Upload folder using huggingface_hub
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1454 -229
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -29,13 +29,13 @@
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
|
|
|
|
|
|
| 32 |
"v_proj",
|
| 33 |
-
"down_proj",
|
| 34 |
"k_proj",
|
| 35 |
-
"gate_proj",
|
| 36 |
-
"q_proj",
|
| 37 |
"o_proj",
|
| 38 |
-
"
|
|
|
|
| 39 |
],
|
| 40 |
"target_parameters": null,
|
| 41 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 29 |
"rank_pattern": {},
|
| 30 |
"revision": null,
|
| 31 |
"target_modules": [
|
| 32 |
+
"up_proj",
|
| 33 |
+
"q_proj",
|
| 34 |
"v_proj",
|
|
|
|
| 35 |
"k_proj",
|
|
|
|
|
|
|
| 36 |
"o_proj",
|
| 37 |
+
"gate_proj",
|
| 38 |
+
"down_proj"
|
| 39 |
],
|
| 40 |
"target_parameters": null,
|
| 41 |
"task_type": "CAUSAL_LM",
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 65674128
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:177986ef3ec8853c7a100227b013bafdfdbc9f14d6e71712e9b5b6b09d660d9f
|
| 3 |
size 65674128
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 119618359
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6edcecd4461e9f55f55a4512bd1689007e5060d969210db70448fc069e731825
|
| 3 |
size 119618359
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17cd930da9783ca70bad4b9cdeee6a06c0acea8f34645a333c93341f487f66a3
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3625ea8ae5237ce569b51fa44782ffb65e60dddba324dce72a6c0c9a368ee983
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,543 +2,1768 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 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": 0.04,
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
-
"learning_rate": 0.
|
| 16 |
-
"loss": 0.
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.08,
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
-
"learning_rate": 0.
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.12,
|
| 28 |
-
"grad_norm": 0.
|
| 29 |
-
"learning_rate": 0.
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.16,
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.2,
|
| 42 |
-
"grad_norm": 0.
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.24,
|
| 49 |
-
"grad_norm": 0.
|
| 50 |
-
"learning_rate": 0.
|
| 51 |
"loss": 0.2002,
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.28,
|
| 56 |
-
"grad_norm": 0.
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.32,
|
| 63 |
-
"grad_norm": 0.
|
| 64 |
-
"learning_rate": 0.
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.36,
|
| 70 |
-
"grad_norm": 0.
|
| 71 |
-
"learning_rate": 0.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.4,
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.44,
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.48,
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
-
"learning_rate": 0.
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 120
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.52,
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
-
"learning_rate": 0.
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 130
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.56,
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
-
"learning_rate": 0.
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 140
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.6,
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
-
"learning_rate": 0.
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 150
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.64,
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
-
"learning_rate": 0.
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 160
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.68,
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
-
"learning_rate": 0.
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 170
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.72,
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
-
"learning_rate": 0.
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 180
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.76,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
-
"learning_rate": 0.
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 190
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.8,
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
-
"learning_rate": 0.
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 200
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.84,
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
-
"learning_rate": 0.
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 210
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.88,
|
| 161 |
-
"grad_norm": 0.
|
| 162 |
-
"learning_rate": 0.
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 220
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.92,
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
-
"learning_rate": 0.
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 230
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.96,
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
-
"learning_rate": 0.
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 240
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 1.0,
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
-
"learning_rate": 0.
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 1.04,
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
-
"learning_rate": 0.
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 260
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 1.08,
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
-
"learning_rate": 0.
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 270
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 1.12,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate": 0.
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 280
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 1.16,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate": 0.
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 290
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 1.2,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate": 0.
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 300
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 1.24,
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate": 0.
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 310
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 1.28,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate": 0.
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 320
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 1.32,
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate": 0.
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 330
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 1.3599999999999999,
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
-
"learning_rate": 0.
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 340
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 1.4,
|
| 252 |
-
"grad_norm": 0.
|
| 253 |
-
"learning_rate": 0.
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 350
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.44,
|
| 259 |
-
"grad_norm": 0.
|
| 260 |
-
"learning_rate": 0.
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 360
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 1.48,
|
| 266 |
-
"grad_norm": 0.
|
| 267 |
-
"learning_rate": 0.
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 370
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 1.52,
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 380
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 1.56,
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 390
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 1.6,
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 1.6400000000000001,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 410
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.6800000000000002,
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss": 0.
|
| 304 |
"step": 420
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 1.72,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 430
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 1.76,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss": 0.
|
| 318 |
"step": 440
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 1.8,
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 450
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.8399999999999999,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 460
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.88,
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 470
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.92,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 480
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.96,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 490
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 2.0,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 2.04,
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 510
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 2.08,
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 520
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 2.12,
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 530
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 2.16,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 540
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 2.2,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 550
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 2.24,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 560
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 2.2800000000000002,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
-
"learning_rate":
|
| 408 |
-
"loss": 0.
|
| 409 |
"step": 570
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 2.32,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss": 0.
|
| 416 |
"step": 580
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 2.36,
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 590
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 2.4,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 600
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 2.44,
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
-
"learning_rate":
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 610
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 2.48,
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
-
"learning_rate":
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 620
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 2.52,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 630
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 2.56,
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 640
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 2.6,
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 650
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 2.64,
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 660
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 2.68,
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 670
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 2.7199999999999998,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss": 0.
|
| 486 |
"step": 680
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 2.76,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 690
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 2.8,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss": 0.
|
| 500 |
"step": 700
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 2.84,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 710
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 2.88,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
-
"learning_rate":
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 720
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 2.92,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
-
"learning_rate":
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 730
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 2.96,
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
-
"learning_rate":
|
| 527 |
-
"loss": 0.
|
| 528 |
"step": 740
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 3.0,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
-
"learning_rate":
|
| 534 |
-
"loss": 0.
