+
"loss": 0.2525,
|
| 443 |
"step": 305
|
| 444 |
},
|
| 445 |
{
|
| 446 |
+
"epoch": 0.3489026449071469,
|
| 447 |
+
"grad_norm": 0.857624351978302,
|
| 448 |
+
"learning_rate": 0.00016456278515588024,
|
| 449 |
+
"loss": 0.3339,
|
| 450 |
"step": 310
|
| 451 |
},
|
| 452 |
{
|
| 453 |
+
"epoch": 0.3545301069217783,
|
| 454 |
+
"grad_norm": 0.33093103766441345,
|
| 455 |
+
"learning_rate": 0.00016304910138891597,
|
| 456 |
+
"loss": 0.3434,
|
| 457 |
"step": 315
|
| 458 |
},
|
| 459 |
{
|
| 460 |
+
"epoch": 0.36015756893640966,
|
| 461 |
+
"grad_norm": 0.44036629796028137,
|
| 462 |
+
"learning_rate": 0.00016151105011443314,
|
| 463 |
+
"loss": 0.229,
|
| 464 |
"step": 320
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.3657850309510411,
|
| 468 |
+
"grad_norm": 0.3742360770702362,
|
| 469 |
+
"learning_rate": 0.00015994922576558263,
|
| 470 |
+
"loss": 0.3026,
|
| 471 |
"step": 325
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 0.37141249296567247,
|
| 475 |
+
"grad_norm": 0.3356841802597046,
|
| 476 |
+
"learning_rate": 0.0001583642319634426,
|
| 477 |
+
"loss": 0.2458,
|
| 478 |
"step": 330
|
| 479 |
},
|
| 480 |
{
|
| 481 |
+
"epoch": 0.3770399549803039,
|
| 482 |
+
"grad_norm": 0.48720481991767883,
|
| 483 |
+
"learning_rate": 0.00015675668128372854,
|
| 484 |
+
"loss": 0.2653,
|
| 485 |
"step": 335
|
| 486 |
},
|
| 487 |
{
|
| 488 |
+
"epoch": 0.3826674169949353,
|
| 489 |
+
"grad_norm": 0.6053282022476196,
|
| 490 |
+
"learning_rate": 0.00015512719502004197,
|
| 491 |
+
"loss": 0.3231,
|
| 492 |
"step": 340
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"epoch": 0.38829487900956666,
|
| 496 |
+
"grad_norm": 0.41004592180252075,
|
| 497 |
+
"learning_rate": 0.00015347640294375005,
|
| 498 |
+
"loss": 0.3181,
|
| 499 |
"step": 345
|
| 500 |
},
|
| 501 |
{
|
| 502 |
+
"epoch": 0.3939223410241981,
|
| 503 |
+
"grad_norm": 0.6467001438140869,
|
| 504 |
+
"learning_rate": 0.0001518049430605887,
|
| 505 |
+
"loss": 0.5039,
|
| 506 |
"step": 350
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.3995498030388295,
|
| 510 |
+
"grad_norm": 0.4072398543357849,
|
| 511 |
+
"learning_rate": 0.0001501134613640832,
|
| 512 |
+
"loss": 0.22,
|
| 513 |
"step": 355
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 0.4051772650534609,
|
| 517 |
+
"grad_norm": 0.7158648371696472,
|
| 518 |
+
"learning_rate": 0.0001484026115858815,
|
| 519 |
+
"loss": 0.2842,
|
| 520 |
"step": 360
|
| 521 |
},
|
| 522 |
{
|
| 523 |
+
"epoch": 0.4108047270680923,
|
| 524 |
+
"grad_norm": 0.5195010304450989,
|
| 525 |
+
"learning_rate": 0.00014667305494309727,
|
| 526 |
+
"loss": 0.2622,
|
| 527 |
"step": 365
|
| 528 |
},
|
| 529 |
{
|
| 530 |
+
"epoch": 0.41643218908272367,
|
| 531 |
+
"grad_norm": 0.5050496459007263,
|
| 532 |
+
"learning_rate": 0.00014492545988275933,
|
| 533 |
+
"loss": 0.3665,
|
| 534 |
"step": 370
|
| 535 |
},
|
| 536 |
{
|
| 537 |
+
"epoch": 0.4220596510973551,
|
| 538 |
+
"grad_norm": 0.27042141556739807,
|
| 539 |
+
"learning_rate": 0.00014316050182346733,
|
| 540 |
+
"loss": 0.3115,
|
| 541 |
"step": 375
|
| 542 |
},
|
| 543 |
{
|
| 544 |
+
"epoch": 0.4276871131119865,
|
| 545 |
+
"grad_norm": 0.35634538531303406,
|
| 546 |
+
"learning_rate": 0.00014137886289435295,
|
| 547 |
+
"loss": 0.2217,
|
| 548 |
"step": 380
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 0.4333145751266179,
