Training in progress, epoch 1
Browse files- all_results.json +6 -6
- confusion_matrix.jpg +0 -0
- model.safetensors +1 -1
- runs/Aug10_19-30-42_prod3/events.out.tfevents.1754826715.prod3.154612.1 +3 -0
- runs/Aug10_21-42-01_prod3/events.out.tfevents.1754833329.prod3.218391.0 +3 -0
- test_results.json +6 -6
- trainer_state.json +630 -469
- training_args.bin +1 -1
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 9.
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 1.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 14.
|
| 7 |
-
"eval_steps_per_second": 3.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.1,
|
| 3 |
+
"eval_accuracy": 0.7073170731707317,
|
| 4 |
+
"eval_loss": 1.3381569385528564,
|
| 5 |
+
"eval_runtime": 5.7056,
|
| 6 |
+
"eval_samples_per_second": 14.372,
|
| 7 |
+
"eval_steps_per_second": 3.681
|
| 8 |
}
|
confusion_matrix.jpg
CHANGED
|
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 344952716
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9513b66cfc09ace925efe252e9955ed06ee40f437c8f1975e1f56f053b7b739
|
| 3 |
size 344952716
|
runs/Aug10_19-30-42_prod3/events.out.tfevents.1754826715.prod3.154612.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44df82968bde756362af43cef07a927835f52990371356d5e99a6051a78cf1e8
|
| 3 |
+
size 411
|
runs/Aug10_21-42-01_prod3/events.out.tfevents.1754833329.prod3.218391.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:57f3c1c66be807e038821ee95bd7a098f743743e83992208cb90f4c55d609b36
|
| 3 |
+
size 10884
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 9.
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 1.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 14.
|
| 7 |
-
"eval_steps_per_second": 3.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.1,
|
| 3 |
+
"eval_accuracy": 0.7073170731707317,
|
| 4 |
+
"eval_loss": 1.3381569385528564,
|
| 5 |
+
"eval_runtime": 5.7056,
|
| 6 |
+
"eval_samples_per_second": 14.372,
|
| 7 |
+
"eval_steps_per_second": 3.681
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,769 +1,930 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint": "./Models/Matcha_clips_224_fintuned_4/checkpoint-
|
| 5 |
-
"epoch": 9.
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss": 1.
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss": 1.
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 1.
|
| 30 |
-
"loss": 1.
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm": 7.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 1.
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 2.
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 1.
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 1.
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 1.
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 1.
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"
|
| 78 |
-
"
|
| 79 |
-
"
|
| 80 |
-
"eval_samples_per_second": 13.518,
|
| 81 |
-
"eval_steps_per_second": 3.476,
|
| 82 |
-
"step": 93
|
| 83 |
-
},
|
| 84 |
-
{
|
| 85 |
-
"epoch": 1.007608695652174,
|
| 86 |
-
"grad_norm": 14.384367942810059,
|
| 87 |
-
"learning_rate": 4.957729468599034e-05,
|
| 88 |
-
"loss": 1.1631,
|
| 89 |
"step": 100
|
| 90 |
},
|
| 91 |
{
|
| 92 |
-
"epoch":
|
| 93 |
-
"grad_norm":
|
| 94 |
-
"learning_rate": 4.
|
| 95 |
-
"loss": 1.
|
| 96 |
"step": 110
|
| 97 |
},
|
| 98 |
{
|
| 99 |
-
"epoch": 1
|
| 100 |
-
"
|
| 101 |
-
"
|
| 102 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
"step": 120
|
| 104 |
},
|
| 105 |
{
|
| 106 |
-
"epoch": 1.
|
| 107 |
-
"grad_norm": 10.
|
| 108 |
-
"learning_rate": 4.
|
| 109 |
-
"loss":
|
| 110 |
"step": 130
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch": 1.
|
| 114 |
-
"grad_norm":
|
| 115 |
-
"learning_rate": 4.
|
| 116 |
-
"loss":
|
| 117 |
"step": 140
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 1.
|
| 121 |
-
"grad_norm":
|
| 122 |
-
"learning_rate": 4.
|
| 123 |
-
"loss": 0.
|
| 124 |
"step": 150
|
| 125 |
},
|
| 126 |
{
|
| 127 |
-
"epoch": 1.
|
| 128 |
-
"grad_norm":
|
| 129 |
-
"learning_rate": 4.
|
| 130 |
-
"loss":
|
| 131 |
"step": 160
|
| 132 |
},
|
| 133 |
{
|
| 134 |
-
"epoch": 1.
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 4.
|
| 137 |
-
"loss": 0.
|
| 138 |
"step": 170
|
| 139 |
},
|
| 140 |
{
|
| 141 |
-
"epoch": 1.
|
| 142 |
-
"grad_norm":
|
| 143 |
-
"learning_rate": 4.
|
| 144 |
-
"loss": 0.
|
| 145 |
"step": 180
|
| 146 |
},
|
| 147 |
{
|
| 148 |
-
"epoch": 1.
|
| 149 |
-
"
|
| 150 |
-
"
|
| 151 |
-
"
|
| 152 |
-
"eval_samples_per_second": 13.854,
|
| 153 |
-
"eval_steps_per_second": 3.563,
|
| 154 |
-
"step": 186
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 2.0043478260869567,
|
| 158 |
-
"grad_norm": 10.325079917907715,
|
| 159 |
-
"learning_rate": 4.414251207729469e-05,
|
| 160 |
-
"loss": 0.5439,
|
| 161 |
"step": 190
|
| 162 |
},
|
| 163 |
{
|
| 164 |
-
"epoch":
|
| 165 |
-
"grad_norm":
|
| 166 |
-
"learning_rate": 4.
|
| 167 |
-
"loss": 0.
|
| 168 |
"step": 200
|
| 169 |
},
|
| 170 |
{
|
| 171 |
-
"epoch":
|
| 172 |
-
"grad_norm":
|
| 173 |
-
"learning_rate": 4.
|
| 174 |
-
"loss":
|
| 175 |
"step": 210
|
| 176 |
},
|
| 177 |
{
|
| 178 |
-
"epoch":
|
| 179 |
-
"grad_norm":
|
| 180 |
-
"learning_rate": 4.
|
| 181 |
-
"loss": 0.
|
| 182 |
"step": 220
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch":
|
| 186 |
-
"grad_norm":
|
| 187 |
-
"learning_rate": 4.
|
| 188 |
-
"loss": 0.
|
| 189 |
"step": 230
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch":
|
| 193 |
-
"
|
| 194 |
-
"
|
| 195 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 196 |
"step": 240
|
| 197 |
},
|
| 198 |
{
|
| 199 |
-
"epoch": 2.
|
| 200 |
-
"grad_norm":
|
| 201 |
-
"learning_rate": 4.
|
| 202 |
-
"loss": 0.
|
| 203 |
"step": 250
|
| 204 |
},
|
| 205 |
{
|
| 206 |
-
"epoch": 2.
|
| 207 |
-
"grad_norm":
|
| 208 |
-
"learning_rate":
|
| 209 |
-
"loss": 0.
|
| 210 |
"step": 260
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"epoch": 2.
|
| 214 |
-
"grad_norm":
|
| 215 |
-
"learning_rate":
|
| 216 |
-
"loss": 0.
|
| 217 |
"step": 270
|
| 218 |
},
|
| 219 |
{
|
| 220 |
-
"epoch": 2.
|
| 221 |
-
"
|
| 222 |
-
"
|
| 223 |
-
"
|
| 224 |
-
"eval_samples_per_second": 13.437,
|
| 225 |
-
"eval_steps_per_second": 3.455,
|
| 226 |
-
"step": 279
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 3.001086956521739,
|
| 230 |
-
"grad_norm": 2.6006569862365723,
|
| 231 |
-
"learning_rate": 3.870772946859904e-05,
|
| 232 |
-
"loss": 0.226,
|
| 233 |
"step": 280
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch":
|
| 237 |
-
"grad_norm": 1.
|
| 238 |
-
"learning_rate":
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 290
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch":
|
| 244 |
-
"grad_norm":
|
| 245 |
-
"learning_rate":
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 300
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch":
|
| 251 |
-
"grad_norm":
|
| 252 |
-
"learning_rate":
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 310
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch":
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 320
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
-
"grad_norm":
|
| 266 |
-
"learning_rate": 3.
