Upload 12 files
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +285 -3085
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -20,12 +20,12 @@
|
|
| 20 |
"revision": null,
|
| 21 |
"target_modules": [
|
| 22 |
"q_proj",
|
| 23 |
-
"down_proj",
|
| 24 |
-
"gate_proj",
|
| 25 |
-
"up_proj",
|
| 26 |
"v_proj",
|
|
|
|
|
|
|
| 27 |
"k_proj",
|
| 28 |
-
"
|
|
|
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"use_rslora": false
|
|
|
|
| 20 |
"revision": null,
|
| 21 |
"target_modules": [
|
| 22 |
"q_proj",
|
|
|
|
|
|
|
|
|
|
| 23 |
"v_proj",
|
| 24 |
+
"gate_proj",
|
| 25 |
+
"o_proj",
|
| 26 |
"k_proj",
|
| 27 |
+
"down_proj",
|
| 28 |
+
"up_proj"
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"use_rslora": false
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 100059752
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db3692a4b95774f2053760024a50fe3e61f64ed72e81fad46448118c57b87026
|
| 3 |
size 100059752
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5299966a3e9ee146dacc89587e580e10540d080929659b4850cf35c9024a1da9
|
| 3 |
+
size 50545780
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f3e429bde762c36daf33ddcd863fe8e4a9386fce953bf5dd0a2dc17c1e3bc5a
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f076786bfa235d4d3bba40d361bfdd13d4dc22aa2b171ee38b02843b1000a93
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,3520 +1,720 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 1.0,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 0.
|
| 15 |
-
"loss": 5.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 2.0,
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 0.
|
| 22 |
-
"loss": 5.
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 3.0,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate": 0.
|
| 29 |
-
"loss": 5.
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 4.0,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 0.
|
| 36 |
-
"loss":
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 5.0,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 0.
|
| 43 |
-
"loss": 4.
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 6.0,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss":
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 7.0,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 0.
|
| 57 |
-
"loss": 3.
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 8.0,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate": 0.
|
| 64 |
-
"loss": 3.
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 9.0,
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate": 0.
|
| 71 |
-
"loss": 2.
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 10.0,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate": 0.
|
| 78 |
-
"loss": 2.
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 11.0,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate": 0.
|
| 85 |
-
"loss":
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 12.0,
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 0.
|
| 92 |
-
"loss":
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 13.0,
|
| 97 |
-
"grad_norm": 17.
|
| 98 |
-
"learning_rate": 0.
|
| 99 |
-
"loss":
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 14.0,
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate": 0.
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 15.0,
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate": 0.
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 16.0,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate": 0.
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 17.0,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 0.
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 18.0,
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate": 0.
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 19.0,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate": 0.
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 20.0,
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
-
"learning_rate": 0.
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 21.0,
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
-
"learning_rate": 0.
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 22.0,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
-
"learning_rate": 0.
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 23.0,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
-
"learning_rate": 0.
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 24.0,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
-
"learning_rate": 0.
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 25.0,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate": 0.
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 26.0,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate": 0.
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 27.0,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate": 0.
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 28.0,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
-
"learning_rate": 0.
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 29.0,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate": 0.
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 30.0,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate": 0.
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 31.0,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
-
"learning_rate": 0.
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 32.0,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
-
"learning_rate": 0.
|
| 232 |
"loss": 0.0048,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 33.0,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate": 0.
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 34.0,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate": 0.
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 35.0,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
-
"learning_rate": 0.
|
| 253 |
"loss": 0.0048,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 36.0,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
-
"learning_rate": 0.
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 37.0,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
-
"learning_rate": 0.
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 38.0,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate": 0.
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 39.0,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
-
"learning_rate": 0.
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 40.0,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
-
"learning_rate": 0.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 41.0,
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
-
"learning_rate": 0.
|
| 295 |
"loss": 0.0048,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 42.0,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
-
"learning_rate": 0.
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 43.0,
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
-
"learning_rate": 0.
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 44.0,
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
-
"learning_rate": 0.
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 45.0,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate": 0.
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 46.0,
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
-
"learning_rate": 0.
|
| 330 |
"loss": 0.005,
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 47.0,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate": 0.
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 48.0,
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
-
"learning_rate": 0.
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 49.0,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
-
"learning_rate": 0.
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 50.0,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
-
"learning_rate": 0.
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 51.0,
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
-
"learning_rate": 0.
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 52.0,
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
-
"learning_rate": 0.
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 53.0,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
-
"learning_rate":
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 54.0,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 55.0,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
-
"learning_rate":
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 56.0,
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
-
"learning_rate":
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 57.0,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate":
|
| 407 |
"loss": 0.0047,
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 58.0,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
-
"learning_rate":
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 59.0,
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
-
"learning_rate":
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 60.0,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 61.0,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate":
|
| 435 |
"loss": 0.0047,
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 62.0,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate":
|
| 442 |
"loss": 0.0047,
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 63.0,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
-
"learning_rate":
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 64.0,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
-
"learning_rate":
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 65.0,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 66.0,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 67.0,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
-
"learning_rate":
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 68.0,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
-
"learning_rate":
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 69.0,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
-
"learning_rate":
|
| 491 |
"loss": 0.0047,
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 70.0,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
-
"learning_rate":
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 71.0,
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
-
"learning_rate":
|
| 505 |
"loss": 0.0047,
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 72.0,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 73.0,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
-
"learning_rate":
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 74.0,
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
-
"learning_rate":
|
| 526 |
"loss": 0.0047,
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 75.0,
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
-
"learning_rate":
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 76.0,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
-
"learning_rate":
|
| 540 |
"loss": 0.0047,
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 77.0,
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 78.0,
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
-
"learning_rate":
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 79.0,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
-
"learning_rate":
|
| 561 |
"loss": 0.0047,
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 80.0,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
-
"learning_rate":
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 81.0,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
-
"learning_rate":
|
| 575 |
"loss": 0.0047,
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 82.0,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
-
"learning_rate":
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 83.0,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
-
"learning_rate":
|
| 589 |
"loss": 0.0047,
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 84.0,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 85.0,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
-
"learning_rate":
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 86.0,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
-
"learning_rate":
|
| 610 |
"loss": 0.0047,
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 87.0,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 88.0,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
-
"learning_rate":
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 89.0,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
-
"learning_rate":
|
| 631 |
"loss": 0.0045,
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 90.0,
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 91.0,
|
| 643 |
-
"grad_norm": 0.
|
| 644 |
-
"learning_rate":
|
| 645 |
"loss": 0.0047,
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 92.0,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
-
"learning_rate":
|
| 652 |
"loss": 0.0047,
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 93.0,
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
-
"learning_rate":
|
| 659 |
"loss": 0.0047,
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 94.0,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
-
"learning_rate":
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 95.0,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
-
"learning_rate":
|
| 673 |
"loss": 0.0045,
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 96.0,
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
-
"learning_rate":
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 97.0,
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
-
"learning_rate":
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 98.0,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
-
"learning_rate":
|
| 694 |
-
"loss": 0.
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 99.0,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 100.0,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
-
"learning_rate":
|
| 708 |
-
"loss": 0.0048,
|
| 709 |
-
"step": 100
|
| 710 |
-
},
|
| 711 |
-
{
|
| 712 |
-
"epoch": 101.0,
|
| 713 |
-
"grad_norm": 0.007468740921467543,
|
| 714 |
-
"learning_rate": 0.0001610441767068273,
|
| 715 |
-
"loss": 0.0047,
|
| 716 |
-
"step": 101
|
| 717 |
-
},
|
| 718 |
-
{
|
| 719 |
-
"epoch": 102.0,
|
| 720 |
-
"grad_norm": 0.052220698446035385,
|
| 721 |
-
"learning_rate": 0.0001606425702811245,
|
| 722 |
-
"loss": 0.0051,
|
| 723 |
-
"step": 102
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 103.0,
|
| 727 |
-
"grad_norm": 0.010211293585598469,
|
| 728 |
-
"learning_rate": 0.0001602409638554217,
|
| 729 |
-
"loss": 0.0047,
|
| 730 |
-
"step": 103
|
| 731 |
-
},
|
| 732 |
-
{
|
| 733 |
-