|
| 552 |
+
"grad_norm": 0.336165189743042,
|
| 553 |
+
"learning_rate": 0.00013958123167144733,
|
| 554 |
+
"loss": 0.2716,
|
| 555 |
"step": 385
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 0.4389420371412493,
|
| 559 |
+
"grad_norm": 0.5529112219810486,
|
| 560 |
+
"learning_rate": 0.00013776830291155703,
|
| 561 |
+
"loss": 0.3433,
|
| 562 |
"step": 390
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"epoch": 0.4445694991558807,
|
| 566 |
+
"grad_norm": 0.3461211025714874,
|
| 567 |
+
"learning_rate": 0.00013594077728375128,
|
| 568 |
+
"loss": 0.2916,
|
| 569 |
"step": 395
|
| 570 |
},
|
| 571 |
{
|
| 572 |
+
"epoch": 0.4501969611705121,
|
| 573 |
+
"grad_norm": 0.6085760593414307,
|
| 574 |
+
"learning_rate": 0.00013409936109856424,
|
| 575 |
+
"loss": 0.2442,
|
| 576 |
"step": 400
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 0.4558244231851435,
|
| 580 |
+
"grad_norm": 0.601696252822876,
|
| 581 |
+
"learning_rate": 0.00013224476603501662,
|
| 582 |
+
"loss": 0.2316,
|
| 583 |
"step": 405
|
| 584 |
},
|
| 585 |
{
|
| 586 |
+
"epoch": 0.4614518851997749,
|
| 587 |
+
"grad_norm": 0.44390204548835754,
|
| 588 |
+
"learning_rate": 0.00013037770886556294,
|
| 589 |
+
"loss": 0.3902,
|
| 590 |
"step": 410
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 0.4670793472144063,
|
| 594 |
+
"grad_norm": 0.3032180368900299,
|
| 595 |
+
"learning_rate": 0.00012849891117906978,
|
| 596 |
+
"loss": 0.2376,
|
| 597 |
"step": 415
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 0.4727068092290377,
|
| 601 |
+
"grad_norm": 1.0338925123214722,
|
| 602 |
+
"learning_rate": 0.00012660909910193303,
|
| 603 |
+
"loss": 0.2524,
|
| 604 |
"step": 420
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 0.4783342712436691,
|
| 608 |
+
"grad_norm": 0.29725944995880127,
|
| 609 |
+
"learning_rate": 0.000124709003017441,
|
| 610 |
+
"loss": 0.2789,
|
| 611 |
"step": 425
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 0.4839617332583005,
|
| 615 |
+
"grad_norm": 0.5633781552314758,
|
| 616 |
+
"learning_rate": 0.0001227993572834926,
|
| 617 |
+
"loss": 0.2129,
|
| 618 |
"step": 430
|
| 619 |
},
|
| 620 |
{
|
| 621 |
+
"epoch": 0.48958919527293193,
|
| 622 |
+
"grad_norm": 0.4558543562889099,
|
| 623 |
+
"learning_rate": 0.0001208808999487793,
|
| 624 |
+
"loss": 0.2721,
|
| 625 |
"step": 435
|
| 626 |
},
|
| 627 |
{
|
| 628 |
+
"epoch": 0.4952166572875633,
|
| 629 |
+
"grad_norm": 0.41546952724456787,
|
| 630 |
+
"learning_rate": 0.00011895437246754074,
|
| 631 |
+
"loss": 0.2707,
|
| 632 |
"step": 440
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 0.5008441193021947,
|
| 636 |
+
"grad_norm": 0.5292194485664368,
|
| 637 |
+
"learning_rate": 0.00011702051941300396,
|
| 638 |
+
"loss": 0.2852,
|
| 639 |
"step": 445
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 0.5064715813168261,
|
| 643 |
+
"grad_norm": 0.44494009017944336,
|
| 644 |
+
"learning_rate": 0.00011508008818961731,
|
| 645 |
+
"loss": 0.1887,
|
| 646 |
"step": 450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
+
"epoch": 0.5120990433314575,
|
| 650 |
+
"grad_norm": 2.672574043273926,
|
| 651 |
+
"learning_rate": 0.00011313382874419031,
|
| 652 |
+
"loss": 0.2296,
|
| 653 |
"step": 455
|
| 654 |
},
|
| 655 |
{
|
| 656 |
+
"epoch": 0.5177265053460889,
