Upload folder using huggingface_hub
Browse files- checkpoint-2000/model.safetensors +1 -1
- checkpoint-2000/optimizer.pt +1 -1
- checkpoint-2000/scaler.pt +1 -1
- checkpoint-2000/scheduler.pt +1 -1
- checkpoint-2000/trainer_state.json +556 -556
- checkpoint-2000/training_args.bin +1 -1
- checkpoint-4000/model.safetensors +1 -1
- checkpoint-4000/optimizer.pt +1 -1
- checkpoint-4000/scaler.pt +1 -1
- checkpoint-4000/scheduler.pt +1 -1
- checkpoint-4000/trainer_state.json +0 -0
- checkpoint-4000/training_args.bin +1 -1
- checkpoint-6000/model.safetensors +1 -1
- checkpoint-6000/optimizer.pt +1 -1
- checkpoint-6000/scaler.pt +1 -1
- checkpoint-6000/scheduler.pt +1 -1
- checkpoint-6000/trainer_state.json +0 -0
- checkpoint-6000/training_args.bin +1 -1
checkpoint-2000/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 966995080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d40c39cd85f23600479692bbe8769ab089b513b869b6ac238f39647119b1963
|
| 3 |
size 966995080
|
checkpoint-2000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1925071165
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c875f33e4a4f446a410bbcf73fb532ec9c21822feca94268c6630ca66107bd98
|
| 3 |
size 1925071165
|
checkpoint-2000/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4aa03f6e0cd07cf67ce1fbe3101d545f5771ef9148b9debf02b11cf6948da5c
|
| 3 |
size 1383
|
checkpoint-2000/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:860788b065bc0123a723da86ef983cbd14bd7a9e97885b45e526dc094743e431
|
| 3 |
size 1465
|
checkpoint-2000/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 2000,
|
| 3 |
-
"best_metric": 1.
|
| 4 |
"best_model_checkpoint": "./SALAMA_NEW8/checkpoint-2000",
|
| 5 |
"epoch": 0.798881565807869,
|
| 6 |
"eval_steps": 2000,
|
|
@@ -11,1416 +11,1416 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.003994407829039345,
|
| 14 |
-
"grad_norm":
|
| 15 |
"learning_rate": 1.8e-07,
|
| 16 |
-
"loss": 0.
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.00798881565807869,
|
| 21 |
-
"grad_norm":
|
| 22 |
"learning_rate": 3.8e-07,
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.011983223487118035,
|
| 28 |
-
"grad_norm": 2.
|
| 29 |
"learning_rate": 5.800000000000001e-07,
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.01597763131615738,
|
| 35 |
-
"grad_norm": 2.
|
| 36 |
"learning_rate": 7.8e-07,
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.019972039145196723,
|
| 42 |
-
"grad_norm":
|
| 43 |
"learning_rate": 9.800000000000001e-07,
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.02396644697423607,
|
| 49 |
-
"grad_norm":
|
| 50 |
"learning_rate": 1.1800000000000001e-06,
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.027960854803275415,
|
| 56 |
-
"grad_norm":
|
| 57 |
"learning_rate": 1.3800000000000001e-06,
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.03195526263231476,
|
| 63 |
-
"grad_norm":
|
| 64 |
"learning_rate": 1.5800000000000001e-06,
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.035949670461354104,
|
| 70 |
-
"grad_norm": 2.
|
| 71 |
"learning_rate": 1.7800000000000001e-06,
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.039944078290393446,
|
| 77 |
-
"grad_norm":
|
| 78 |
"learning_rate": 1.98e-06,
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.043938486119432796,
|
| 84 |
-
"grad_norm":
|
| 85 |
"learning_rate": 2.1800000000000003e-06,
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.04793289394847214,
|
| 91 |
-
"grad_norm":
|
| 92 |
"learning_rate": 2.38e-06,
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 120
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.05192730177751148,
|
| 98 |
-
"grad_norm": 2.