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 330
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch":
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate": 3.
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 340
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch":
|
| 279 |
-
"
|
| 280 |
-
"
|
| 281 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 282 |
"step": 350
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 3.
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
-
"learning_rate": 3.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 360
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 3.
|
| 293 |
-
"grad_norm":
|
| 294 |
-
"learning_rate": 3.
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 370
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 3.
|
| 300 |
-
"
|
| 301 |
-
"
|
| 302 |
-
"
|
| 303 |
-
"eval_samples_per_second": 13.236,
|
| 304 |
-
"eval_steps_per_second": 3.404,
|
| 305 |
-
"step": 372
|
| 306 |
-
},
|
| 307 |
-
{
|
| 308 |
-
"epoch": 4.008695652173913,
|
| 309 |
-
"grad_norm": 0.04769546538591385,
|
| 310 |
-
"learning_rate": 3.2669082125603866e-05,
|
| 311 |
-
"loss": 0.0041,
|
| 312 |
"step": 380
|
| 313 |
},
|
| 314 |
{
|
| 315 |
-
"epoch":
|
| 316 |
-
"grad_norm": 0.
|
| 317 |
-
"learning_rate": 3.
|
| 318 |
-
"loss": 0.
|
| 319 |
"step": 390
|
| 320 |
},
|
| 321 |
{
|
| 322 |
-
"epoch":
|
| 323 |
-
"grad_norm": 0.
|
| 324 |
-
"learning_rate": 3.
|
| 325 |
-
"loss": 0.
|
| 326 |
"step": 400
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch":
|
| 330 |
-
"grad_norm": 0.
|
| 331 |
-
"learning_rate": 3.
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 410
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch":
|
| 337 |
-
"grad_norm": 0.
|
| 338 |
-
"learning_rate": 3.
|
| 339 |
-
"loss": 0.
|
| 340 |
"step": 420
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"epoch":
|
| 344 |
-
"grad_norm": 0.
|
| 345 |
-
"learning_rate":
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 430
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch":
|
| 351 |
-
"grad_norm": 0.
|
| 352 |
-
"learning_rate":
|
| 353 |
-
"loss": 0.
|
| 354 |
"step": 440
|
| 355 |
},
|
| 356 |
{
|
| 357 |
-
"epoch":
|
| 358 |
-
"grad_norm": 0.
|
| 359 |
-
"learning_rate":
|
| 360 |
-
"loss": 0.
|
| 361 |
"step": 450
|
| 362 |
},
|
| 363 |
{
|
| 364 |
-
"epoch":
|
| 365 |
-
"grad_norm": 0.
|
| 366 |
-
"learning_rate":
|
| 367 |
-
"loss": 0.
|
| 368 |
"step": 460
|
| 369 |
},
|
| 370 |
{
|
| 371 |
-
"epoch":
|
| 372 |
-
"eval_accuracy": 0.
|
| 373 |
-
"eval_loss": 1.
|
| 374 |
-
"eval_runtime":
|
| 375 |
-
"eval_samples_per_second":
|
| 376 |
-
"eval_steps_per_second": 3.
|
| 377 |
-
"step":
|
| 378 |
},
|
| 379 |
{
|
| 380 |
-
"epoch":
|
| 381 |
-
"grad_norm":
|
| 382 |
-
"learning_rate":
|
| 383 |
-
"loss": 0.
|
| 384 |
"step": 470
|
| 385 |
},
|
| 386 |
{
|
| 387 |
-
"epoch":
|
| 388 |
-
"grad_norm":
|
| 389 |
-
"learning_rate":
|
| 390 |
-
"loss": 0.
|
| 391 |
"step": 480
|
| 392 |
},
|
| 393 |
{
|
| 394 |
-
"epoch":
|
| 395 |
-
"grad_norm": 0.
|
| 396 |
-
"learning_rate":
|
| 397 |
-
"loss": 0.
|
| 398 |
"step": 490
|
| 399 |
},
|
| 400 |
{
|
| 401 |
-
"epoch":
|
| 402 |
-
"grad_norm": 0.
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 500
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
-
"grad_norm": 0.
|
| 410 |
-
"learning_rate":
|
| 411 |
-
"loss": 0.
|
| 412 |
"step": 510
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"epoch":
|
| 416 |
-
"grad_norm":
|
| 417 |
-
"learning_rate":
|
| 418 |
-
"loss": 0.
|
| 419 |
"step": 520
|
| 420 |
},
|
| 421 |
{
|
| 422 |
-
"epoch":
|
| 423 |
-
"grad_norm": 0.
|
| 424 |
-
"learning_rate":
|
| 425 |
-
"loss": 0.
|
| 426 |
"step": 530
|
| 427 |
},
|
| 428 |
{
|
| 429 |
-
"epoch":
|
| 430 |
-
"grad_norm": 0.
|
| 431 |
-
"learning_rate": 2.
|
| 432 |
-
"loss": 0.
|
| 433 |
"step": 540
|
| 434 |
},
|
| 435 |
{
|
| 436 |
-
"epoch":
|
| 437 |
-
"grad_norm": 0.
|
| 438 |
-
"learning_rate": 2.
|
| 439 |
-
"loss": 0.
|
| 440 |
"step": 550
|
| 441 |
},
|
| 442 |
{
|
| 443 |
-
"epoch":
|
| 444 |
-
"
|
| 445 |
-
"
|
| 446 |
-
"
|
| 447 |
-
"eval_samples_per_second": 13.562,
|
| 448 |
-
"eval_steps_per_second": 3.487,
|
| 449 |
-
"step": 558
|
| 450 |
-
},
|
| 451 |
-
{
|
| 452 |
-
"epoch": 6.002173913043478,
|
| 453 |
-
"grad_norm": 0.017682170495390892,
|
| 454 |
-
"learning_rate": 2.1799516908212562e-05,
|
| 455 |
-
"loss": 0.0078,
|
| 456 |
"step": 560
|
| 457 |
},
|
| 458 |
{
|
| 459 |
-
"epoch":
|
| 460 |
-
"grad_norm": 0.
|
| 461 |
-
"learning_rate": 2.
|
| 462 |
-
"loss": 0.
|
| 463 |
"step": 570
|
| 464 |
},
|
| 465 |
{
|
| 466 |
-
"epoch":
|
| 467 |
-
"
|
| 468 |
-
"
|
| 469 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 470 |
"step": 580
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch":
|
| 474 |
-
"grad_norm": 0.
|
| 475 |
-
"learning_rate":
|
| 476 |
-
"loss": 0.
|
| 477 |
"step": 590
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
-
"grad_norm": 0.
|
| 482 |
-
"learning_rate":
|
| 483 |
-
"loss": 0.
|
| 484 |
"step": 600
|
| 485 |
},
|
| 486 |
{
|
| 487 |
-
"epoch":
|
| 488 |
-
"grad_norm": 0.
|
| 489 |
-
"learning_rate":
|
| 490 |
-
"loss": 0.
|
| 491 |
"step": 610
|
| 492 |
},
|
| 493 |
{
|
| 494 |
-
"epoch":
|
| 495 |
-
"grad_norm": 0.
|
| 496 |
-
"learning_rate":
|
| 497 |
-
"loss": 0.
|
| 498 |
"step": 620
|
| 499 |
},
|
| 500 |
{
|
| 501 |
-
"epoch":
|
| 502 |
-
"grad_norm": 0.
|
| 503 |
-
"learning_rate":
|
| 504 |
-
"loss": 0.
|
| 505 |
"step": 630
|
| 506 |
},
|
| 507 |
{
|
| 508 |
-
"epoch":
|
| 509 |
-
"grad_norm": 0.