|
| 657 |
+
"grad_norm": 0.4071643054485321,
|
| 658 |
+
"learning_rate": 0.00011118249327605055,
|
| 659 |
+
"loss": 0.2921,
|
| 660 |
"step": 460
|
| 661 |
},
|
| 662 |
{
|
| 663 |
+
"epoch": 0.5233539673607203,
|
| 664 |
+
"grad_norm": 0.41754430532455444,
|
| 665 |
+
"learning_rate": 0.00010922683594633021,
|
| 666 |
+
"loss": 0.2266,
|
| 667 |
"step": 465
|
| 668 |
},
|
| 669 |
{
|
| 670 |
+
"epoch": 0.5289814293753518,
|
| 671 |
+
"grad_norm": 0.27264049649238586,
|
| 672 |
+
"learning_rate": 0.00010726761258649461,
|
| 673 |
+
"loss": 0.239,
|
| 674 |
"step": 470
|
| 675 |
},
|
| 676 |
{
|
| 677 |
+
"epoch": 0.5346088913899831,
|
| 678 |
+
"grad_norm": 0.5015512704849243,
|
| 679 |
+
"learning_rate": 0.00010530558040622472,
|
| 680 |
+
"loss": 0.3052,
|
| 681 |
"step": 475
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 0.5402363534046145,
|
| 685 |
+
"grad_norm": 0.3031107187271118,
|
| 686 |
+
"learning_rate": 0.00010334149770076747,
|
| 687 |
+
"loss": 0.3044,
|
| 688 |
"step": 480
|
| 689 |
},
|
| 690 |
{
|
| 691 |
+
"epoch": 0.5458638154192459,
|
| 692 |
+
"grad_norm": 0.75295090675354,
|
| 693 |
+
"learning_rate": 0.00010137612355786618,
|
| 694 |
+
"loss": 0.2545,
|
| 695 |
"step": 485
|
| 696 |
},
|
| 697 |
{
|
| 698 |
+
"epoch": 0.5514912774338773,
|
| 699 |
+
"grad_norm": 0.5242462754249573,
|
| 700 |
+
"learning_rate": 9.941021756438488e-05,
|
| 701 |
+
"loss": 0.2483,
|
| 702 |
"step": 490
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"epoch": 0.5571187394485088,
|
| 706 |
+
"grad_norm": 0.30885180830955505,
|
| 707 |
+
"learning_rate": 9.744453951273968e-05,
|
| 708 |
+
"loss": 0.279,
|
| 709 |
"step": 495
|
| 710 |
},
|
| 711 |
{
|
| 712 |
+
"epoch": 0.5627462014631401,
|
| 713 |
+
"grad_norm": 0.39063045382499695,
|
| 714 |
+
"learning_rate": 9.547984910725064e-05,
|
| 715 |
+
"loss": 0.211,
|
| 716 |
"step": 500
|
| 717 |
},
|
| 718 |
{
|
| 719 |
+
"epoch": 0.5683736634777715,
|
| 720 |
+
"grad_norm": 0.3500135838985443,
|
| 721 |
+
"learning_rate": 9.35169056705278e-05,
|
| 722 |
+
"loss": 0.2148,
|
| 723 |
"step": 505
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 0.5740011254924029,
|
| 727 |
+
"grad_norm": 0.34409910440444946,
|
| 728 |
+
"learning_rate": 9.155646785000467e-05,
|
| 729 |
+
"loss": 0.2509,
|
| 730 |
"step": 510
|
| 731 |
},
|
| 732 |
{
|
| 733 |
+
"epoch": 0.5796285875070343,
|
| 734 |
+
"grad_norm": 0.3541821539402008,
|
| 735 |
+
"learning_rate": 8.959929332473262e-05,
|
| 736 |
+
"loss": 0.3381,
|
| 737 |
"step": 515
|
| 738 |
},
|
| 739 |
{
|
| 740 |
+
"epoch": 0.5852560495216658,
|
| 741 |
+
"grad_norm": 0.8181503415107727,
|
| 742 |
+
"learning_rate": 8.764613851254968e-05,
|
| 743 |
+
"loss": 0.2808,
|
| 744 |
"step": 520
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"epoch": 0.5908835115362971,
|
| 748 |
+
"grad_norm": 0.36616384983062744,
|
| 749 |
+
"learning_rate": 8.569775827773656e-05,
|
| 750 |
+
"loss": 0.1762,
|
| 751 |
"step": 525
|
| 752 |
},
|
| 753 |
{
|
| 754 |
+
"epoch": 0.5965109735509285,
|
| 755 |
+
"grad_norm": 0.5440443754196167,
|
| 756 |
+
"learning_rate": 8.375490563927328e-05,
|
| 757 |
+
"loss": 0.2909,
|
| 758 |
"step": 530
|
| 759 |
},
|
| 760 |
{
|
| 761 |
+
"epoch": 0.6021384355655599,
|
| 762 |
+
"grad_norm": 0.4618743360042572,
|
| 763 |
+
"learning_rate": 8.181833147980894e-05,