|
| 535 |
"step": 750
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 536 |
}
|
| 537 |
],
|
| 538 |
"logging_steps": 10,
|
| 539 |
-
"max_steps":
|
| 540 |
"num_input_tokens_seen": 0,
|
| 541 |
-
"num_train_epochs":
|
| 542 |
"save_steps": 500,
|
| 543 |
"stateful_callbacks": {
|
| 544 |
"TrainerControl": {
|
|
@@ -552,7 +1777,7 @@
|
|
| 552 |
"attributes": {}
|
| 553 |
}
|
| 554 |
},
|
| 555 |
-
"total_flos":
|
| 556 |
"train_batch_size": 20,
|
| 557 |
"trial_name": null,
|
| 558 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 2500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.04,
|
| 14 |
+
"grad_norm": 0.7923470139503479,
|
| 15 |
+
"learning_rate": 0.00019967935871743488,
|
| 16 |
+
"loss": 0.4612,
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.08,
|
| 21 |
+
"grad_norm": 0.5407691597938538,
|
| 22 |
+
"learning_rate": 0.00019887775551102204,
|
| 23 |
+
"loss": 0.2353,
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.12,
|
| 28 |
+
"grad_norm": 0.5324429869651794,
|
| 29 |
+
"learning_rate": 0.00019807615230460924,
|
| 30 |
+
"loss": 0.2157,
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.16,
|
| 35 |
+
"grad_norm": 0.5783653259277344,
|
| 36 |
+
"learning_rate": 0.0001972745490981964,
|
| 37 |
+
"loss": 0.2017,
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.2,
|
| 42 |
+
"grad_norm": 0.47596609592437744,
|
| 43 |
+
"learning_rate": 0.00019647294589178357,
|
| 44 |
+
"loss": 0.1951,
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.24,
|
| 49 |
+
"grad_norm": 0.4201284945011139,
|
| 50 |
+
"learning_rate": 0.00019567134268537074,
|
| 51 |
"loss": 0.2002,
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.28,
|
| 56 |
+
"grad_norm": 0.4552278518676758,
|
| 57 |
+
"learning_rate": 0.00019486973947895793,
|
| 58 |
+
"loss": 0.1904,
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.32,
|
| 63 |
+
"grad_norm": 0.7762654423713684,
|
| 64 |
+
"learning_rate": 0.0001940681362725451,
|
| 65 |
+
"loss": 0.1789,
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.36,
|
| 70 |
+
"grad_norm": 0.48717886209487915,
|
| 71 |
+
"learning_rate": 0.00019326653306613227,
|
| 72 |
+
"loss": 0.1754,
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.4,
|
| 77 |
+
"grad_norm": 0.5642560720443726,
|
| 78 |
+
"learning_rate": 0.00019246492985971943,
|
| 79 |
+
"loss": 0.1807,
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.44,
|
| 84 |
+
"grad_norm": 0.5994827151298523,
|
| 85 |
+
"learning_rate": 0.00019166332665330663,
|
| 86 |
+
"loss": 0.1865,
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.48,
|
| 91 |
+
"grad_norm": 0.5326362252235413,
|
| 92 |
+
"learning_rate": 0.0001908617234468938,
|
| 93 |
+
"loss": 0.1845,
|
| 94 |
"step": 120
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.52,
|
| 98 |
+
"grad_norm": 0.8340044021606445,
|
| 99 |
+
"learning_rate": 0.00019006012024048096,
|
| 100 |
+
"loss": 0.1696,
|
| 101 |
"step": 130
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.56,
|
| 105 |
+
"grad_norm": 0.7146855592727661,
|
| 106 |
+
"learning_rate": 0.00018925851703406813,
|
| 107 |
+
"loss": 0.1844,
|
| 108 |
"step": 140
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.6,
|
| 112 |
+
"grad_norm": 0.5249161720275879,
|
| 113 |
+
"learning_rate": 0.00018845691382765532,
|
| 114 |
+
"loss": 0.1757,
|
| 115 |
"step": 150
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.64,
|
| 119 |
+
"grad_norm": 0.7083638310432434,
|
| 120 |
+
"learning_rate": 0.0001876553106212425,
|
| 121 |
+
"loss": 0.1918,
|
| 122 |
"step": 160
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.68,
|
| 126 |
+
"grad_norm": 0.40141209959983826,
|
| 127 |
+
"learning_rate": 0.00018685370741482966,
|
| 128 |
+
"loss": 0.1868,
|
| 129 |
"step": 170
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.72,
|
| 133 |
+
"grad_norm": 0.4578303098678589,
|
| 134 |
+
"learning_rate": 0.00018605210420841683,
|
| 135 |
+
"loss": 0.1809,
|
| 136 |
"step": 180
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.76,
|
| 140 |
+
"grad_norm": 0.49453985691070557,
|
| 141 |
+
"learning_rate": 0.00018525050100200402,
|
| 142 |
+
"loss": 0.1747,
|
| 143 |
"step": 190
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.8,
|
| 147 |
+
"grad_norm": 0.5069108009338379,
|
| 148 |
+
"learning_rate": 0.0001844488977955912,
|
| 149 |
+
"loss": 0.1781,
|
| 150 |
"step": 200
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.84,
|
| 154 |
+
"grad_norm": 0.6718438267707825,
|
| 155 |
+
"learning_rate": 0.00018364729458917838,
|
| 156 |
+
"loss": 0.1772,
|
| 157 |
"step": 210
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.88,
|
| 161 |
+
"grad_norm": 0.4564284682273865,
|
| 162 |
+
"learning_rate": 0.00018284569138276555,
|
| 163 |
+
"loss": 0.177,
|
| 164 |
"step": 220
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.92,
|
| 168 |
+
"grad_norm": 0.47330281138420105,
|
| 169 |
+
"learning_rate": 0.00018204408817635271,
|
| 170 |
+
"loss": 0.1715,
|
| 171 |
"step": 230
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.96,
|
| 175 |
+
"grad_norm": 0.4193236827850342,
|
| 176 |
+
"learning_rate": 0.0001812424849699399,
|
| 177 |
+
"loss": 0.1749,
|
| 178 |
"step": 240
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 1.0,
|
| 182 |
+
"grad_norm": 0.5325985550880432,
|
| 183 |
+
"learning_rate": 0.00018044088176352708,
|
| 184 |
+
"loss": 0.18,
|
| 185 |
"step": 250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 1.04,
|
| 189 |
+
"grad_norm": 0.5181522965431213,
|
| 190 |
+
"learning_rate": 0.00017963927855711424,
|
| 191 |
+
"loss": 0.1537,
|
| 192 |
"step": 260
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 1.08,
|
| 196 |
+
"grad_norm": 0.4463769197463989,
|
| 197 |
+
"learning_rate": 0.0001788376753507014,
|
| 198 |
+
"loss": 0.1432,
|
| 199 |
"step": 270
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 1.12,
|
| 203 |
+
"grad_norm": 0.5249760746955872,
|
| 204 |
+
"learning_rate": 0.00017803607214428858,
|
| 205 |
+
"loss": 0.1607,
|
| 206 |
"step": 280
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 1.16,
|
| 210 |
+
"grad_norm": 0.42788001894950867,
|
| 211 |
+
"learning_rate": 0.00017723446893787577,
|
| 212 |
+
"loss": 0.1473,
|
| 213 |
"step": 290
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 1.2,
|
| 217 |
+
"grad_norm": 0.6883480548858643,
|
| 218 |
+
"learning_rate": 0.00017643286573146294,
|
| 219 |
+
"loss": 0.1481,
|
| 220 |
"step": 300
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 1.24,
|
| 224 |
+
"grad_norm": 0.5434200167655945,
|
| 225 |
+
"learning_rate": 0.0001756312625250501,
|
| 226 |
+
"loss": 0.1557,
|
| 227 |
"step": 310
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 1.28,
|
| 231 |
+
"grad_norm": 0.770155668258667,
|
| 232 |
+
"learning_rate": 0.00017482965931863727,
|
| 233 |
+
"loss": 0.1397,
|
| 234 |
"step": 320
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 1.32,
|
| 238 |
+
"grad_norm": 0.6204828023910522,
|
| 239 |
+
"learning_rate": 0.00017402805611222447,