"epoch": 104.0,
|
| 734 |
-
"grad_norm": 0.0423286072909832,
|
| 735 |
-
"learning_rate": 0.00015983935742971888,
|
| 736 |
-
"loss": 0.0048,
|
| 737 |
-
"step": 104
|
| 738 |
-
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 105.0,
|
| 741 |
-
"grad_norm": 0.02506270445883274,
|
| 742 |
-
"learning_rate": 0.00015943775100401607,
|
| 743 |
-
"loss": 0.0048,
|
| 744 |
-
"step": 105
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 106.0,
|
| 748 |
-
"grad_norm": 0.04458456113934517,
|
| 749 |
-
"learning_rate": 0.00015903614457831326,
|
| 750 |
-
"loss": 0.0048,
|
| 751 |
-
"step": 106
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 107.0,
|
| 755 |
-
"grad_norm": 0.02753458172082901,
|
| 756 |
-
"learning_rate": 0.00015863453815261045,
|
| 757 |
-
"loss": 0.0048,
|
| 758 |
-
"step": 107
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 108.0,
|
| 762 |
-
"grad_norm": 0.04234497249126434,
|
| 763 |
-
"learning_rate": 0.00015823293172690763,
|
| 764 |
-
"loss": 0.0048,
|
| 765 |
-
"step": 108
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 109.0,
|
| 769 |
-
"grad_norm": 0.03287286311388016,
|
| 770 |
-
"learning_rate": 0.00015783132530120482,
|
| 771 |
-
"loss": 0.0046,
|
| 772 |
-
"step": 109
|
| 773 |
-
},
|
| 774 |
-
{
|
| 775 |
-
"epoch": 110.0,
|
| 776 |
-
"grad_norm": 0.03671009838581085,
|
| 777 |
-
"learning_rate": 0.000157429718875502,
|
| 778 |
-
"loss": 0.0046,
|
| 779 |
-
"step": 110
|
| 780 |
-
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 111.0,
|
| 783 |
-
"grad_norm": 0.044372282922267914,
|
| 784 |
-
"learning_rate": 0.0001570281124497992,
|
| 785 |
-
"loss": 0.0048,
|
| 786 |
-
"step": 111
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 112.0,
|
| 790 |
-
"grad_norm": 0.03433435037732124,
|
| 791 |
-
"learning_rate": 0.0001566265060240964,
|
| 792 |
-
"loss": 0.005,
|
| 793 |
-
"step": 112
|
| 794 |
-
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 113.0,
|
| 797 |
-
"grad_norm": 0.04201051965355873,
|
| 798 |
-
"learning_rate": 0.00015622489959839358,
|
| 799 |
-
"loss": 0.0048,
|
| 800 |
-
"step": 113
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 114.0,
|
| 804 |
-
"grad_norm": 0.027228495106101036,
|
| 805 |
-
"learning_rate": 0.00015582329317269077,
|
| 806 |
-
"loss": 0.0048,
|
| 807 |
-
"step": 114
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 115.0,
|
| 811 |
-
"grad_norm": 0.04395684599876404,
|
| 812 |
-
"learning_rate": 0.00015542168674698796,
|
| 813 |
-
"loss": 0.0048,
|
| 814 |
-
"step": 115
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 116.0,
|
| 818 |
-
"grad_norm": 0.0418144129216671,
|
| 819 |
-
"learning_rate": 0.00015502008032128515,
|
| 820 |
-
"loss": 0.0048,
|
| 821 |
-
"step": 116
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 117.0,
|
| 825 |
-
"grad_norm": 0.015051459893584251,
|
| 826 |
-
"learning_rate": 0.00015461847389558234,
|
| 827 |
-
"loss": 0.0045,
|
| 828 |
-
"step": 117
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 118.0,
|
| 832 |
-
"grad_norm": 0.010102898813784122,
|
| 833 |
-
"learning_rate": 0.00015421686746987953,
|
| 834 |
-
"loss": 0.0047,
|
| 835 |
-
"step": 118
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 119.0,
|
| 839 |
-
"grad_norm": 0.026845330372452736,
|
| 840 |
-
"learning_rate": 0.00015381526104417672,
|
| 841 |
-
"loss": 0.0048,
|
| 842 |
-
"step": 119
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 120.0,
|
| 846 |
-
"grad_norm": 0.010000316426157951,
|
| 847 |
-
"learning_rate": 0.0001534136546184739,
|
| 848 |
-
"loss": 0.0047,
|
| 849 |
-
"step": 120
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 121.0,
|
| 853 |
-
"grad_norm": 0.041023530066013336,
|
| 854 |
-
"learning_rate": 0.0001530120481927711,
|
| 855 |
-
"loss": 0.0048,
|
| 856 |
-
"step": 121
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 122.0,
|
| 860 |
-
"grad_norm": 0.0072840056382119656,
|
| 861 |
-
"learning_rate": 0.00015261044176706828,
|
| 862 |
-
"loss": 0.0047,
|
| 863 |
-
"step": 122
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 123.0,
|
| 867 |
-
"grad_norm": 0.02653086557984352,
|
| 868 |
-
"learning_rate": 0.00015220883534136547,
|
| 869 |
-
"loss": 0.0048,
|
| 870 |
-
"step": 123
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 124.0,
|
| 874 |
-
"grad_norm": 0.01739220879971981,
|
| 875 |
-
"learning_rate": 0.00015180722891566266,
|
| 876 |
-
"loss": 0.005,
|
| 877 |
-
"step": 124
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 125.0,
|
| 881 |
-
"grad_norm": 0.007263750769197941,
|
| 882 |
-
"learning_rate": 0.00015140562248995985,
|
| 883 |
-
"loss": 0.0047,
|
| 884 |
-
"step": 125
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 126.0,
|
| 888 |
-
"grad_norm": 0.04079929739236832,
|
| 889 |
-
"learning_rate": 0.00015100401606425701,
|
| 890 |
-
"loss": 0.0048,
|
| 891 |
-
"step": 126
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 127.0,
|
| 895 |
-
"grad_norm": 0.009831869974732399,
|
| 896 |
-
"learning_rate": 0.00015060240963855423,
|
| 897 |
-
"loss": 0.0047,
|
| 898 |
-
"step": 127
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 128.0,
|
| 902 |
-
"grad_norm": 0.026599382981657982,
|
| 903 |
-
"learning_rate": 0.00015020080321285142,
|
| 904 |
-
"loss": 0.0048,
|
| 905 |
-
"step": 128
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 129.0,
|
| 909 |
-
"grad_norm": 0.018224092200398445,
|
| 910 |
-
"learning_rate": 0.0001497991967871486,
|
| 911 |
-
"loss": 0.0047,
|
| 912 |
-
"step": 129
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 130.0,
|
| 916 |
-
"grad_norm": 0.02792244777083397,
|
| 917 |
-
"learning_rate": 0.00014939759036144577,
|
| 918 |
-
"loss": 0.0047,
|
| 919 |
-
"step": 130
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 131.0,
|
| 923 |
-
"grad_norm": 0.024047773331403732,
|
| 924 |
-
"learning_rate": 0.000148995983935743,
|
| 925 |
-
"loss": 0.0048,
|
| 926 |
-
"step": 131
|
| 927 |
-
},
|
| 928 |
-
{
|
| 929 |
-
"epoch": 132.0,
|
| 930 |
-
"grad_norm": 0.018245236948132515,
|
| 931 |
-
"learning_rate": 0.00014859437751004018,
|
| 932 |
-
"loss": 0.0047,
|
| 933 |
-
"step": 132
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 133.0,
|
| 937 |
-
"grad_norm": 0.026581475511193275,
|
| 938 |
-
"learning_rate": 0.00014819277108433737,
|
| 939 |
-
"loss": 0.0048,
|
| 940 |
-
"step": 133
|
| 941 |
-
},
|
| 942 |
-
{
|
| 943 |
-
"epoch": 134.0,
|
| 944 |
-
"grad_norm": 0.007153503131121397,
|
| 945 |
-
"learning_rate": 0.00014779116465863453,
|
| 946 |
-
"loss": 0.0047,
|
| 947 |
-
"step": 134
|
| 948 |
-
},
|
| 949 |
-
{
|
| 950 |
-
"epoch": 135.0,
|
| 951 |
-
"grad_norm": 0.0323675237596035,
|
| 952 |
-
"learning_rate": 0.00014738955823293172,
|
| 953 |
-
"loss": 0.0048,
|
| 954 |
-
"step": 135
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 136.0,
|
| 958 |
-
"grad_norm": 0.005924216937273741,
|
| 959 |
-
"learning_rate": 0.00014698795180722893,
|
| 960 |
-
"loss": 0.0046,
|
| 961 |
-
"step": 136
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 137.0,
|
| 965 |
-
"grad_norm": 0.026371264830231667,
|
| 966 |
-
"learning_rate": 0.00014658634538152612,
|
| 967 |
-
"loss": 0.0048,
|
| 968 |
-
"step": 137
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 138.0,
|
| 972 |
-
"grad_norm": 0.002094594296067953,
|
| 973 |
-
"learning_rate": 0.00014618473895582328,
|
| 974 |
-
"loss": 0.0047,
|
| 975 |
-
"step": 138
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 139.0,
|
| 979 |
-
"grad_norm": 0.023835647851228714,
|
| 980 |
-
"learning_rate": 0.00014578313253012047,
|
| 981 |
-
"loss": 0.0048,
|
| 982 |
-
"step": 139
|
| 983 |
-
},
|
| 984 |
-
{
|
| 985 |
-
"epoch": 140.0,
|
| 986 |
-
"grad_norm": 0.007189361844211817,
|
| 987 |
-
"learning_rate": 0.0001453815261044177,
|
| 988 |
"loss": 0.0047,
|
| 989 |
-
"step":
|
| 990 |
-
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 141.0,
|
| 993 |
-
"grad_norm": 0.021903127431869507,
|
| 994 |
-
"learning_rate": 0.00014497991967871488,
|
| 995 |
-
"loss": 0.0049,
|
| 996 |
-
"step": 141
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 142.0,
|
| 1000 |
-
"grad_norm": 0.009862157516181469,
|
| 1001 |
-
"learning_rate": 0.00014457831325301204,
|
| 1002 |
-
"loss": 0.0047,
|
| 1003 |
-
"step": 142
|
| 1004 |
-
},
|
| 1005 |
-
{
|
| 1006 |
-
"epoch": 143.0,
|
| 1007 |
-
"grad_norm": 0.01557657215744257,
|
| 1008 |
-
"learning_rate": 0.00014417670682730923,
|
| 1009 |
-
"loss": 0.0047,
|
| 1010 |
-
"step": 143
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 144.0,
|
| 1014 |
-
"grad_norm": 0.01178675051778555,
|
| 1015 |
-
"learning_rate": 0.00014377510040160642,
|
| 1016 |
-
"loss": 0.0048,
|
| 1017 |
-
"step": 144
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 145.0,
|
| 1021 |
-
"grad_norm": 0.009945346042513847,
|
| 1022 |
-
"learning_rate": 0.00014337349397590364,
|
| 1023 |
-
"loss": 0.0047,
|
| 1024 |
-
"step": 145
|
| 1025 |
-
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 146.0,
|
| 1028 |
-
"grad_norm": 0.013681800104677677,
|
| 1029 |
-
"learning_rate": 0.0001429718875502008,
|
| 1030 |
-
"loss": 0.0048,
|
| 1031 |
-
"step": 146
|
| 1032 |
-
},
|
| 1033 |
-
{
|
| 1034 |
-
"epoch": 147.0,
|
| 1035 |
-
"grad_norm": 0.0072199697606265545,
|
| 1036 |
-
"learning_rate": 0.000142570281124498,
|
| 1037 |
-
"loss": 0.0047,
|
| 1038 |
-
"step": 147
|
| 1039 |
-
},
|
| 1040 |
-
{
|
| 1041 |
-
"epoch": 148.0,
|
| 1042 |
-
"grad_norm": 0.015607825480401516,
|
| 1043 |
-
"learning_rate": 0.00014216867469879518,
|
| 1044 |
-
"loss": 0.0047,
|
| 1045 |
-
"step": 148
|
| 1046 |
-
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 149.0,
|
| 1049 |
-
"grad_norm": 0.00990898534655571,
|
| 1050 |
-
"learning_rate": 0.0001417670682730924,
|
| 1051 |
-
"loss": 0.0047,
|
| 1052 |
-
"step": 149
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 150.0,
|
| 1056 |
-
"grad_norm": 0.009972168132662773,
|
| 1057 |
-
"learning_rate": 0.00014136546184738956,
|
| 1058 |
-
"loss": 0.0047,
|
| 1059 |
-
"step": 150
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 151.0,
|
| 1063 |
-
"grad_norm": 0.007324350066483021,
|
| 1064 |
-
"learning_rate": 0.00014096385542168674,
|
| 1065 |
-
"loss": 0.0047,
|
| 1066 |
-
"step": 151
|
| 1067 |
-
},
|
| 1068 |
-
{
|
| 1069 |
-
"epoch": 152.0,
|
| 1070 |
-
"grad_norm": 0.007397031411528587,
|
| 1071 |
-
"learning_rate": 0.00014056224899598393,
|
| 1072 |
-
"loss": 0.0047,
|
| 1073 |
-
"step": 152
|
| 1074 |
-
},
|
| 1075 |
-
{
|
| 1076 |
-
"epoch": 153.0,
|
| 1077 |
-
"grad_norm": 0.00632756482809782,
|
| 1078 |
-
"learning_rate": 0.00014016064257028115,
|
| 1079 |
-
"loss": 0.0046,
|
| 1080 |
-
"step": 153
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 154.0,
|
| 1084 |
-
"grad_norm": 0.0021140226162970066,
|
| 1085 |
-
"learning_rate": 0.00013975903614457834,
|
| 1086 |
-
"loss": 0.0047,
|
| 1087 |
-
"step": 154
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 155.0,
|
| 1091 |
-
"grad_norm": 0.0021610369440168142,
|
| 1092 |
-
"learning_rate": 0.0001393574297188755,
|
| 1093 |
-
"loss": 0.0047,
|
| 1094 |
-
"step": 155
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 156.0,
|
| 1098 |
-
"grad_norm": 0.004180264193564653,
|
| 1099 |
-
"learning_rate": 0.0001389558232931727,
|
| 1100 |
-
"loss": 0.0048,
|
| 1101 |
-
"step": 156
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 157.0,
|
| 1105 |
-
"grad_norm": 0.002081150421872735,
|
| 1106 |
-
"learning_rate": 0.00013855421686746988,
|
| 1107 |
-
"loss": 0.0047,
|
| 1108 |
-
"step": 157
|
| 1109 |
-
},
|
| 1110 |
-
{
|
| 1111 |
-
"epoch": 158.0,
|
| 1112 |
-
"grad_norm": 0.004214874934405088,
|
| 1113 |
-
"learning_rate": 0.0001381526104417671,
|
| 1114 |
-
"loss": 0.0048,
|
| 1115 |
-
"step": 158
|
| 1116 |
-
},
|
| 1117 |
-
{
|
| 1118 |
-
"epoch": 159.0,
|
| 1119 |
-
"grad_norm": 0.002199581591412425,
|
| 1120 |
-
"learning_rate": 0.00013775100401606426,
|
| 1121 |
-
"loss": 0.0047,
|
| 1122 |
-
"step": 159
|
| 1123 |
-
},
|
| 1124 |
-
{
|
| 1125 |
-
"epoch": 160.0,
|
| 1126 |
-
"grad_norm": 0.010448895394802094,
|
| 1127 |
-
"learning_rate": 0.00013734939759036145,
|
| 1128 |
-
"loss": 0.0047,
|
| 1129 |
-
"step": 160
|
| 1130 |
-
},
|
| 1131 |
-
{
|
| 1132 |
-
"epoch": 161.0,
|
| 1133 |
-
"grad_norm": 0.007638930808752775,
|
| 1134 |
-
"learning_rate": 0.00013694779116465864,
|
| 1135 |
-
"loss": 0.0047,
|
| 1136 |
-
"step": 161
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 162.0,
|
| 1140 |
-
"grad_norm": 0.007661975454539061,
|
| 1141 |
-
"learning_rate": 0.00013654618473895585,
|
| 1142 |
-
"loss": 0.0047,
|
| 1143 |
-
"step": 162
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 163.0,
|
| 1147 |
-
"grad_norm": 0.0014610282378271222,
|
| 1148 |
-
"learning_rate": 0.00013614457831325302,
|
| 1149 |
-
"loss": 0.0047,
|
| 1150 |
-
"step": 163
|
| 1151 |
-
},
|
| 1152 |
-
{
|
| 1153 |
-
"epoch": 164.0,
|
| 1154 |
-
"grad_norm": 0.01049406360834837,
|
| 1155 |
-
"learning_rate": 0.0001357429718875502,
|
| 1156 |
-
"loss": 0.0047,
|
| 1157 |
-
"step": 164
|
| 1158 |
-
},
|
| 1159 |
-
{
|
| 1160 |
-
"epoch": 165.0,
|
| 1161 |
-
"grad_norm": 0.0021653317380696535,
|
| 1162 |
-
"learning_rate": 0.0001353413654618474,
|
| 1163 |
-
"loss": 0.0047,
|
| 1164 |
-
"step": 165
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 166.0,
|
| 1168 |
-
"grad_norm": 0.016612282022833824,
|
| 1169 |
-
"learning_rate": 0.00013493975903614458,
|
| 1170 |
-
"loss": 0.0047,
|
| 1171 |
-
"step": 166
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 167.0,
|
| 1175 |
-
"grad_norm": 0.005656089633703232,
|
| 1176 |
-
"learning_rate": 0.00013453815261044177,
|
| 1177 |
-
"loss": 0.0048,
|
| 1178 |
-
"step": 167
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 168.0,
|
| 1182 |
-
"grad_norm": 0.015557671897113323,
|
| 1183 |
-
"learning_rate": 0.00013413654618473896,
|
| 1184 |
-
"loss": 0.0046,
|
| 1185 |
-
"step": 168
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 169.0,
|
| 1189 |
-
"grad_norm": 0.00793896708637476,
|
| 1190 |
-
"learning_rate": 0.00013373493975903615,
|
| 1191 |
-
"loss": 0.0047,
|
| 1192 |
-
"step": 169
|
| 1193 |
-
},
|
| 1194 |
-
{
|
| 1195 |
-
"epoch": 170.0,
|
| 1196 |
-
"grad_norm": 0.016965791583061218,
|
| 1197 |
-
"learning_rate": 0.00013333333333333334,
|
| 1198 |
-
"loss": 0.0047,
|
| 1199 |
-
"step": 170
|
| 1200 |
-
},
|
| 1201 |
-
{
|
| 1202 |
-
"epoch": 171.0,
|
| 1203 |
-
"grad_norm": 0.010896142572164536,
|
| 1204 |
-
"learning_rate": 0.00013293172690763053,
|
| 1205 |
-
"loss": 0.0047,
|
| 1206 |
-
"step": 171
|
| 1207 |
-
},
|
| 1208 |
-
{
|
| 1209 |
-
"epoch": 172.0,
|
| 1210 |
-
"grad_norm": 0.014977889135479927,
|
| 1211 |
-
"learning_rate": 0.00013253012048192772,
|
| 1212 |
-
"loss": 0.0048,
|
| 1213 |
-
"step": 172
|
| 1214 |
-
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 173.0,
|
| 1217 |
-
"grad_norm": 0.007990415208041668,
|
| 1218 |
-
"learning_rate": 0.0001321285140562249,
|
| 1219 |
-
"loss": 0.0047,
|
| 1220 |
-
"step": 173
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 174.0,