|
| 764 |
+
"loss": 0.3326,
|
| 765 |
"step": 535
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 0.6077658975801913,
|
| 769 |
+
"grad_norm": 0.604923665523529,
|
| 770 |
+
"learning_rate": 7.98887842554572e-05,
|
| 771 |
+
"loss": 0.2639,
|
| 772 |
"step": 540
|
| 773 |
},
|
| 774 |
{
|
| 775 |
+
"epoch": 0.6133933595948228,
|
| 776 |
+
"grad_norm": 0.46064886450767517,
|
| 777 |
+
"learning_rate": 7.796700970652932e-05,
|
| 778 |
+
"loss": 0.2933,
|
| 779 |
"step": 545
|
| 780 |
},
|
| 781 |
{
|
| 782 |
+
"epoch": 0.6190208216094542,
|
| 783 |
+
"grad_norm": 0.4644724428653717,
|
| 784 |
+
"learning_rate": 7.605375056931712e-05,
|
| 785 |
+
"loss": 0.2487,
|
| 786 |
"step": 550
|
| 787 |
},
|
| 788 |
{
|
| 789 |
+
"epoch": 0.6246482836240855,
|
| 790 |
+
"grad_norm": 0.5088258981704712,
|
| 791 |
+
"learning_rate": 7.41497462890369e-05,
|
| 792 |
+
"loss": 0.2957,
|
| 793 |
"step": 555
|
| 794 |
},
|
| 795 |
{
|
| 796 |
+
"epoch": 0.6302757456387169,
|
| 797 |
+
"grad_norm": 0.6130680441856384,
|
| 798 |
+
"learning_rate": 7.225573273404513e-05,
|
| 799 |
+
"loss": 0.2588,
|
| 800 |
"step": 560
|
| 801 |
},
|
| 802 |
{
|
| 803 |
+
"epoch": 0.6359032076533483,
|
| 804 |
+
"grad_norm": 0.45168906450271606,
|
| 805 |
+
"learning_rate": 7.037244191143661e-05,
|
| 806 |
+
"loss": 0.2477,
|
| 807 |
"step": 565
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 0.6415306696679798,
|
| 811 |
+
"grad_norm": 0.2619435787200928,
|
| 812 |
+
"learning_rate": 6.850060168413518e-05,
|
| 813 |
+
"loss": 0.1566,
|
| 814 |
"step": 570
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"epoch": 0.6471581316826112,
|
| 818 |
+
"grad_norm": 0.36049678921699524,
|
| 819 |
+
"learning_rate": 6.66409354895857e-05,
|
| 820 |
+
"loss": 0.3076,
|
| 821 |
"step": 575
|
| 822 |
},
|
| 823 |
{
|
| 824 |
+
"epoch": 0.6527855936972425,
|
| 825 |
+
"grad_norm": 0.20779716968536377,
|
| 826 |
+
"learning_rate": 6.479416206015679e-05,
|
| 827 |
+
"loss": 0.1849,
|
| 828 |
"step": 580
|
| 829 |
},
|
| 830 |
{
|
| 831 |
+
"epoch": 0.6584130557118739,
|
| 832 |
+
"grad_norm": 0.5497414469718933,
|
| 833 |
+
"learning_rate": 6.296099514536167e-05,
|
| 834 |
+
"loss": 0.2734,
|
| 835 |
"step": 585
|
| 836 |
},
|
| 837 |
{
|
| 838 |
+
"epoch": 0.6640405177265053,
|
| 839 |
+
"grad_norm": 0.3398374319076538,
|
| 840 |
+
"learning_rate": 6.114214323600504e-05,
|
| 841 |
+
"loss": 0.2271,
|
| 842 |
"step": 590
|
| 843 |
},
|
| 844 |
{
|
| 845 |
+
"epoch": 0.6696679797411368,
|
| 846 |
+
"grad_norm": 0.2531116306781769,
|
| 847 |
+
"learning_rate": 5.9338309290362324e-05,
|
| 848 |
+
"loss": 0.1864,
|
| 849 |
"step": 595
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 0.6752954417557682,
|
| 853 |
+
"grad_norm": 0.38855037093162537,
|
| 854 |
+
"learning_rate": 5.7550190462496946e-05,
|
| 855 |
+
"loss": 0.2591,
|
| 856 |
"step": 600
|
| 857 |
},
|
| 858 |
{
|
| 859 |
+
"epoch": 0.6809229037703995,
|
| 860 |
+
"grad_norm": 0.4525040090084076,
|
| 861 |
+
"learning_rate": 5.577847783282122e-05,
|
| 862 |
+
"loss": 0.3319,
|
| 863 |
"step": 605
|
| 864 |
},
|
| 865 |
{
|
| 866 |
+
"epoch": 0.6865503657850309,
|
| 867 |
+
"grad_norm": 0.5147562623023987,
|
| 868 |
+
"learning_rate": 5.4023856141004236e-05,
|
| 869 |
+
"loss": 0.2328,
|
| 870 |
"step": 610
|
| 871 |
},
|
| 872 |
{
|
| 873 |