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss": 0.
|
| 512 |
"step": 640
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch":
|
| 516 |
-
"grad_norm": 0.
|
| 517 |
-
"learning_rate":
|
| 518 |
-
"loss": 0.
|
| 519 |
"step": 650
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"
|
| 524 |
-
"
|
| 525 |
-
"
|
| 526 |
-
"eval_samples_per_second": 13.425,
|
| 527 |
-
"eval_steps_per_second": 3.452,
|
| 528 |
-
"step": 651
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 7.0097826086956525,
|
| 532 |
-
"grad_norm": 0.013392376713454723,
|
| 533 |
-
"learning_rate": 1.5760869565217393e-05,
|
| 534 |
-
"loss": 0.0009,
|
| 535 |
"step": 660
|
| 536 |
},
|
| 537 |
{
|
| 538 |
-
"epoch":
|
| 539 |
-
"grad_norm": 0.
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 670
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch":
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 680
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch":
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
-
"learning_rate":
|
| 555 |
-
"loss": 0.
|
| 556 |
"step": 690
|
| 557 |
},
|
| 558 |
{
|
| 559 |
-
"epoch":
|
| 560 |
-
"
|
| 561 |
-
"
|
| 562 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 563 |
"step": 700
|
| 564 |
},
|
| 565 |
{
|
| 566 |
-
"epoch":
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
-
"learning_rate":
|
| 569 |
-
"loss": 0.
|
| 570 |
"step": 710
|
| 571 |
},
|
| 572 |
{
|
| 573 |
-
"epoch":
|
| 574 |
-
"grad_norm": 0.
|
| 575 |
-
"learning_rate":
|
| 576 |
-
"loss": 0.
|
| 577 |
"step": 720
|
| 578 |
},
|
| 579 |
{
|
| 580 |
-
"epoch":
|
| 581 |
-
"grad_norm": 0.
|
| 582 |
-
"learning_rate":
|
| 583 |
-
"loss": 0.
|
| 584 |
"step": 730
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch":
|
| 588 |
-
"grad_norm": 0.
|
| 589 |
-
"learning_rate": 1.
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 740
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
-
"
|
| 596 |
-
"
|
| 597 |
-
"
|
| 598 |
-
"eval_samples_per_second": 13.669,
|
| 599 |
-
"eval_steps_per_second": 3.515,
|
| 600 |
-
"step": 744
|
| 601 |
-
},
|
| 602 |
-
{
|
| 603 |
-
"epoch": 8.006521739130434,
|
| 604 |
-
"grad_norm": 0.016901936382055283,
|
| 605 |
-
"learning_rate": 1.032608695652174e-05,
|
| 606 |
-
"loss": 0.0008,
|
| 607 |
"step": 750
|
| 608 |
},
|
| 609 |
{
|
| 610 |
-
"epoch":
|
| 611 |
-
"grad_norm": 0.
|
| 612 |
-
"learning_rate":
|
| 613 |
-
"loss": 0.
|
| 614 |
"step": 760
|
| 615 |
},
|
| 616 |
{
|
| 617 |
-
"epoch":
|
| 618 |
-
"grad_norm": 0.
|
| 619 |
-
"learning_rate":
|
| 620 |
-
"loss": 0.
|
| 621 |
"step": 770
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch":
|
| 625 |
-
"grad_norm": 0.
|
| 626 |
-
"learning_rate":
|
| 627 |
-
"loss": 0.
|
| 628 |
"step": 780
|
| 629 |
},
|
| 630 |
{
|
| 631 |
-
"epoch":
|
| 632 |
-
"grad_norm": 0.
|
| 633 |
-
"learning_rate":
|
| 634 |
-
"loss": 0.
|
| 635 |
"step": 790
|
| 636 |
},
|
| 637 |
{
|
| 638 |
-
"epoch":
|
| 639 |
-
"grad_norm": 0.
|
| 640 |
-
"learning_rate":
|
| 641 |
-
"loss": 0.
|
| 642 |
"step": 800
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"epoch":
|
| 646 |
-
"
|
| 647 |
-
"
|
| 648 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 649 |
"step": 810
|
| 650 |
},
|
| 651 |
{
|
| 652 |
-
"epoch":
|
| 653 |
-
"grad_norm": 0.
|
| 654 |
-
"learning_rate":
|
| 655 |
"loss": 0.0009,
|
| 656 |
"step": 820
|
| 657 |
},
|
| 658 |
{
|
| 659 |
-
"epoch":
|
| 660 |
-
"grad_norm": 0.
|
| 661 |
-
"learning_rate":
|
| 662 |
-
"loss": 0.
|
| 663 |
"step": 830
|
| 664 |
},
|
| 665 |
{
|
| 666 |
-
"epoch":
|
| 667 |
-
"
|
| 668 |
-
"
|
| 669 |
-
"
|
| 670 |
-
"eval_samples_per_second": 13.549,
|
| 671 |
-
"eval_steps_per_second": 3.484,
|
| 672 |
-
"step": 837
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 9.003260869565217,
|
| 676 |
-
"grad_norm": 0.009687897749245167,
|
| 677 |
-
"learning_rate": 4.891304347826087e-06,
|
| 678 |
-
"loss": 0.0006,
|
| 679 |
"step": 840
|
| 680 |
},
|
| 681 |
{
|
| 682 |
-
"epoch":
|
| 683 |
-
"grad_norm": 0.
|
| 684 |
-
"learning_rate":
|
| 685 |
"loss": 0.0008,
|
| 686 |
"step": 850
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch":
|
| 690 |
-
"grad_norm": 0.
|
| 691 |
-
"learning_rate":
|
| 692 |
"loss": 0.0008,
|
| 693 |
"step": 860
|
| 694 |
},
|
| 695 |
{
|
| 696 |
-
"epoch":
|
| 697 |
-
"grad_norm": 0.
|
| 698 |
-
"learning_rate":
|
| 699 |
-
"loss": 0.
|
| 700 |
"step": 870
|
| 701 |
},
|
| 702 |
{
|
| 703 |
-
"epoch":
|
| 704 |
-
"grad_norm": 0.
|
| 705 |
-
"learning_rate":
|
| 706 |
-
"loss": 0.
|
| 707 |
"step": 880
|
| 708 |
},
|
| 709 |
{
|
| 710 |
-
"epoch":
|
| 711 |
-
"grad_norm": 0.
|
| 712 |
-
"learning_rate": 1.
|
| 713 |
-
"loss": 0.
|
| 714 |
"step": 890
|
| 715 |
},
|
| 716 |
{
|
| 717 |
-
"epoch":
|
| 718 |
-
"grad_norm": 0.
|
| 719 |
-
"learning_rate": 1.
|
| 720 |
-
"loss": 0.
|
| 721 |
"step": 900
|
| 722 |
},
|
| 723 |
{
|
| 724 |
-
"epoch":
|
| 725 |
-
"grad_norm": 0.
|
| 726 |
-
"learning_rate":
|
| 727 |
"loss": 0.0007,
|
| 728 |
"step": 910
|
| 729 |
},
|
| 730 |
{
|
| 731 |
-
"epoch":
|
| 732 |
-
"grad_norm": 0.
|
| 733 |
-
"learning_rate":
|
| 734 |
"loss": 0.0008,
|
| 735 |
"step": 920
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch":
|
| 739 |
-
"eval_accuracy": 0.
|
| 740 |
-
"eval_loss": 1.
|
| 741 |
-
"eval_runtime":
|
| 742 |
-
"eval_samples_per_second":
|
| 743 |
-
"eval_steps_per_second": 3.