|
| 1224 |
-
"grad_norm": 0.01746082492172718,
|
| 1225 |
-
"learning_rate": 0.0001317269076305221,
|
| 1226 |
-
"loss": 0.0047,
|
| 1227 |
-
"step": 174
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 175.0,
|
| 1231 |
-
"grad_norm": 0.011155808344483376,
|
| 1232 |
-
"learning_rate": 0.00013132530120481929,
|
| 1233 |
-
"loss": 0.0047,
|
| 1234 |
-
"step": 175
|
| 1235 |
-
},
|
| 1236 |
-
{
|
| 1237 |
-
"epoch": 176.0,
|
| 1238 |
-
"grad_norm": 0.020374253392219543,
|
| 1239 |
-
"learning_rate": 0.00013092369477911648,
|
| 1240 |
-
"loss": 0.0047,
|
| 1241 |
-
"step": 176
|
| 1242 |
-
},
|
| 1243 |
-
{
|
| 1244 |
-
"epoch": 177.0,
|
| 1245 |
-
"grad_norm": 0.027189314365386963,
|
| 1246 |
-
"learning_rate": 0.00013052208835341366,
|
| 1247 |
-
"loss": 0.0048,
|
| 1248 |
-
"step": 177
|
| 1249 |
-
},
|
| 1250 |
-
{
|
| 1251 |
-
"epoch": 178.0,
|
| 1252 |
-
"grad_norm": 0.0035853274166584015,
|
| 1253 |
-
"learning_rate": 0.00013012048192771085,
|
| 1254 |
-
"loss": 0.0046,
|
| 1255 |
-
"step": 178
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 179.0,
|
| 1259 |
-
"grad_norm": 0.02057839184999466,
|
| 1260 |
-
"learning_rate": 0.00012971887550200804,
|
| 1261 |
-
"loss": 0.0047,
|
| 1262 |
-
"step": 179
|
| 1263 |
-
},
|
| 1264 |
-
{
|
| 1265 |
-
"epoch": 180.0,
|
| 1266 |
-
"grad_norm": 0.004144900944083929,
|
| 1267 |
-
"learning_rate": 0.00012931726907630523,
|
| 1268 |
-
"loss": 0.0047,
|
| 1269 |
-
"step": 180
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 181.0,
|
| 1273 |
-
"grad_norm": 0.027396870777010918,
|
| 1274 |
-
"learning_rate": 0.00012891566265060242,
|
| 1275 |
-
"loss": 0.0048,
|
| 1276 |
-
"step": 181
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 182.0,
|
| 1280 |
-
"grad_norm": 0.004257265478372574,
|
| 1281 |
-
"learning_rate": 0.0001285140562248996,
|
| 1282 |
-
"loss": 0.0048,
|
| 1283 |
-
"step": 182
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 183.0,
|
| 1287 |
-
"grad_norm": 0.029999705031514168,
|
| 1288 |
-
"learning_rate": 0.0001281124497991968,
|
| 1289 |
-
"loss": 0.0048,
|
| 1290 |
-
"step": 183
|
| 1291 |
-
},
|
| 1292 |
-
{
|
| 1293 |
-
"epoch": 184.0,
|
| 1294 |
-
"grad_norm": 0.012998619116842747,
|
| 1295 |
-
"learning_rate": 0.00012771084337349396,
|
| 1296 |
-
"loss": 0.0047,
|
| 1297 |
-
"step": 184
|
| 1298 |
-
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 185.0,
|
| 1301 |
-
"grad_norm": 0.027277518063783646,
|
| 1302 |
-
"learning_rate": 0.00012730923694779118,
|
| 1303 |
-
"loss": 0.0048,
|
| 1304 |
-
"step": 185
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 186.0,
|
| 1308 |
-
"grad_norm": 0.020436229184269905,
|
| 1309 |
-
"learning_rate": 0.00012690763052208837,
|
| 1310 |
-
"loss": 0.0047,
|
| 1311 |
-
"step": 186
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 187.0,
|
| 1315 |
-
"grad_norm": 0.017912449315190315,
|
| 1316 |
-
"learning_rate": 0.00012650602409638556,
|
| 1317 |
-
"loss": 0.0048,
|
| 1318 |
-
"step": 187
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 188.0,
|
| 1322 |
-
"grad_norm": 0.022624023258686066,
|
| 1323 |
-
"learning_rate": 0.00012610441767068272,
|
| 1324 |
-
"loss": 0.0047,
|
| 1325 |
-
"step": 188
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 189.0,
|
| 1329 |
-
"grad_norm": 0.00817954819649458,
|
| 1330 |
-
"learning_rate": 0.00012570281124497994,
|
| 1331 |
-
"loss": 0.0047,
|
| 1332 |
-
"step": 189
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 190.0,
|
| 1336 |
-
"grad_norm": 0.01837238110601902,
|
| 1337 |
-
"learning_rate": 0.00012530120481927712,
|
| 1338 |
-
"loss": 0.0047,
|
| 1339 |
-
"step": 190
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 191.0,
|
| 1343 |
-
"grad_norm": 0.006303516216576099,
|
| 1344 |
-
"learning_rate": 0.0001248995983935743,
|
| 1345 |
-
"loss": 0.0047,
|
| 1346 |
-
"step": 191
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 192.0,
|
| 1350 |
-
"grad_norm": 0.02228759415447712,
|
| 1351 |
-
"learning_rate": 0.00012449799196787148,
|
| 1352 |
-
"loss": 0.0047,
|
| 1353 |
-
"step": 192
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 193.0,
|
| 1357 |
-
"grad_norm": 0.004886090289801359,
|
| 1358 |
-
"learning_rate": 0.0001240963855421687,
|
| 1359 |
-
"loss": 0.0047,
|
| 1360 |
-
"step": 193
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 194.0,
|
| 1364 |
-
"grad_norm": 0.017957722768187523,
|
| 1365 |
-
"learning_rate": 0.00012369477911646588,
|
| 1366 |
-
"loss": 0.0048,
|
| 1367 |
-
"step": 194
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 195.0,
|
| 1371 |
-
"grad_norm": 0.008203946985304356,
|
| 1372 |
-
"learning_rate": 0.00012329317269076307,
|
| 1373 |
-
"loss": 0.0047,
|
| 1374 |
-
"step": 195
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 196.0,
|
| 1378 |
-
"grad_norm": 0.013346477411687374,
|
| 1379 |
-
"learning_rate": 0.00012289156626506023,
|
| 1380 |
-
"loss": 0.0047,
|
| 1381 |
-
"step": 196
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 197.0,
|
| 1385 |
-
"grad_norm": 0.008646669797599316,
|
| 1386 |
-
"learning_rate": 0.00012248995983935742,
|
| 1387 |
-
"loss": 0.0048,
|
| 1388 |
-
"step": 197
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 198.0,
|
| 1392 |
-
"grad_norm": 0.011337646283209324,
|
| 1393 |
-
"learning_rate": 0.00012208835341365464,
|
| 1394 |
-
"loss": 0.0047,
|
| 1395 |
-
"step": 198
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 199.0,
|
| 1399 |
-
"grad_norm": 0.01342825498431921,
|
| 1400 |
-
"learning_rate": 0.00012168674698795181,
|
| 1401 |
-
"loss": 0.0047,
|
| 1402 |
-
"step": 199
|
| 1403 |
-
},
|
| 1404 |
-
{
|
| 1405 |
-
"epoch": 200.0,
|
| 1406 |
-
"grad_norm": 0.006050780415534973,
|
| 1407 |
-
"learning_rate": 0.000121285140562249,
|
| 1408 |
-
"loss": 0.0048,
|
| 1409 |
-
"step": 200
|
| 1410 |
-
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 201.0,
|
| 1413 |
-
"grad_norm": 0.01836731843650341,
|
| 1414 |
-
"learning_rate": 0.00012088353413654618,
|
| 1415 |
-
"loss": 0.0048,
|
| 1416 |
-
"step": 201
|
| 1417 |
-
},
|
| 1418 |
-
{
|
| 1419 |
-
"epoch": 202.0,
|
| 1420 |
-
"grad_norm": 0.008505699224770069,
|
| 1421 |
-
"learning_rate": 0.0001204819277108434,
|
| 1422 |
-
"loss": 0.0047,
|
| 1423 |
-
"step": 202
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 203.0,
|
| 1427 |
-
"grad_norm": 0.008753190748393536,
|
| 1428 |
-
"learning_rate": 0.00012008032128514057,
|
| 1429 |
-
"loss": 0.0047,
|
| 1430 |
-
"step": 203
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 204.0,
|
| 1434 |
-
"grad_norm": 0.009133870713412762,
|
| 1435 |
-
"learning_rate": 0.00011967871485943776,
|
| 1436 |
-
"loss": 0.0048,
|
| 1437 |
-
"step": 204
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 205.0,
|
| 1441 |
-
"grad_norm": 0.0029124633874744177,
|
| 1442 |
-
"learning_rate": 0.00011927710843373494,
|
| 1443 |
-
"loss": 0.0048,
|
| 1444 |
-
"step": 205
|
| 1445 |
-
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 206.0,
|
| 1448 |
-
"grad_norm": 0.00847614649683237,
|
| 1449 |
-
"learning_rate": 0.00011887550200803212,
|
| 1450 |
-
"loss": 0.0047,
|
| 1451 |
-
"step": 206
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 207.0,
|
| 1455 |
-
"grad_norm": 0.0028652322944253683,
|
| 1456 |
-
"learning_rate": 0.00011847389558232933,
|
| 1457 |
-
"loss": 0.0047,
|
| 1458 |
-
"step": 207
|
| 1459 |
-
},
|
| 1460 |
-
{
|
| 1461 |
-
"epoch": 208.0,
|
| 1462 |
-
"grad_norm": 0.009550940245389938,
|
| 1463 |
-
"learning_rate": 0.00011807228915662652,
|
| 1464 |
-
"loss": 0.0047,
|
| 1465 |
-
"step": 208
|
| 1466 |
-
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 209.0,
|
| 1469 |
-
"grad_norm": 0.008687314577400684,
|
| 1470 |
-
"learning_rate": 0.00011767068273092369,
|
| 1471 |
-
"loss": 0.0047,
|
| 1472 |
-
"step": 209
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 210.0,
|
| 1476 |
-
"grad_norm": 0.004487687721848488,
|
| 1477 |
-
"learning_rate": 0.00011726907630522088,
|
| 1478 |
-
"loss": 0.0047,
|
| 1479 |
-
"step": 210
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 211.0,
|
| 1483 |
-
"grad_norm": 0.009615003131330013,
|
| 1484 |
-
"learning_rate": 0.00011686746987951808,
|
| 1485 |
-
"loss": 0.0047,
|
| 1486 |
-
"step": 211
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 212.0,
|
| 1490 |
-
"grad_norm": 0.001644686795771122,
|
| 1491 |
-
"learning_rate": 0.00011646586345381527,
|
| 1492 |
-
"loss": 0.0047,
|
| 1493 |
-
"step": 212
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 213.0,
|
| 1497 |
-
"grad_norm": 0.012544874101877213,
|
| 1498 |
-
"learning_rate": 0.00011606425702811245,
|
| 1499 |
-
"loss": 0.0047,
|
| 1500 |
-
"step": 213
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 214.0,
|
| 1504 |
-
"grad_norm": 0.009784480556845665,
|
| 1505 |
-
"learning_rate": 0.00011566265060240964,
|
| 1506 |
-
"loss": 0.0047,
|
| 1507 |
-
"step": 214
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 215.0,
|
| 1511 |
-
"grad_norm": 0.008885402232408524,
|
| 1512 |
-
"learning_rate": 0.00011526104417670683,
|
| 1513 |
-
"loss": 0.0047,
|
| 1514 |
-
"step": 215
|
| 1515 |
-
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 216.0,
|
| 1518 |
-
"grad_norm": 0.015116319991648197,
|
| 1519 |
-
"learning_rate": 0.00011485943775100403,
|
| 1520 |
-
"loss": 0.0047,
|
| 1521 |
-
"step": 216
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 217.0,
|
| 1525 |
-
"grad_norm": 0.0030038722325116396,
|
| 1526 |
-
"learning_rate": 0.0001144578313253012,
|
| 1527 |
-
"loss": 0.0047,
|
| 1528 |
-
"step": 217
|
| 1529 |
-
},
|
| 1530 |
-
{
|
| 1531 |
-
"epoch": 218.0,
|
| 1532 |
-
"grad_norm": 0.014432215131819248,
|
| 1533 |
-
"learning_rate": 0.0001140562248995984,
|
| 1534 |
-
"loss": 0.0047,
|
| 1535 |
-
"step": 218
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 219.0,
|
| 1539 |
-
"grad_norm": 0.012555493041872978,
|
| 1540 |
-
"learning_rate": 0.00011365461847389558,
|
| 1541 |
-
"loss": 0.0047,
|
| 1542 |
-
"step": 219
|
| 1543 |
-
},
|
| 1544 |
-
{
|
| 1545 |
-
"epoch": 220.0,
|
| 1546 |
-
"grad_norm": 0.004350865725427866,
|
| 1547 |
-
"learning_rate": 0.00011325301204819279,
|
| 1548 |
-
"loss": 0.0047,
|
| 1549 |
-
"step": 220
|
| 1550 |
-
},
|
| 1551 |
-
{
|
| 1552 |
-
"epoch": 221.0,
|
| 1553 |
-
"grad_norm": 0.011415142565965652,
|
| 1554 |
-
"learning_rate": 0.00011285140562248996,
|
| 1555 |
-
"loss": 0.0047,
|
| 1556 |
-
"step": 221
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 222.0,
|
| 1560 |
-
"grad_norm": 0.002292638411745429,
|
| 1561 |
-
"learning_rate": 0.00011244979919678715,
|
| 1562 |
-
"loss": 0.0047,
|
| 1563 |
-
"step": 222
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 223.0,
|
| 1567 |
-
"grad_norm": 0.011944664642214775,
|
| 1568 |
-
"learning_rate": 0.00011204819277108434,
|
| 1569 |
-
"loss": 0.0047,
|
| 1570 |
-
"step": 223
|
| 1571 |
-
},
|
| 1572 |
-
{
|
| 1573 |
-
"epoch": 224.0,
|
| 1574 |
-
"grad_norm": 0.009497747756540775,
|
| 1575 |
-
"learning_rate": 0.00011164658634538152,
|
| 1576 |
-
"loss": 0.0047,
|
| 1577 |
-
"step": 224
|
| 1578 |
-
},
|
| 1579 |
-
{
|
| 1580 |
-
"epoch": 225.0,
|
| 1581 |
-
"grad_norm": 0.004834707360714674,
|
| 1582 |
-
"learning_rate": 0.00011124497991967872,
|
| 1583 |
-
"loss": 0.0047,
|
| 1584 |
-
"step": 225
|
| 1585 |
-
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 226.0,
|
| 1588 |
-
"grad_norm": 0.012152622453868389,
|
| 1589 |
-
"learning_rate": 0.00011084337349397591,
|
| 1590 |
-
"loss": 0.0048,
|
| 1591 |
-
"step": 226
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 227.0,
|
| 1595 |
-
"grad_norm": 0.005508288741111755,
|
| 1596 |
-
"learning_rate": 0.0001104417670682731,
|
| 1597 |
-
"loss": 0.0047,
|
| 1598 |
-
"step": 227
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 228.0,
|
| 1602 |
-
"grad_norm": 0.004662630148231983,
|
| 1603 |
-
"learning_rate": 0.00011004016064257027,
|
| 1604 |
-
"loss": 0.0047,
|
| 1605 |
-
"step": 228
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 229.0,
|
| 1609 |
-
"grad_norm": 0.0046459161676466465,
|
| 1610 |
-
"learning_rate": 0.00010963855421686749,
|
| 1611 |
-
"loss": 0.0047,
|
| 1612 |
-
"step": 229
|
| 1613 |
-
},
|
| 1614 |
-
{
|
| 1615 |
-
"epoch": 230.0,
|
| 1616 |
-
"grad_norm": 0.003354718443006277,
|
| 1617 |
-
"learning_rate": 0.00010923694779116467,
|
| 1618 |
-
"loss": 0.0047,
|
| 1619 |
-
"step": 230
|
| 1620 |
-
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 231.0,
|
| 1623 |
-
"grad_norm": 0.005278999917209148,
|
| 1624 |
-
"learning_rate": 0.00010883534136546186,
|
| 1625 |
-
"loss": 0.0048,
|
| 1626 |
-
"step": 231
|
| 1627 |
-
},
|
| 1628 |
-
{
|
| 1629 |
-
"epoch": 232.0,
|
| 1630 |
-
"grad_norm": 0.006012341473251581,
|
| 1631 |
-
"learning_rate": 0.00010843373493975903,
|
| 1632 |
-
"loss": 0.0047,
|
| 1633 |
-
"step": 232
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 233.0,
|
| 1637 |
-
"grad_norm": 0.0016911854036152363,
|
| 1638 |
-
"learning_rate": 0.00010803212851405625,
|
| 1639 |
-
"loss": 0.0047,
|
| 1640 |
-
"step": 233
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 234.0,
|
| 1644 |
-
"grad_norm": 0.004478626884520054,
|
| 1645 |
-
"learning_rate": 0.00010763052208835342,
|
| 1646 |
-
"loss": 0.0047,
|
| 1647 |
-
"step": 234
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 235.0,
|
| 1651 |
-
"grad_norm": 0.005508603993803263,
|
| 1652 |
-
"learning_rate": 0.00010722891566265061,
|
| 1653 |
-
"loss": 0.0047,
|
| 1654 |
-
"step": 235
|
| 1655 |
-
},
|
| 1656 |
-
{
|
| 1657 |
-
"epoch": 236.0,
|
| 1658 |
-
"grad_norm": 0.0032062928657978773,
|
| 1659 |
-
"learning_rate": 0.00010682730923694779,
|
| 1660 |
-
"loss": 0.0047,
|
| 1661 |
-
"step": 236
|
| 1662 |
-
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 237.0,
|
| 1665 |
-
"grad_norm": 0.006789966020733118,
|
| 1666 |
-
"learning_rate": 0.00010642570281124498,
|
| 1667 |
-
"loss": 0.0048,
|
| 1668 |
-
"step": 237
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 238.0,
|
| 1672 |
-
"grad_norm": 0.005682968068867922,
|
| 1673 |
-
"learning_rate": 0.00010602409638554218,
|
| 1674 |
-
"loss": 0.0047,
|
| 1675 |
-
"step": 238
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 239.0,
|
| 1679 |
-
"grad_norm": 0.00296304514631629,
|
| 1680 |
-
"learning_rate": 0.00010562248995983937,
|
| 1681 |
-
"loss": 0.0047,
|
| 1682 |
-
"step": 239
|
| 1683 |
-
},
|
| 1684 |
-
{
|
| 1685 |
-
"epoch": 240.0,
|
| 1686 |
-
"grad_norm": 0.006762088742107153,
|
| 1687 |
-
"learning_rate": 0.00010522088353413654,
|
| 1688 |
-
"loss": 0.0047,
|
| 1689 |
-
"step": 240
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 241.0,
|
| 1693 |
-
"grad_norm": 0.0028762409929186106,
|
| 1694 |
-
"learning_rate": 0.00010481927710843373,
|
| 1695 |
-
"loss": 0.0047,
|
| 1696 |
-
"step": 241
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 242.0,
|
| 1700 |
-
"grad_norm": 0.009846026077866554,
|
| 1701 |
-
"learning_rate": 0.00010441767068273094,
|
| 1702 |
-
"loss": 0.0047,
|
| 1703 |
-
"step": 242
|
| 1704 |
-
},
|
| 1705 |
-
{
|