|
| 240 |
+
"loss": 0.1569,
|
| 241 |
"step": 330
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 1.3599999999999999,
|
| 245 |
+
"grad_norm": 0.7150607705116272,
|
| 246 |
+
"learning_rate": 0.00017322645290581163,
|
| 247 |
+
"loss": 0.1523,
|
| 248 |
"step": 340
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 1.4,
|
| 252 |
+
"grad_norm": 0.5075599551200867,
|
| 253 |
+
"learning_rate": 0.0001724248496993988,
|
| 254 |
+
"loss": 0.1631,
|
| 255 |
"step": 350
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.44,
|
| 259 |
+
"grad_norm": 0.6147515177726746,
|
| 260 |
+
"learning_rate": 0.00017162324649298597,
|
| 261 |
+
"loss": 0.1476,
|
| 262 |
"step": 360
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 1.48,
|
| 266 |
+
"grad_norm": 0.6109094023704529,
|
| 267 |
+
"learning_rate": 0.00017082164328657316,
|
| 268 |
+
"loss": 0.1522,
|
| 269 |
"step": 370
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 1.52,
|
| 273 |
+
"grad_norm": 0.5690982341766357,
|
| 274 |
+
"learning_rate": 0.00017002004008016033,
|
| 275 |
+
"loss": 0.1555,
|
| 276 |
"step": 380
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 1.56,
|
| 280 |
+
"grad_norm": 0.44981974363327026,
|
| 281 |
+
"learning_rate": 0.0001692184368737475,
|
| 282 |
+
"loss": 0.1592,
|
| 283 |
"step": 390
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 1.6,
|
| 287 |
+
"grad_norm": 0.4784778952598572,
|
| 288 |
+
"learning_rate": 0.00016841683366733466,
|
| 289 |
+
"loss": 0.1636,
|
| 290 |
"step": 400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 1.6400000000000001,
|
| 294 |
+
"grad_norm": 0.5931491851806641,
|
| 295 |
+
"learning_rate": 0.00016761523046092186,
|
| 296 |
+
"loss": 0.152,
|
| 297 |
"step": 410
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.6800000000000002,
|
| 301 |
+
"grad_norm": 0.663811206817627,
|
| 302 |
+
"learning_rate": 0.00016681362725450903,
|
| 303 |
+
"loss": 0.1704,
|
| 304 |
"step": 420
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 1.72,
|
| 308 |
+
"grad_norm": 0.4538479447364807,
|
| 309 |
+
"learning_rate": 0.0001660120240480962,
|
| 310 |
+
"loss": 0.1482,
|
| 311 |
"step": 430
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 1.76,
|
| 315 |
+
"grad_norm": 0.6272454261779785,
|
| 316 |
+
"learning_rate": 0.00016521042084168336,
|
| 317 |
+
"loss": 0.1545,
|
| 318 |
"step": 440
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 1.8,
|
| 322 |
+
"grad_norm": 0.5804396271705627,
|
| 323 |
+
"learning_rate": 0.00016440881763527055,
|
| 324 |
+
"loss": 0.1622,
|
| 325 |
"step": 450
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.8399999999999999,
|
| 329 |
+
"grad_norm": 0.5440978407859802,
|
| 330 |
+
"learning_rate": 0.00016360721442885772,
|
| 331 |
+
"loss": 0.1607,
|
| 332 |
"step": 460
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.88,
|
| 336 |
+
"grad_norm": 0.6250975131988525,
|
| 337 |
+
"learning_rate": 0.0001628056112224449,
|
| 338 |
+
"loss": 0.1547,
|
| 339 |
"step": 470
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.92,
|
| 343 |
+
"grad_norm": 0.41538259387016296,
|
| 344 |
+
"learning_rate": 0.00016200400801603208,
|
| 345 |
+
"loss": 0.146,
|
| 346 |
"step": 480
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.96,
|
| 350 |
+
"grad_norm": 0.5187687277793884,
|
| 351 |
+
"learning_rate": 0.00016120240480961925,
|
| 352 |
+
"loss": 0.1663,
|
| 353 |
"step": 490
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 2.0,
|
| 357 |
+
"grad_norm": 0.392794132232666,
|
| 358 |
+
"learning_rate": 0.00016040080160320644,
|
| 359 |
+
"loss": 0.1492,
|
| 360 |
"step": 500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 2.04,
|
| 364 |
+
"grad_norm": 0.5112284421920776,
|
| 365 |
+
"learning_rate": 0.0001595991983967936,
|
| 366 |
+
"loss": 0.1276,
|
| 367 |
"step": 510
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 2.08,
|
| 371 |
+
"grad_norm": 0.577057421207428,
|
| 372 |
+
"learning_rate": 0.00015879759519038078,
|
| 373 |
+
"loss": 0.118,
|
| 374 |
"step": 520
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 2.12,
|
| 378 |
+
"grad_norm": 0.4589303433895111,
|
| 379 |
+
"learning_rate": 0.00015799599198396794,
|
| 380 |
+
"loss": 0.1251,
|
| 381 |
"step": 530
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 2.16,
|
| 385 |
+
"grad_norm": 0.4399261176586151,
|
| 386 |
+
"learning_rate": 0.0001571943887775551,
|
| 387 |
+
"loss": 0.1174,
|
| 388 |
"step": 540
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 2.2,
|
| 392 |
+
"grad_norm": 0.47207799553871155,
|
| 393 |
+
"learning_rate": 0.0001563927855711423,
|
| 394 |
+
"loss": 0.1301,
|
| 395 |
"step": 550
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 2.24,
|
| 399 |
+
"grad_norm": 0.5883737802505493,
|
| 400 |
+
"learning_rate": 0.00015559118236472947,
|
| 401 |
+
"loss": 0.1285,
|
| 402 |
"step": 560
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 2.2800000000000002,
|
| 406 |
+
"grad_norm": 0.5191317796707153,
|
| 407 |
+
"learning_rate": 0.00015478957915831664,
|
| 408 |
+
"loss": 0.1232,
|
| 409 |
"step": 570
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 2.32,
|
| 413 |
+
"grad_norm": 0.546741247177124,
|
| 414 |
+
"learning_rate": 0.0001539879759519038,
|
| 415 |
+
"loss": 0.1177,
|
| 416 |
"step": 580
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 2.36,
|
| 420 |
+
"grad_norm": 0.5307457447052002,
|
| 421 |
+
"learning_rate": 0.000153186372745491,
|
| 422 |
+
"loss": 0.1226,
|
| 423 |
"step": 590
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 2.4,
|
| 427 |
+
"grad_norm": 0.49284040927886963,
|
| 428 |
+
"learning_rate": 0.00015238476953907817,
|
| 429 |
+
"loss": 0.1398,
|
| 430 |
"step": 600
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 2.44,
|
| 434 |
+
"grad_norm": 0.46590954065322876,
|
| 435 |
+
"learning_rate": 0.00015158316633266534,
|
| 436 |
+
"loss": 0.1287,
|
| 437 |
"step": 610
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 2.48,
|
| 441 |
+
"grad_norm": 0.5511562824249268,
|
| 442 |
+
"learning_rate": 0.0001507815631262525,
|
| 443 |
+
"loss": 0.118,
|
| 444 |
"step": 620
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 2.52,
|
| 448 |
+
"grad_norm": 0.5350615978240967,
|
| 449 |
+
"learning_rate": 0.0001499799599198397,
|
| 450 |
+
"loss": 0.1295,
|
| 451 |
"step": 630
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 2.56,
|
| 455 |
+
"grad_norm": 0.5491256713867188,
|
| 456 |
+
"learning_rate": 0.00014917835671342686,
|
| 457 |
+
"loss": 0.129,
|
| 458 |
"step": 640
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 2.6,
|
| 462 |
+
"grad_norm": 0.5817452669143677,
|
| 463 |
+
"learning_rate": 0.00014837675350701403,
|
| 464 |
+
"loss": 0.1373,
|
| 465 |
"step": 650
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 2.64,
|
| 469 |
+
"grad_norm": 0.4481925964355469,
|
| 470 |
+
"learning_rate": 0.0001475751503006012,
|
| 471 |
+
"loss": 0.1389,
|
| 472 |
"step": 660
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 2.68,