|
| 744 |
"step": 920
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch":
|
| 748 |
-
"
|
| 749 |
-
"
|
| 750 |
-
"
|
| 751 |
-
"
|
| 752 |
-
"train_samples_per_second": 3.799,
|
| 753 |
-
"train_steps_per_second": 0.95
|
| 754 |
},
|
| 755 |
{
|
| 756 |
-
"epoch":
|
| 757 |
-
"
|
| 758 |
-
"
|
| 759 |
-
"
|
| 760 |
-
"
|
| 761 |
-
|
| 762 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 763 |
}
|
| 764 |
],
|
| 765 |
"logging_steps": 10,
|
| 766 |
-
"max_steps":
|
| 767 |
"num_input_tokens_seen": 0,
|
| 768 |
"num_train_epochs": 9223372036854775807,
|
| 769 |
"save_steps": 500,
|
|
@@ -779,7 +940,7 @@
|
|
| 779 |
"attributes": {}
|
| 780 |
}
|
| 781 |
},
|
| 782 |
-
"total_flos":
|
| 783 |
"train_batch_size": 4,
|
| 784 |
"trial_name": null,
|
| 785 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 575,
|
| 3 |
+
"best_metric": 0.7073170731707317,
|
| 4 |
+
"best_model_checkpoint": "./Models/Matcha_clips_224_fintuned_4/checkpoint-575",
|
| 5 |
+
"epoch": 9.1,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1150,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.008695652173913044,
|
| 14 |
+
"grad_norm": 6.172690391540527,
|
| 15 |
+
"learning_rate": 3.91304347826087e-06,
|
| 16 |
+
"loss": 1.9556,
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.017391304347826087,
|
| 21 |
+
"grad_norm": 5.285815238952637,
|
| 22 |
+
"learning_rate": 8.26086956521739e-06,
|
| 23 |
+
"loss": 1.9446,
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.02608695652173913,
|
| 28 |
+
"grad_norm": 7.100526809692383,
|
| 29 |
+
"learning_rate": 1.2608695652173914e-05,
|
| 30 |
+
"loss": 1.8525,
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.034782608695652174,
|
| 35 |
+
"grad_norm": 7.057590961456299,
|
| 36 |
+
"learning_rate": 1.6956521739130433e-05,
|
| 37 |
+
"loss": 1.8414,
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.043478260869565216,
|
| 42 |
+
"grad_norm": 6.360028266906738,
|
| 43 |
+
"learning_rate": 2.1304347826086958e-05,
|
| 44 |
+
"loss": 1.6835,
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.05217391304347826,
|
| 49 |
+
"grad_norm": 6.762235164642334,
|
| 50 |
+
"learning_rate": 2.5652173913043483e-05,
|
| 51 |
+
"loss": 1.8069,
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.06086956521739131,
|
| 56 |
+
"grad_norm": 9.110674858093262,
|
| 57 |
+
"learning_rate": 3e-05,
|
| 58 |
+
"loss": 1.6742,
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.06956521739130435,
|
| 63 |
+
"grad_norm": 5.7140631675720215,
|
| 64 |
+
"learning_rate": 3.4347826086956526e-05,
|
| 65 |
+
"loss": 1.675,
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.0782608695652174,
|
| 70 |
+
"grad_norm": 7.441013336181641,
|
| 71 |
+
"learning_rate": 3.869565217391305e-05,
|
| 72 |
+
"loss": 1.7402,
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.08695652173913043,
|
| 77 |
+
"grad_norm": 8.591211318969727,
|
| 78 |
+
"learning_rate": 4.304347826086957e-05,
|
| 79 |
+
"loss": 1.6256,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.09565217391304348,
|
| 84 |
+
"grad_norm": 11.78553581237793,
|
| 85 |
+
"learning_rate": 4.739130434782609e-05,
|
| 86 |
+
"loss": 1.4198,
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.1,
|
| 91 |
+
"eval_accuracy": 0.4634146341463415,
|
| 92 |
+
"eval_loss": 1.4963948726654053,
|
| 93 |
+
"eval_runtime": 6.2391,
|
| 94 |
+
"eval_samples_per_second": 13.143,
|
| 95 |
+
"eval_steps_per_second": 3.366,
|
| 96 |
+
"step": 115
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.0043478260869565,
|
| 100 |
+
"grad_norm": 9.049060821533203,
|
| 101 |
+
"learning_rate": 4.980676328502415e-05,
|
| 102 |
+
"loss": 1.3079,
|
| 103 |
"step": 120
|
| 104 |
},
|
| 105 |
{
|
| 106 |
+
"epoch": 1.0130434782608695,
|
| 107 |
+
"grad_norm": 10.772295951843262,
|
| 108 |
+
"learning_rate": 4.932367149758454e-05,
|
| 109 |
+
"loss": 1.0031,
|
| 110 |
"step": 130
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 1.0217391304347827,
|
| 114 |
+
"grad_norm": 20.79416275024414,
|
| 115 |
+
"learning_rate": 4.884057971014493e-05,
|
| 116 |
+
"loss": 0.9924,
|
| 117 |
"step": 140
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 1.0304347826086957,
|
| 121 |
+
"grad_norm": 6.840089321136475,
|
| 122 |
+
"learning_rate": 4.8357487922705316e-05,
|
| 123 |
+
"loss": 0.7422,
|
| 124 |
"step": 150
|
| 125 |
},
|
| 126 |
{
|
| 127 |
+
"epoch": 1.0391304347826087,
|
| 128 |
+
"grad_norm": 9.69269847869873,
|
| 129 |
+
"learning_rate": 4.7874396135265706e-05,
|
| 130 |
+
"loss": 1.0388,
|
| 131 |
"step": 160
|
| 132 |
},
|
| 133 |
{
|
| 134 |
+
"epoch": 1.0478260869565217,
|
| 135 |
+
"grad_norm": 9.885075569152832,
|
| 136 |
+
"learning_rate": 4.739130434782609e-05,
|
| 137 |
+
"loss": 0.9266,
|
| 138 |
"step": 170
|
| 139 |
},
|
| 140 |
{
|
| 141 |
+
"epoch": 1.0565217391304347,
|
| 142 |
+
"grad_norm": 16.479467391967773,
|
| 143 |
+
"learning_rate": 4.690821256038648e-05,
|
| 144 |
+
"loss": 0.8315,
|
| 145 |
"step": 180
|
| 146 |
},
|
| 147 |
{
|
| 148 |
+
"epoch": 1.065217391304348,
|
| 149 |
+
"grad_norm": 22.804908752441406,
|
| 150 |
+
"learning_rate": 4.642512077294686e-05,
|
| 151 |
+
"loss": 0.736,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 152 |
"step": 190
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 1.0739130434782609,
|
| 156 |
+
"grad_norm": 12.778817176818848,
|
| 157 |
+
"learning_rate": 4.594202898550725e-05,
|
| 158 |
+
"loss": 0.8893,
|
| 159 |
"step": 200
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 1.0826086956521739,
|
| 163 |
+
"grad_norm": 18.733671188354492,
|
| 164 |
+
"learning_rate": 4.545893719806764e-05,
|
| 165 |
+
"loss": 1.3588,
|
| 166 |
"step": 210
|
| 167 |
},
|
| 168 |
{
|
| 169 |
+
"epoch": 1.0913043478260869,
|
| 170 |
+
"grad_norm": 19.088119506835938,
|
| 171 |
+
"learning_rate": 4.497584541062802e-05,
|
| 172 |
+
"loss": 0.7065,
|
| 173 |
"step": 220
|
| 174 |
},
|
| 175 |
{
|
| 176 |
+
"epoch": 1.1,
|
| 177 |
+
"grad_norm": 10.45570182800293,
|
| 178 |
+
"learning_rate": 4.449275362318841e-05,
|
| 179 |
+
"loss": 0.6674,
|
| 180 |