+
"epoch": 0.6921778277996623,
|
| 874 |
+
"grad_norm": 0.48170509934425354,
|
| 875 |
+
"learning_rate": 5.228700352133071e-05,
|
| 876 |
+
"loss": 0.3459,
|
| 877 |
"step": 615
|
| 878 |
},
|
| 879 |
{
|
| 880 |
+
"epoch": 0.6978052898142938,
|
| 881 |
+
"grad_norm": 0.3423648178577423,
|
| 882 |
+
"learning_rate": 5.05685912406123e-05,
|
| 883 |
+
"loss": 0.2372,
|
| 884 |
"step": 620
|
| 885 |
},
|
| 886 |
{
|
| 887 |
+
"epoch": 0.7034327518289252,
|
| 888 |
+
"grad_norm": 0.34861063957214355,
|
| 889 |
+
"learning_rate": 4.886928343875341e-05,
|
| 890 |
+
"loss": 0.2686,
|
| 891 |
"step": 625
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 0.7090602138435566,
|
| 895 |
+
"grad_norm": 0.5629642009735107,
|
| 896 |
+
"learning_rate": 4.71897368720714e-05,
|
| 897 |
+
"loss": 0.1908,
|
| 898 |
"step": 630
|
| 899 |
},
|
| 900 |
{
|
| 901 |
+
"epoch": 0.7146876758581879,
|
| 902 |
+
"grad_norm": 0.47687891125679016,
|
| 903 |
+
"learning_rate": 4.553060065947013e-05,
|
| 904 |
+
"loss": 0.3051,
|
| 905 |
"step": 635
|
| 906 |
},
|
| 907 |
{
|
| 908 |
+
"epoch": 0.7203151378728193,
|
| 909 |
+
"grad_norm": 0.3600665032863617,
|
| 910 |
+
"learning_rate": 4.3892516031565954e-05,
|
| 911 |
+
"loss": 0.216,
|
| 912 |
"step": 640
|
| 913 |
},
|
| 914 |
{
|
| 915 |
+
"epoch": 0.7259425998874508,
|
| 916 |
+
"grad_norm": 0.3238658905029297,
|
| 917 |
+
"learning_rate": 4.227611608286147e-05,
|
| 918 |
+
"loss": 0.2727,
|
| 919 |
"step": 645
|
| 920 |
},
|
| 921 |
{
|
| 922 |
+
"epoch": 0.7315700619020822,
|
| 923 |
+
"grad_norm": 0.5159012675285339,
|
| 924 |
+
"learning_rate": 4.0682025527064486e-05,
|
| 925 |
+
"loss": 0.2408,
|
| 926 |
"step": 650
|
| 927 |
},
|
| 928 |
{
|
| 929 |
+
"epoch": 0.7371975239167136,
|
| 930 |
+
"grad_norm": 0.5588057041168213,
|
| 931 |
+
"learning_rate": 3.911086045564575e-05,
|
| 932 |
+
"loss": 0.2084,
|
| 933 |
"step": 655
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 0.7428249859313449,
|
| 937 |
+
"grad_norm": 0.33754897117614746,
|
| 938 |
+
"learning_rate": 3.756322809972905e-05,
|
| 939 |
+
"loss": 0.2405,
|
| 940 |
"step": 660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
+
"epoch": 0.7484524479459763,
|
| 944 |
+
"grad_norm": 0.3047542870044708,
|
| 945 |
+
"learning_rate": 3.6039726595405755e-05,
|
| 946 |
+
"loss": 0.2054,
|
| 947 |
"step": 665
|
| 948 |
},
|
| 949 |
{
|
| 950 |
+
"epoch": 0.7540799099606078,
|
| 951 |
+
"grad_norm": 0.5348085761070251,
|
| 952 |
+
"learning_rate": 3.4540944752564406e-05,
|
| 953 |
+
"loss": 0.215,
|
| 954 |
"step": 670
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"epoch": 0.7597073719752392,
|
| 958 |
+
"grad_norm": 0.6096937656402588,
|
| 959 |
+
"learning_rate": 3.3067461827324755e-05,
|
| 960 |
+
"loss": 0.3615,
|
| 961 |
"step": 675
|
| 962 |
},
|
| 963 |
{
|
| 964 |
+
"epoch": 0.7653348339898706,
|
| 965 |
+
"grad_norm": 0.29212549328804016,
|
| 966 |
+
"learning_rate": 3.161984729816415e-05,
|
| 967 |
+
"loss": 0.2056,
|
| 968 |
"step": 680
|
| 969 |
},
|
| 970 |
{
|
| 971 |
+
"epoch": 0.770962296004502,
|
| 972 |
+
"grad_norm": 0.40820997953414917,
|
| 973 |
+
"learning_rate": 3.0198660645822985e-05,
|
| 974 |
+
"loss": 0.2244,
|
| 975 |
"step": 685
|
| 976 |
},
|
| 977 |
{
|
| 978 |
+
"epoch": 0.7765897580191333,
|
| 979 |
+
"grad_norm": 0.35913002490997314,
|