|
| 476 |
+
"grad_norm": 0.5628023743629456,
|
| 477 |
+
"learning_rate": 0.0001467735470941884,
|
| 478 |
+
"loss": 0.1238,
|
| 479 |
"step": 670
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 2.7199999999999998,
|
| 483 |
+
"grad_norm": 0.678969144821167,
|
| 484 |
+
"learning_rate": 0.00014597194388777556,
|
| 485 |
+
"loss": 0.1318,
|
| 486 |
"step": 680
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 2.76,
|
| 490 |
+
"grad_norm": 0.5662192106246948,
|
| 491 |
+
"learning_rate": 0.00014517034068136273,
|
| 492 |
+
"loss": 0.1242,
|
| 493 |
"step": 690
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 2.8,
|
| 497 |
+
"grad_norm": 0.532526433467865,
|
| 498 |
+
"learning_rate": 0.0001443687374749499,
|
| 499 |
+
"loss": 0.1279,
|
| 500 |
"step": 700
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 2.84,
|
| 504 |
+
"grad_norm": 0.4268568158149719,
|
| 505 |
+
"learning_rate": 0.00014356713426853706,
|
| 506 |
+
"loss": 0.1342,
|
| 507 |
"step": 710
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 2.88,
|
| 511 |
+
"grad_norm": 0.6674071550369263,
|
| 512 |
+
"learning_rate": 0.00014276553106212425,
|
| 513 |
+
"loss": 0.1374,
|
| 514 |
"step": 720
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 2.92,
|
| 518 |
+
"grad_norm": 0.5499129891395569,
|
| 519 |
+
"learning_rate": 0.00014196392785571142,
|
| 520 |
+
"loss": 0.1226,
|
| 521 |
"step": 730
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 2.96,
|
| 525 |
+
"grad_norm": 0.5989218354225159,
|
| 526 |
+
"learning_rate": 0.0001411623246492986,
|
| 527 |
+
"loss": 0.1282,
|
| 528 |
"step": 740
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 3.0,
|
| 532 |
+
"grad_norm": 0.5111316442489624,
|
| 533 |
+
"learning_rate": 0.00014036072144288576,
|
| 534 |
+
"loss": 0.1241,
|
| 535 |
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 3.04,
|
| 539 |
+
"grad_norm": 0.4813462197780609,
|
| 540 |
+
"learning_rate": 0.00013955911823647295,
|
| 541 |
+
"loss": 0.0941,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.08,
|
| 546 |
+
"grad_norm": 0.5328338742256165,
|
| 547 |
+
"learning_rate": 0.00013875751503006014,
|
| 548 |
+
"loss": 0.0914,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 3.12,
|
| 553 |
+
"grad_norm": 0.5423403382301331,
|
| 554 |
+
"learning_rate": 0.0001379559118236473,
|
| 555 |
+
"loss": 0.0988,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.16,
|
| 560 |
+
"grad_norm": 0.6739408373832703,
|
| 561 |
+
"learning_rate": 0.00013715430861723448,
|
| 562 |
+
"loss": 0.0995,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.2,
|
| 567 |
+
"grad_norm": 0.5204668045043945,
|
| 568 |
+
"learning_rate": 0.00013635270541082165,
|
| 569 |
+
"loss": 0.0973,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.24,
|
| 574 |
+
"grad_norm": 0.6864869594573975,
|
| 575 |
+
"learning_rate": 0.00013555110220440884,
|
| 576 |
+
"loss": 0.0911,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 3.2800000000000002,
|
| 581 |
+
"grad_norm": 0.587958037853241,
|
| 582 |
+
"learning_rate": 0.000134749498997996,
|
| 583 |
+
"loss": 0.103,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.32,
|
| 588 |
+
"grad_norm": 0.6141840219497681,
|
| 589 |
+
"learning_rate": 0.00013394789579158317,
|
| 590 |
+
"loss": 0.0925,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 3.36,
|
| 595 |
+
"grad_norm": 0.5647754073143005,
|
| 596 |
+
"learning_rate": 0.00013314629258517034,
|
| 597 |
+
"loss": 0.1056,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 3.4,
|
| 602 |
+
"grad_norm": 0.5085099339485168,
|
| 603 |
+
"learning_rate": 0.00013234468937875754,
|
| 604 |
+
"loss": 0.1071,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 3.44,
|
| 609 |
+
"grad_norm": 0.6452666521072388,
|
| 610 |
+
"learning_rate": 0.0001315430861723447,
|
| 611 |
+
"loss": 0.1074,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 3.48,
|
| 616 |
+
"grad_norm": 0.5613518357276917,
|
| 617 |
+
"learning_rate": 0.00013074148296593187,
|
| 618 |
+
"loss": 0.1066,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 3.52,
|
| 623 |
+
"grad_norm": 0.5741850137710571,
|
| 624 |
+
"learning_rate": 0.00012993987975951904,
|
| 625 |
+
"loss": 0.1085,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.56,
|
| 630 |
+
"grad_norm": 0.63754802942276,
|
| 631 |
+
"learning_rate": 0.00012913827655310623,
|
| 632 |
+
"loss": 0.1053,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 3.6,
|
| 637 |
+
"grad_norm": 0.5717872381210327,
|
| 638 |
+
"learning_rate": 0.0001283366733466934,
|
| 639 |
+
"loss": 0.1058,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 3.64,
|
| 644 |
+
"grad_norm": 0.5148450136184692,
|
| 645 |
+
"learning_rate": 0.00012753507014028056,
|
| 646 |
+
"loss": 0.1038,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 3.68,
|
| 651 |
+
"grad_norm": 0.5819758772850037,
|
| 652 |
+
"learning_rate": 0.00012673346693386773,
|
| 653 |
+
"loss": 0.1119,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.7199999999999998,
|
| 658 |
+
"grad_norm": 0.6014116406440735,
|
| 659 |
+
"learning_rate": 0.00012593186372745493,
|
| 660 |
+
"loss": 0.1081,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 3.76,
|
| 665 |
+
"grad_norm": 0.574530303478241,
|
| 666 |
+
"learning_rate": 0.0001251302605210421,
|
| 667 |
+
"loss": 0.1023,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.8,
|
| 672 |
+
"grad_norm": 0.6252711415290833,
|
| 673 |
+
"learning_rate": 0.00012432865731462926,
|
| 674 |
+
"loss": 0.0977,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 3.84,
|
| 679 |
+
"grad_norm": 0.587404727935791,
|
| 680 |
+
"learning_rate": 0.00012352705410821643,
|
| 681 |
+
"loss": 0.0996,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 3.88,
|
| 686 |
+
"grad_norm": 0.47545889019966125,
|
| 687 |
+
"learning_rate": 0.0001227254509018036,
|
| 688 |
+
"loss": 0.105,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 3.92,
|
| 693 |
+
"grad_norm": 0.6459314823150635,
|
| 694 |
+
"learning_rate": 0.00012192384769539077,
|
| 695 |
+
"loss": 0.1003,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 3.96,
|
| 700 |
+
"grad_norm": 0.5450368523597717,
|
| 701 |
+
"learning_rate": 0.00012112224448897796,
|
| 702 |
+
"loss": 0.1124,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 4.0,
|
| 707 |
+
"grad_norm": 0.5874629616737366,
|
| 708 |
+
"learning_rate": 0.00012032064128256512,
|
| 709 |
+
"loss": 0.1089,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.04,
|
| 714 |
+
"grad_norm": 0.6516720056533813,
|
| 715 |
+
"learning_rate": 0.0001195190380761523,
|
| 716 |
+
"loss": 0.0842,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 4.08,
|
| 721 |
+
"grad_norm": 0.70643550157547,
|
| 722 |
+
"learning_rate": 0.00011871743486973947,
|
| 723 |
+
"loss": 0.075,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 4.12,
|
| 728 |
+
"grad_norm": 0.7798948287963867,
|
| 729 |
+
"learning_rate": 0.00011791583166332665,
|
| 730 |
+
"loss": 0.079,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 4.16,