| 1706 |
-
"epoch": 243.0,
|
| 1707 |
-
"grad_norm": 0.01071733795106411,
|
| 1708 |
-
"learning_rate": 0.00010401606425702813,
|
| 1709 |
-
"loss": 0.0047,
|
| 1710 |
-
"step": 243
|
| 1711 |
-
},
|
| 1712 |
-
{
|
| 1713 |
-
"epoch": 244.0,
|
| 1714 |
-
"grad_norm": 0.001721803448162973,
|
| 1715 |
-
"learning_rate": 0.0001036144578313253,
|
| 1716 |
-
"loss": 0.0047,
|
| 1717 |
-
"step": 244
|
| 1718 |
-
},
|
| 1719 |
-
{
|
| 1720 |
-
"epoch": 245.0,
|
| 1721 |
-
"grad_norm": 0.015101822093129158,
|
| 1722 |
-
"learning_rate": 0.00010321285140562249,
|
| 1723 |
-
"loss": 0.0047,
|
| 1724 |
-
"step": 245
|
| 1725 |
-
},
|
| 1726 |
-
{
|
| 1727 |
-
"epoch": 246.0,
|
| 1728 |
-
"grad_norm": 0.012098951265215874,
|
| 1729 |
-
"learning_rate": 0.00010281124497991968,
|
| 1730 |
-
"loss": 0.0047,
|
| 1731 |
-
"step": 246
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 247.0,
|
| 1735 |
-
"grad_norm": 0.006853122264146805,
|
| 1736 |
-
"learning_rate": 0.00010240963855421688,
|
| 1737 |
-
"loss": 0.0047,
|
| 1738 |
-
"step": 247
|
| 1739 |
-
},
|
| 1740 |
-
{
|
| 1741 |
-
"epoch": 248.0,
|
| 1742 |
-
"grad_norm": 0.016420654952526093,
|
| 1743 |
-
"learning_rate": 0.00010200803212851406,
|
| 1744 |
-
"loss": 0.0047,
|
| 1745 |
-
"step": 248
|
| 1746 |
-
},
|
| 1747 |
-
{
|
| 1748 |
-
"epoch": 249.0,
|
| 1749 |
-
"grad_norm": 0.00433447165414691,
|
| 1750 |
-
"learning_rate": 0.00010160642570281125,
|
| 1751 |
-
"loss": 0.0047,
|
| 1752 |
-
"step": 249
|
| 1753 |
-
},
|
| 1754 |
-
{
|
| 1755 |
-
"epoch": 250.0,
|
| 1756 |
-
"grad_norm": 0.012108503840863705,
|
| 1757 |
-
"learning_rate": 0.00010120481927710844,
|
| 1758 |
-
"loss": 0.0047,
|
| 1759 |
-
"step": 250
|
| 1760 |
-
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 251.0,
|
| 1763 |
-
"grad_norm": 0.008616012521088123,
|
| 1764 |
-
"learning_rate": 0.00010080321285140564,
|
| 1765 |
-
"loss": 0.0047,
|
| 1766 |
-
"step": 251
|
| 1767 |
-
},
|
| 1768 |
-
{
|
| 1769 |
-
"epoch": 252.0,
|
| 1770 |
-
"grad_norm": 0.008652638643980026,
|
| 1771 |
-
"learning_rate": 0.00010040160642570282,
|
| 1772 |
-
"loss": 0.0047,
|
| 1773 |
-
"step": 252
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 253.0,
|
| 1777 |
-
"grad_norm": 0.01692233793437481,
|
| 1778 |
-
"learning_rate": 0.0001,
|
| 1779 |
-
"loss": 0.0047,
|
| 1780 |
-
"step": 253
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 254.0,
|
| 1784 |
-
"grad_norm": 0.006091867107897997,
|
| 1785 |
-
"learning_rate": 9.95983935742972e-05,
|
| 1786 |
-
"loss": 0.0047,
|
| 1787 |
-
"step": 254
|
| 1788 |
-
},
|
| 1789 |
-
{
|
| 1790 |
-
"epoch": 255.0,
|
| 1791 |
-
"grad_norm": 0.011222707107663155,
|
| 1792 |
-
"learning_rate": 9.919678714859438e-05,
|
| 1793 |
-
"loss": 0.0047,
|
| 1794 |
-
"step": 255
|
| 1795 |
-
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 256.0,
|
| 1798 |
-
"grad_norm": 0.011931284330785275,
|
| 1799 |
-
"learning_rate": 9.879518072289157e-05,
|
| 1800 |
-
"loss": 0.0047,
|
| 1801 |
-
"step": 256
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 257.0,
|
| 1805 |
-
"grad_norm": 0.004332751967012882,
|
| 1806 |
-
"learning_rate": 9.839357429718876e-05,
|
| 1807 |
-
"loss": 0.0048,
|
| 1808 |
-
"step": 257
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 258.0,
|
| 1812 |
-
"grad_norm": 0.013347852043807507,
|
| 1813 |
-
"learning_rate": 9.799196787148595e-05,
|
| 1814 |
-
"loss": 0.0047,
|
| 1815 |
-
"step": 258
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 259.0,
|
| 1819 |
-
"grad_norm": 0.006608230993151665,
|
| 1820 |
-
"learning_rate": 9.759036144578314e-05,
|
| 1821 |
-
"loss": 0.0047,
|
| 1822 |
-
"step": 259
|
| 1823 |
-
},
|
| 1824 |
-
{
|
| 1825 |
-
"epoch": 260.0,
|
| 1826 |
-
"grad_norm": 0.004314239602535963,
|
| 1827 |
-
"learning_rate": 9.718875502008033e-05,
|
| 1828 |
-
"loss": 0.0048,
|
| 1829 |
-
"step": 260
|
| 1830 |
-
},
|
| 1831 |
-
{
|
| 1832 |
-
"epoch": 261.0,
|
| 1833 |
-
"grad_norm": 0.0060504162684082985,
|
| 1834 |
-
"learning_rate": 9.678714859437752e-05,
|
| 1835 |
-
"loss": 0.0048,
|
| 1836 |
-
"step": 261
|
| 1837 |
-
},
|
| 1838 |
-
{
|
| 1839 |
-
"epoch": 262.0,
|
| 1840 |
-
"grad_norm": 0.0033889045007526875,
|
| 1841 |
-
"learning_rate": 9.638554216867471e-05,
|
| 1842 |
-
"loss": 0.0047,
|
| 1843 |
-
"step": 262
|
| 1844 |
-
},
|
| 1845 |
-
{
|
| 1846 |
-
"epoch": 263.0,
|
| 1847 |
-
"grad_norm": 0.008607766591012478,
|
| 1848 |
-
"learning_rate": 9.598393574297188e-05,
|
| 1849 |
-
"loss": 0.0048,
|
| 1850 |
-
"step": 263
|
| 1851 |
-
},
|
| 1852 |
-
{
|
| 1853 |
-
"epoch": 264.0,
|
| 1854 |
-
"grad_norm": 0.001759338192641735,
|
| 1855 |
-
"learning_rate": 9.558232931726909e-05,
|
| 1856 |
-
"loss": 0.0047,
|
| 1857 |
-
"step": 264
|
| 1858 |
-
},
|
| 1859 |
-
{
|
| 1860 |
-
"epoch": 265.0,
|
| 1861 |
-
"grad_norm": 0.008092672564089298,
|
| 1862 |
-
"learning_rate": 9.518072289156626e-05,
|
| 1863 |
-
"loss": 0.0047,
|
| 1864 |
-
"step": 265
|
| 1865 |
-
},
|
| 1866 |
-
{
|
| 1867 |
-
"epoch": 266.0,
|
| 1868 |
-
"grad_norm": 0.0017322164494544268,
|
| 1869 |
-
"learning_rate": 9.477911646586346e-05,
|
| 1870 |
-
"loss": 0.0047,
|
| 1871 |
-
"step": 266
|
| 1872 |
-
},
|
| 1873 |
-
{
|
| 1874 |
-
"epoch": 267.0,
|
| 1875 |
-
"grad_norm": 0.0017360730562359095,
|
| 1876 |
-
"learning_rate": 9.437751004016064e-05,
|
| 1877 |
-
"loss": 0.0047,
|
| 1878 |
-
"step": 267
|
| 1879 |
-
},
|
| 1880 |
-
{
|
| 1881 |
-
"epoch": 268.0,
|
| 1882 |
-
"grad_norm": 0.0017291605472564697,
|
| 1883 |
-
"learning_rate": 9.397590361445784e-05,
|
| 1884 |
-
"loss": 0.0047,
|
| 1885 |
-
"step": 268
|
| 1886 |
-
},
|
| 1887 |
-
{
|
| 1888 |
-
"epoch": 269.0,
|
| 1889 |
-
"grad_norm": 0.0016873552231118083,
|
| 1890 |
-
"learning_rate": 9.357429718875502e-05,
|
| 1891 |
-
"loss": 0.0047,
|
| 1892 |
-
"step": 269
|
| 1893 |
-
},
|
| 1894 |
-
{
|
| 1895 |
-
"epoch": 270.0,
|
| 1896 |
-
"grad_norm": 0.007920237258076668,
|
| 1897 |
-
"learning_rate": 9.317269076305222e-05,
|
| 1898 |
-
"loss": 0.0047,
|
| 1899 |
-
"step": 270
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 271.0,
|
| 1903 |
-
"grad_norm": 0.0036651096306741238,
|
| 1904 |
-
"learning_rate": 9.27710843373494e-05,
|
| 1905 |
-
"loss": 0.0048,
|
| 1906 |
-
"step": 271
|
| 1907 |
-
},
|
| 1908 |
-
{
|
| 1909 |
-
"epoch": 272.0,
|
| 1910 |
-
"grad_norm": 0.01999688521027565,
|
| 1911 |
-
"learning_rate": 9.23694779116466e-05,
|
| 1912 |
-
"loss": 0.0047,
|
| 1913 |
-
"step": 272
|
| 1914 |
-
},
|
| 1915 |
-
{
|
| 1916 |
-
"epoch": 273.0,
|
| 1917 |
-
"grad_norm": 0.007801530417054892,
|
| 1918 |
-
"learning_rate": 9.196787148594378e-05,
|
| 1919 |
-
"loss": 0.0047,
|
| 1920 |
-
"step": 273
|
| 1921 |
-
},
|
| 1922 |
-
{
|
| 1923 |
-
"epoch": 274.0,
|
| 1924 |
-
"grad_norm": 0.012862344272434711,
|
| 1925 |
-
"learning_rate": 9.156626506024096e-05,
|
| 1926 |
-
"loss": 0.0048,
|
| 1927 |
-
"step": 274
|
| 1928 |
-
},
|
| 1929 |
-
{
|
| 1930 |
-
"epoch": 275.0,
|
| 1931 |
-
"grad_norm": 0.0016800492303445935,
|
| 1932 |
-
"learning_rate": 9.116465863453815e-05,
|
| 1933 |
-
"loss": 0.0047,
|
| 1934 |
-
"step": 275
|
| 1935 |
-
},
|
| 1936 |
-
{
|
| 1937 |
-
"epoch": 276.0,
|
| 1938 |
-
"grad_norm": 0.019683390855789185,
|
| 1939 |
-
"learning_rate": 9.076305220883534e-05,
|
| 1940 |
-
"loss": 0.0047,
|
| 1941 |
-
"step": 276
|
| 1942 |
-
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 277.0,
|
| 1945 |
-
"grad_norm": 0.0017172418301925063,
|
| 1946 |
-
"learning_rate": 9.036144578313253e-05,
|
| 1947 |
-
"loss": 0.0047,
|
| 1948 |
-
"step": 277
|
| 1949 |
-
},
|
| 1950 |
-
{
|
| 1951 |
-
"epoch": 278.0,
|
| 1952 |
-
"grad_norm": 0.030774248763918877,
|
| 1953 |
-
"learning_rate": 8.995983935742972e-05,
|
| 1954 |
-
"loss": 0.0049,
|
| 1955 |
-
"step": 278
|
| 1956 |
-
},
|
| 1957 |
-
{
|
| 1958 |
-
"epoch": 279.0,
|
| 1959 |
-
"grad_norm": 0.0035772966220974922,
|
| 1960 |
-
"learning_rate": 8.955823293172691e-05,
|
| 1961 |
-
"loss": 0.0048,
|
| 1962 |
-
"step": 279
|
| 1963 |
-
},
|
| 1964 |
-
{
|
| 1965 |
-
"epoch": 280.0,
|
| 1966 |
-
"grad_norm": 0.03694264590740204,
|
| 1967 |
-
"learning_rate": 8.91566265060241e-05,
|
| 1968 |
-
"loss": 0.0048,
|
| 1969 |
-
"step": 280
|
| 1970 |
-
},
|
| 1971 |
-
{
|
| 1972 |
-
"epoch": 281.0,
|
| 1973 |
-
"grad_norm": 0.0031574727036058903,
|
| 1974 |
-
"learning_rate": 8.875502008032129e-05,
|
| 1975 |
-
"loss": 0.0045,
|
| 1976 |
-
"step": 281
|
| 1977 |
-
},
|
| 1978 |
-
{
|
| 1979 |
-
"epoch": 282.0,
|
| 1980 |
-
"grad_norm": 0.03539184108376503,
|
| 1981 |
-
"learning_rate": 8.835341365461848e-05,
|
| 1982 |
-
"loss": 0.005,
|
| 1983 |
-
"step": 282
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 283.0,
|
| 1987 |
-
"grad_norm": 0.010516048409044743,
|
| 1988 |
-
"learning_rate": 8.795180722891567e-05,
|
| 1989 |
-
"loss": 0.0047,
|
| 1990 |
-
"step": 283
|
| 1991 |
-
},
|
| 1992 |
-
{
|
| 1993 |
-
"epoch": 284.0,
|
| 1994 |
-
"grad_norm": 0.010487216524779797,
|
| 1995 |
-
"learning_rate": 8.755020080321286e-05,
|
| 1996 |
-
"loss": 0.0047,
|
| 1997 |
-
"step": 284
|
| 1998 |
-
},
|
| 1999 |
-
{
|
| 2000 |
-
"epoch": 285.0,
|
| 2001 |
-
"grad_norm": 0.010434217751026154,
|
| 2002 |
-
"learning_rate": 8.714859437751005e-05,
|
| 2003 |
-
"loss": 0.0047,
|
| 2004 |
-
"step": 285
|
| 2005 |
-
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 286.0,
|
| 2008 |
-
"grad_norm": 0.007612716872245073,
|
| 2009 |
-
"learning_rate": 8.674698795180724e-05,
|
| 2010 |
-
"loss": 0.0047,
|
| 2011 |
-
"step": 286
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 287.0,
|
| 2015 |
-
"grad_norm": 0.025456130504608154,
|
| 2016 |
-
"learning_rate": 8.634538152610442e-05,
|
| 2017 |
-
"loss": 0.0048,
|
| 2018 |
-
"step": 287
|
| 2019 |
-
},
|
| 2020 |
-
{
|
| 2021 |
-
"epoch": 288.0,
|
| 2022 |
-
"grad_norm": 0.010371106676757336,
|
| 2023 |
-
"learning_rate": 8.594377510040161e-05,
|
| 2024 |
-
"loss": 0.0047,
|
| 2025 |
-
"step": 288
|
| 2026 |
-
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 289.0,
|
| 2029 |
-
"grad_norm": 0.028378885239362717,
|
| 2030 |
-
"learning_rate": 8.55421686746988e-05,
|
| 2031 |
-
"loss": 0.0048,
|
| 2032 |
-
"step": 289
|
| 2033 |
-
},
|
| 2034 |
-
{
|
| 2035 |
-
"epoch": 290.0,
|
| 2036 |
-
"grad_norm": 0.007674542721360922,
|
| 2037 |
-
"learning_rate": 8.514056224899599e-05,
|
| 2038 |
-
"loss": 0.0047,
|
| 2039 |
-
"step": 290
|
| 2040 |
-
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 291.0,
|
| 2043 |
-
"grad_norm": 0.025507695972919464,
|
| 2044 |
-
"learning_rate": 8.473895582329318e-05,
|
| 2045 |
-
"loss": 0.0048,
|
| 2046 |
-
"step": 291
|
| 2047 |
-
},
|
| 2048 |
-
{
|
| 2049 |
-
"epoch": 292.0,
|
| 2050 |
-
"grad_norm": 0.0009824644075706601,
|
| 2051 |
-
"learning_rate": 8.433734939759037e-05,
|
| 2052 |
-
"loss": 0.0049,
|
| 2053 |
-
"step": 292
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 293.0,
|
| 2057 |
-
"grad_norm": 0.03781874477863312,
|
| 2058 |
-
"learning_rate": 8.393574297188756e-05,
|
| 2059 |
-
"loss": 0.0046,
|
| 2060 |
-
"step": 293
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 294.0,
|
| 2064 |
-
"grad_norm": 0.007710340432822704,
|
| 2065 |
-
"learning_rate": 8.353413654618474e-05,
|
| 2066 |
-
"loss": 0.0047,
|
| 2067 |
-
"step": 294
|
| 2068 |
-
},
|
| 2069 |
-
{
|
| 2070 |
-
"epoch": 295.0,
|
| 2071 |
-
"grad_norm": 0.03508257865905762,
|
| 2072 |
-
"learning_rate": 8.313253012048194e-05,
|
| 2073 |
-
"loss": 0.005,
|
| 2074 |
-
"step": 295
|
| 2075 |
-
},
|
| 2076 |
-
{
|
| 2077 |
-
"epoch": 296.0,
|
| 2078 |
-
"grad_norm": 0.007686258759349585,
|
| 2079 |
-
"learning_rate": 8.273092369477911e-05,
|
| 2080 |
-
"loss": 0.0047,
|
| 2081 |
-
"step": 296
|
| 2082 |
-
},
|
| 2083 |
-
{
|
| 2084 |
-
"epoch": 297.0,
|
| 2085 |
-
"grad_norm": 0.04559750109910965,
|
| 2086 |
-
"learning_rate": 8.232931726907632e-05,
|
| 2087 |
-
"loss": 0.0048,
|
| 2088 |
-
"step": 297
|
| 2089 |
-
},
|
| 2090 |
-
{
|
| 2091 |
-
"epoch": 298.0,
|
| 2092 |
-
"grad_norm": 0.010225856676697731,
|
| 2093 |
-
"learning_rate": 8.192771084337349e-05,
|
| 2094 |
-
"loss": 0.0047,
|
| 2095 |
-
"step": 298
|
| 2096 |
-
},
|
| 2097 |
-
{
|
| 2098 |
-
"epoch": 299.0,
|
| 2099 |
-
"grad_norm": 0.02517218515276909,
|
| 2100 |
-
"learning_rate": 8.15261044176707e-05,
|
| 2101 |
-
"loss": 0.0048,
|
| 2102 |
-
"step": 299
|
| 2103 |
-
},
|
| 2104 |
-
{
|
| 2105 |
-
"epoch": 300.0,
|
| 2106 |
-
"grad_norm": 0.025102809071540833,
|
| 2107 |
-
"learning_rate": 8.112449799196787e-05,
|
| 2108 |
-
"loss": 0.0048,
|
| 2109 |
-
"step": 300
|
| 2110 |
-
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 301.0,
|
| 2113 |
-
"grad_norm": 0.0025317783001810312,
|
| 2114 |
-
"learning_rate": 8.072289156626507e-05,
|
| 2115 |
-
"loss": 0.0045,
|
| 2116 |
-
"step": 301
|
| 2117 |
-
},
|
| 2118 |
-
{
|
| 2119 |
-
"epoch": 302.0,
|
| 2120 |
-
"grad_norm": 0.0447322279214859,
|
| 2121 |
-
"learning_rate": 8.032128514056225e-05,
|
| 2122 |
-
"loss": 0.0048,
|
| 2123 |
-
"step": 302
|
| 2124 |
-
},
|
| 2125 |
-
{
|
| 2126 |
-
"epoch": 303.0,
|
| 2127 |
-
"grad_norm": 0.0074921357445418835,
|
| 2128 |
-
"learning_rate": 7.991967871485944e-05,
|
| 2129 |
-
"loss": 0.0047,
|
| 2130 |
-
"step": 303
|
| 2131 |
-
},
|
| 2132 |
-
{
|
| 2133 |
-
"epoch": 304.0,
|
| 2134 |
-
"grad_norm": 0.02473229542374611,
|
| 2135 |
-
"learning_rate": 7.951807228915663e-05,
|
| 2136 |
-
"loss": 0.0048,
|
| 2137 |
-
"step": 304
|
| 2138 |
-
},
|
| 2139 |
-
{
|
| 2140 |
-
"epoch": 305.0,
|
| 2141 |
-
"grad_norm": 0.007424222771078348,
|
| 2142 |
-
"learning_rate": 7.911646586345382e-05,
|
| 2143 |
-
"loss": 0.0047,
|
| 2144 |
-
"step": 305
|
| 2145 |
-
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 306.0,
|
| 2148 |
-
"grad_norm": 0.027306661009788513,
|
| 2149 |
-
"learning_rate": 7.8714859437751e-05,
|
| 2150 |
-
"loss": 0.0048,
|
| 2151 |
-
"step": 306
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 307.0,
|
| 2155 |
-
"grad_norm": 0.007345912978053093,
|
| 2156 |
-
"learning_rate": 7.83132530120482e-05,
|
| 2157 |
-
"loss": 0.0047,
|
| 2158 |
-
"step": 307
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 308.0,
|
| 2162 |
-
"grad_norm": 0.02444930374622345,
|
| 2163 |
-
"learning_rate": 7.791164658634539e-05,
|
| 2164 |
-
"loss": 0.0048,
|
| 2165 |
-
"step": 308
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 309.0,
|
| 2169 |
-
"grad_norm": 0.02708006091415882,
|
| 2170 |
-
"learning_rate": 7.751004016064257e-05,
|
| 2171 |
-
"loss": 0.0048,
|
| 2172 |
-
"step": 309
|
| 2173 |
-
},
|
| 2174 |
-
{
|
| 2175 |
-
"epoch": 310.0,
|
| 2176 |
-
"grad_norm": 0.007334452122449875,
|
| 2177 |
-
"learning_rate": 7.710843373493976e-05,
|
| 2178 |
-
"loss": 0.0047,
|
| 2179 |
-
"step": 310
|
| 2180 |
-
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 311.0,
|
| 2183 |
-
"grad_norm": 0.009786078706383705,
|
| 2184 |
-
"learning_rate": 7.670682730923695e-05,
|
| 2185 |
-
"loss": 0.0047,
|
| 2186 |
-
"step": 311
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 312.0,
|
| 2190 |
-
"grad_norm": 0.007255980744957924,
|
| 2191 |