| 980 |
+
"learning_rate": 2.880445113707384e-05,
|
| 981 |
+
"loss": 0.2536,
|
| 982 |
"step": 690
|
| 983 |
},
|
| 984 |
{
|
| 985 |
+
"epoch": 0.7822172200337648,
|
| 986 |
+
"grad_norm": 0.36664295196533203,
|
| 987 |
+
"learning_rate": 2.743775761243843e-05,
|
| 988 |
+
"loss": 0.266,
|
| 989 |
"step": 695
|
| 990 |
},
|
| 991 |
{
|
| 992 |
+
"epoch": 0.7878446820483962,
|
| 993 |
+
"grad_norm": 0.3958778381347656,
|
| 994 |
+
"learning_rate": 2.6099108277934103e-05,
|
| 995 |
+
"loss": 0.2483,
|
| 996 |
"step": 700
|
| 997 |
},
|
| 998 |
{
|
| 999 |
+
"epoch": 0.7934721440630276,
|
| 1000 |
+
"grad_norm": 0.4310711920261383,
|
| 1001 |
+
"learning_rate": 2.4789020500930095e-05,
|
| 1002 |
+
"loss": 0.179,
|
| 1003 |
"step": 705
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
+
"epoch": 0.799099606077659,
|
| 1007 |
+
"grad_norm": 1.0834096670150757,
|
| 1008 |
+
"learning_rate": 2.3508000610193258e-05,
|
| 1009 |
+
"loss": 0.3142,
|
| 1010 |
"step": 710
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
+
"epoch": 0.8047270680922903,
|
| 1014 |
+
"grad_norm": 1.0267940759658813,
|
| 1015 |
+
"learning_rate": 2.2256543700199685e-05,
|
| 1016 |
+
"loss": 0.2546,
|
| 1017 |
"step": 715
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
+
"epoch": 0.8103545301069218,
|
| 1021 |
+
"grad_norm": 0.2950025498867035,
|
| 1022 |
+
"learning_rate": 2.1035133439788236e-05,
|
| 1023 |
+
"loss": 0.151,
|
| 1024 |
"step": 720
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"epoch": 0.8159819921215532,
|
| 1028 |
+
"grad_norm": 0.3759666085243225,
|
| 1029 |
+
"learning_rate": 1.9844241885230163e-05,
|
| 1030 |
+
"loss": 0.2239,
|
| 1031 |
"step": 725
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
+
"epoch": 0.8216094541361846,
|
| 1035 |
+
"grad_norm": 0.34732529520988464,
|
| 1036 |
+
"learning_rate": 1.8684329297786453e-05,
|
| 1037 |
+
"loss": 0.2362,
|
| 1038 |
"step": 730
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
+
"epoch": 0.827236916150816,
|
| 1042 |
+
"grad_norm": 0.2747817635536194,
|
| 1043 |
+
"learning_rate": 1.7555843965823992e-05,
|
| 1044 |
+
"loss": 0.2227,
|
| 1045 |
"step": 735
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
+
"epoch": 0.8328643781654473,
|
| 1049 |
+
"grad_norm": 0.5855769515037537,
|
| 1050 |
+
"learning_rate": 1.6459222031558974e-05,
|
| 1051 |
+
"loss": 0.2677,
|
| 1052 |
"step": 740
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
+
"epoch": 0.8384918401800788,
|
| 1056 |
+
"grad_norm": 0.49194827675819397,
|
| 1057 |
+
"learning_rate": 1.5394887322494732e-05,
|
| 1058 |
+
"loss": 0.2306,
|
| 1059 |
"step": 745
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 0.8441193021947102,
|
| 1063 |
+
"grad_norm": 0.475304514169693,
|
| 1064 |
+
"learning_rate": 1.4363251187618854e-05,
|
| 1065 |
+
"loss": 0.3053,
|
| 1066 |
"step": 750
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
+
"epoch": 0.8497467642093416,
|
| 1070 |
+
"grad_norm": 0.37457942962646484,
|
| 1071 |
+
"learning_rate": 1.3364712338423214e-05,
|
| 1072 |
+
"loss": 0.1884,
|
| 1073 |
"step": 755
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
+
"epoch": 0.855374226223973,
|
| 1077 |
+
"grad_norm": 0.4119722247123718,
|
| 1078 |
+
"learning_rate": 1.2399656694807971e-05,
|
| 1079 |
+
"loss": 0.3746,
|
| 1080 |
"step": 760