"step": 230
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"epoch": 1.1,
|
| 184 |
+
"eval_accuracy": 0.6585365853658537,
|
| 185 |
+
"eval_loss": 1.0704329013824463,
|
| 186 |
+
"eval_runtime": 6.1407,
|
| 187 |
+
"eval_samples_per_second": 13.354,
|
| 188 |
+
"eval_steps_per_second": 3.42,
|
| 189 |
+
"step": 230
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 2.008695652173913,
|
| 193 |
+
"grad_norm": 7.579151153564453,
|
| 194 |
+
"learning_rate": 4.4009661835748794e-05,
|
| 195 |
+
"loss": 0.2858,
|
| 196 |
"step": 240
|
| 197 |
},
|
| 198 |
{
|
| 199 |
+
"epoch": 2.017391304347826,
|
| 200 |
+
"grad_norm": 9.255645751953125,
|
| 201 |
+
"learning_rate": 4.352657004830918e-05,
|
| 202 |
+
"loss": 0.5303,
|
| 203 |
"step": 250
|
| 204 |
},
|
| 205 |
{
|
| 206 |
+
"epoch": 2.026086956521739,
|
| 207 |
+
"grad_norm": 1.6530203819274902,
|
| 208 |
+
"learning_rate": 4.304347826086957e-05,
|
| 209 |
+
"loss": 0.2094,
|
| 210 |
"step": 260
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"epoch": 2.034782608695652,
|
| 214 |
+
"grad_norm": 13.71268367767334,
|
| 215 |
+
"learning_rate": 4.256038647342995e-05,
|
| 216 |
+
"loss": 0.2021,
|
| 217 |
"step": 270
|
| 218 |
},
|
| 219 |
{
|
| 220 |
+
"epoch": 2.0434782608695654,
|
| 221 |
+
"grad_norm": 2.5825889110565186,
|
| 222 |
+
"learning_rate": 4.207729468599034e-05,
|
| 223 |
+
"loss": 0.1329,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 224 |
"step": 280
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 2.0521739130434784,
|
| 228 |
+
"grad_norm": 1.7412315607070923,
|
| 229 |
+
"learning_rate": 4.1594202898550726e-05,
|
| 230 |
+
"loss": 0.1377,
|
| 231 |
"step": 290
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 2.0608695652173914,
|
| 235 |
+
"grad_norm": 18.217609405517578,
|
| 236 |
+
"learning_rate": 4.111111111111111e-05,
|
| 237 |
+
"loss": 0.2466,
|
| 238 |
"step": 300
|
| 239 |
},
|
| 240 |
{
|
| 241 |
+
"epoch": 2.0695652173913044,
|
| 242 |
+
"grad_norm": 7.633990287780762,
|
| 243 |
+
"learning_rate": 4.06280193236715e-05,
|
| 244 |
+
"loss": 0.3553,
|
| 245 |
"step": 310
|
| 246 |
},
|
| 247 |
{
|
| 248 |
+
"epoch": 2.0782608695652174,
|
| 249 |
+
"grad_norm": 19.997325897216797,
|
| 250 |
+
"learning_rate": 4.014492753623188e-05,
|
| 251 |
+
"loss": 0.3189,
|
| 252 |
"step": 320
|
| 253 |
},
|
| 254 |
{
|
| 255 |
+
"epoch": 2.0869565217391304,
|
| 256 |
+
"grad_norm": 1.3608593940734863,
|
| 257 |
+
"learning_rate": 3.966183574879227e-05,
|
| 258 |
+
"loss": 0.1477,
|
| 259 |
"step": 330
|
| 260 |
},
|
| 261 |
{
|
| 262 |
+
"epoch": 2.0956521739130434,
|
| 263 |
+
"grad_norm": 6.2525739669799805,
|
| 264 |
+
"learning_rate": 3.9178743961352657e-05,
|
| 265 |
+
"loss": 0.4521,
|
| 266 |
"step": 340
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 2.1,
|
| 270 |
+
"eval_accuracy": 0.6341463414634146,
|
| 271 |
+
"eval_loss": 0.9644566774368286,
|
| 272 |
+
"eval_runtime": 6.2182,
|
| 273 |
+
"eval_samples_per_second": 13.187,
|
| 274 |
+
"eval_steps_per_second": 3.377,
|
| 275 |
+
"step": 345
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 3.0043478260869567,
|
| 279 |
+
"grad_norm": 0.2062215656042099,
|
| 280 |
+
"learning_rate": 3.869565217391305e-05,
|
| 281 |
+
"loss": 0.0714,
|
| 282 |
"step": 350
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 3.0130434782608697,
|
| 286 |
+
"grad_norm": 0.07909725606441498,
|
| 287 |
+
"learning_rate": 3.821256038647344e-05,
|
| 288 |
+
"loss": 0.0455,
|
| 289 |
"step": 360
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 3.0217391304347827,
|
| 293 |
+
"grad_norm": 2.4341115951538086,
|
| 294 |
+
"learning_rate": 3.772946859903382e-05,
|
| 295 |
+
"loss": 0.0473,
|
| 296 |
"step": 370
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 3.0304347826086957,
|
| 300 |
+
"grad_norm": 0.42377185821533203,
|
| 301 |
+
"learning_rate": 3.7246376811594204e-05,
|
| 302 |
+
"loss": 0.111,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 303 |
"step": 380
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 3.0391304347826087,
|
| 307 |
+
"grad_norm": 0.19808898866176605,
|
| 308 |
+
"learning_rate": 3.6763285024154594e-05,
|
| 309 |
+
"loss": 0.0683,
|
| 310 |
"step": 390
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 3.0478260869565217,
|
| 314 |
+
"grad_norm": 0.8478032946586609,
|
| 315 |
+
"learning_rate": 3.628019323671498e-05,
|
| 316 |
+
"loss": 0.0441,
|
| 317 |
"step": 400
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 3.0565217391304347,
|
| 321 |
+
"grad_norm": 0.15200255811214447,
|
| 322 |
+
"learning_rate": 3.579710144927537e-05,
|
| 323 |
+
"loss": 0.0142,
|
| 324 |
"step": 410
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 3.0652173913043477,
|
| 328 |
+
"grad_norm": 0.1159711629152298,
|
| 329 |
+
"learning_rate": 3.531400966183575e-05,
|
| 330 |
+
"loss": 0.0203,
|
| 331 |
"step": 420
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 3.0739130434782607,
|
| 335 |
+
"grad_norm": 0.1384093165397644,
|
| 336 |
+
"learning_rate": 3.4830917874396135e-05,
|
| 337 |
+
"loss": 0.1054,
|
| 338 |
"step": 430
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 3.082608695652174,
|
| 342 |
+
"grad_norm": 0.5738388895988464,
|
| 343 |
+
"learning_rate": 3.4347826086956526e-05,
|
| 344 |
+
"loss": 0.0707,
|
| 345 |
"step": 440
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 3.091304347826087,
|
| 349 |
+
"grad_norm": 0.16567471623420715,
|
| 350 |
+
"learning_rate": 3.386473429951691e-05,
|
| 351 |
+
"loss": 0.1002,
|
| 352 |
"step": 450
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 3.1,
|
| 356 |
+
"grad_norm": 0.6657196283340454,
|
| 357 |
+
"learning_rate": 3.338164251207729e-05,
|
| 358 |
+
"loss": 0.1047,
|
| 359 |
"step": 460
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 3.1,
|
| 363 |
+
"eval_accuracy": 0.6463414634146342,
|
| 364 |
+
"eval_loss": 1.2479301691055298,
|
| 365 |
+
"eval_runtime": 6.3163,
|
| 366 |
+
"eval_samples_per_second": 12.982,
|
| 367 |
+
"eval_steps_per_second": 3.325,
|
| 368 |
+
"step": 460