|
| 735 |
+
"grad_norm": 0.7087392807006836,
|
| 736 |
+
"learning_rate": 0.00011711422845691385,
|
| 737 |
+
"loss": 0.0759,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 4.2,
|
| 742 |
+
"grad_norm": 0.5795313715934753,
|
| 743 |
+
"learning_rate": 0.00011631262525050101,
|
| 744 |
+
"loss": 0.0781,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 4.24,
|
| 749 |
+
"grad_norm": 0.6577343940734863,
|
| 750 |
+
"learning_rate": 0.00011551102204408819,
|
| 751 |
+
"loss": 0.0802,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.28,
|
| 756 |
+
"grad_norm": 0.6984887719154358,
|
| 757 |
+
"learning_rate": 0.00011470941883767536,
|
| 758 |
+
"loss": 0.0745,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.32,
|
| 763 |
+
"grad_norm": 0.7415528893470764,
|
| 764 |
+
"learning_rate": 0.00011390781563126254,
|
| 765 |
+
"loss": 0.0776,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 4.36,
|
| 770 |
+
"grad_norm": 0.46346089243888855,
|
| 771 |
+
"learning_rate": 0.00011310621242484971,
|
| 772 |
+
"loss": 0.0793,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 4.4,
|
| 777 |
+
"grad_norm": 0.5514031052589417,
|
| 778 |
+
"learning_rate": 0.00011230460921843689,
|
| 779 |
+
"loss": 0.0766,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 4.44,
|
| 784 |
+
"grad_norm": 0.5561518669128418,
|
| 785 |
+
"learning_rate": 0.00011150300601202406,
|
| 786 |
+
"loss": 0.069,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 4.48,
|
| 791 |
+
"grad_norm": 0.6889087557792664,
|
| 792 |
+
"learning_rate": 0.00011070140280561124,
|
| 793 |
+
"loss": 0.0809,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.52,
|
| 798 |
+
"grad_norm": 0.6279156804084778,
|
| 799 |
+
"learning_rate": 0.0001098997995991984,
|
| 800 |
+
"loss": 0.0684,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 4.5600000000000005,
|
| 805 |
+
"grad_norm": 0.4698103964328766,
|
| 806 |
+
"learning_rate": 0.00010909819639278558,
|
| 807 |
+
"loss": 0.0862,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 4.6,
|
| 812 |
+
"grad_norm": 0.5812392234802246,
|
| 813 |
+
"learning_rate": 0.00010829659318637275,
|
| 814 |
+
"loss": 0.0827,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 4.64,
|
| 819 |
+
"grad_norm": 0.6872307658195496,
|
| 820 |
+
"learning_rate": 0.00010749498997995993,
|
| 821 |
+
"loss": 0.0768,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 4.68,
|
| 826 |
+
"grad_norm": 0.6259503364562988,
|
| 827 |
+
"learning_rate": 0.0001066933867735471,
|
| 828 |
+
"loss": 0.0824,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.72,
|
| 833 |
+
"grad_norm": 0.5298041701316833,
|
| 834 |
+
"learning_rate": 0.00010589178356713428,
|
| 835 |
+
"loss": 0.0679,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.76,
|
| 840 |
+
"grad_norm": 0.8088281750679016,
|
| 841 |
+
"learning_rate": 0.00010509018036072145,
|
| 842 |
+
"loss": 0.0805,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.8,
|
| 847 |
+
"grad_norm": 0.49759751558303833,
|
| 848 |
+
"learning_rate": 0.00010428857715430861,
|
| 849 |
+
"loss": 0.0761,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.84,
|
| 854 |
+
"grad_norm": 0.4452104866504669,
|
| 855 |
+
"learning_rate": 0.0001034869739478958,
|
| 856 |
+
"loss": 0.0836,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.88,
|
| 861 |
+
"grad_norm": 0.4896758794784546,
|
| 862 |
+
"learning_rate": 0.00010268537074148296,
|
| 863 |
+
"loss": 0.0709,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.92,
|
| 868 |
+
"grad_norm": 0.6835769414901733,
|
| 869 |
+
"learning_rate": 0.00010188376753507014,
|
| 870 |
+
"loss": 0.0799,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.96,
|
| 875 |
+
"grad_norm": 0.7380080819129944,
|
| 876 |
+
"learning_rate": 0.00010108216432865731,
|
| 877 |
+
"loss": 0.0764,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 5.0,
|
| 882 |
+
"grad_norm": 0.7432966828346252,
|
| 883 |
+
"learning_rate": 0.00010028056112224449,
|
| 884 |
+
"loss": 0.079,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 5.04,
|
| 889 |
+
"grad_norm": 0.6639225482940674,
|
| 890 |
+
"learning_rate": 9.947895791583167e-05,
|
| 891 |
+
"loss": 0.0554,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 5.08,
|
| 896 |
+
"grad_norm": 0.5844371914863586,
|
| 897 |
+
"learning_rate": 9.867735470941885e-05,
|
| 898 |
+
"loss": 0.0502,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 5.12,
|
| 903 |
+
"grad_norm": 0.4597128629684448,
|
| 904 |
+
"learning_rate": 9.787575150300602e-05,
|
| 905 |
+
"loss": 0.0596,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 5.16,
|
| 910 |
+
"grad_norm": 0.7378055453300476,
|
| 911 |
+
"learning_rate": 9.70741482965932e-05,
|
| 912 |
+
"loss": 0.0556,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 5.2,
|
| 917 |
+
"grad_norm": 0.5333693027496338,
|
| 918 |
+
"learning_rate": 9.627254509018037e-05,
|
| 919 |
+
"loss": 0.0485,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.24,
|
| 924 |
+
"grad_norm": 0.51535564661026,
|
| 925 |
+
"learning_rate": 9.547094188376755e-05,
|
| 926 |
+
"loss": 0.0506,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 5.28,
|
| 931 |
+
"grad_norm": 0.5654531121253967,
|
| 932 |
+
"learning_rate": 9.466933867735471e-05,
|
| 933 |
+
"loss": 0.0556,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 5.32,
|
| 938 |
+
"grad_norm": 0.6834219694137573,
|
| 939 |
+
"learning_rate": 9.386773547094188e-05,
|
| 940 |
+
"loss": 0.0587,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 5.36,
|
| 945 |
+
"grad_norm": 0.5739651322364807,
|
| 946 |
+
"learning_rate": 9.306613226452906e-05,
|
| 947 |
+
"loss": 0.0533,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 5.4,
|
| 952 |
+
"grad_norm": 0.5924126505851746,
|
| 953 |
+
"learning_rate": 9.226452905811623e-05,
|
| 954 |
+
"loss": 0.0566,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 5.44,
|
| 959 |
+
"grad_norm": 0.6210141181945801,
|
| 960 |
+
"learning_rate": 9.146292585170341e-05,
|
| 961 |
+
"loss": 0.058,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 5.48,
|
| 966 |
+
"grad_norm": 0.445516437292099,
|
| 967 |
+
"learning_rate": 9.066132264529058e-05,
|
| 968 |
+
"loss": 0.0547,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 5.52,
|
| 973 |
+
"grad_norm": 0.5603981614112854,
|
| 974 |
+
"learning_rate": 8.985971943887777e-05,
|
| 975 |
+
"loss": 0.0565,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 5.5600000000000005,
|
| 980 |
+
"grad_norm": 0.6049801707267761,
|
| 981 |
+
"learning_rate": 8.905811623246494e-05,
|
| 982 |
+
"loss": 0.0579,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 5.6,
|
| 987 |
+
"grad_norm": 0.506960391998291,
|
| 988 |
+
"learning_rate": 8.825651302605212e-05,
|
| 989 |
+
"loss": 0.0493,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 5.64,
|
| 994 |
+
"grad_norm": 0.6527524590492249,
|
| 995 |
+
"learning_rate": 8.745490981963928e-05,
|
| 996 |
+
"loss": 0.0596,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 5.68,