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
+
"epoch": 0.8610016882386043,
|
| 1084 |
+
"grad_norm": 0.5167866945266724,
|
| 1085 |
+
"learning_rate": 1.1468457235929597e-05,
|
| 1086 |
+
"loss": 0.2426,
|
| 1087 |
"step": 765
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
+
"epoch": 0.8666291502532358,
|
| 1091 |
+
"grad_norm": 0.41178828477859497,
|
| 1092 |
+
"learning_rate": 1.0571473856050107e-05,
|
| 1093 |
+
"loss": 0.2886,
|
| 1094 |
"step": 770
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
+
"epoch": 0.8722566122678672,
|
| 1098 |
+
"grad_norm": 0.43694284558296204,
|
| 1099 |
+
"learning_rate": 9.709053225443487e-06,
|
| 1100 |
+
"loss": 0.2708,
|
| 1101 |
"step": 775
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 0.8778840742824986,
|
| 1105 |
+
"grad_norm": 0.22772341966629028,
|
| 1106 |
+
"learning_rate": 8.881528656412963e-06,
|
| 1107 |
+
"loss": 0.1469,
|
| 1108 |
"step": 780
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
+
"epoch": 0.88351153629713,
|
| 1112 |
+
"grad_norm": 0.38090091943740845,
|
| 1113 |
+
"learning_rate": 8.08921997447094e-06,
|
| 1114 |
+
"loss": 0.3346,
|
| 1115 |
"step": 785
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
+
"epoch": 0.8891389983117614,
|
| 1119 |
+
"grad_norm": 0.3263999819755554,
|
| 1120 |
+
"learning_rate": 7.332433394731331e-06,
|
| 1121 |
+
"loss": 0.2386,
|
| 1122 |
"step": 790
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
+
"epoch": 0.8947664603263928,
|
| 1126 |
+
"grad_norm": 0.30097389221191406,
|
| 1127 |
+
"learning_rate": 6.611461403562147e-06,
|
| 1128 |
+
"loss": 0.1905,
|
| 1129 |
"step": 795
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
+
"epoch": 0.9003939223410242,
|
| 1133 |
+
"grad_norm": 0.41428399085998535,
|
| 1134 |
+
"learning_rate": 5.92658264554401e-06,
|
| 1135 |
+
"loss": 0.2399,
|
| 1136 |
"step": 800
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"epoch": 0.9060213843556556,
|
| 1140 |
+
"grad_norm": 0.4272648096084595,
|
| 1141 |
+
"learning_rate": 5.278061815778313e-06,
|
| 1142 |
+
"loss": 0.195,
|
| 1143 |
"step": 805
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
+
"epoch": 0.911648846370287,
|
| 1147 |
+
"grad_norm": 0.44488322734832764,
|
| 1148 |
+
"learning_rate": 4.666149557586697e-06,
|
| 1149 |
+
"loss": 0.2342,
|
| 1150 |
"step": 810
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
+
"epoch": 0.9172763083849184,
|
| 1154 |
+
"grad_norm": 0.3607059717178345,
|
| 1155 |
+
"learning_rate": 4.091082365641085e-06,
|
| 1156 |
+
"loss": 0.2702,
|
| 1157 |
"step": 815
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
+
"epoch": 0.9229037703995498,
|
| 1161 |
+
"grad_norm": 0.30268949270248413,
|
| 1162 |
+
"learning_rate": 3.5530824945623542e-06,
|
| 1163 |
+
"loss": 0.2509,
|
| 1164 |
"step": 820
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
+
"epoch": 0.9285312324141812,
|
| 1168 |
+
"grad_norm": 0.478462815284729,
|
| 1169 |
+
"learning_rate": 3.0523578730221713e-06,
|
| 1170 |
+
"loss": 0.2529,
|
| 1171 |
"step": 825
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
+
"epoch": 0.9341586944288126,
|
| 1175 |
+
"grad_norm": 0.36823543906211853,
|
| 1176 |
+
"learning_rate": 2.589102023381895e-06,
|
| 1177 |
+
"loss": 0.2497,
|
| 1178 |
"step": 830
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