-
"learning_rate": 7.630522088353414e-05,
|
| 2192 |
-
"loss": 0.0047,
|
| 2193 |
-
"step": 312
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 313.0,
|
| 2197 |
-
"grad_norm": 0.014863966032862663,
|
| 2198 |
-
"learning_rate": 7.590361445783133e-05,
|
| 2199 |
-
"loss": 0.0045,
|
| 2200 |
-
"step": 313
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 314.0,
|
| 2204 |
-
"grad_norm": 0.009730237536132336,
|
| 2205 |
-
"learning_rate": 7.550200803212851e-05,
|
| 2206 |
-
"loss": 0.0047,
|
| 2207 |
-
"step": 314
|
| 2208 |
-
},
|
| 2209 |
-
{
|
| 2210 |
-
"epoch": 315.0,
|
| 2211 |
-
"grad_norm": 0.00974965188652277,
|
| 2212 |
-
"learning_rate": 7.510040160642571e-05,
|
| 2213 |
-
"loss": 0.0047,
|
| 2214 |
-
"step": 315
|
| 2215 |
-
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 316.0,
|
| 2218 |
-
"grad_norm": 0.0008471576729789376,
|
| 2219 |
-
"learning_rate": 7.469879518072289e-05,
|
| 2220 |
-
"loss": 0.0049,
|
| 2221 |
-
"step": 316
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 317.0,
|
| 2225 |
-
"grad_norm": 0.007247288711369038,
|
| 2226 |
-
"learning_rate": 7.429718875502009e-05,
|
| 2227 |
-
"loss": 0.0047,
|
| 2228 |
-
"step": 317
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 318.0,
|
| 2232 |
-
"grad_norm": 0.00724770175293088,
|
| 2233 |
-
"learning_rate": 7.389558232931726e-05,
|
| 2234 |
-
"loss": 0.0047,
|
| 2235 |
-
"step": 318
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 319.0,
|
| 2239 |
-
"grad_norm": 0.009726744145154953,
|
| 2240 |
-
"learning_rate": 7.349397590361447e-05,
|
| 2241 |
-
"loss": 0.0047,
|
| 2242 |
-
"step": 319
|
| 2243 |
-
},
|
| 2244 |
-
{
|
| 2245 |
-
"epoch": 320.0,
|
| 2246 |
-
"grad_norm": 0.002370405476540327,
|
| 2247 |
-
"learning_rate": 7.309236947791164e-05,
|
| 2248 |
-
"loss": 0.0045,
|
| 2249 |
-
"step": 320
|
| 2250 |
-
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 321.0,
|
| 2253 |
-
"grad_norm": 0.007258490659296513,
|
| 2254 |
-
"learning_rate": 7.269076305220885e-05,
|
| 2255 |
-
"loss": 0.0047,
|
| 2256 |
-
"step": 321
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 322.0,
|
| 2260 |
-
"grad_norm": 0.009763582609593868,
|
| 2261 |
-
"learning_rate": 7.228915662650602e-05,
|
| 2262 |
-
"loss": 0.0047,
|
| 2263 |
-
"step": 322
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 323.0,
|
| 2267 |
-
"grad_norm": 0.0072934250347316265,
|
| 2268 |
-
"learning_rate": 7.188755020080321e-05,
|
| 2269 |
-
"loss": 0.0047,
|
| 2270 |
-
"step": 323
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 324.0,
|
| 2274 |
-
"grad_norm": 0.007325456012040377,
|
| 2275 |
-
"learning_rate": 7.14859437751004e-05,
|
| 2276 |
-
"loss": 0.0047,
|
| 2277 |
-
"step": 324
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 325.0,
|
| 2281 |
-
"grad_norm": 0.02716483175754547,
|
| 2282 |
-
"learning_rate": 7.108433734939759e-05,
|
| 2283 |
-
"loss": 0.0048,
|
| 2284 |
-
"step": 325
|
| 2285 |
-
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 326.0,
|
| 2288 |
-
"grad_norm": 0.007358341012150049,
|
| 2289 |
-
"learning_rate": 7.068273092369478e-05,
|
| 2290 |
-
"loss": 0.0047,
|
| 2291 |
-
"step": 326
|
| 2292 |
-
},
|
| 2293 |
-
{
|
| 2294 |
-
"epoch": 327.0,
|
| 2295 |
-
"grad_norm": 0.04177234321832657,
|
| 2296 |
-
"learning_rate": 7.028112449799197e-05,
|
| 2297 |
-
"loss": 0.0048,
|
| 2298 |
-
"step": 327
|
| 2299 |
-
},
|
| 2300 |
-
{
|
| 2301 |
-
"epoch": 328.0,
|
| 2302 |
-
"grad_norm": 0.027355682104825974,
|
| 2303 |
-
"learning_rate": 6.987951807228917e-05,
|
| 2304 |
-
"loss": 0.0048,
|
| 2305 |
-
"step": 328
|
| 2306 |
-
},
|
| 2307 |
-
{
|
| 2308 |
-
"epoch": 329.0,
|
| 2309 |
-
"grad_norm": 0.02739114686846733,
|
| 2310 |
-
"learning_rate": 6.947791164658635e-05,
|
| 2311 |
-
"loss": 0.0048,
|
| 2312 |
-
"step": 329
|
| 2313 |
-
},
|
| 2314 |
-
{
|
| 2315 |
-
"epoch": 330.0,
|
| 2316 |
-
"grad_norm": 0.024521242827177048,
|
| 2317 |
-
"learning_rate": 6.907630522088355e-05,
|
| 2318 |
-
"loss": 0.0048,
|
| 2319 |
-
"step": 330
|
| 2320 |
-
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 331.0,
|
| 2323 |
-
"grad_norm": 0.007352576591074467,
|
| 2324 |
-
"learning_rate": 6.867469879518072e-05,
|
| 2325 |
-
"loss": 0.0047,
|
| 2326 |
-
"step": 331
|
| 2327 |
-
},
|
| 2328 |
-
{
|
| 2329 |
-
"epoch": 332.0,
|
| 2330 |
-
"grad_norm": 0.0023073432967066765,
|
| 2331 |
-
"learning_rate": 6.827309236947793e-05,
|
| 2332 |
-
"loss": 0.0045,
|
| 2333 |
-
"step": 332
|
| 2334 |
-
},
|
| 2335 |
-
{
|
| 2336 |
-
"epoch": 333.0,
|
| 2337 |
-
"grad_norm": 0.009870602749288082,
|
| 2338 |
-
"learning_rate": 6.78714859437751e-05,
|
| 2339 |
-
"loss": 0.0047,
|
| 2340 |
-
"step": 333
|
| 2341 |
-
},
|
| 2342 |
-
{
|
| 2343 |
-
"epoch": 334.0,
|
| 2344 |
-
"grad_norm": 0.00988033413887024,
|
| 2345 |
-
"learning_rate": 6.746987951807229e-05,
|
| 2346 |
-
"loss": 0.0047,
|
| 2347 |
-
"step": 334
|
| 2348 |
-
},
|
| 2349 |
-
{
|
| 2350 |
-
"epoch": 335.0,
|
| 2351 |
-
"grad_norm": 0.024536525830626488,
|
| 2352 |
-
"learning_rate": 6.706827309236948e-05,
|
| 2353 |
-
"loss": 0.0048,
|
| 2354 |
-
"step": 335
|
| 2355 |
-
},
|
| 2356 |
-
{
|
| 2357 |
-
"epoch": 336.0,
|
| 2358 |
-
"grad_norm": 0.009867388755083084,
|
| 2359 |
-
"learning_rate": 6.666666666666667e-05,
|
| 2360 |
-
"loss": 0.0047,
|
| 2361 |
-
"step": 336
|
| 2362 |
-
},
|
| 2363 |
-
{
|
| 2364 |
-
"epoch": 337.0,
|
| 2365 |
-
"grad_norm": 0.00988020095974207,
|
| 2366 |
-
"learning_rate": 6.626506024096386e-05,
|
| 2367 |
-
"loss": 0.0047,
|
| 2368 |
-
"step": 337
|
| 2369 |
-
},
|
| 2370 |
-
{
|
| 2371 |
-
"epoch": 338.0,
|
| 2372 |
-
"grad_norm": 0.007355022244155407,
|
| 2373 |
-
"learning_rate": 6.586345381526105e-05,
|
| 2374 |
-
"loss": 0.0047,
|
| 2375 |
-
"step": 338
|
| 2376 |
-
},
|
| 2377 |
-
{
|
| 2378 |
-
"epoch": 339.0,
|
| 2379 |
-
"grad_norm": 0.007368543650954962,
|
| 2380 |
-
"learning_rate": 6.546184738955824e-05,
|
| 2381 |
-
"loss": 0.0047,
|
| 2382 |
-
"step": 339
|
| 2383 |
-
},
|
| 2384 |
-
{
|
| 2385 |
-
"epoch": 340.0,
|
| 2386 |
-
"grad_norm": 0.0073932805098593235,
|
| 2387 |
-
"learning_rate": 6.506024096385543e-05,
|
| 2388 |
-
"loss": 0.0047,
|
| 2389 |
-
"step": 340
|
| 2390 |
-
},
|
| 2391 |
-
{
|
| 2392 |
-
"epoch": 341.0,
|
| 2393 |
-
"grad_norm": 0.027419747784733772,
|
| 2394 |
-
"learning_rate": 6.465863453815262e-05,
|
| 2395 |
-
"loss": 0.0048,
|
| 2396 |
-
"step": 341
|
| 2397 |
-
},
|
| 2398 |
-
{
|
| 2399 |
-
"epoch": 342.0,
|
| 2400 |
-
"grad_norm": 0.009928101673722267,
|
| 2401 |
-
"learning_rate": 6.42570281124498e-05,
|
| 2402 |
-
"loss": 0.0047,
|
| 2403 |
-
"step": 342
|
| 2404 |
-
},
|
| 2405 |
-
{
|
| 2406 |
-
"epoch": 343.0,
|
| 2407 |
-
"grad_norm": 0.03391006961464882,
|
| 2408 |
-
"learning_rate": 6.385542168674698e-05,
|
| 2409 |
-
"loss": 0.005,
|
| 2410 |
-
"step": 343
|
| 2411 |
-
},
|
| 2412 |
-
{
|
| 2413 |
-
"epoch": 344.0,
|
| 2414 |
-
"grad_norm": 0.024691926315426826,
|
| 2415 |
-
"learning_rate": 6.345381526104418e-05,
|
| 2416 |
-
"loss": 0.0048,
|
| 2417 |
-
"step": 344
|
| 2418 |
-
},
|
| 2419 |
-
{
|
| 2420 |
-
"epoch": 345.0,
|
| 2421 |
-
"grad_norm": 0.036551907658576965,
|
| 2422 |
-
"learning_rate": 6.305220883534136e-05,
|
| 2423 |
-
"loss": 0.0046,
|
| 2424 |
-
"step": 345
|
| 2425 |
-
},
|
| 2426 |
-
{
|
| 2427 |
-
"epoch": 346.0,
|
| 2428 |
-
"grad_norm": 0.027336876839399338,
|
| 2429 |
-
"learning_rate": 6.265060240963856e-05,
|
| 2430 |
-
"loss": 0.0048,
|
| 2431 |
-
"step": 346
|
| 2432 |
-
},
|
| 2433 |
-
{
|
| 2434 |
-
"epoch": 347.0,
|
| 2435 |
-
"grad_norm": 0.016788320615887642,
|
| 2436 |
-
"learning_rate": 6.224899598393574e-05,
|
| 2437 |
-
"loss": 0.005,
|
| 2438 |
-
"step": 347
|
| 2439 |
-
},
|
| 2440 |
-
{
|
| 2441 |
-
"epoch": 348.0,
|
| 2442 |
-
"grad_norm": 0.0243778508156538,
|
| 2443 |
-
"learning_rate": 6.184738955823294e-05,
|
| 2444 |
-
"loss": 0.0048,
|
| 2445 |
-
"step": 348
|
| 2446 |
-
},
|
| 2447 |
-
{
|
| 2448 |
-
"epoch": 349.0,
|
| 2449 |
-
"grad_norm": 0.007300014141947031,
|
| 2450 |
-
"learning_rate": 6.144578313253012e-05,
|
| 2451 |
-
"loss": 0.0047,
|
| 2452 |
-
"step": 349
|
| 2453 |
-
},
|
| 2454 |
-
{
|
| 2455 |
-
"epoch": 350.0,
|
| 2456 |
-
"grad_norm": 0.01766958087682724,
|
| 2457 |
-
"learning_rate": 6.104417670682732e-05,
|
| 2458 |
-
"loss": 0.005,
|
| 2459 |
-
"step": 350
|
| 2460 |
-
},
|
| 2461 |
-
{
|
| 2462 |
-
"epoch": 351.0,
|
| 2463 |
-
"grad_norm": 0.026927420869469643,
|
| 2464 |
-
"learning_rate": 6.06425702811245e-05,
|
| 2465 |
-
"loss": 0.0048,
|
| 2466 |
-
"step": 351
|
| 2467 |
-
},
|
| 2468 |
-
{
|
| 2469 |
-
"epoch": 352.0,
|
| 2470 |
-
"grad_norm": 0.016552148386836052,
|
| 2471 |
-
"learning_rate": 6.02409638554217e-05,
|
| 2472 |
-
"loss": 0.005,
|
| 2473 |
-
"step": 352
|
| 2474 |
-
},
|
| 2475 |
-
{
|
| 2476 |
-
"epoch": 353.0,
|
| 2477 |
-
"grad_norm": 0.02408471331000328,
|
| 2478 |
-
"learning_rate": 5.983935742971888e-05,
|
| 2479 |
-
"loss": 0.0048,
|
| 2480 |
-
"step": 353
|
| 2481 |
-
},
|
| 2482 |
-
{
|
| 2483 |
-
"epoch": 354.0,
|
| 2484 |
-
"grad_norm": 0.0007216089288704097,
|
| 2485 |
-
"learning_rate": 5.943775100401606e-05,
|
| 2486 |
-
"loss": 0.0049,
|
| 2487 |
-
"step": 354
|
| 2488 |
-
},
|
| 2489 |
-
{
|
| 2490 |
-
"epoch": 355.0,
|
| 2491 |
-
"grad_norm": 0.009755726903676987,
|
| 2492 |
-
"learning_rate": 5.903614457831326e-05,
|
| 2493 |
-
"loss": 0.0047,
|
| 2494 |
-
"step": 355
|
| 2495 |
-
},
|
| 2496 |
-
{
|
| 2497 |
-
"epoch": 356.0,
|
| 2498 |
-
"grad_norm": 0.026764798909425735,
|
| 2499 |
-
"learning_rate": 5.863453815261044e-05,
|
| 2500 |
-
"loss": 0.0048,
|
| 2501 |
-
"step": 356
|
| 2502 |
-
},
|
| 2503 |
-
{
|
| 2504 |
-
"epoch": 357.0,
|
| 2505 |
-
"grad_norm": 0.007179186213761568,
|
| 2506 |
-
"learning_rate": 5.823293172690764e-05,
|
| 2507 |
-
"loss": 0.0047,
|
| 2508 |
-
"step": 357
|
| 2509 |
-
},
|
| 2510 |
-
{
|
| 2511 |
-
"epoch": 358.0,
|
| 2512 |
-
"grad_norm": 0.023883482441306114,
|
| 2513 |
-
"learning_rate": 5.783132530120482e-05,
|
| 2514 |
-
"loss": 0.0048,
|
| 2515 |
-
"step": 358
|
| 2516 |
-
},
|
| 2517 |
-
{
|
| 2518 |
-
"epoch": 359.0,
|
| 2519 |
-
"grad_norm": 0.007125131320208311,
|
| 2520 |
-
"learning_rate": 5.7429718875502015e-05,
|
| 2521 |
-
"loss": 0.0047,
|
| 2522 |
-
"step": 359
|
| 2523 |
-
},
|
| 2524 |
-
{
|
| 2525 |
-
"epoch": 360.0,
|
| 2526 |
-
"grad_norm": 0.026448102667927742,
|
| 2527 |
-
"learning_rate": 5.70281124497992e-05,
|
| 2528 |
-
"loss": 0.0048,
|
| 2529 |
-
"step": 360
|
| 2530 |
-
},
|
| 2531 |
-
{
|
| 2532 |
-
"epoch": 361.0,
|
| 2533 |
-
"grad_norm": 0.009603764861822128,
|
| 2534 |
-
"learning_rate": 5.6626506024096394e-05,
|
| 2535 |
-
"loss": 0.0047,
|
| 2536 |
-
"step": 361
|
| 2537 |
-
},
|
| 2538 |
-
{
|
| 2539 |
-
"epoch": 362.0,
|
| 2540 |
-
"grad_norm": 0.023749757558107376,
|
| 2541 |
-
"learning_rate": 5.6224899598393576e-05,
|
| 2542 |
-
"loss": 0.0048,
|
| 2543 |
-
"step": 362
|
| 2544 |
-
},
|
| 2545 |
-
{
|
| 2546 |
-
"epoch": 363.0,
|
| 2547 |
-
"grad_norm": 0.007082722615450621,
|
| 2548 |
-
"learning_rate": 5.582329317269076e-05,
|
| 2549 |
-
"loss": 0.0047,
|
| 2550 |
-
"step": 363
|
| 2551 |
-
},
|
| 2552 |
-
{
|
| 2553 |
-
"epoch": 364.0,
|
| 2554 |
-
"grad_norm": 0.0006819483824074268,
|
| 2555 |
-
"learning_rate": 5.5421686746987955e-05,
|
| 2556 |
-
"loss": 0.0049,
|
| 2557 |
-
"step": 364
|
| 2558 |
-
},
|
| 2559 |
-
{
|
| 2560 |
-
"epoch": 365.0,
|
| 2561 |
-
"grad_norm": 0.026299143210053444,
|
| 2562 |
-
"learning_rate": 5.502008032128514e-05,
|
| 2563 |
-
"loss": 0.0048,
|
| 2564 |
-
"step": 365
|
| 2565 |
-
},
|
| 2566 |
-
{
|
| 2567 |
-
"epoch": 366.0,
|
| 2568 |
-
"grad_norm": 0.007077427115291357,
|
| 2569 |
-
"learning_rate": 5.461847389558233e-05,
|
| 2570 |
-
"loss": 0.0047,
|
| 2571 |
-
"step": 366
|
| 2572 |
-
},
|
| 2573 |
-
{
|
| 2574 |
-
"epoch": 367.0,
|
| 2575 |
-
"grad_norm": 0.0070725963450968266,
|
| 2576 |
-
"learning_rate": 5.4216867469879516e-05,
|
| 2577 |
-
"loss": 0.0047,
|
| 2578 |
-
"step": 367
|
| 2579 |
-
},
|
| 2580 |
-
{
|
| 2581 |
-
"epoch": 368.0,
|
| 2582 |
-
"grad_norm": 0.007042643614113331,
|
| 2583 |
-
"learning_rate": 5.381526104417671e-05,
|
| 2584 |
-
"loss": 0.0047,
|
| 2585 |
-
"step": 368
|
| 2586 |
-
},
|
| 2587 |
-
{
|
| 2588 |
-
"epoch": 369.0,
|
| 2589 |
-
"grad_norm": 0.009521468542516232,
|
| 2590 |
-
"learning_rate": 5.3413654618473894e-05,
|
| 2591 |
-
"loss": 0.0047,
|
| 2592 |
-
"step": 369
|
| 2593 |
-
},
|
| 2594 |
-
{
|
| 2595 |
-
"epoch": 370.0,
|
| 2596 |
-
"grad_norm": 0.0022259766701608896,
|
| 2597 |
-
"learning_rate": 5.301204819277109e-05,
|
| 2598 |
-
"loss": 0.0045,
|
| 2599 |
-
"step": 370
|
| 2600 |
-
},
|
| 2601 |
-
{
|
| 2602 |
-
"epoch": 371.0,
|
| 2603 |
-
"grad_norm": 0.023667145520448685,
|
| 2604 |
-
"learning_rate": 5.261044176706827e-05,
|
| 2605 |
-
"loss": 0.0048,
|
| 2606 |
-
"step": 371
|
| 2607 |
-
},
|
| 2608 |
-
{
|
| 2609 |
-
"epoch": 372.0,
|
| 2610 |
-
"grad_norm": 0.009567582048475742,
|
| 2611 |
-
"learning_rate": 5.220883534136547e-05,
|
| 2612 |
-
"loss": 0.0047,
|
| 2613 |
-
"step": 372
|
| 2614 |
-
},
|
| 2615 |
-
{
|
| 2616 |
-
"epoch": 373.0,
|
| 2617 |
-
"grad_norm": 0.00958697684109211,
|
| 2618 |
-
"learning_rate": 5.180722891566265e-05,
|
| 2619 |
-
"loss": 0.0047,
|
| 2620 |
-
"step": 373
|
| 2621 |
-
},
|
| 2622 |
-
{
|
| 2623 |
-
"epoch": 374.0,
|
| 2624 |
-
"grad_norm": 0.00959386583417654,
|
| 2625 |
-
"learning_rate": 5.140562248995984e-05,
|
| 2626 |
-
"loss": 0.0047,
|
| 2627 |
-
"step": 374
|
| 2628 |
-
},
|
| 2629 |
-
{
|
| 2630 |
-
"epoch": 375.0,
|
| 2631 |
-
"grad_norm": 0.00957813672721386,
|
| 2632 |
-
"learning_rate": 5.100401606425703e-05,
|
| 2633 |
-
"loss": 0.0047,
|
| 2634 |
-
"step": 375
|
| 2635 |
-
},
|
| 2636 |
-
{
|
| 2637 |
-
"epoch": 376.0,
|
| 2638 |
-
"grad_norm": 0.016183258965611458,
|
| 2639 |
-
"learning_rate": 5.060240963855422e-05,
|
| 2640 |
-
"loss": 0.005,
|
| 2641 |
-
"step": 376
|
| 2642 |
-
},
|
| 2643 |
-
{
|
| 2644 |
-
"epoch": 377.0,
|
| 2645 |
-
"grad_norm": 0.007078688126057386,
|
| 2646 |
-
"learning_rate": 5.020080321285141e-05,
|
| 2647 |
-
"loss": 0.0047,
|
| 2648 |
-
"step": 377
|
| 2649 |
-
},
|
| 2650 |
-
{
|
| 2651 |
-
"epoch": 378.0,
|
| 2652 |
-
"grad_norm": 0.007077342830598354,
|
| 2653 |
-
"learning_rate": 4.97991967871486e-05,
|
| 2654 |
-
"loss": 0.0047,
|
| 2655 |
-
"step": 378
|
| 2656 |
-
},
|
| 2657 |
-
{
|
| 2658 |
-
"epoch": 379.0,
|
| 2659 |
-
"grad_norm": 0.002201406517997384,
|
| 2660 |
-
"learning_rate": 4.9397590361445786e-05,
|
| 2661 |
-
"loss": 0.0045,
|
| 2662 |
-
"step": 379
|
| 2663 |
-
},
|
| 2664 |
-
{
|
| 2665 |
-
"epoch": 380.0,
|
| 2666 |
-
"grad_norm": 0.009655999019742012,
|
| 2667 |
-
"learning_rate": 4.8995983935742975e-05,
|
| 2668 |
-
"loss": 0.0047,
|
| 2669 |
-
"step": 380
|
| 2670 |
-
},
|
| 2671 |
-
{
|
| 2672 |
-
"epoch": 381.0,
|
| 2673 |
-
"grad_norm": 0.017296917736530304,
|
| 2674 |
-
"learning_rate": 4.8594377510040165e-05,
|
| 2675 |
-
"loss": 0.005,
|
| 2676 |
-
"step": 381
|
| 2677 |
-
},
|
| 2678 |
-
{
|
| 2679 |
-
"epoch": 382.0,