|
| 1001 |
+
"grad_norm": 0.6459997296333313,
|
| 1002 |
+
"learning_rate": 8.665330661322647e-05,
|
| 1003 |
+
"loss": 0.0512,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 5.72,
|
| 1008 |
+
"grad_norm": 0.4932115375995636,
|
| 1009 |
+
"learning_rate": 8.585170340681363e-05,
|
| 1010 |
+
"loss": 0.0573,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 5.76,
|
| 1015 |
+
"grad_norm": 0.7221292853355408,
|
| 1016 |
+
"learning_rate": 8.50501002004008e-05,
|
| 1017 |
+
"loss": 0.0605,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 5.8,
|
| 1022 |
+
"grad_norm": 0.43518489599227905,
|
| 1023 |
+
"learning_rate": 8.424849699398798e-05,
|
| 1024 |
+
"loss": 0.0519,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 5.84,
|
| 1029 |
+
"grad_norm": 0.7757974863052368,
|
| 1030 |
+
"learning_rate": 8.344689378757515e-05,
|
| 1031 |
+
"loss": 0.0567,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 5.88,
|
| 1036 |
+
"grad_norm": 0.6453976631164551,
|
| 1037 |
+
"learning_rate": 8.264529058116233e-05,
|
| 1038 |
+
"loss": 0.0596,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 5.92,
|
| 1043 |
+
"grad_norm": 0.6767393946647644,
|
| 1044 |
+
"learning_rate": 8.18436873747495e-05,
|
| 1045 |
+
"loss": 0.0597,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 5.96,
|
| 1050 |
+
"grad_norm": 0.6432074904441833,
|
| 1051 |
+
"learning_rate": 8.104208416833668e-05,
|
| 1052 |
+
"loss": 0.0608,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 6.0,
|
| 1057 |
+
"grad_norm": 0.4562494456768036,
|
| 1058 |
+
"learning_rate": 8.024048096192384e-05,
|
| 1059 |
+
"loss": 0.0519,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 6.04,
|
| 1064 |
+
"grad_norm": 0.5894383788108826,
|
| 1065 |
+
"learning_rate": 7.943887775551102e-05,
|
| 1066 |
+
"loss": 0.0376,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 6.08,
|
| 1071 |
+
"grad_norm": 0.5043510794639587,
|
| 1072 |
+
"learning_rate": 7.86372745490982e-05,
|
| 1073 |
+
"loss": 0.0364,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 6.12,
|
| 1078 |
+
"grad_norm": 0.753901481628418,
|
| 1079 |
+
"learning_rate": 7.783567134268538e-05,
|
| 1080 |
+
"loss": 0.0387,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 6.16,
|
| 1085 |
+
"grad_norm": 0.5726089477539062,
|
| 1086 |
+
"learning_rate": 7.703406813627255e-05,
|
| 1087 |
+
"loss": 0.0362,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 6.2,
|
| 1092 |
+
"grad_norm": 0.7004730701446533,
|
| 1093 |
+
"learning_rate": 7.623246492985973e-05,
|
| 1094 |
+
"loss": 0.0327,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 6.24,
|
| 1099 |
+
"grad_norm": 0.522422194480896,
|
| 1100 |
+
"learning_rate": 7.54308617234469e-05,
|
| 1101 |
+
"loss": 0.0326,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 6.28,
|
| 1106 |
+
"grad_norm": 0.7951876521110535,
|
| 1107 |
+
"learning_rate": 7.462925851703407e-05,
|
| 1108 |
+
"loss": 0.0396,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 6.32,
|
| 1113 |
+
"grad_norm": 0.5453311800956726,
|
| 1114 |
+
"learning_rate": 7.382765531062125e-05,
|
| 1115 |
+
"loss": 0.0366,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 6.36,
|
| 1120 |
+
"grad_norm": 0.7363042235374451,
|
| 1121 |
+
"learning_rate": 7.302605210420841e-05,
|
| 1122 |
+
"loss": 0.0403,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 6.4,
|
| 1127 |
+
"grad_norm": 0.7100352048873901,
|
| 1128 |
+
"learning_rate": 7.22244488977956e-05,
|
| 1129 |
+
"loss": 0.0392,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 6.44,
|
| 1134 |
+
"grad_norm": 0.9634444713592529,
|
| 1135 |
+
"learning_rate": 7.142284569138276e-05,
|
| 1136 |
+
"loss": 0.0398,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 6.48,
|
| 1141 |
+
"grad_norm": 0.6527013182640076,
|
| 1142 |
+
"learning_rate": 7.062124248496994e-05,
|
| 1143 |
+
"loss": 0.0369,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 6.52,
|
| 1148 |
+
"grad_norm": 0.5168381333351135,
|
| 1149 |
+
"learning_rate": 6.981963927855711e-05,
|
| 1150 |
+
"loss": 0.0399,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 6.5600000000000005,
|
| 1155 |
+
"grad_norm": 0.5673239827156067,
|
| 1156 |
+
"learning_rate": 6.901803607214429e-05,
|
| 1157 |
+
"loss": 0.0364,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 6.6,
|
| 1162 |
+
"grad_norm": 0.4419175684452057,
|
| 1163 |
+
"learning_rate": 6.821643286573146e-05,
|
| 1164 |
+
"loss": 0.034,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 6.64,
|
| 1169 |
+
"grad_norm": 0.5721241235733032,
|
| 1170 |
+
"learning_rate": 6.741482965931865e-05,
|
| 1171 |
+
"loss": 0.0407,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 6.68,
|
| 1176 |
+
"grad_norm": 0.8612061738967896,
|
| 1177 |
+
"learning_rate": 6.661322645290582e-05,
|
| 1178 |
+
"loss": 0.0369,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 6.72,
|
| 1183 |
+
"grad_norm": 0.6673168540000916,
|
| 1184 |
+
"learning_rate": 6.581162324649299e-05,
|
| 1185 |
+
"loss": 0.034,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 6.76,
|
| 1190 |
+
"grad_norm": 0.816121518611908,
|
| 1191 |
+
"learning_rate": 6.501002004008017e-05,
|
| 1192 |
+
"loss": 0.0372,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 6.8,
|
| 1197 |
+
"grad_norm": 0.5812684893608093,
|
| 1198 |
+
"learning_rate": 6.420841683366733e-05,
|
| 1199 |
+
"loss": 0.0379,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 6.84,
|
| 1204 |
+
"grad_norm": 0.6159153580665588,
|
| 1205 |
+
"learning_rate": 6.340681362725451e-05,
|
| 1206 |
+
"loss": 0.0346,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 6.88,
|
| 1211 |
+
"grad_norm": 0.809228241443634,
|
| 1212 |
+
"learning_rate": 6.260521042084168e-05,
|
| 1213 |
+
"loss": 0.0486,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 6.92,
|
| 1218 |
+
"grad_norm": 0.6421244740486145,
|
| 1219 |
+
"learning_rate": 6.180360721442886e-05,
|
| 1220 |
+
"loss": 0.0387,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 6.96,
|
| 1225 |
+
"grad_norm": 0.7321845889091492,
|
| 1226 |
+
"learning_rate": 6.1002004008016036e-05,
|
| 1227 |
+
"loss": 0.0395,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 7.0,
|
| 1232 |
+
"grad_norm": 0.7744494676589966,
|
| 1233 |
+
"learning_rate": 6.020040080160321e-05,
|
| 1234 |
+
"loss": 0.0434,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 7.04,
|
| 1239 |
+
"grad_norm": 0.6629056334495544,
|
| 1240 |
+
"learning_rate": 5.9398797595190384e-05,
|
| 1241 |
+
"loss": 0.0229,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 7.08,
|
| 1246 |
+
"grad_norm": 0.5707855224609375,
|
| 1247 |
+
"learning_rate": 5.859719438877756e-05,
|
| 1248 |
+
"loss": 0.0261,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 7.12,
|
| 1253 |
+
"grad_norm": 0.5904133319854736,
|
| 1254 |
+
"learning_rate": 5.7795591182364725e-05,
|
| 1255 |
+
"loss": 0.0239,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 7.16,
|
| 1260 |
+
"grad_norm": 0.4738862216472626,
|
| 1261 |
+
"learning_rate": 5.69939879759519e-05,