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 4.008695652173913,
|
| 372 |
+
"grad_norm": 2.4936015605926514,
|
| 373 |
+
"learning_rate": 3.289855072463768e-05,
|
| 374 |
+
"loss": 0.0126,
|
| 375 |
"step": 470
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 4.017391304347826,
|
| 379 |
+
"grad_norm": 25.129255294799805,
|
| 380 |
+
"learning_rate": 3.2415458937198066e-05,
|
| 381 |
+
"loss": 0.0839,
|
| 382 |
"step": 480
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"epoch": 4.026086956521739,
|
| 386 |
+
"grad_norm": 0.05726484954357147,
|
| 387 |
+
"learning_rate": 3.1932367149758457e-05,
|
| 388 |
+
"loss": 0.0028,
|
| 389 |
"step": 490
|
| 390 |
},
|
| 391 |
{
|
| 392 |
+
"epoch": 4.034782608695652,
|
| 393 |
+
"grad_norm": 0.034710388630628586,
|
| 394 |
+
"learning_rate": 3.144927536231884e-05,
|
| 395 |
+
"loss": 0.0035,
|
| 396 |
"step": 500
|
| 397 |
},
|
| 398 |
{
|
| 399 |
+
"epoch": 4.043478260869565,
|
| 400 |
+
"grad_norm": 0.04727836325764656,
|
| 401 |
+
"learning_rate": 3.0966183574879224e-05,
|
| 402 |
+
"loss": 0.0505,
|
| 403 |
"step": 510
|
| 404 |
},
|
| 405 |
{
|
| 406 |
+
"epoch": 4.052173913043478,
|
| 407 |
+
"grad_norm": 48.47065353393555,
|
| 408 |
+
"learning_rate": 3.0483091787439617e-05,
|
| 409 |
+
"loss": 0.1449,
|
| 410 |
"step": 520
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 4.060869565217391,
|
| 414 |
+
"grad_norm": 0.09721066057682037,
|
| 415 |
+
"learning_rate": 3e-05,
|
| 416 |
+
"loss": 0.0098,
|
| 417 |
"step": 530
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 4.069565217391304,
|
| 421 |
+
"grad_norm": 0.3183492422103882,
|
| 422 |
+
"learning_rate": 2.951690821256039e-05,
|
| 423 |
+
"loss": 0.0041,
|
| 424 |
"step": 540
|
| 425 |
},
|
| 426 |
{
|
| 427 |
+
"epoch": 4.078260869565217,
|
| 428 |
+
"grad_norm": 0.02137056551873684,
|
| 429 |
+
"learning_rate": 2.9033816425120775e-05,
|
| 430 |
+
"loss": 0.0696,
|
| 431 |
"step": 550
|
| 432 |
},
|
| 433 |
{
|
| 434 |
+
"epoch": 4.086956521739131,
|
| 435 |
+
"grad_norm": 0.04266425967216492,
|
| 436 |
+
"learning_rate": 2.8550724637681158e-05,
|
| 437 |
+
"loss": 0.004,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 438 |
"step": 560
|
| 439 |
},
|
| 440 |
{
|
| 441 |
+
"epoch": 4.095652173913043,
|
| 442 |
+
"grad_norm": 0.04853034391999245,
|
| 443 |
+
"learning_rate": 2.806763285024155e-05,
|
| 444 |
+
"loss": 0.0315,
|
| 445 |
"step": 570
|
| 446 |
},
|
| 447 |
{
|
| 448 |
+
"epoch": 4.1,
|
| 449 |
+
"eval_accuracy": 0.7073170731707317,
|
| 450 |
+
"eval_loss": 1.338157057762146,
|
| 451 |
+
"eval_runtime": 6.3184,
|
| 452 |
+
"eval_samples_per_second": 12.978,
|
| 453 |
+
"eval_steps_per_second": 3.324,
|
| 454 |
+
"step": 575
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 5.004347826086956,
|
| 458 |
+
"grad_norm": 0.028169039636850357,
|
| 459 |
+
"learning_rate": 2.7584541062801932e-05,
|
| 460 |
+
"loss": 0.0025,
|
| 461 |
"step": 580
|
| 462 |
},
|
| 463 |
{
|
| 464 |
+
"epoch": 5.01304347826087,
|
| 465 |
+
"grad_norm": 0.04628513753414154,
|
| 466 |
+
"learning_rate": 2.7101449275362322e-05,
|
| 467 |
+
"loss": 0.0033,
|
| 468 |
"step": 590
|
| 469 |
},
|
| 470 |
{
|
| 471 |
+
"epoch": 5.021739130434782,
|
| 472 |
+
"grad_norm": 0.19218483567237854,
|
| 473 |
+
"learning_rate": 2.6618357487922706e-05,
|
| 474 |
+
"loss": 0.0018,
|
| 475 |
"step": 600
|
| 476 |
},
|
| 477 |
{
|
| 478 |
+
"epoch": 5.030434782608696,
|
| 479 |
+
"grad_norm": 0.04920341819524765,
|
| 480 |
+
"learning_rate": 2.6135265700483093e-05,
|
| 481 |
+
"loss": 0.0011,
|
| 482 |
"step": 610
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 5.039130434782609,
|
| 486 |
+
"grad_norm": 0.015601912513375282,
|
| 487 |
+
"learning_rate": 2.5652173913043483e-05,
|
| 488 |
+
"loss": 0.0029,
|
| 489 |
"step": 620
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 5.047826086956522,
|
| 493 |
+
"grad_norm": 0.041672300547361374,
|
| 494 |
+
"learning_rate": 2.5169082125603866e-05,
|
| 495 |
+
"loss": 0.0095,
|
| 496 |
"step": 630
|
| 497 |
},
|
| 498 |
{
|
| 499 |
+
"epoch": 5.056521739130435,
|
| 500 |
+
"grad_norm": 0.031528741121292114,
|
| 501 |
+
"learning_rate": 2.4685990338164253e-05,
|
| 502 |
+
"loss": 0.0053,
|
| 503 |
"step": 640
|
| 504 |
},
|
| 505 |
{
|
| 506 |
+
"epoch": 5.065217391304348,
|
| 507 |
+
"grad_norm": 0.06605325639247894,
|
| 508 |
+
"learning_rate": 2.420289855072464e-05,
|
| 509 |
+
"loss": 0.0013,
|
| 510 |
"step": 650
|
| 511 |
},
|
| 512 |
{
|
| 513 |
+
"epoch": 5.073913043478261,
|
| 514 |
+
"grad_norm": 0.05129764974117279,
|
| 515 |
+
"learning_rate": 2.3719806763285024e-05,
|
| 516 |
+
"loss": 0.0011,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 517 |
"step": 660
|
| 518 |
},
|
| 519 |
{
|
| 520 |
+
"epoch": 5.082608695652174,
|
| 521 |
+
"grad_norm": 0.06676599383354187,
|
| 522 |
+
"learning_rate": 2.323671497584541e-05,
|
| 523 |
+
"loss": 0.0014,
|
| 524 |
"step": 670
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 5.091304347826087,
|
| 528 |
+
"grad_norm": 0.03804684802889824,
|
| 529 |
+
"learning_rate": 2.2753623188405797e-05,
|
| 530 |
+
"loss": 0.0115,
|
| 531 |
"step": 680
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 5.1,
|
| 535 |
+
"grad_norm": 0.027047328650951385,
|
| 536 |
+
"learning_rate": 2.2270531400966184e-05,
|
| 537 |
+
"loss": 0.0012,
|
| 538 |
"step": 690
|
| 539 |
},
|
| 540 |
{
|
| 541 |
+
"epoch": 5.1,
|
| 542 |
+
"eval_accuracy": 0.6829268292682927,
|
| 543 |
+
"eval_loss": 1.3492704629898071,
|
| 544 |
+
"eval_runtime": 6.0789,
|
| 545 |
+
"eval_samples_per_second": 13.489,
|
| 546 |
+
"eval_steps_per_second": 3.455,
|
| 547 |
+
"step": 690
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 6.008695652173913,
|
| 551 |
+
"grad_norm": 0.0289983619004488,
|
| 552 |
+
"learning_rate": 2.178743961352657e-05,
|
| 553 |
+
"loss": 0.001,
|
| 554 |
"step": 700
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 6.017391304347826,
|
| 558 |
+
"grad_norm": 0.015507020987570286,
|
| 559 |
+
"learning_rate": 2.1304347826086958e-05,
|
| 560 |
+
"loss": 0.0014,
|
| 561 |
"step": 710