|
| 2680 |
-
"grad_norm": 0.007125664968043566,
|
| 2681 |
-
"learning_rate": 4.8192771084337354e-05,
|
| 2682 |
-
"loss": 0.0047,
|
| 2683 |
-
"step": 382
|
| 2684 |
-
},
|
| 2685 |
-
{
|
| 2686 |
-
"epoch": 383.0,
|
| 2687 |
-
"grad_norm": 0.0238560251891613,
|
| 2688 |
-
"learning_rate": 4.779116465863454e-05,
|
| 2689 |
-
"loss": 0.0048,
|
| 2690 |
-
"step": 383
|
| 2691 |
-
},
|
| 2692 |
-
{
|
| 2693 |
-
"epoch": 384.0,
|
| 2694 |
-
"grad_norm": 0.009672388434410095,
|
| 2695 |
-
"learning_rate": 4.738955823293173e-05,
|
| 2696 |
-
"loss": 0.0047,
|
| 2697 |
-
"step": 384
|
| 2698 |
-
},
|
| 2699 |
-
{
|
| 2700 |
-
"epoch": 385.0,
|
| 2701 |
-
"grad_norm": 0.0022134389728307724,
|
| 2702 |
-
"learning_rate": 4.698795180722892e-05,
|
| 2703 |
-
"loss": 0.0045,
|
| 2704 |
-
"step": 385
|
| 2705 |
-
},
|
| 2706 |
-
{
|
| 2707 |
-
"epoch": 386.0,
|
| 2708 |
-
"grad_norm": 0.00969780795276165,
|
| 2709 |
-
"learning_rate": 4.658634538152611e-05,
|
| 2710 |
-
"loss": 0.0047,
|
| 2711 |
-
"step": 386
|
| 2712 |
-
},
|
| 2713 |
-
{
|
| 2714 |
-
"epoch": 387.0,
|
| 2715 |
-
"grad_norm": 0.007160300388932228,
|
| 2716 |
-
"learning_rate": 4.61847389558233e-05,
|
| 2717 |
-
"loss": 0.0047,
|
| 2718 |
-
"step": 387
|
| 2719 |
-
},
|
| 2720 |
-
{
|
| 2721 |
-
"epoch": 388.0,
|
| 2722 |
-
"grad_norm": 0.007146279327571392,
|
| 2723 |
-
"learning_rate": 4.578313253012048e-05,
|
| 2724 |
-
"loss": 0.0047,
|
| 2725 |
-
"step": 388
|
| 2726 |
-
},
|
| 2727 |
-
{
|
| 2728 |
-
"epoch": 389.0,
|
| 2729 |
-
"grad_norm": 0.002192781073972583,
|
| 2730 |
-
"learning_rate": 4.538152610441767e-05,
|
| 2731 |
-
"loss": 0.0045,
|
| 2732 |
-
"step": 389
|
| 2733 |
-
},
|
| 2734 |
-
{
|
| 2735 |
-
"epoch": 390.0,
|
| 2736 |
-
"grad_norm": 0.009754459373652935,
|
| 2737 |
-
"learning_rate": 4.497991967871486e-05,
|
| 2738 |
-
"loss": 0.0047,
|
| 2739 |
-
"step": 390
|
| 2740 |
-
},
|
| 2741 |
-
{
|
| 2742 |
-
"epoch": 391.0,
|
| 2743 |
-
"grad_norm": 0.00970425084233284,
|
| 2744 |
-
"learning_rate": 4.457831325301205e-05,
|
| 2745 |
-
"loss": 0.0047,
|
| 2746 |
-
"step": 391
|
| 2747 |
-
},
|
| 2748 |
-
{
|
| 2749 |
-
"epoch": 392.0,
|
| 2750 |
-
"grad_norm": 0.007177860010415316,
|
| 2751 |
-
"learning_rate": 4.417670682730924e-05,
|
| 2752 |
-
"loss": 0.0047,
|
| 2753 |
-
"step": 392
|
| 2754 |
-
},
|
| 2755 |
-
{
|
| 2756 |
-
"epoch": 393.0,
|
| 2757 |
-
"grad_norm": 0.024075862020254135,
|
| 2758 |
-
"learning_rate": 4.377510040160643e-05,
|
| 2759 |
-
"loss": 0.0048,
|
| 2760 |
-
"step": 393
|
| 2761 |
-
},
|
| 2762 |
-
{
|
| 2763 |
-
"epoch": 394.0,
|
| 2764 |
-
"grad_norm": 0.009754209779202938,
|
| 2765 |
-
"learning_rate": 4.337349397590362e-05,
|
| 2766 |
-
"loss": 0.0047,
|
| 2767 |
-
"step": 394
|
| 2768 |
-
},
|
| 2769 |
-
{
|
| 2770 |
-
"epoch": 395.0,
|
| 2771 |
-
"grad_norm": 0.009791336953639984,
|
| 2772 |
-
"learning_rate": 4.297188755020081e-05,
|
| 2773 |
-
"loss": 0.0047,
|
| 2774 |
-
"step": 395
|
| 2775 |
-
},
|
| 2776 |
-
{
|
| 2777 |
-
"epoch": 396.0,
|
| 2778 |
-
"grad_norm": 0.009780234657227993,
|
| 2779 |
-
"learning_rate": 4.2570281124497996e-05,
|
| 2780 |
-
"loss": 0.0047,
|
| 2781 |
-
"step": 396
|
| 2782 |
-
},
|
| 2783 |
-
{
|
| 2784 |
-
"epoch": 397.0,
|
| 2785 |
-
"grad_norm": 0.007180201821029186,
|
| 2786 |
-
"learning_rate": 4.2168674698795186e-05,
|
| 2787 |
-
"loss": 0.0047,
|
| 2788 |
-
"step": 397
|
| 2789 |
-
},
|
| 2790 |
-
{
|
| 2791 |
-
"epoch": 398.0,
|
| 2792 |
-
"grad_norm": 0.007196042221039534,
|
| 2793 |
-
"learning_rate": 4.176706827309237e-05,
|
| 2794 |
-
"loss": 0.0047,
|
| 2795 |
-
"step": 398
|
| 2796 |
-
},
|
| 2797 |
-
{
|
| 2798 |
-
"epoch": 399.0,
|
| 2799 |
-
"grad_norm": 0.00722030783072114,
|
| 2800 |
-
"learning_rate": 4.136546184738956e-05,
|
| 2801 |
-
"loss": 0.0047,
|
| 2802 |
-
"step": 399
|
| 2803 |
-
},
|
| 2804 |
-
{
|
| 2805 |
-
"epoch": 400.0,
|
| 2806 |
-
"grad_norm": 0.009809617884457111,
|
| 2807 |
-
"learning_rate": 4.0963855421686746e-05,
|
| 2808 |
-
"loss": 0.0047,
|
| 2809 |
-
"step": 400
|
| 2810 |
-
},
|
| 2811 |
-
{
|
| 2812 |
-
"epoch": 401.0,
|
| 2813 |
-
"grad_norm": 0.007194210775196552,
|
| 2814 |
-
"learning_rate": 4.0562248995983936e-05,
|
| 2815 |
-
"loss": 0.0047,
|
| 2816 |
-
"step": 401
|
| 2817 |
-
},
|
| 2818 |
-
{
|
| 2819 |
-
"epoch": 402.0,
|
| 2820 |
-
"grad_norm": 0.00984671525657177,
|
| 2821 |
-
"learning_rate": 4.0160642570281125e-05,
|
| 2822 |
-
"loss": 0.0047,
|
| 2823 |
-
"step": 402
|
| 2824 |
-
},
|
| 2825 |
-
{
|
| 2826 |
-
"epoch": 403.0,
|
| 2827 |
-
"grad_norm": 0.00983081478625536,
|
| 2828 |
-
"learning_rate": 3.9759036144578314e-05,
|
| 2829 |
-
"loss": 0.0047,
|
| 2830 |
-
"step": 403
|
| 2831 |
-
},
|
| 2832 |
-
{
|
| 2833 |
-
"epoch": 404.0,
|
| 2834 |
-
"grad_norm": 0.007195820100605488,
|
| 2835 |
-
"learning_rate": 3.93574297188755e-05,
|
| 2836 |
-
"loss": 0.0047,
|
| 2837 |
-
"step": 404
|
| 2838 |
-
},
|
| 2839 |
-
{
|
| 2840 |
-
"epoch": 405.0,
|
| 2841 |
-
"grad_norm": 0.007182563189417124,
|
| 2842 |
-
"learning_rate": 3.895582329317269e-05,
|
| 2843 |
-
"loss": 0.0047,
|
| 2844 |
-
"step": 405
|
| 2845 |
-
},
|
| 2846 |
-
{
|
| 2847 |
-
"epoch": 406.0,
|
| 2848 |
-
"grad_norm": 0.007222824264317751,
|
| 2849 |
-
"learning_rate": 3.855421686746988e-05,
|
| 2850 |
-
"loss": 0.0047,
|
| 2851 |
-
"step": 406
|
| 2852 |
-
},
|
| 2853 |
-
{
|
| 2854 |
-
"epoch": 407.0,
|
| 2855 |
-
"grad_norm": 0.007211462128907442,
|
| 2856 |
-
"learning_rate": 3.815261044176707e-05,
|
| 2857 |
-
"loss": 0.0047,
|
| 2858 |
-
"step": 407
|
| 2859 |
-
},
|
| 2860 |
-
{
|
| 2861 |
-
"epoch": 408.0,
|
| 2862 |
-
"grad_norm": 0.009921679273247719,
|
| 2863 |
-
"learning_rate": 3.7751004016064253e-05,
|
| 2864 |
-
"loss": 0.0047,
|
| 2865 |
-
"step": 408
|
| 2866 |
-
},
|
| 2867 |
-
{
|
| 2868 |
-
"epoch": 409.0,
|
| 2869 |
-
"grad_norm": 0.02687370777130127,
|
| 2870 |
-
"learning_rate": 3.734939759036144e-05,
|
| 2871 |
-
"loss": 0.0048,
|
| 2872 |
-
"step": 409
|
| 2873 |
-
},
|
| 2874 |
-
{
|
| 2875 |
-
"epoch": 410.0,
|
| 2876 |
-
"grad_norm": 0.009853348135948181,
|
| 2877 |
-
"learning_rate": 3.694779116465863e-05,
|
| 2878 |
-
"loss": 0.0047,
|
| 2879 |
-
"step": 410
|
| 2880 |
-
},
|
| 2881 |
-
{
|
| 2882 |
-
"epoch": 411.0,
|
| 2883 |
-
"grad_norm": 0.024240443482995033,
|
| 2884 |
-
"learning_rate": 3.654618473895582e-05,
|
| 2885 |
-
"loss": 0.0048,
|
| 2886 |
-
"step": 411
|
| 2887 |
-
},
|
| 2888 |
-
{
|
| 2889 |
-
"epoch": 412.0,
|
| 2890 |
-
"grad_norm": 0.04131508618593216,
|
| 2891 |
-
"learning_rate": 3.614457831325301e-05,
|
| 2892 |
-
"loss": 0.0048,
|
| 2893 |
-
"step": 412
|
| 2894 |
-
},
|
| 2895 |
-
{
|
| 2896 |
-
"epoch": 413.0,
|
| 2897 |
-
"grad_norm": 0.009910643100738525,
|
| 2898 |
-
"learning_rate": 3.57429718875502e-05,
|
| 2899 |
-
"loss": 0.0047,
|
| 2900 |
-
"step": 413
|
| 2901 |
-
},
|
| 2902 |
-
{
|
| 2903 |
-
"epoch": 414.0,
|
| 2904 |
-
"grad_norm": 0.03583821654319763,
|
| 2905 |
-
"learning_rate": 3.534136546184739e-05,
|
| 2906 |
-
"loss": 0.0046,
|
| 2907 |
-
"step": 414
|
| 2908 |
-
},
|
| 2909 |
-
{
|
| 2910 |
-
"epoch": 415.0,
|
| 2911 |
-
"grad_norm": 0.017576098442077637,
|
| 2912 |
-
"learning_rate": 3.4939759036144585e-05,
|
| 2913 |
-
"loss": 0.005,
|
| 2914 |
-
"step": 415
|
| 2915 |
-
},
|
| 2916 |
-
{
|
| 2917 |
-
"epoch": 416.0,
|
| 2918 |
-
"grad_norm": 0.01653764583170414,
|
| 2919 |
-
"learning_rate": 3.4538152610441774e-05,
|
| 2920 |
-
"loss": 0.005,
|
| 2921 |
-
"step": 416
|
| 2922 |
-
},
|
| 2923 |
-
{
|
| 2924 |
-
"epoch": 417.0,
|
| 2925 |
-
"grad_norm": 0.024247560650110245,
|
| 2926 |
-
"learning_rate": 3.413654618473896e-05,
|
| 2927 |
-
"loss": 0.0048,
|
| 2928 |
-
"step": 417
|
| 2929 |
-
},
|
| 2930 |
-
{
|
| 2931 |
-
"epoch": 418.0,
|
| 2932 |
-
"grad_norm": 0.007164428010582924,
|
| 2933 |
-
"learning_rate": 3.3734939759036146e-05,
|
| 2934 |
-
"loss": 0.0047,
|
| 2935 |
-
"step": 418
|
| 2936 |
-
},
|
| 2937 |
-
{
|
| 2938 |
-
"epoch": 419.0,
|
| 2939 |
-
"grad_norm": 0.0098257539793849,
|
| 2940 |
-
"learning_rate": 3.3333333333333335e-05,
|
| 2941 |
-
"loss": 0.0047,
|
| 2942 |
-
"step": 419
|
| 2943 |
-
},
|
| 2944 |
-
{
|
| 2945 |
-
"epoch": 420.0,
|
| 2946 |
-
"grad_norm": 0.017520597204566002,
|
| 2947 |
-
"learning_rate": 3.2931726907630524e-05,
|
| 2948 |
-
"loss": 0.005,
|
| 2949 |
-
"step": 420
|
| 2950 |
-
},
|
| 2951 |
-
{
|
| 2952 |
-
"epoch": 421.0,
|
| 2953 |
-
"grad_norm": 0.018933523446321487,
|
| 2954 |
-
"learning_rate": 3.253012048192771e-05,
|
| 2955 |
-
"loss": 0.0045,
|
| 2956 |
-
"step": 421
|
| 2957 |
-
},
|
| 2958 |
-
{
|
| 2959 |
-
"epoch": 422.0,
|
| 2960 |
-
"grad_norm": 0.007097120396792889,
|
| 2961 |
-
"learning_rate": 3.21285140562249e-05,
|
| 2962 |
-
"loss": 0.0047,
|
| 2963 |
-
"step": 422
|
| 2964 |
-
},
|
| 2965 |
-
{
|
| 2966 |
-
"epoch": 423.0,
|
| 2967 |
-
"grad_norm": 0.024037552997469902,
|
| 2968 |
-
"learning_rate": 3.172690763052209e-05,
|
| 2969 |
-
"loss": 0.0048,
|
| 2970 |
-
"step": 423
|
| 2971 |
-
},
|
| 2972 |
-
{
|
| 2973 |
-
"epoch": 424.0,
|
| 2974 |
-
"grad_norm": 0.024066118523478508,
|
| 2975 |
-
"learning_rate": 3.132530120481928e-05,
|
| 2976 |
-
"loss": 0.0048,
|
| 2977 |
-
"step": 424
|
| 2978 |
-
},
|
| 2979 |
-
{
|
| 2980 |
-
"epoch": 425.0,
|
| 2981 |
-
"grad_norm": 0.0006517537985928357,
|
| 2982 |
-
"learning_rate": 3.092369477911647e-05,
|
| 2983 |
-
"loss": 0.0049,
|
| 2984 |
-
"step": 425
|
| 2985 |
-
},
|
| 2986 |
-
{
|
| 2987 |
-
"epoch": 426.0,
|
| 2988 |
-
"grad_norm": 0.034052085131406784,
|
| 2989 |
-
"learning_rate": 3.052208835341366e-05,
|
| 2990 |
-
"loss": 0.005,
|
| 2991 |
-
"step": 426
|
| 2992 |
-
},
|
| 2993 |
-
{
|
| 2994 |
-
"epoch": 427.0,
|
| 2995 |
-
"grad_norm": 0.017422359436750412,
|
| 2996 |
-
"learning_rate": 3.012048192771085e-05,
|
| 2997 |
-
"loss": 0.005,
|
| 2998 |
-
"step": 427
|
| 2999 |
-
},
|
| 3000 |
-
{
|
| 3001 |
-
"epoch": 428.0,
|
| 3002 |
-
"grad_norm": 0.009725161828100681,
|
| 3003 |
-
"learning_rate": 2.971887550200803e-05,
|
| 3004 |
-
"loss": 0.0047,
|
| 3005 |
-
"step": 428
|
| 3006 |
-
},
|
| 3007 |
-
{
|
| 3008 |
-
"epoch": 429.0,
|
| 3009 |
-
"grad_norm": 0.023919757455587387,
|
| 3010 |
-
"learning_rate": 2.931726907630522e-05,
|
| 3011 |
-
"loss": 0.0048,
|
| 3012 |
-
"step": 429
|
| 3013 |
-
},
|
| 3014 |
-
{
|
| 3015 |
-
"epoch": 430.0,
|
| 3016 |
-
"grad_norm": 0.04076967015862465,
|
| 3017 |
-
"learning_rate": 2.891566265060241e-05,
|
| 3018 |
-
"loss": 0.0048,
|
| 3019 |
-
"step": 430
|
| 3020 |
-
},
|
| 3021 |
-
{
|
| 3022 |
-
"epoch": 431.0,
|
| 3023 |
-
"grad_norm": 0.002202529925853014,
|
| 3024 |
-
"learning_rate": 2.85140562248996e-05,
|
| 3025 |
-
"loss": 0.0045,
|
| 3026 |
-
"step": 431
|
| 3027 |
-
},
|
| 3028 |
-
{
|
| 3029 |
-
"epoch": 432.0,
|
| 3030 |
-
"grad_norm": 0.026460327208042145,
|
| 3031 |
-
"learning_rate": 2.8112449799196788e-05,
|
| 3032 |
-
"loss": 0.0048,
|
| 3033 |
-
"step": 432
|
| 3034 |
-
},
|
| 3035 |
-
{
|
| 3036 |
-
"epoch": 433.0,
|
| 3037 |
-
"grad_norm": 0.026399288326501846,
|
| 3038 |
-
"learning_rate": 2.7710843373493977e-05,
|
| 3039 |
-
"loss": 0.0048,
|
| 3040 |
-
"step": 433
|
| 3041 |
-
},
|
| 3042 |
-
{
|
| 3043 |
-
"epoch": 434.0,
|
| 3044 |
-
"grad_norm": 0.007069241255521774,
|
| 3045 |
-
"learning_rate": 2.7309236947791167e-05,
|
| 3046 |
-
"loss": 0.0047,
|
| 3047 |
-
"step": 434
|
| 3048 |
-
},
|
| 3049 |
-
{
|
| 3050 |
-
"epoch": 435.0,
|
| 3051 |
-
"grad_norm": 0.023836608976125717,
|
| 3052 |
-
"learning_rate": 2.6907630522088356e-05,
|
| 3053 |
-
"loss": 0.0048,
|
| 3054 |
-
"step": 435
|
| 3055 |
-
},
|
| 3056 |
-
{
|
| 3057 |
-
"epoch": 436.0,
|
| 3058 |
-
"grad_norm": 0.007101175840944052,
|
| 3059 |
-
"learning_rate": 2.6506024096385545e-05,
|
| 3060 |
-
"loss": 0.0047,
|
| 3061 |
-
"step": 436
|
| 3062 |
-
},
|
| 3063 |
-
{
|
| 3064 |
-
"epoch": 437.0,
|
| 3065 |
-
"grad_norm": 0.009633008390665054,
|
| 3066 |
-
"learning_rate": 2.6104417670682734e-05,
|
| 3067 |
-
"loss": 0.0047,
|
| 3068 |
-
"step": 437
|
| 3069 |
-
},
|
| 3070 |
-
{
|
| 3071 |
-
"epoch": 438.0,
|
| 3072 |
-
"grad_norm": 0.00965973362326622,
|
| 3073 |
-
"learning_rate": 2.570281124497992e-05,
|
| 3074 |
-
"loss": 0.0047,
|
| 3075 |
-
"step": 438
|
| 3076 |
-
},
|
| 3077 |
-
{
|
| 3078 |
-
"epoch": 439.0,
|
| 3079 |
-
"grad_norm": 0.009659296832978725,
|
| 3080 |
-
"learning_rate": 2.530120481927711e-05,
|
| 3081 |
-
"loss": 0.0047,
|
| 3082 |
-
"step": 439
|
| 3083 |
-
},
|
| 3084 |
-
{
|
| 3085 |
-
"epoch": 440.0,
|
| 3086 |
-
"grad_norm": 0.00703496765345335,
|
| 3087 |
-
"learning_rate": 2.48995983935743e-05,
|
| 3088 |
-
"loss": 0.0047,
|
| 3089 |
-
"step": 440
|
| 3090 |
-
},
|
| 3091 |
-
{
|
| 3092 |
-
"epoch": 441.0,
|
| 3093 |
-
"grad_norm": 0.007030785549432039,
|
| 3094 |
-
"learning_rate": 2.4497991967871488e-05,
|
| 3095 |
-
"loss": 0.0047,
|
| 3096 |
-
"step": 441
|
| 3097 |
-
},
|
| 3098 |
-
{
|
| 3099 |
-
"epoch": 442.0,
|
| 3100 |
-
"grad_norm": 0.007043666671961546,
|
| 3101 |
-
"learning_rate": 2.4096385542168677e-05,
|
| 3102 |
-
"loss": 0.0047,
|
| 3103 |
-
"step": 442
|
| 3104 |
-
},
|
| 3105 |
-
{
|
| 3106 |
-
"epoch": 443.0,
|
| 3107 |
-
"grad_norm": 0.007040888071060181,
|
| 3108 |
-
"learning_rate": 2.3694779116465866e-05,
|
| 3109 |
-
"loss": 0.0047,
|
| 3110 |
-
"step": 443
|
| 3111 |
-
},
|
| 3112 |
-
{
|
| 3113 |
-
"epoch": 444.0,
|
| 3114 |
-
"grad_norm": 0.009699525311589241,
|
| 3115 |
-
"learning_rate": 2.3293172690763055e-05,
|
| 3116 |
-
"loss": 0.0047,
|
| 3117 |
-
"step": 444
|
| 3118 |
-
},
|
| 3119 |
-
{
|
| 3120 |
-
"epoch": 445.0,
|
| 3121 |
-
"grad_norm": 0.0007015722803771496,
|
| 3122 |
-
"learning_rate": 2.289156626506024e-05,
|
| 3123 |
-
"loss": 0.0049,
|
| 3124 |
-
"step": 445
|
| 3125 |
-
},
|
| 3126 |
-
{
|
| 3127 |
-
"epoch": 446.0,
|
| 3128 |
-
"grad_norm": 0.00966339185833931,
|
| 3129 |
-
"learning_rate": 2.248995983935743e-05,
|
| 3130 |
-
"loss": 0.0047,
|
| 3131 |
-
"step": 446
|
| 3132 |
-
},
|
| 3133 |
-
{
|
| 3134 |
-
"epoch": 447.0,
|
| 3135 |
-
"grad_norm": 0.009668245911598206,
|
| 3136 |
-
"learning_rate": 2.208835341365462e-05,
|
| 3137 |
-
"loss": 0.0047,
|
| 3138 |
-
"step": 447
|
| 3139 |
-
},
|
| 3140 |
-
{
|
| 3141 |
-
"epoch": 448.0,
|
| 3142 |
-
"grad_norm": 0.007046937942504883,
|
| 3143 |
-
"learning_rate": 2.168674698795181e-05,
|
| 3144 |
-
"loss": 0.0047,
|
| 3145 |
-
"step": 448
|
| 3146 |
-
},
|
| 3147 |
-
{
|
| 3148 |
-
"epoch": 449.0,
|
| 3149 |
-
"grad_norm": 0.0070527587085962296,
|
| 3150 |
-
"learning_rate": 2.1285140562248998e-05,
|
| 3151 |
-
"loss": 0.0047,
|
| 3152 |
-
"step": 449
|
| 3153 |
-
},
|
| 3154 |
-
{
|
| 3155 |
-
"epoch": 450.0,
|
| 3156 |
-
"grad_norm": 0.007038488052785397,
|
| 3157 |
-
"learning_rate": 2.0883534136546184e-05,
|
| 3158 |
-
"loss": 0.0047,
|
| 3159 |
-
"step": 450
|
| 3160 |
-
},
|
| 3161 |
-
{
|
| 3162 |
-
"epoch": 451.0,
|
| 3163 |
-
"grad_norm": 0.0021306683775037527,
|
| 3164 |
-
"learning_rate": 2.0481927710843373e-05,
|
| 3165 |
-
"loss": 0.0045,
|
| 3166 |
-
"step": 451
|
| 3167 |
-
},
|
| 3168 |
-
{
|
| 3169 |