|
| 1262 |
+
"loss": 0.0205,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 7.2,
|
| 1267 |
+
"grad_norm": 0.5075474381446838,
|
| 1268 |
+
"learning_rate": 5.6192384769539086e-05,
|
| 1269 |
+
"loss": 0.025,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 7.24,
|
| 1274 |
+
"grad_norm": 0.5726251602172852,
|
| 1275 |
+
"learning_rate": 5.539078156312626e-05,
|
| 1276 |
+
"loss": 0.0247,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 7.28,
|
| 1281 |
+
"grad_norm": 0.5094057321548462,
|
| 1282 |
+
"learning_rate": 5.4589178356713434e-05,
|
| 1283 |
+
"loss": 0.0189,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 7.32,
|
| 1288 |
+
"grad_norm": 0.47997888922691345,
|
| 1289 |
+
"learning_rate": 5.378757515030061e-05,
|
| 1290 |
+
"loss": 0.0221,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 7.36,
|
| 1295 |
+
"grad_norm": 0.5335679650306702,
|
| 1296 |
+
"learning_rate": 5.298597194388778e-05,
|
| 1297 |
+
"loss": 0.0239,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 7.4,
|
| 1302 |
+
"grad_norm": 0.5913345217704773,
|
| 1303 |
+
"learning_rate": 5.2184368737474955e-05,
|
| 1304 |
+
"loss": 0.0237,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 7.44,
|
| 1309 |
+
"grad_norm": 0.41921791434288025,
|
| 1310 |
+
"learning_rate": 5.138276553106213e-05,
|
| 1311 |
+
"loss": 0.0294,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 7.48,
|
| 1316 |
+
"grad_norm": 0.610205352306366,
|
| 1317 |
+
"learning_rate": 5.05811623246493e-05,
|
| 1318 |
+
"loss": 0.0237,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 7.52,
|
| 1323 |
+
"grad_norm": 0.5666776299476624,
|
| 1324 |
+
"learning_rate": 4.977955911823648e-05,
|
| 1325 |
+
"loss": 0.0236,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 7.5600000000000005,
|
| 1330 |
+
"grad_norm": 0.42802920937538147,
|
| 1331 |
+
"learning_rate": 4.897795591182365e-05,
|
| 1332 |
+
"loss": 0.0239,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 7.6,
|
| 1337 |
+
"grad_norm": 0.748887836933136,
|
| 1338 |
+
"learning_rate": 4.8176352705410824e-05,
|
| 1339 |
+
"loss": 0.0269,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 7.64,
|
| 1344 |
+
"grad_norm": 0.43109023571014404,
|
| 1345 |
+
"learning_rate": 4.7374749498998e-05,
|
| 1346 |
+
"loss": 0.0246,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 7.68,
|
| 1351 |
+
"grad_norm": 0.43403562903404236,
|
| 1352 |
+
"learning_rate": 4.657314629258517e-05,
|
| 1353 |
+
"loss": 0.0219,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 7.72,
|
| 1358 |
+
"grad_norm": 0.5174989104270935,
|
| 1359 |
+
"learning_rate": 4.5771543086172346e-05,
|
| 1360 |
+
"loss": 0.0229,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 7.76,
|
| 1365 |
+
"grad_norm": 0.5523115396499634,
|
| 1366 |
+
"learning_rate": 4.496993987975952e-05,
|
| 1367 |
+
"loss": 0.0242,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 7.8,
|
| 1372 |
+
"grad_norm": 0.4592013359069824,
|
| 1373 |
+
"learning_rate": 4.4168336673346694e-05,
|
| 1374 |
+
"loss": 0.0237,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 7.84,
|
| 1379 |
+
"grad_norm": 0.574874997138977,
|
| 1380 |
+
"learning_rate": 4.336673346693387e-05,
|
| 1381 |
+
"loss": 0.0207,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 7.88,
|
| 1386 |
+
"grad_norm": 0.6085746884346008,
|
| 1387 |
+
"learning_rate": 4.256513026052105e-05,
|
| 1388 |
+
"loss": 0.0213,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 7.92,
|
| 1393 |
+
"grad_norm": 0.5086420178413391,
|
| 1394 |
+
"learning_rate": 4.176352705410822e-05,
|
| 1395 |
+
"loss": 0.0234,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 7.96,
|
| 1400 |
+
"grad_norm": 0.4850587844848633,
|
| 1401 |
+
"learning_rate": 4.0961923847695396e-05,
|
| 1402 |
+
"loss": 0.0227,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 8.0,
|
| 1407 |
+
"grad_norm": 0.469855934381485,
|
| 1408 |
+
"learning_rate": 4.016032064128257e-05,
|
| 1409 |
+
"loss": 0.026,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 8.04,
|
| 1414 |
+
"grad_norm": 0.3705706000328064,
|
| 1415 |
+
"learning_rate": 3.9358717434869744e-05,
|
| 1416 |
+
"loss": 0.0156,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 8.08,
|
| 1421 |
+
"grad_norm": 0.5124081969261169,
|
| 1422 |
+
"learning_rate": 3.855711422845692e-05,
|
| 1423 |
+
"loss": 0.0146,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 8.12,
|
| 1428 |
+
"grad_norm": 0.3145318627357483,
|
| 1429 |
+
"learning_rate": 3.7755511022044085e-05,
|
| 1430 |
+
"loss": 0.0156,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 8.16,
|
| 1435 |
+
"grad_norm": 0.47414782643318176,
|
| 1436 |
+
"learning_rate": 3.6953907815631265e-05,
|
| 1437 |
+
"loss": 0.015,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 8.2,
|
| 1442 |
+
"grad_norm": 0.3775177597999573,
|
| 1443 |
+
"learning_rate": 3.615230460921844e-05,
|
| 1444 |
+
"loss": 0.0167,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 8.24,
|
| 1449 |
+
"grad_norm": 0.43829330801963806,
|
| 1450 |
+
"learning_rate": 3.535070140280561e-05,
|
| 1451 |
+
"loss": 0.0135,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 8.28,
|
| 1456 |
+
"grad_norm": 0.5022293329238892,
|
| 1457 |
+
"learning_rate": 3.454909819639279e-05,
|
| 1458 |
+
"loss": 0.013,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 8.32,
|
| 1463 |
+
"grad_norm": 0.27599331736564636,
|
| 1464 |
+
"learning_rate": 3.374749498997996e-05,
|
| 1465 |
+
"loss": 0.0153,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 8.36,
|
| 1470 |
+
"grad_norm": 0.82197105884552,
|
| 1471 |
+
"learning_rate": 3.2945891783567135e-05,
|
| 1472 |
+
"loss": 0.0164,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 8.4,
|
| 1477 |
+
"grad_norm": 0.20758652687072754,
|
| 1478 |
+
"learning_rate": 3.214428857715431e-05,
|
| 1479 |
+
"loss": 0.0135,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 8.44,
|
| 1484 |
+
"grad_norm": 0.29413196444511414,
|
| 1485 |
+
"learning_rate": 3.134268537074149e-05,
|
| 1486 |
+
"loss": 0.0122,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 8.48,
|
| 1491 |
+
"grad_norm": 0.38946202397346497,
|
| 1492 |
+
"learning_rate": 3.054108216432866e-05,
|
| 1493 |
+
"loss": 0.0124,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 8.52,
|
| 1498 |
+
"grad_norm": 0.3506149649620056,
|
| 1499 |
+
"learning_rate": 2.9739478957915833e-05,
|
| 1500 |
+
"loss": 0.0146,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 8.56,
|
| 1505 |
+
"grad_norm": 0.31000277400016785,
|
| 1506 |
+
"learning_rate": 2.8937875751503007e-05,
|
| 1507 |
+
"loss": 0.0143,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 8.6,
|
| 1512 |
+
"grad_norm": 0.3397505581378937,
|
| 1513 |
+
"learning_rate": 2.813627254509018e-05,
|
| 1514 |
+
"loss": 0.0146,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 8.64,
|
| 1519 |
+
"grad_norm": 0.46700355410575867,
|
| 1520 |
+
"learning_rate": 2.7334669338677355e-05,
|
| 1521 |
+
"loss": 0.0148,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 8.68,