|
| 99 |
"learning_rate": 2.5800000000000003e-06,
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 130
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.05592170960655083,
|
| 105 |
-
"grad_norm":
|
| 106 |
"learning_rate": 2.7800000000000005e-06,
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 140
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.05991611743559017,
|
| 112 |
-
"grad_norm":
|
| 113 |
"learning_rate": 2.9800000000000003e-06,
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 150
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.06391052526462952,
|
| 119 |
-
"grad_norm":
|
| 120 |
"learning_rate": 3.1800000000000005e-06,
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 160
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.06790493309366886,
|
| 126 |
-
"grad_norm": 2.
|
| 127 |
"learning_rate": 3.3800000000000007e-06,
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 170
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.07189934092270821,
|
| 133 |
-
"grad_norm": 2.
|
| 134 |
"learning_rate": 3.58e-06,
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 180
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.07589374875174755,
|
| 140 |
-
"grad_norm": 2.
|
| 141 |
"learning_rate": 3.7800000000000002e-06,
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 190
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.07988815658078689,
|
| 147 |
-
"grad_norm":
|
| 148 |
"learning_rate": 3.980000000000001e-06,
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 200
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.08388256440982625,
|
| 154 |
-
"grad_norm": 1.
|
| 155 |
"learning_rate": 4.18e-06,
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 210
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.08787697223886559,
|
| 161 |
-
"grad_norm": 3.
|
| 162 |
"learning_rate": 4.38e-06,
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 220
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.09187138006790493,
|
| 168 |
-
"grad_norm":
|
| 169 |
"learning_rate": 4.58e-06,
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 230
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.09586578789694428,
|
| 175 |
-
"grad_norm":
|
| 176 |
"learning_rate": 4.78e-06,
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 240
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.09986019572598362,
|
| 182 |
-
"grad_norm":
|
| 183 |
"learning_rate": 4.980000000000001e-06,
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.10385460355502296,
|
| 189 |
-
"grad_norm":
|
| 190 |
"learning_rate": 5.18e-06,
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 260
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.10784901138406232,
|
| 196 |
-
"grad_norm":
|
| 197 |
"learning_rate": 5.380000000000001e-06,
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 270
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.11184341921310166,
|
| 203 |
-
"grad_norm":
|
| 204 |
"learning_rate": 5.580000000000001e-06,
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 280
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.115837827042141,
|
| 210 |
-
"grad_norm":
|
| 211 |
"learning_rate": 5.78e-06,
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 290
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.11983223487118035,
|
| 217 |
-
"grad_norm": 2.
|
| 218 |
"learning_rate": 5.98e-06,
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 300
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.12382664270021969,
|
| 224 |
-
"grad_norm":
|
| 225 |
"learning_rate": 6.18e-06,
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 310
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.12782105052925904,
|
| 231 |
-
"grad_norm":
|
| 232 |
"learning_rate": 6.380000000000001e-06,
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 320
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.1318154583582984,
|
| 238 |
-
"grad_norm":
|
| 239 |
"learning_rate": 6.5800000000000005e-06,
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 330
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.13580986618733773,
|
| 245 |
-
"grad_norm":
|
| 246 |
"learning_rate": 6.780000000000001e-06,
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 340
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.13980427401637707,
|
| 252 |
-
"grad_norm": 2.
|
| 253 |
"learning_rate": 6.98e-06,
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 350
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.14379868184541642,
|
| 259 |
-
"grad_norm": 2.
|
| 260 |
"learning_rate": 7.180000000000001e-06,
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 360
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.14779308967445576,
|
| 266 |
-
"grad_norm": 1.
|
| 267 |
"learning_rate": 7.3800000000000005e-06,
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 370
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.1517874975034951,
|
| 273 |
-
"grad_norm":
|
| 274 |
"learning_rate": 7.58e-06,
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 380
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.15578190533253444,
|
| 280 |
-
"grad_norm":
|
| 281 |
"learning_rate": 7.78e-06,
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 390
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.15977631316157379,
|
| 287 |
-
"grad_norm": 2.
|
| 288 |
"learning_rate": 7.980000000000002e-06,
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.16377072099061313,
|
| 294 |
-
"grad_norm":
|
| 295 |
"learning_rate": 8.18e-06,
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 410
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.1677651288196525,
|
| 301 |
-
"grad_norm":
|
| 302 |
"learning_rate": 8.380000000000001e-06,
|
| 303 |
-
"loss": 0.