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 6.026086956521739,
|
| 565 |
+
"grad_norm": 0.026568621397018433,
|
| 566 |
+
"learning_rate": 2.0821256038647345e-05,
|
| 567 |
+
"loss": 0.0011,
|
| 568 |
"step": 720
|
| 569 |
},
|
| 570 |
{
|
| 571 |
+
"epoch": 6.034782608695652,
|
| 572 |
+
"grad_norm": 0.029902568086981773,
|
| 573 |
+
"learning_rate": 2.0338164251207732e-05,
|
| 574 |
+
"loss": 0.0009,
|
| 575 |
"step": 730
|
| 576 |
},
|
| 577 |
{
|
| 578 |
+
"epoch": 6.043478260869565,
|
| 579 |
+
"grad_norm": 0.01757122576236725,
|
| 580 |
+
"learning_rate": 1.985507246376812e-05,
|
| 581 |
+
"loss": 0.0008,
|
| 582 |
"step": 740
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"epoch": 6.052173913043478,
|
| 586 |
+
"grad_norm": 0.02430758625268936,
|
| 587 |
+
"learning_rate": 1.9371980676328502e-05,
|
| 588 |
+
"loss": 0.0041,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 589 |
"step": 750
|
| 590 |
},
|
| 591 |
{
|
| 592 |
+
"epoch": 6.060869565217391,
|
| 593 |
+
"grad_norm": 0.013940953649580479,
|
| 594 |
+
"learning_rate": 1.888888888888889e-05,
|
| 595 |
+
"loss": 0.0019,
|
| 596 |
"step": 760
|
| 597 |
},
|
| 598 |
{
|
| 599 |
+
"epoch": 6.069565217391304,
|
| 600 |
+
"grad_norm": 0.01289789192378521,
|
| 601 |
+
"learning_rate": 1.8405797101449276e-05,
|
| 602 |
+
"loss": 0.001,
|
| 603 |
"step": 770
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 6.078260869565217,
|
| 607 |
+
"grad_norm": 0.012633569538593292,
|
| 608 |
+
"learning_rate": 1.7922705314009663e-05,
|
| 609 |
+
"loss": 0.0008,
|
| 610 |
"step": 780
|
| 611 |
},
|
| 612 |
{
|
| 613 |
+
"epoch": 6.086956521739131,
|
| 614 |
+
"grad_norm": 0.02719777263700962,
|
| 615 |
+
"learning_rate": 1.7439613526570046e-05,
|
| 616 |
+
"loss": 0.0013,
|
| 617 |
"step": 790
|
| 618 |
},
|
| 619 |
{
|
| 620 |
+
"epoch": 6.095652173913043,
|
| 621 |
+
"grad_norm": 0.01544960681349039,
|
| 622 |
+
"learning_rate": 1.6956521739130433e-05,
|
| 623 |
+
"loss": 0.0009,
|
| 624 |
"step": 800
|
| 625 |
},
|
| 626 |
{
|
| 627 |
+
"epoch": 6.1,
|
| 628 |
+
"eval_accuracy": 0.7073170731707317,
|
| 629 |
+
"eval_loss": 1.3622010946273804,
|
| 630 |
+
"eval_runtime": 6.1913,
|
| 631 |
+
"eval_samples_per_second": 13.244,
|
| 632 |
+
"eval_steps_per_second": 3.392,
|
| 633 |
+
"step": 805
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 7.004347826086956,
|
| 637 |
+
"grad_norm": 0.014636457897722721,
|
| 638 |
+
"learning_rate": 1.6473429951690824e-05,
|
| 639 |
+
"loss": 0.0022,
|
| 640 |
"step": 810
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 7.01304347826087,
|
| 644 |
+
"grad_norm": 0.011960177682340145,
|
| 645 |
+
"learning_rate": 1.599033816425121e-05,
|
| 646 |
"loss": 0.0009,
|
| 647 |
"step": 820
|
| 648 |
},
|
| 649 |
{
|
| 650 |
+
"epoch": 7.021739130434782,
|
| 651 |
+
"grad_norm": 0.025420162826776505,
|
| 652 |
+
"learning_rate": 1.5507246376811597e-05,
|
| 653 |
+
"loss": 0.0008,
|
| 654 |
"step": 830
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 7.030434782608696,
|
| 658 |
+
"grad_norm": 0.015812881290912628,
|
| 659 |
+
"learning_rate": 1.5024154589371981e-05,
|
| 660 |
+
"loss": 0.0009,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 661 |
"step": 840
|
| 662 |
},
|
| 663 |
{
|
| 664 |
+
"epoch": 7.039130434782609,
|
| 665 |
+
"grad_norm": 0.012759621255099773,
|
| 666 |
+
"learning_rate": 1.4541062801932368e-05,
|
| 667 |
"loss": 0.0008,
|
| 668 |
"step": 850
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 7.047826086956522,
|
| 672 |
+
"grad_norm": 0.013558686710894108,
|
| 673 |
+
"learning_rate": 1.4057971014492755e-05,
|
| 674 |
"loss": 0.0008,
|
| 675 |
"step": 860
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 7.056521739130435,
|
| 679 |
+
"grad_norm": 0.009738407097756863,
|
| 680 |
+
"learning_rate": 1.3574879227053142e-05,
|
| 681 |
+
"loss": 0.0008,
|
| 682 |
"step": 870
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 7.065217391304348,
|
| 686 |
+
"grad_norm": 0.020090168341994286,
|
| 687 |
+
"learning_rate": 1.3091787439613527e-05,
|
| 688 |
+
"loss": 0.0008,
|
| 689 |
"step": 880
|
| 690 |
},
|
| 691 |
{
|
| 692 |
+
"epoch": 7.073913043478261,
|
| 693 |
+
"grad_norm": 0.00872560404241085,
|
| 694 |
+
"learning_rate": 1.2608695652173914e-05,
|
| 695 |
+
"loss": 0.0008,
|
| 696 |
"step": 890
|
| 697 |
},
|
| 698 |
{
|
| 699 |
+
"epoch": 7.082608695652174,
|
| 700 |
+
"grad_norm": 0.0459994301199913,
|
| 701 |
+
"learning_rate": 1.21256038647343e-05,
|
| 702 |
+
"loss": 0.0009,
|
| 703 |
"step": 900
|
| 704 |
},
|
| 705 |
{
|
| 706 |
+
"epoch": 7.091304347826087,
|
| 707 |
+
"grad_norm": 0.015241548418998718,
|
| 708 |
+
"learning_rate": 1.1642512077294687e-05,
|
| 709 |
"loss": 0.0007,
|
| 710 |
"step": 910
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 7.1,
|
| 714 |
+
"grad_norm": 0.019985787570476532,
|
| 715 |
+
"learning_rate": 1.1159420289855073e-05,
|
| 716 |
"loss": 0.0008,
|
| 717 |
"step": 920
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 7.1,
|
| 721 |
+
"eval_accuracy": 0.7073170731707317,
|
| 722 |
+
"eval_loss": 1.3999933004379272,
|
| 723 |
+
"eval_runtime": 6.3918,
|
| 724 |
+
"eval_samples_per_second": 12.829,
|
| 725 |
+
"eval_steps_per_second": 3.285,
|
| 726 |
"step": 920
|
| 727 |
},
|
| 728 |
{
|
| 729 |
+
"epoch": 8.008695652173913,
|
| 730 |
+
"grad_norm": 0.009398338384926319,
|
| 731 |
+
"learning_rate": 1.067632850241546e-05,
|
| 732 |
+
"loss": 0.0007,
|
| 733 |
+
"step": 930
|
|
|
|
|
|
|
| 734 |
},
|
| 735 |
{
|
| 736 |
+
"epoch": 8.017391304347827,
|
| 737 |
+
"grad_norm": 0.026491643860936165,
|
| 738 |
+
"learning_rate": 1.0193236714975846e-05,
|
| 739 |
+
"loss": 0.0007,
|
| 740 |
+
"step": 940
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 8.02608695652174,
|
| 744 |
+
"grad_norm": 0.015245695598423481,
|
| 745 |
+
"learning_rate": 9.710144927536233e-06,
|
| 746 |
+
"loss": 0.0007,
|
| 747 |
+
"step": 950
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 8.034782608695652,
|
| 751 |
+
"grad_norm": 0.014611267484724522,
|
| 752 |
+