+
"epoch": 0.939786156443444,
|
| 1182 |
+
"grad_norm": 0.6508424878120422,
|
| 1183 |
+
"learning_rate": 2.1634939868990235e-06,
|
| 1184 |
+
"loss": 0.4272,
|
| 1185 |
"step": 835
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 0.9454136184580754,
|
| 1189 |
+
"grad_norm": 0.4297642409801483,
|
| 1190 |
+
"learning_rate": 1.7756982545306443e-06,
|
| 1191 |
+
"loss": 0.2261,
|
| 1192 |
"step": 840
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
+
"epoch": 0.9510410804727069,
|
| 1196 |
+
"grad_norm": 0.2725857198238373,
|
| 1197 |
+
"learning_rate": 1.4258647033601024e-06,
|
| 1198 |
+
"loss": 0.1526,
|
| 1199 |
"step": 845
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
+
"epoch": 0.9566685424873382,
|
| 1203 |
+
"grad_norm": 0.366515189409256,
|
| 1204 |
+
"learning_rate": 1.1141285386718437e-06,
|
| 1205 |
+
"loss": 0.2365,
|
| 1206 |
"step": 850
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
+
"epoch": 0.9622960045019696,
|
| 1210 |
+
"grad_norm": 0.4519096910953522,
|
| 1211 |
+
"learning_rate": 8.406102416967043e-07,
|
| 1212 |
+
"loss": 0.2357,
|
| 1213 |
"step": 855
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
+
"epoch": 0.967923466516601,
|
| 1217 |
+
"grad_norm": 0.36480775475502014,
|
| 1218 |
+
"learning_rate": 6.054155230476699e-07,
|
| 1219 |
+
"loss": 0.2632,
|
| 1220 |
"step": 860
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
+
"epoch": 0.9735509285312324,
|
| 1224 |
+
"grad_norm": 0.29836106300354004,
|
| 1225 |
+
"learning_rate": 4.0863528186445564e-07,
|
| 1226 |
+
"loss": 0.2645,
|
| 1227 |
"step": 865
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
+
"epoch": 0.9791783905458639,
|
| 1231 |
+
"grad_norm": 0.33390042185783386,
|
| 1232 |
+
"learning_rate": 2.50345570682331e-07,
|
| 1233 |
+
"loss": 0.2836,
|
| 1234 |
"step": 870
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
+
"epoch": 0.9848058525604952,
|
| 1238 |
+
"grad_norm": 0.5138782262802124,
|
| 1239 |
+
"learning_rate": 1.3060756603897605e-07,
|
| 1240 |
+
"loss": 0.2035,
|
| 1241 |
"step": 875
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
+
"epoch": 0.9904333145751266,
|
| 1245 |
+
"grad_norm": 0.47627323865890503,
|
| 1246 |
+
"learning_rate": 4.946754483071692e-08,
|
| 1247 |
+
"loss": 0.2365,
|
| 1248 |
"step": 880
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
+
"epoch": 0.996060776589758,
|
| 1252 |
+
"grad_norm": 0.2622200548648834,
|
| 1253 |
+
"learning_rate": 6.95686642719906e-09,
|
| 1254 |
+
"loss": 0.1815,
|
| 1255 |
"step": 885
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
+
"epoch": 0.9994372537985369,
|
| 1259 |
+
"step": 888,
|
| 1260 |
+
"total_flos": 8.639715708407644e+17,
|
| 1261 |
+
"train_loss": 0.2980908542066007,
|
| 1262 |
+
"train_runtime": 5659.1413,
|
| 1263 |
+
"train_samples_per_second": 2.512,
|
| 1264 |
+
"train_steps_per_second": 0.157
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1265 |
}
|
| 1266 |
],
|
| 1267 |
"logging_steps": 5,
|
| 1268 |
+
"max_steps": 888,
|
| 1269 |
"num_input_tokens_seen": 0,
|
| 1270 |
"num_train_epochs": 1,
|
| 1271 |
"save_steps": 50,
|
|
|
|
| 1281 |
"attributes": {}
|
| 1282 |
}
|
| 1283 |
},
|
| 1284 |
+
"total_flos": 8.639715708407644e+17,
|
| 1285 |
"train_batch_size": 4,
|
| 1286 |
"trial_name": null,
|
| 1287 |
"trial_params": null
|