|
| 304 |
"step": 420
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.17175953664869184,
|
| 308 |
-
"grad_norm":
|
| 309 |
"learning_rate": 8.580000000000001e-06,
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 430
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.17575394447773118,
|
| 315 |
-
"grad_norm":
|
| 316 |
"learning_rate": 8.78e-06,
|
| 317 |
-
"loss": 0.
|
| 318 |
"step": 440
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.17974835230677053,
|
| 322 |
-
"grad_norm":
|
| 323 |
"learning_rate": 8.98e-06,
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 450
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.18374276013580987,
|
| 329 |
-
"grad_norm":
|
| 330 |
"learning_rate": 9.180000000000002e-06,
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 460
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.1877371679648492,
|
| 336 |
-
"grad_norm":
|
| 337 |
"learning_rate": 9.38e-06,
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 470
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.19173157579388855,
|
| 343 |
-
"grad_norm":
|
| 344 |
"learning_rate": 9.58e-06,
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 480
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.1957259836229279,
|
| 350 |
-
"grad_norm": 3.
|
| 351 |
"learning_rate": 9.780000000000001e-06,
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 490
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.19972039145196724,
|
| 357 |
-
"grad_norm": 1.
|
| 358 |
"learning_rate": 9.980000000000001e-06,
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.20371479928100658,
|
| 364 |
-
"grad_norm":
|
| 365 |
-
"learning_rate": 9.
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 510
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.20770920711004592,
|
| 371 |
-
"grad_norm":
|
| 372 |
-
"learning_rate": 9.
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 520
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.2117036149390853,
|
| 378 |
-
"grad_norm":
|
| 379 |
-
"learning_rate": 9.
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 530
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.21569802276812464,
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate": 9.
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 540
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.21969243059716398,
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate": 9.
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 550
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.22368683842620332,
|
| 399 |
-
"grad_norm":
|
| 400 |
-
"learning_rate": 9.
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 560
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.22768124625524266,
|
| 406 |
-
"grad_norm": 2.
|
| 407 |
-
"learning_rate": 9.
|
| 408 |
-
"loss": 0.
|
| 409 |
"step": 570
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.231675654084282,
|
| 413 |
-
"grad_norm": 2.
|
| 414 |
-
"learning_rate": 9.
|
| 415 |
-
"loss": 0.
|
| 416 |
"step": 580
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.23567006191332135,
|
| 420 |
-
"grad_norm":
|
| 421 |
-
"learning_rate": 9.
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 590
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.2396644697423607,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 9.
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 600
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.24365887757140003,
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate": 9.
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 610
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.24765328540043938,
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate": 9.
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 620
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.2516476932294787,
|
| 448 |
-
"grad_norm": 3.
|
| 449 |
-
"learning_rate": 9.
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 630
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.2556421010585181,
|
| 455 |
-
"grad_norm":
|
| 456 |
-
"learning_rate": 9.
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 640
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.2596365088875574,
|
| 462 |
-
"grad_norm":
|
| 463 |
-
"learning_rate": 9.
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 650
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.2636309167165968,
|
| 469 |
-
"grad_norm":
|
| 470 |
-
"learning_rate": 9.
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 660
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.2676253245456361,
|
| 476 |
-
"grad_norm":
|
| 477 |
-
"learning_rate": 9.
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 670
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.27161973237467546,
|
| 483 |
-
"grad_norm":
|
| 484 |
-
"learning_rate": 9.
|
| 485 |
-
"loss": 0.
|
| 486 |
"step": 680
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.2756141402037148,
|
| 490 |
-
"grad_norm":
|
| 491 |
-
"learning_rate": 9.
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 690
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.27960854803275415,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate": 9.
|
| 499 |
-
"loss": 0.
|
| 500 |
"step": 700
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.2836029558617935,
|
| 504 |
-
"grad_norm": 2.
|
| 505 |
-
"learning_rate": 9.
|
| 506 |
"loss": 0.0292,
|
| 507 |
"step": 710
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.28759736369083283,
|
| 511 |
-
"grad_norm":
|
| 512 |
-
"learning_rate": 9.