"learning_rate": 9.227053140096618e-06,
|
| 753 |
+
"loss": 0.0008,
|
| 754 |
+
"step": 960
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 8.043478260869565,
|
| 758 |
+
"grad_norm": 0.023189160972833633,
|
| 759 |
+
"learning_rate": 8.743961352657005e-06,
|
| 760 |
+
"loss": 0.0008,
|
| 761 |
+
"step": 970
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 8.052173913043479,
|
| 765 |
+
"grad_norm": 0.01567276194691658,
|
| 766 |
+
"learning_rate": 8.26086956521739e-06,
|
| 767 |
+
"loss": 0.0008,
|
| 768 |
+
"step": 980
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 8.060869565217391,
|
| 772 |
+
"grad_norm": 0.0071034678258001804,
|
| 773 |
+
"learning_rate": 7.777777777777777e-06,
|
| 774 |
+
"loss": 0.0006,
|
| 775 |
+
"step": 990
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 8.069565217391304,
|
| 779 |
+
"grad_norm": 0.00896136648952961,
|
| 780 |
+
"learning_rate": 7.294685990338164e-06,
|
| 781 |
+
"loss": 0.0006,
|
| 782 |
+
"step": 1000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 8.078260869565218,
|
| 786 |
+
"grad_norm": 0.012383806519210339,
|
| 787 |
+
"learning_rate": 6.811594202898551e-06,
|
| 788 |
+
"loss": 0.0008,
|
| 789 |
+
"step": 1010
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 8.08695652173913,
|
| 793 |
+
"grad_norm": 0.02067715860903263,
|
| 794 |
+
"learning_rate": 6.328502415458938e-06,
|
| 795 |
+
"loss": 0.0007,
|
| 796 |
+
"step": 1020
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 8.095652173913043,
|
| 800 |
+
"grad_norm": 0.02547420747578144,
|
| 801 |
+
"learning_rate": 5.845410628019324e-06,
|
| 802 |
+
"loss": 0.0007,
|
| 803 |
+
"step": 1030
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 8.1,
|
| 807 |
+
"eval_accuracy": 0.7073170731707317,
|
| 808 |
+
"eval_loss": 1.401489496231079,
|
| 809 |
+
"eval_runtime": 6.1399,
|
| 810 |
+
"eval_samples_per_second": 13.355,
|
| 811 |
+
"eval_steps_per_second": 3.42,
|
| 812 |
+
"step": 1035
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 9.004347826086956,
|
| 816 |
+
"grad_norm": 0.012581247836351395,
|
| 817 |
+
"learning_rate": 5.36231884057971e-06,
|
| 818 |
+
"loss": 0.0007,
|
| 819 |
+
"step": 1040
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 9.013043478260869,
|
| 823 |
+
"grad_norm": 0.016566790640354156,
|
| 824 |
+
"learning_rate": 4.879227053140096e-06,
|
| 825 |
+
"loss": 0.0006,
|
| 826 |
+
"step": 1050
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 9.021739130434783,
|
| 830 |
+
"grad_norm": 0.03303120657801628,
|
| 831 |
+
"learning_rate": 4.396135265700483e-06,
|
| 832 |
+
"loss": 0.0008,
|
| 833 |
+
"step": 1060
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 9.030434782608696,
|
| 837 |
+
"grad_norm": 0.015636784955859184,
|
| 838 |
+
"learning_rate": 3.91304347826087e-06,
|
| 839 |
+
"loss": 0.0007,
|
| 840 |
+
"step": 1070
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 9.039130434782608,
|
| 844 |
+
"grad_norm": 0.015506476163864136,
|
| 845 |
+
"learning_rate": 3.4299516908212565e-06,
|
| 846 |
+
"loss": 0.0008,
|
| 847 |
+
"step": 1080
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 9.047826086956523,
|
| 851 |
+
"grad_norm": 0.01221100240945816,
|
| 852 |
+
"learning_rate": 2.9468599033816426e-06,
|
| 853 |
+
"loss": 0.0007,
|
| 854 |
+
"step": 1090
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 9.056521739130435,
|
| 858 |
+
"grad_norm": 0.011791340075433254,
|
| 859 |
+
"learning_rate": 2.463768115942029e-06,
|
| 860 |
+
"loss": 0.0007,
|
| 861 |
+
"step": 1100
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 9.065217391304348,
|
| 865 |
+
"grad_norm": 0.025301584973931313,
|
| 866 |
+
"learning_rate": 1.9806763285024155e-06,
|
| 867 |
+
"loss": 0.0008,
|
| 868 |
+
"step": 1110
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 9.07391304347826,
|
| 872 |
+
"grad_norm": 0.014775657095015049,
|
| 873 |
+
"learning_rate": 1.497584541062802e-06,
|
| 874 |
+
"loss": 0.0007,
|
| 875 |
+
"step": 1120
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 9.082608695652175,
|
| 879 |
+
"grad_norm": 0.025027761235833168,
|
| 880 |
+
"learning_rate": 1.0144927536231885e-06,
|
| 881 |
+
"loss": 0.0007,
|
| 882 |
+
"step": 1130
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 9.091304347826087,
|
| 886 |
+
"grad_norm": 0.007070675026625395,
|
| 887 |
+
"learning_rate": 5.314009661835749e-07,
|
| 888 |
+
"loss": 0.0006,
|
| 889 |
+
"step": 1140
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 9.1,
|
| 893 |
+
"grad_norm": 0.011773703619837761,
|
| 894 |
+
"learning_rate": 4.8309178743961356e-08,
|
| 895 |
+
"loss": 0.0007,
|
| 896 |
+
"step": 1150
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 9.1,
|
| 900 |
+
"eval_accuracy": 0.7073170731707317,
|
| 901 |
+
"eval_loss": 1.4007967710494995,
|
| 902 |
+
"eval_runtime": 6.2852,
|
| 903 |
+
"eval_samples_per_second": 13.046,
|
| 904 |
+
"eval_steps_per_second": 3.341,
|
| 905 |
+
"step": 1150
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 9.1,
|
| 909 |
+
"step": 1150,
|
| 910 |
+
"total_flos": 5.732152700888678e+18,
|
| 911 |
+
"train_loss": 0.30212519713110575,
|
| 912 |
+
"train_runtime": 1217.5297,
|
| 913 |
+
"train_samples_per_second": 3.778,
|
| 914 |
+
"train_steps_per_second": 0.945
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 9.1,
|
| 918 |
+
"eval_accuracy": 0.7073170731707317,
|
| 919 |
+
"eval_loss": 1.3381569385528564,
|
| 920 |
+
"eval_runtime": 5.7056,
|
| 921 |
+
"eval_samples_per_second": 14.372,
|
| 922 |
+
"eval_steps_per_second": 3.681,
|
| 923 |
+
"step": 1150
|
| 924 |
}
|
| 925 |
],
|
| 926 |
"logging_steps": 10,
|
| 927 |
+
"max_steps": 1150,
|
| 928 |
"num_input_tokens_seen": 0,
|
| 929 |
"num_train_epochs": 9223372036854775807,
|
| 930 |
"save_steps": 500,
|
|
|
|
| 940 |
"attributes": {}
|
| 941 |
}
|
| 942 |
},
|
| 943 |
+
"total_flos": 5.732152700888678e+18,
|
| 944 |
"train_batch_size": 4,
|
| 945 |
"trial_name": null,
|
| 946 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9dcd726e58dbf901b06a4f42284e673960d4b6d9185a8974cbd9733ea2202cde
|
| 3 |
size 5368
|