-
"epoch": 452.0,
|
| 3170 |
-
"grad_norm": 0.007048788480460644,
|
| 3171 |
-
"learning_rate": 2.0080321285140562e-05,
|
| 3172 |
-
"loss": 0.0047,
|
| 3173 |
-
"step": 452
|
| 3174 |
-
},
|
| 3175 |
-
{
|
| 3176 |
-
"epoch": 453.0,
|
| 3177 |
-
"grad_norm": 0.0006838923436589539,
|
| 3178 |
-
"learning_rate": 1.967871485943775e-05,
|
| 3179 |
-
"loss": 0.0049,
|
| 3180 |
-
"step": 453
|
| 3181 |
-
},
|
| 3182 |
-
{
|
| 3183 |
-
"epoch": 454.0,
|
| 3184 |
-
"grad_norm": 0.00969866942614317,
|
| 3185 |
-
"learning_rate": 1.927710843373494e-05,
|
| 3186 |
-
"loss": 0.0047,
|
| 3187 |
-
"step": 454
|
| 3188 |
-
},
|
| 3189 |
-
{
|
| 3190 |
-
"epoch": 455.0,
|
| 3191 |
-
"grad_norm": 0.009712344966828823,
|
| 3192 |
-
"learning_rate": 1.8875502008032127e-05,
|
| 3193 |
-
"loss": 0.0047,
|
| 3194 |
-
"step": 455
|
| 3195 |
-
},
|
| 3196 |
-
{
|
| 3197 |
-
"epoch": 456.0,
|
| 3198 |
-
"grad_norm": 0.009688866324722767,
|
| 3199 |
-
"learning_rate": 1.8473895582329316e-05,
|
| 3200 |
-
"loss": 0.0047,
|
| 3201 |
-
"step": 456
|
| 3202 |
-
},
|
| 3203 |
-
{
|
| 3204 |
-
"epoch": 457.0,
|
| 3205 |
-
"grad_norm": 0.007071357686072588,
|
| 3206 |
-
"learning_rate": 1.8072289156626505e-05,
|
| 3207 |
-
"loss": 0.0047,
|
| 3208 |
-
"step": 457
|
| 3209 |
-
},
|
| 3210 |
-
{
|
| 3211 |
-
"epoch": 458.0,
|
| 3212 |
-
"grad_norm": 0.014747700653970242,
|
| 3213 |
-
"learning_rate": 1.7670682730923694e-05,
|
| 3214 |
-
"loss": 0.0045,
|
| 3215 |
-
"step": 458
|
| 3216 |
-
},
|
| 3217 |
-
{
|
| 3218 |
-
"epoch": 459.0,
|
| 3219 |
-
"grad_norm": 0.007095323875546455,
|
| 3220 |
-
"learning_rate": 1.7269076305220887e-05,
|
| 3221 |
-
"loss": 0.0047,
|
| 3222 |
-
"step": 459
|
| 3223 |
-
},
|
| 3224 |
-
{
|
| 3225 |
-
"epoch": 460.0,
|
| 3226 |
-
"grad_norm": 0.0021590902470052242,
|
| 3227 |
-
"learning_rate": 1.6867469879518073e-05,
|
| 3228 |
-
"loss": 0.0045,
|
| 3229 |
-
"step": 460
|
| 3230 |
-
},
|
| 3231 |
-
{
|
| 3232 |
-
"epoch": 461.0,
|
| 3233 |
-
"grad_norm": 0.0006770145264454186,
|
| 3234 |
-
"learning_rate": 1.6465863453815262e-05,
|
| 3235 |
-
"loss": 0.0049,
|
| 3236 |
-
"step": 461
|
| 3237 |
-
},
|
| 3238 |
-
{
|
| 3239 |
-
"epoch": 462.0,
|
| 3240 |
-
"grad_norm": 0.0006719469674862921,
|
| 3241 |
-
"learning_rate": 1.606425702811245e-05,
|
| 3242 |
-
"loss": 0.0049,
|
| 3243 |
-
"step": 462
|
| 3244 |
-
},
|
| 3245 |
-
{
|
| 3246 |
-
"epoch": 463.0,
|
| 3247 |
-
"grad_norm": 0.009731734171509743,
|
| 3248 |
-
"learning_rate": 1.566265060240964e-05,
|
| 3249 |
-
"loss": 0.0047,
|
| 3250 |
-
"step": 463
|
| 3251 |
-
},
|
| 3252 |
-
{
|
| 3253 |
-
"epoch": 464.0,
|
| 3254 |
-
"grad_norm": 0.009717794135212898,
|
| 3255 |
-
"learning_rate": 1.526104417670683e-05,
|
| 3256 |
-
"loss": 0.0047,
|
| 3257 |
-
"step": 464
|
| 3258 |
-
},
|
| 3259 |
-
{
|
| 3260 |
-
"epoch": 465.0,
|
| 3261 |
-
"grad_norm": 0.009746178984642029,
|
| 3262 |
-
"learning_rate": 1.4859437751004016e-05,
|
| 3263 |
-
"loss": 0.0047,
|
| 3264 |
-
"step": 465
|
| 3265 |
-
},
|
| 3266 |
-
{
|
| 3267 |
-
"epoch": 466.0,
|
| 3268 |
-
"grad_norm": 0.007102122530341148,
|
| 3269 |
-
"learning_rate": 1.4457831325301205e-05,
|
| 3270 |
-
"loss": 0.0047,
|
| 3271 |
-
"step": 466
|
| 3272 |
-
},
|
| 3273 |
-
{
|
| 3274 |
-
"epoch": 467.0,
|
| 3275 |
-
"grad_norm": 0.00708032725378871,
|
| 3276 |
-
"learning_rate": 1.4056224899598394e-05,
|
| 3277 |
-
"loss": 0.0047,
|
| 3278 |
-
"step": 467
|
| 3279 |
-
},
|
| 3280 |
-
{
|
| 3281 |
-
"epoch": 468.0,
|
| 3282 |
-
"grad_norm": 0.007086113095283508,
|
| 3283 |
-
"learning_rate": 1.3654618473895583e-05,
|
| 3284 |
-
"loss": 0.0047,
|
| 3285 |
-
"step": 468
|
| 3286 |
-
},
|
| 3287 |
-
{
|
| 3288 |
-
"epoch": 469.0,
|
| 3289 |
-
"grad_norm": 0.014743141829967499,
|
| 3290 |
-
"learning_rate": 1.3253012048192772e-05,
|
| 3291 |
-
"loss": 0.0045,
|
| 3292 |
-
"step": 469
|
| 3293 |
-
},
|
| 3294 |
-
{
|
| 3295 |
-
"epoch": 470.0,
|
| 3296 |
-
"grad_norm": 0.0021814049687236547,
|
| 3297 |
-
"learning_rate": 1.285140562248996e-05,
|
| 3298 |
-
"loss": 0.0045,
|
| 3299 |
-
"step": 470
|
| 3300 |
-
},
|
| 3301 |
-
{
|
| 3302 |
-
"epoch": 471.0,
|
| 3303 |
-
"grad_norm": 0.007091291714459658,
|
| 3304 |
-
"learning_rate": 1.244979919678715e-05,
|
| 3305 |
-
"loss": 0.0047,
|
| 3306 |
-
"step": 471
|
| 3307 |
-
},
|
| 3308 |
-
{
|
| 3309 |
-
"epoch": 472.0,
|
| 3310 |
-
"grad_norm": 0.0071353972889482975,
|
| 3311 |
-
"learning_rate": 1.2048192771084338e-05,
|
| 3312 |
-
"loss": 0.0047,
|
| 3313 |
-
"step": 472
|
| 3314 |
-
},
|
| 3315 |
-
{
|
| 3316 |
-
"epoch": 473.0,
|
| 3317 |
-
"grad_norm": 0.009766222909092903,
|
| 3318 |
-
"learning_rate": 1.1646586345381528e-05,
|
| 3319 |
-
"loss": 0.0047,
|
| 3320 |
-
"step": 473
|
| 3321 |
-
},
|
| 3322 |
-
{
|
| 3323 |
-
"epoch": 474.0,
|
| 3324 |
-
"grad_norm": 0.017397722229361534,
|
| 3325 |
-
"learning_rate": 1.1244979919678715e-05,
|
| 3326 |
-
"loss": 0.005,
|
| 3327 |
-
"step": 474
|
| 3328 |
-
},
|
| 3329 |
-
{
|
| 3330 |
-
"epoch": 475.0,
|
| 3331 |
-
"grad_norm": 0.026444077491760254,
|
| 3332 |
-
"learning_rate": 1.0843373493975904e-05,
|
| 3333 |
-
"loss": 0.0048,
|
| 3334 |
-
"step": 475
|
| 3335 |
-
},
|
| 3336 |
-
{
|
| 3337 |
-
"epoch": 476.0,
|
| 3338 |
-
"grad_norm": 0.009789888747036457,
|
| 3339 |
-
"learning_rate": 1.0441767068273092e-05,
|
| 3340 |
-
"loss": 0.0047,
|
| 3341 |
-
"step": 476
|
| 3342 |
-
},
|
| 3343 |
-
{
|
| 3344 |
-
"epoch": 477.0,
|
| 3345 |
-
"grad_norm": 0.007101455237716436,
|
| 3346 |
-
"learning_rate": 1.0040160642570281e-05,
|
| 3347 |
-
"loss": 0.0047,
|
| 3348 |
-
"step": 477
|
| 3349 |
-
},
|
| 3350 |
-
{
|
| 3351 |
-
"epoch": 478.0,
|
| 3352 |
-
"grad_norm": 0.007093328982591629,
|
| 3353 |
-
"learning_rate": 9.63855421686747e-06,
|
| 3354 |
-
"loss": 0.0047,
|
| 3355 |
-
"step": 478
|
| 3356 |
-
},
|
| 3357 |
-
{
|
| 3358 |
-
"epoch": 479.0,
|
| 3359 |
-
"grad_norm": 0.014785111881792545,
|
| 3360 |
-
"learning_rate": 9.236947791164658e-06,
|
| 3361 |
-
"loss": 0.0045,
|
| 3362 |
-
"step": 479
|
| 3363 |
-
},
|
| 3364 |
-
{
|
| 3365 |
-
"epoch": 480.0,
|
| 3366 |
-
"grad_norm": 0.014800012111663818,
|
| 3367 |
-
"learning_rate": 8.835341365461847e-06,
|
| 3368 |
-
"loss": 0.0045,
|
| 3369 |
-
"step": 480
|
| 3370 |
-
},
|
| 3371 |
-
{
|
| 3372 |
-
"epoch": 481.0,
|
| 3373 |
-
"grad_norm": 0.00710406294092536,
|
| 3374 |
-
"learning_rate": 8.433734939759036e-06,
|
| 3375 |
-
"loss": 0.0047,
|
| 3376 |
-
"step": 481
|
| 3377 |
-
},
|
| 3378 |
-
{
|
| 3379 |
-
"epoch": 482.0,
|
| 3380 |
-
"grad_norm": 0.0163735318928957,
|
| 3381 |
-
"learning_rate": 8.032128514056226e-06,
|
| 3382 |
-
"loss": 0.005,
|
| 3383 |
-
"step": 482
|
| 3384 |
-
},
|
| 3385 |
-
{
|
| 3386 |
-
"epoch": 483.0,
|
| 3387 |
-
"grad_norm": 0.00218359031714499,
|
| 3388 |
-
"learning_rate": 7.630522088353415e-06,
|
| 3389 |
-
"loss": 0.0045,
|
| 3390 |
-
"step": 483
|
| 3391 |
-
},
|
| 3392 |
-
{
|
| 3393 |
-
"epoch": 484.0,
|
| 3394 |
-
"grad_norm": 0.009749580174684525,
|
| 3395 |
-
"learning_rate": 7.228915662650602e-06,
|
| 3396 |
-
"loss": 0.0047,
|
| 3397 |
-
"step": 484
|
| 3398 |
-
},
|
| 3399 |
-
{
|
| 3400 |
-
"epoch": 485.0,
|
| 3401 |
-
"grad_norm": 0.009765205904841423,
|
| 3402 |
-
"learning_rate": 6.827309236947792e-06,
|
| 3403 |
-
"loss": 0.0047,
|
| 3404 |
-
"step": 485
|
| 3405 |
-
},
|
| 3406 |
-
{
|
| 3407 |
-
"epoch": 486.0,
|
| 3408 |
-
"grad_norm": 0.000685015635099262,
|
| 3409 |
-
"learning_rate": 6.42570281124498e-06,
|
| 3410 |
-
"loss": 0.0049,
|
| 3411 |
-
"step": 486
|
| 3412 |
-
},
|
| 3413 |
-
{
|
| 3414 |
-
"epoch": 487.0,
|
| 3415 |
-
"grad_norm": 0.00976163987070322,
|
| 3416 |
-
"learning_rate": 6.024096385542169e-06,
|
| 3417 |
-
"loss": 0.0047,
|
| 3418 |
-
"step": 487
|
| 3419 |
-
},
|
| 3420 |
-
{
|
| 3421 |
-
"epoch": 488.0,
|
| 3422 |
-
"grad_norm": 0.0006887281779199839,
|
| 3423 |
-
"learning_rate": 5.622489959839358e-06,
|
| 3424 |
-
"loss": 0.0049,
|
| 3425 |
-
"step": 488
|
| 3426 |
-
},
|
| 3427 |
-
{
|
| 3428 |
-
"epoch": 489.0,
|
| 3429 |
-
"grad_norm": 0.009809976443648338,
|
| 3430 |
-
"learning_rate": 5.220883534136546e-06,
|
| 3431 |
-
"loss": 0.0047,
|
| 3432 |
-
"step": 489
|
| 3433 |
-
},
|
| 3434 |
-
{
|
| 3435 |
-
"epoch": 490.0,
|
| 3436 |
-
"grad_norm": 0.009795522317290306,
|
| 3437 |
-
"learning_rate": 4.819277108433735e-06,
|
| 3438 |
-
"loss": 0.0047,
|
| 3439 |
-
"step": 490
|
| 3440 |
-
},
|
| 3441 |
-
{
|
| 3442 |
-
"epoch": 491.0,
|
| 3443 |
-
"grad_norm": 0.009749695658683777,
|
| 3444 |
-
"learning_rate": 4.417670682730924e-06,
|
| 3445 |
-
"loss": 0.0047,
|
| 3446 |
-
"step": 491
|
| 3447 |
-
},
|
| 3448 |
-
{
|
| 3449 |
-
"epoch": 492.0,
|
| 3450 |
-
"grad_norm": 0.0021824862342327833,
|
| 3451 |
-
"learning_rate": 4.016064257028113e-06,
|
| 3452 |
-
"loss": 0.0045,
|
| 3453 |
-
"step": 492
|
| 3454 |
-
},
|
| 3455 |
-
{
|
| 3456 |
-
"epoch": 493.0,
|
| 3457 |
-
"grad_norm": 0.007140113040804863,
|
| 3458 |
-
"learning_rate": 3.614457831325301e-06,
|
| 3459 |
-
"loss": 0.0047,
|
| 3460 |
-
"step": 493
|
| 3461 |
-
},
|
| 3462 |
-
{
|
| 3463 |
-
"epoch": 494.0,
|
| 3464 |
-
"grad_norm": 0.007124903611838818,
|
| 3465 |
-
"learning_rate": 3.21285140562249e-06,
|
| 3466 |
-
"loss": 0.0047,
|
| 3467 |
-
"step": 494
|
| 3468 |
-
},
|
| 3469 |
-
{
|
| 3470 |
-
"epoch": 495.0,
|
| 3471 |
-
"grad_norm": 0.0071158865466713905,
|
| 3472 |
-
"learning_rate": 2.811244979919679e-06,
|
| 3473 |
-
"loss": 0.0047,
|
| 3474 |
-
"step": 495
|
| 3475 |
-
},
|
| 3476 |
-
{
|
| 3477 |
-
"epoch": 496.0,
|
| 3478 |
-
"grad_norm": 0.007119073532521725,
|
| 3479 |
-
"learning_rate": 2.4096385542168676e-06,
|
| 3480 |
-
"loss": 0.0047,
|
| 3481 |
-
"step": 496
|
| 3482 |
-
},
|
| 3483 |
-
{
|
| 3484 |
-
"epoch": 497.0,
|
| 3485 |
-
"grad_norm": 0.007112898863852024,
|
| 3486 |
-
"learning_rate": 2.0080321285140564e-06,
|
| 3487 |
-
"loss": 0.0047,
|
| 3488 |
-
"step": 497
|
| 3489 |
-
},
|
| 3490 |
-
{
|
| 3491 |
-
"epoch": 498.0,
|
| 3492 |
-
"grad_norm": 0.014866248704493046,
|
| 3493 |
-
"learning_rate": 1.606425702811245e-06,
|
| 3494 |
-
"loss": 0.0045,
|
| 3495 |
-
"step": 498
|
| 3496 |
-
},
|
| 3497 |
-
{
|
| 3498 |
-
"epoch": 499.0,
|
| 3499 |
-
"grad_norm": 0.007127212826162577,
|
| 3500 |
-
"learning_rate": 1.2048192771084338e-06,
|
| 3501 |
-
"loss": 0.0047,
|
| 3502 |
-
"step": 499
|
| 3503 |
-
},
|
| 3504 |
-
{
|
| 3505 |
-
"epoch": 500.0,
|
| 3506 |
-
"grad_norm": 0.007142535876482725,
|
| 3507 |
-
"learning_rate": 8.032128514056225e-07,
|
| 3508 |
-
"loss": 0.0047,
|
| 3509 |
-
"step": 500
|
| 3510 |
}
|
| 3511 |
],
|
| 3512 |
"logging_steps": 1,
|
| 3513 |
-
"max_steps":
|
| 3514 |
"num_input_tokens_seen": 0,
|
| 3515 |
-
"num_train_epochs":
|
| 3516 |
"save_steps": 500,
|
| 3517 |
-
"total_flos":
|
| 3518 |
"train_batch_size": 1,
|
| 3519 |
"trial_name": null,
|
| 3520 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 100.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 100,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 1.0,
|
| 13 |
+
"grad_norm": 3.3439955711364746,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 5.4154,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 2.0,
|
| 20 |
+
"grad_norm": 3.3439955711364746,
|
| 21 |
+
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 5.4154,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 3.0,
|
| 27 |
+
"grad_norm": 4.64585542678833,
|
| 28 |
+
"learning_rate": 0.00019795918367346938,
|
| 29 |
+
"loss": 5.1115,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 4.0,
|
| 34 |
+
"grad_norm": 7.641846656799316,
|
| 35 |
+
"learning_rate": 0.0001959183673469388,
|
| 36 |
+
"loss": 4.5815,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 5.0,
|
| 41 |
+
"grad_norm": 10.534232139587402,
|
| 42 |
+
"learning_rate": 0.00019387755102040816,
|
| 43 |
+
"loss": 4.1388,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 6.0,
|
| 48 |
+
"grad_norm": Infinity,
|
| 49 |
+
"learning_rate": 0.00019387755102040816,
|
| 50 |
+
"loss": 3.7196,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 7.0,
|
| 55 |
+
"grad_norm": 13.133545875549316,
|
| 56 |
+
"learning_rate": 0.00019183673469387756,
|
| 57 |
+
"loss": 3.7196,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 8.0,
|
| 62 |
+
"grad_norm": 15.314064025878906,
|
| 63 |
+
"learning_rate": 0.00018979591836734697,
|
| 64 |
+
"loss": 3.2758,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 9.0,
|
| 69 |
+
"grad_norm": 16.992698669433594,
|
| 70 |
+
"learning_rate": 0.00018775510204081634,
|
| 71 |
+
"loss": 2.8057,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 10.0,
|
| 76 |
+
"grad_norm": 18.105308532714844,
|
| 77 |
+
"learning_rate": 0.00018571428571428572,
|
| 78 |
+
"loss": 2.3037,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 11.0,
|
| 83 |
+
"grad_norm": 18.624324798583984,
|
| 84 |
+
"learning_rate": 0.00018367346938775512,
|
| 85 |
+
"loss": 1.7698,
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 12.0,
|
| 90 |
+
"grad_norm": 18.554916381835938,
|
| 91 |
+
"learning_rate": 0.0001816326530612245,
|
| 92 |
+
"loss": 1.2219,
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 13.0,
|
| 97 |
+
"grad_norm": 17.93878746032715,
|
| 98 |
+
"learning_rate": 0.0001795918367346939,
|
| 99 |
+
"loss": 0.6767,
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 14.0,
|
| 104 |
+
"grad_norm": 2.917646884918213,
|
| 105 |
+
"learning_rate": 0.00017755102040816327,
|
| 106 |
+
"loss": 0.1873,
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 15.0,
|
| 111 |
+
"grad_norm": 3.2019948959350586,
|
| 112 |
+
"learning_rate": 0.00017551020408163265,
|
| 113 |
+
"loss": 0.1488,
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 16.0,
|
| 118 |
+
"grad_norm": 3.1802170276641846,
|
| 119 |
+
"learning_rate": 0.00017346938775510205,
|
| 120 |
+
"loss": 0.1134,
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 17.0,
|
| 125 |
+
"grad_norm": 1.5858300924301147,
|
| 126 |
+
"learning_rate": 0.00017142857142857143,
|
| 127 |
+
"loss": 0.0832,
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 18.0,
|
| 132 |
+
"grad_norm": 0.5080211758613586,
|
| 133 |
+
"learning_rate": 0.00016938775510204083,
|
| 134 |
+
"loss": 0.0701,
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 19.0,
|
| 139 |
+
"grad_norm": 0.5199990272521973,
|
| 140 |
+
"learning_rate": 0.00016734693877551023,
|
| 141 |
+
"loss": 0.0623,
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 20.0,
|
| 146 |
+
"grad_norm": 0.5190378427505493,
|
| 147 |
+
"learning_rate": 0.0001653061224489796,
|
| 148 |
+
"loss": 0.0552,
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 21.0,
|
| 153 |
+
"grad_norm": 0.5189256072044373,
|
| 154 |
+
"learning_rate": 0.00016326530612244898,
|
| 155 |
+
"loss": 0.0456,
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 22.0,
|
| 160 |
+
"grad_norm": 0.5494657754898071,
|
| 161 |
+
"learning_rate": 0.00016122448979591838,
|
| 162 |
+
"loss": 0.0356,
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 23.0,
|
| 167 |
+
"grad_norm": 0.5577398538589478,
|
| 168 |
+