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 720
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.2915917715198722,
|
| 518 |
-
"grad_norm":
|
| 519 |
-
"learning_rate": 9.
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 730
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.2955861793489115,
|
| 525 |
-
"grad_norm":
|
| 526 |
-
"learning_rate": 9.
|
| 527 |
-
"loss": 0.
|
| 528 |
"step": 740
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.2995805871779509,
|
| 532 |
-
"grad_norm":
|
| 533 |
-
"learning_rate": 9.
|
| 534 |
-
"loss": 0.
|
| 535 |
"step": 750
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.3035749950069902,
|
| 539 |
-
"grad_norm":
|
| 540 |
-
"learning_rate": 9.
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 760
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.30756940283602957,
|
| 546 |
-
"grad_norm":
|
| 547 |
-
"learning_rate": 9.
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 770
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.3115638106650689,
|
| 553 |
-
"grad_norm": 2.
|
| 554 |
-
"learning_rate": 9.
|
| 555 |
-
"loss": 0.
|
| 556 |
"step": 780
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.31555821849410826,
|
| 560 |
-
"grad_norm":
|
| 561 |
-
"learning_rate": 9.
|
| 562 |
-
"loss": 0.
|
| 563 |
"step": 790
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.31955262632314757,
|
| 567 |
-
"grad_norm": 2.
|
| 568 |
-
"learning_rate": 9.
|
| 569 |
-
"loss": 0.
|
| 570 |
"step": 800
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.32354703415218694,
|
| 574 |
-
"grad_norm": 2.
|
| 575 |
-
"learning_rate": 9.
|
| 576 |
-
"loss": 0.
|
| 577 |
"step": 810
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.32754144198122626,
|
| 581 |
-
"grad_norm": 2.
|
| 582 |
-
"learning_rate": 9.
|
| 583 |
-
"loss": 0.
|
| 584 |
"step": 820
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.3315358498102656,
|
| 588 |
-
"grad_norm":
|
| 589 |
-
"learning_rate": 9.
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 830
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.335530257639305,
|
| 595 |
-
"grad_norm": 2.
|
| 596 |
-
"learning_rate": 9.
|
| 597 |
-
"loss": 0.
|
| 598 |
"step": 840
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.3395246654683443,
|
| 602 |
-
"grad_norm": 3.
|
| 603 |
-
"learning_rate": 9.
|
| 604 |
-
"loss": 0.
|
| 605 |
"step": 850
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.3435190732973837,
|
| 609 |
-
"grad_norm": 2.
|
| 610 |
-
"learning_rate": 9.
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 860
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.347513481126423,
|
| 616 |
-
"grad_norm":
|
| 617 |
-
"learning_rate": 9.
|
| 618 |
-
"loss": 0.
|
| 619 |
"step": 870
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.35150788895546237,
|
| 623 |
-
"grad_norm":
|
| 624 |
-
"learning_rate": 9.
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 880
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.3555022967845017,
|
| 630 |
-
"grad_norm":
|
| 631 |
-
"learning_rate": 9.
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 890
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.35949670461354105,
|
| 637 |
-
"grad_norm": 2.
|
| 638 |
-
"learning_rate": 9.
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 900
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.36349111244258037,
|
| 644 |
-
"grad_norm": 2.
|
| 645 |
-
"learning_rate": 9.
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 910
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.36748552027161974,
|
| 651 |
-
"grad_norm": 2.
|
| 652 |
-
"learning_rate": 9.
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 920
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.37147992810065905,
|
| 658 |
-
"grad_norm": 3.
|
| 659 |
-
"learning_rate": 9.
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 930
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.3754743359296984,
|
| 665 |
-
"grad_norm":
|
| 666 |
-
"learning_rate": 9.
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 940
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.3794687437587378,
|
| 672 |
-
"grad_norm": 2.
|
| 673 |
-
"learning_rate": 9.
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 950
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.3834631515877771,
|
| 679 |
-
"grad_norm":
|
| 680 |
-
"learning_rate": 9.
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 960
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.3874575594168165,
|
| 686 |
-
"grad_norm":
|
| 687 |
-
"learning_rate": 9.