|
| 1526 |
+
"grad_norm": 0.43519356846809387,
|
| 1527 |
+
"learning_rate": 2.653306613226453e-05,
|
| 1528 |
+
"loss": 0.0116,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 8.72,
|
| 1533 |
+
"grad_norm": 0.3716731369495392,
|
| 1534 |
+
"learning_rate": 2.5731462925851706e-05,
|
| 1535 |
+
"loss": 0.0138,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 8.76,
|
| 1540 |
+
"grad_norm": 0.43048301339149475,
|
| 1541 |
+
"learning_rate": 2.4929859719438877e-05,
|
| 1542 |
+
"loss": 0.0116,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 8.8,
|
| 1547 |
+
"grad_norm": 0.586083173751831,
|
| 1548 |
+
"learning_rate": 2.4128256513026054e-05,
|
| 1549 |
+
"loss": 0.0118,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 8.84,
|
| 1554 |
+
"grad_norm": 0.5221779346466064,
|
| 1555 |
+
"learning_rate": 2.3326653306613228e-05,
|
| 1556 |
+
"loss": 0.013,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 8.88,
|
| 1561 |
+
"grad_norm": 0.4493936598300934,
|
| 1562 |
+
"learning_rate": 2.25250501002004e-05,
|
| 1563 |
+
"loss": 0.0161,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 8.92,
|
| 1568 |
+
"grad_norm": 0.629578173160553,
|
| 1569 |
+
"learning_rate": 2.172344689378758e-05,
|
| 1570 |
+
"loss": 0.0136,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 8.96,
|
| 1575 |
+
"grad_norm": 0.445533812046051,
|
| 1576 |
+
"learning_rate": 2.092184368737475e-05,
|
| 1577 |
+
"loss": 0.0116,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 9.0,
|
| 1582 |
+
"grad_norm": 0.4184821546077728,
|
| 1583 |
+
"learning_rate": 2.0120240480961923e-05,
|
| 1584 |
+
"loss": 0.0108,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 9.04,
|
| 1589 |
+
"grad_norm": 0.18857356905937195,
|
| 1590 |
+
"learning_rate": 1.9318637274549097e-05,
|
| 1591 |
+
"loss": 0.0088,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 9.08,
|
| 1596 |
+
"grad_norm": 0.3190409541130066,
|
| 1597 |
+
"learning_rate": 1.8517034068136274e-05,
|
| 1598 |
+
"loss": 0.01,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 9.12,
|
| 1603 |
+
"grad_norm": 0.23249320685863495,
|
| 1604 |
+
"learning_rate": 1.7715430861723448e-05,
|
| 1605 |
+
"loss": 0.0079,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 9.16,
|
| 1610 |
+
"grad_norm": 0.29095596075057983,
|
| 1611 |
+
"learning_rate": 1.6913827655310622e-05,
|
| 1612 |
+
"loss": 0.0071,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 9.2,
|
| 1617 |
+
"grad_norm": 0.2573925852775574,
|
| 1618 |
+
"learning_rate": 1.6112224448897796e-05,
|
| 1619 |
+
"loss": 0.0069,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 9.24,
|
| 1624 |
+
"grad_norm": 0.14938637614250183,
|
| 1625 |
+
"learning_rate": 1.531062124248497e-05,
|
| 1626 |
+
"loss": 0.0098,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 9.28,
|
| 1631 |
+
"grad_norm": 0.2534259855747223,
|
| 1632 |
+
"learning_rate": 1.4509018036072145e-05,
|
| 1633 |
+
"loss": 0.0082,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 9.32,
|
| 1638 |
+
"grad_norm": 0.22516509890556335,
|
| 1639 |
+
"learning_rate": 1.3707414829659317e-05,
|
| 1640 |
+
"loss": 0.0085,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 9.36,
|
| 1645 |
+
"grad_norm": 0.1505204439163208,
|
| 1646 |
+
"learning_rate": 1.2905811623246495e-05,
|
| 1647 |
+
"loss": 0.0071,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 9.4,
|
| 1652 |
+
"grad_norm": 0.2739148736000061,
|
| 1653 |
+
"learning_rate": 1.2104208416833669e-05,
|
| 1654 |
+
"loss": 0.0096,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 9.44,
|
| 1659 |
+
"grad_norm": 0.6390639543533325,
|
| 1660 |
+
"learning_rate": 1.1302605210420842e-05,
|
| 1661 |
+
"loss": 0.0072,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 9.48,
|
| 1666 |
+
"grad_norm": 0.3169389069080353,
|
| 1667 |
+
"learning_rate": 1.0501002004008016e-05,
|
| 1668 |
+
"loss": 0.0069,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 9.52,
|
| 1673 |
+
"grad_norm": 0.20216785371303558,
|
| 1674 |
+
"learning_rate": 9.699398797595192e-06,
|
| 1675 |
+
"loss": 0.0069,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 9.56,
|
| 1680 |
+
"grad_norm": 0.21013762056827545,
|
| 1681 |
+
"learning_rate": 8.897795591182364e-06,
|
| 1682 |
+
"loss": 0.0084,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 9.6,
|
| 1687 |
+
"grad_norm": 0.21212315559387207,
|
| 1688 |
+
"learning_rate": 8.09619238476954e-06,
|
| 1689 |
+
"loss": 0.0062,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 9.64,
|
| 1694 |
+
"grad_norm": 0.23686246573925018,
|
| 1695 |
+
"learning_rate": 7.294589178356714e-06,
|
| 1696 |
+
"loss": 0.0102,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 9.68,
|
| 1701 |
+
"grad_norm": 0.2729778587818146,
|
| 1702 |
+
"learning_rate": 6.492985971943888e-06,
|
| 1703 |
+
"loss": 0.007,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 9.72,
|
| 1708 |
+
"grad_norm": 0.19149982929229736,
|
| 1709 |
+
"learning_rate": 5.691382765531062e-06,
|
| 1710 |
+
"loss": 0.0087,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 9.76,
|
| 1715 |
+
"grad_norm": 0.20065714418888092,
|
| 1716 |
+
"learning_rate": 4.889779559118237e-06,
|
| 1717 |
+
"loss": 0.0067,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 9.8,
|
| 1722 |
+
"grad_norm": 0.29255425930023193,
|
| 1723 |
+
"learning_rate": 4.0881763527054114e-06,
|
| 1724 |
+
"loss": 0.008,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 9.84,
|
| 1729 |
+
"grad_norm": 0.254088819026947,
|
| 1730 |
+
"learning_rate": 3.2865731462925853e-06,
|
| 1731 |
+
"loss": 0.0065,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 9.88,
|
| 1736 |
+
"grad_norm": 0.2548205554485321,
|
| 1737 |
+
"learning_rate": 2.4849699398797596e-06,
|
| 1738 |
+
"loss": 0.0082,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 9.92,
|
| 1743 |
+
"grad_norm": 0.2607119679450989,
|
| 1744 |
+
"learning_rate": 1.6833667334669339e-06,
|
| 1745 |
+
"loss": 0.0091,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 9.96,
|
| 1750 |
+
"grad_norm": 0.2837667465209961,
|
| 1751 |
+
"learning_rate": 8.817635270541082e-07,
|
| 1752 |
+
"loss": 0.0083,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 10.0,
|
| 1757 |
+
"grad_norm": 0.24088458716869354,
|
| 1758 |
+
"learning_rate": 8.016032064128256e-08,
|
| 1759 |
+
"loss": 0.0089,
|
| 1760 |
+
"step": 2500
|
| 1761 |
}
|
| 1762 |
],
|
| 1763 |
"logging_steps": 10,
|
| 1764 |
+
"max_steps": 2500,
|
| 1765 |
"num_input_tokens_seen": 0,
|
| 1766 |
+
"num_train_epochs": 10,
|
| 1767 |
"save_steps": 500,
|
| 1768 |
"stateful_callbacks": {
|
| 1769 |
"TrainerControl": {
|
|
|
|
| 1777 |
"attributes": {}
|
| 1778 |
}
|
| 1779 |
},
|
| 1780 |
+
"total_flos": 3.17341887028272e+17,
|
| 1781 |
"train_batch_size": 20,
|
| 1782 |
"trial_name": null,
|
| 1783 |
"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 5841
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0b7b74174fe53181dd86ef5b8c6b0f001c2ddbd4fbc4f691eacd79b321210ee
|
| 3 |
size 5841
|