"learning_rate": 0.00015918367346938776,
|
| 169 |
+
"loss": 0.024,
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 24.0,
|
| 174 |
+
"grad_norm": 0.39652666449546814,
|
| 175 |
+
"learning_rate": 0.00015714285714285716,
|
| 176 |
+
"loss": 0.0128,
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 25.0,
|
| 181 |
+
"grad_norm": 0.12145466357469559,
|
| 182 |
+
"learning_rate": 0.00015510204081632654,
|
| 183 |
+
"loss": 0.0064,
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 26.0,
|
| 188 |
+
"grad_norm": 0.16227224469184875,
|
| 189 |
+
"learning_rate": 0.0001530612244897959,
|
| 190 |
+
"loss": 0.0058,
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 27.0,
|
| 195 |
+
"grad_norm": 0.010005966760218143,
|
| 196 |
+
"learning_rate": 0.0001510204081632653,
|
| 197 |
+
"loss": 0.0046,
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 28.0,
|
| 202 |
+
"grad_norm": 0.11974961310625076,
|
| 203 |
+
"learning_rate": 0.00014897959183673472,
|
| 204 |
+
"loss": 0.0053,
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 29.0,
|
| 209 |
+
"grad_norm": 0.026845891028642654,
|
| 210 |
+
"learning_rate": 0.0001469387755102041,
|
| 211 |
+
"loss": 0.0046,
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 30.0,
|
| 216 |
+
"grad_norm": 0.09766196459531784,
|
| 217 |
+
"learning_rate": 0.0001448979591836735,
|
| 218 |
+
"loss": 0.0053,
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 31.0,
|
| 223 |
+
"grad_norm": 0.05061652511358261,
|
| 224 |
+
"learning_rate": 0.00014285714285714287,
|
| 225 |
+
"loss": 0.0046,
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 32.0,
|
| 230 |
+
"grad_norm": 0.0490519218146801,
|
| 231 |
+
"learning_rate": 0.00014081632653061224,
|
| 232 |
"loss": 0.0048,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 33.0,
|
| 237 |
+
"grad_norm": 0.06446336209774017,
|
| 238 |
+
"learning_rate": 0.00013877551020408165,
|
| 239 |
+
"loss": 0.0049,
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 34.0,
|
| 244 |
+
"grad_norm": 0.007198956795036793,
|
| 245 |
+
"learning_rate": 0.00013673469387755102,
|
| 246 |
+
"loss": 0.0045,
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 35.0,
|
| 251 |
+
"grad_norm": 0.06024309992790222,
|
| 252 |
+
"learning_rate": 0.0001346938775510204,
|
| 253 |
"loss": 0.0048,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 36.0,
|
| 258 |
+
"grad_norm": 0.04099668189883232,
|
| 259 |
+
"learning_rate": 0.0001326530612244898,
|
| 260 |
+
"loss": 0.0048,
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 37.0,
|
| 265 |
+
"grad_norm": 0.030607614666223526,
|
| 266 |
+
"learning_rate": 0.00013061224489795917,
|
| 267 |
+
"loss": 0.0047,
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 38.0,
|
| 272 |
+
"grad_norm": 0.04647354409098625,
|
| 273 |
+
"learning_rate": 0.00012857142857142858,
|
| 274 |
+
"loss": 0.0048,
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 39.0,
|
| 279 |
+
"grad_norm": 0.013468354940414429,
|
| 280 |
+
"learning_rate": 0.00012653061224489798,
|
| 281 |
+
"loss": 0.0047,
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 40.0,
|
| 286 |
+
"grad_norm": 0.02858574502170086,
|
| 287 |
+
"learning_rate": 0.00012448979591836735,
|
| 288 |
+
"loss": 0.0045,
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 41.0,
|
| 293 |
+
"grad_norm": 0.039958544075489044,
|
| 294 |
+
"learning_rate": 0.00012244897959183676,
|
| 295 |
"loss": 0.0048,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 42.0,
|
| 300 |
+
"grad_norm": 0.003950281068682671,
|
| 301 |
+
"learning_rate": 0.00012040816326530613,
|
| 302 |
+
"loss": 0.0047,
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 43.0,
|
| 307 |
+
"grad_norm": 0.029744163155555725,
|
| 308 |
+
"learning_rate": 0.00011836734693877552,
|
| 309 |
+
"loss": 0.0047,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 44.0,
|
| 314 |
+
"grad_norm": 0.029589517042040825,
|
| 315 |
+
"learning_rate": 0.0001163265306122449,
|
| 316 |
+
"loss": 0.0047,
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 45.0,
|
| 321 |
+
"grad_norm": 0.0025897289160639048,
|
| 322 |
+
"learning_rate": 0.00011428571428571428,
|
| 323 |
+
"loss": 0.0049,
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 46.0,
|
| 328 |
+
"grad_norm": 0.03239087387919426,
|
| 329 |
+
"learning_rate": 0.00011224489795918367,
|
| 330 |
"loss": 0.005,
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 47.0,
|
| 335 |
+
"grad_norm": 0.021254172548651695,
|
| 336 |
+
"learning_rate": 0.00011020408163265306,
|
| 337 |
+
"loss": 0.0047,
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 48.0,
|
| 342 |
+
"grad_norm": 0.01296665333211422,
|
| 343 |
+
"learning_rate": 0.00010816326530612246,
|
| 344 |
+
"loss": 0.0047,
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 49.0,
|
| 349 |
+
"grad_norm": 0.012932351790368557,
|
| 350 |
+
"learning_rate": 0.00010612244897959185,
|
| 351 |
+
"loss": 0.0047,
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 50.0,
|
| 356 |
+
"grad_norm": 0.0024836526717990637,
|
| 357 |
+
"learning_rate": 0.00010408163265306123,
|
| 358 |
+
"loss": 0.0049,
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 51.0,
|
| 363 |
+
"grad_norm": 0.012770796194672585,
|
| 364 |
+
"learning_rate": 0.00010204081632653062,
|
| 365 |
+
"loss": 0.0047,
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 52.0,
|
| 370 |
+
"grad_norm": 0.010142244398593903,
|
| 371 |
+
"learning_rate": 0.0001,
|
| 372 |
+
"loss": 0.0045,
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 53.0,
|
| 377 |
+
"grad_norm": 0.020766526460647583,
|
| 378 |
+
"learning_rate": 9.79591836734694e-05,
|
| 379 |
+
"loss": 0.0047,
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 54.0,
|
| 384 |
+
"grad_norm": 0.0039475164376199245,
|
| 385 |
+
"learning_rate": 9.591836734693878e-05,
|
| 386 |
+
"loss": 0.0047,
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 55.0,
|
| 391 |
+
"grad_norm": 0.012760567478835583,
|
| 392 |
+
"learning_rate": 9.387755102040817e-05,
|
| 393 |
+
"loss": 0.0047,
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 56.0,
|
| 398 |
+
"grad_norm": 0.028441807255148888,
|
| 399 |
+
"learning_rate": 9.183673469387756e-05,
|
| 400 |
+
"loss": 0.0047,
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 57.0,
|
| 405 |
+
"grad_norm": 0.004045899026095867,
|
| 406 |
+
"learning_rate": 8.979591836734695e-05,
|
| 407 |
"loss": 0.0047,
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 58.0,
|
| 412 |
+
"grad_norm": 0.020622774958610535,
|
| 413 |
+
"learning_rate": 8.775510204081632e-05,
|
| 414 |
+
"loss": 0.0047,
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 59.0,
|
| 419 |
+
"grad_norm": 0.010014132596552372,
|
| 420 |
+
"learning_rate": 8.571428571428571e-05,
|
| 421 |
+
"loss": 0.0045,
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 60.0,
|
| 426 |
+
"grad_norm": 0.0038325833156704903,
|
| 427 |
+
"learning_rate": 8.367346938775511e-05,
|
| 428 |
+
"loss": 0.0047,
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 61.0,
|
| 433 |
+
"grad_norm": 0.01254805363714695,
|
| 434 |
+
"learning_rate": 8.163265306122449e-05,
|
| 435 |
"loss": 0.0047,
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 62.0,
|
| 440 |
+
"grad_norm": 0.012542281299829483,
|
| 441 |
+
"learning_rate": 7.959183673469388e-05,
|
| 442 |
"loss": 0.0047,
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 63.0,
|
| 447 |
+
"grad_norm": 0.006473184563219547,
|
| 448 |
+
"learning_rate": 7.755102040816327e-05,
|
| 449 |
+
"loss": 0.0045,
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 64.0,
|
| 454 |
+
"grad_norm": 0.014362330548465252,
|
| 455 |
+
"learning_rate": 7.551020408163266e-05,
|
| 456 |
+
"loss": 0.0049,
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 65.0,
|
| 461 |
+
"grad_norm": 0.0204475000500679,
|
| 462 |
+
"learning_rate": 7.346938775510205e-05,
|
| 463 |
+
"loss": 0.0047,
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 66.0,
|
| 468 |
+
"grad_norm": 0.0023198979906737804,
|
| 469 |
+
"learning_rate": 7.142857142857143e-05,
|
| 470 |
+
"loss": 0.0049,
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 67.0,
|
| 475 |
+
"grad_norm": 0.018235381692647934,
|
| 476 |
+
"learning_rate": 6.938775510204082e-05,
|
| 477 |
+
"loss": 0.0049,
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 68.0,
|
| 482 |
+
"grad_norm": 0.012397863902151585,
|
| 483 |
+
"learning_rate": 6.73469387755102e-05,
|
| 484 |
+
"loss": 0.0047,
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 69.0,
|
| 489 |
+
"grad_norm": 0.003842939855530858,
|
| 490 |
+
"learning_rate": 6.530612244897959e-05,
|
| 491 |
"loss": 0.0047,
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 70.0,
|
| 496 |
+
"grad_norm": 0.003821918275207281,
|
| 497 |
+
"learning_rate": 6.326530612244899e-05,
|
| 498 |
+
"loss": 0.0047,
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 71.0,
|
| 503 |
+
"grad_norm": 0.0038236272521317005,
|
| 504 |
+
"learning_rate": 6.122448979591838e-05,
|
| 505 |
"loss": 0.0047,
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 72.0,
|
| 510 |
+
"grad_norm": 0.009961229749023914,
|
| 511 |
+
"learning_rate": 5.918367346938776e-05,
|
| 512 |
+
"loss": 0.0045,
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 73.0,
|
| 517 |
+
"grad_norm": 0.0037804129533469677,
|
| 518 |
+
"learning_rate": 5.714285714285714e-05,
|
| 519 |
+
"loss": 0.0047,
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 74.0,
|
| 524 |
+
"grad_norm": 0.003801505547016859,
|
| 525 |
+
"learning_rate": 5.510204081632653e-05,
|
| 526 |
"loss": 0.0047,
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 75.0,
|
| 531 |
+
"grad_norm": 0.012337015941739082,
|
| 532 |
+
"learning_rate": 5.3061224489795926e-05,
|
| 533 |
+
"loss": 0.0047,
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 76.0,
|
| 538 |
+
"grad_norm": 0.012310854159295559,
|
| 539 |
+
"learning_rate": 5.102040816326531e-05,
|
| 540 |
"loss": 0.0047,
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 77.0,
|
| 545 |
+
"grad_norm": 0.0063883536495268345,
|
| 546 |
+
"learning_rate": 4.89795918367347e-05,
|
| 547 |
+
"loss": 0.0045,
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 78.0,
|
| 552 |
+
"grad_norm": 0.0036704791709780693,
|
| 553 |
+
"learning_rate": 4.6938775510204086e-05,
|
| 554 |
+
"loss": 0.0047,
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 79.0,
|
| 559 |
+
"grad_norm": 0.0038380566984415054,
|
| 560 |
+
"learning_rate": 4.4897959183673474e-05,
|
| 561 |
"loss": 0.0047,
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 80.0,
|
| 566 |
+
"grad_norm": 0.013979257084429264,
|
| 567 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 568 |
+
"loss": 0.0049,
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 81.0,
|
| 573 |
+
"grad_norm": 0.003754157805815339,
|
| 574 |
+
"learning_rate": 4.0816326530612245e-05,
|
| 575 |
"loss": 0.0047,
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 82.0,
|
| 580 |
+
"grad_norm": 0.003731819801032543,
|
| 581 |
+
"learning_rate": 3.8775510204081634e-05,
|
| 582 |
+
"loss": 0.0047,
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 83.0,
|
| 587 |
+
"grad_norm": 0.003740091575309634,
|
| 588 |
+
"learning_rate": 3.673469387755102e-05,
|
| 589 |
"loss": 0.0047,
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 84.0,
|
| 594 |
+
"grad_norm": 0.002331367926672101,
|
| 595 |
+
"learning_rate": 3.469387755102041e-05,
|
| 596 |
+
"loss": 0.0049,
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 85.0,
|
| 601 |
+
"grad_norm": 0.006460592150688171,
|
| 602 |
+
"learning_rate": 3.265306122448979e-05,
|
| 603 |
+
"loss": 0.0045,
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 86.0,
|
| 608 |
+
"grad_norm": 0.01226158905774355,
|
| 609 |
+
"learning_rate": 3.061224489795919e-05,
|
| 610 |
"loss": 0.0047,
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 87.0,
|
| 615 |
+
"grad_norm": 0.012233145534992218,
|
| 616 |
+
"learning_rate": 2.857142857142857e-05,
|
| 617 |
+
"loss": 0.0047,
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 88.0,
|
| 622 |
+
"grad_norm": 0.002268604002892971,
|
| 623 |
+
"learning_rate": 2.6530612244897963e-05,
|
| 624 |
+
"loss": 0.0049,
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 89.0,
|
| 629 |
+
"grad_norm": 0.009719762951135635,
|
| 630 |
+
"learning_rate": 2.448979591836735e-05,
|
| 631 |
"loss": 0.0045,
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 90.0,
|
| 636 |
+
"grad_norm": 0.003743007080629468,
|
| 637 |
+
"learning_rate": 2.2448979591836737e-05,
|
| 638 |
+
"loss": 0.0047,
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 91.0,
|
| 643 |
+
"grad_norm": 0.00371920526959002,
|
| 644 |
+
"learning_rate": 2.0408163265306123e-05,
|
| 645 |
"loss": 0.0047,
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 92.0,
|
| 650 |
+
"grad_norm": 0.0036711846478283405,
|
| 651 |
+
"learning_rate": 1.836734693877551e-05,
|
| 652 |
"loss": 0.0047,
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 93.0,
|
| 657 |
+
"grad_norm": 0.020141731947660446,
|
| 658 |
+
"learning_rate": 1.6326530612244897e-05,
|
| 659 |
"loss": 0.0047,
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 94.0,
|
| 664 |
+
"grad_norm": 0.0037132452707737684,
|
| 665 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 666 |
+
"loss": 0.0047,
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 95.0,
|
| 671 |
+
"grad_norm": 0.0063994950614869595,
|
| 672 |
+
"learning_rate": 1.2244897959183674e-05,
|
| 673 |
"loss": 0.0045,
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 96.0,
|
| 678 |
+
"grad_norm": 0.0037392491940408945,
|
| 679 |
+
"learning_rate": 1.0204081632653061e-05,
|
| 680 |
+
"loss": 0.0047,
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 97.0,
|
| 685 |
+
"grad_norm": 0.003673990024253726,
|
| 686 |
+
"learning_rate": 8.163265306122448e-06,
|
| 687 |
+
"loss": 0.0047,
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 98.0,
|
| 692 |
+
"grad_norm": 0.012367010116577148,
|
| 693 |
+
"learning_rate": 6.122448979591837e-06,
|
| 694 |
+
"loss": 0.0047,
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 99.0,
|
| 699 |
+
"grad_norm": 0.002525837393477559,
|
| 700 |
+
"learning_rate": 4.081632653061224e-06,
|
| 701 |
+
"loss": 0.0049,
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 100.0,
|
| 706 |
+
"grad_norm": 0.012382498942315578,
|
| 707 |
+
"learning_rate": 2.040816326530612e-06,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 708 |
"loss": 0.0047,
|
| 709 |
+
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 710 |
}
|
| 711 |
],
|
| 712 |
"logging_steps": 1,
|
| 713 |
+
"max_steps": 100,
|
| 714 |
"num_input_tokens_seen": 0,
|
| 715 |
+
"num_train_epochs": 100,
|
| 716 |
"save_steps": 500,
|
| 717 |
+
"total_flos": 713859830784000.0,
|
| 718 |
"train_batch_size": 1,
|
| 719 |
"trial_name": null,
|
| 720 |
"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 4984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:27d00b17c8f483287b4263e1232460d1b6343148500f5c99d62e6dc5981a857d
|
| 3 |
size 4984
|