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 970
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.3914519672458558,
|
| 693 |
-
"grad_norm":
|
| 694 |
-
"learning_rate": 9.
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 980
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.39544637507489516,
|
| 700 |
-
"grad_norm":
|
| 701 |
-
"learning_rate": 9.
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 990
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.3994407829039345,
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate": 9.
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 1000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.40343519073297385,
|
| 714 |
-
"grad_norm":
|
| 715 |
-
"learning_rate": 9.
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 1010
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.40742959856201316,
|
| 721 |
-
"grad_norm":
|
| 722 |
-
"learning_rate": 9.
|
| 723 |
-
"loss": 0.
|
| 724 |
"step": 1020
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.41142400639105253,
|
| 728 |
-
"grad_norm":
|
| 729 |
-
"learning_rate": 9.
|
| 730 |
-
"loss": 0.
|
| 731 |
"step": 1030
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.41541841422009185,
|
| 735 |
-
"grad_norm":
|
| 736 |
-
"learning_rate": 9.
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 1040
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.4194128220491312,
|
| 742 |
-
"grad_norm":
|
| 743 |
-
"learning_rate": 9.
|
| 744 |
-
"loss": 0.
|
| 745 |
"step": 1050
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.4234072298781706,
|
| 749 |
-
"grad_norm":
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss": 0.
|
| 752 |
"step": 1060
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.4274016377072099,
|
| 756 |
-
"grad_norm": 2.
|
| 757 |
-
"learning_rate":
|
| 758 |
-
"loss": 0.
|
| 759 |
"step": 1070
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.4313960455362493,
|
| 763 |
-
"grad_norm": 2.
|
| 764 |
-
"learning_rate":
|
| 765 |
-
"loss": 0.
|
| 766 |
"step": 1080
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.4353904533652886,
|
| 770 |
-
"grad_norm":
|
| 771 |
-
"learning_rate":
|
| 772 |
-
"loss": 0.
|
| 773 |
"step": 1090
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.43938486119432796,
|
| 777 |
-
"grad_norm":
|
| 778 |
-
"learning_rate":
|
| 779 |
-
"loss": 0.
|
| 780 |
"step": 1100
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.4433792690233673,
|
| 784 |
-
"grad_norm": 2.
|
| 785 |
-
"learning_rate":
|
| 786 |
-
"loss": 0.
|
| 787 |
"step": 1110
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.44737367685240664,
|
| 791 |
-
"grad_norm": 3.
|
| 792 |
-
"learning_rate":
|
| 793 |
-
"loss": 0.
|
| 794 |
"step": 1120
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.45136808468144596,
|
| 798 |
-
"grad_norm": 2.
|
| 799 |
-
"learning_rate":
|
| 800 |
-
"loss": 0.
|
| 801 |
"step": 1130
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.45536249251048533,
|
| 805 |
-
"grad_norm":
|
| 806 |
-
"learning_rate":
|
| 807 |
-
"loss": 0.
|
| 808 |
"step": 1140
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.45935690033952464,
|
| 812 |
-
"grad_norm":
|
| 813 |
-
"learning_rate":
|
| 814 |
-
"loss": 0.
|
| 815 |
"step": 1150
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.463351308168564,
|
| 819 |
-
"grad_norm":
|
| 820 |
-
"learning_rate":
|
| 821 |
-
"loss": 0.
|
| 822 |
"step": 1160
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 0.46734571599760333,
|
| 826 |
-
"grad_norm":
|
| 827 |
-
"learning_rate":
|
| 828 |
-
"loss": 0.
|
| 829 |
"step": 1170
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 0.4713401238266427,
|
| 833 |
-
"grad_norm":
|
| 834 |
-
"learning_rate":
|
| 835 |
-
"loss": 0.
|
| 836 |
"step": 1180
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.47533453165568207,
|
| 840 |
-
"grad_norm":
|
| 841 |
-
"learning_rate":
|
| 842 |
-
"loss": 0.
|
| 843 |
"step": 1190
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 0.4793289394847214,
|
| 847 |
-
"grad_norm":
|
| 848 |
-
"learning_rate":
|
| 849 |
-
"loss": 0.
|
| 850 |
"step": 1200
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 0.48332334731376075,
|
| 854 |
-
"grad_norm": 3.
|
| 855 |
-
"learning_rate": 8.
|
| 856 |
-
"loss": 0.
|
| 857 |
"step": 1210
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.48731775514280007,
|
| 861 |
-
"grad_norm":
|
| 862 |
-
"learning_rate": 8.
|
| 863 |
-
"loss": 0.
|
| 864 |
"step": 1220
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.49131216297183944,
|
| 868 |
-
"grad_norm": 2.
|
| 869 |
-
"learning_rate": 8.
|
| 870 |
-
"loss": 0.
|
| 871 |
"step": 1230
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 0.49530657080087875,
|
| 875 |
-
"grad_norm":
|
| 876 |
-
"learning_rate": 8.
|
| 877 |
-
"loss": 0.
|
| 878 |
"step": 1240
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.4993009786299181,
|
| 882 |
-
"grad_norm": 2.
|
| 883 |
-
"learning_rate": 8.
|
| 884 |
-
"loss": 0.
|
| 885 |
"step": 1250
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 0.5032953864589574,
|
| 889 |
-
"grad_norm":
|
| 890 |
-
"learning_rate": 8.
|
| 891 |
-
"loss": 0.
|
| 892 |
"step": 1260
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 0.5072897942879968,
|
| 896 |
-
"grad_norm":
|
| 897 |
-
"learning_rate": 8.
|
| 898 |
-
"loss": 0.
|
| 899 |
"step": 1270
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 0.5112842021170362,
|
| 903 |
-
"grad_norm":
|
| 904 |
-
"learning_rate": 8.
|
| 905 |
-
"loss": 0.
|
| 906 |
"step": 1280
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 0.5152786099460755,
|
| 910 |
-
"grad_norm":
|
| 911 |
-
"learning_rate": 8.
|
| 912 |
-
"loss": 0.
|
| 913 |
"step": 1290
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 0.5192730177751148,
|
| 917 |
-
"grad_norm": 2.
|
| 918 |
-
"learning_rate": 8.
|
| 919 |
-
"loss": 0.
|
| 920 |
"step": 1300
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 0.5232674256041542,
|
| 924 |
-
"grad_norm": 3.
|
| 925 |
-
"learning_rate": 8.
|
| 926 |
-
"loss": 0.
|
| 927 |
"step": 1310
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.5272618334331935,
|
| 931 |
-
"grad_norm": 2.
|
| 932 |
-
"learning_rate": 8.
|
| 933 |
-
"loss": 0.
|
| 934 |
"step": 1320
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 0.5312562412622329,
|
| 938 |
-
"grad_norm":
|
| 939 |
-
"learning_rate": 8.
|
| 940 |
-
"loss": 0.
|
| 941 |
"step": 1330
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.5352506490912722,
|
| 945 |
-
"grad_norm":
|
| 946 |
-
"learning_rate": 8.
|
| 947 |
-
"loss": 0.
|
| 948 |
"step": 1340
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 0.5392450569203115,
|
| 952 |
-
"grad_norm": 3.
|
| 953 |
-
"learning_rate": 8.
|
| 954 |
-
"loss": 0.
|
| 955 |
"step": 1350
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 0.5432394647493509,
|
| 959 |
-
"grad_norm":
|
| 960 |
-
"learning_rate": 8.
|
| 961 |
-
"loss": 0.
|
| 962 |
"step": 1360
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.5472338725783903,
|
| 966 |
-
"grad_norm": 3.
|
| 967 |
-
"learning_rate": 8.
|
| 968 |
-
"loss": 0.
|
| 969 |
"step": 1370
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.5512282804074295,
|
| 973 |
-
"grad_norm":
|
| 974 |
-
"learning_rate": 8.
|
| 975 |
-
"loss": 0.
|
| 976 |
"step": 1380
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 0.5552226882364689,
|
| 980 |
-
"grad_norm":
|
| 981 |
-
"learning_rate": 8.
|
| 982 |
-
"loss": 0.
|
| 983 |
"step": 1390
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 0.5592170960655083,
|
| 987 |
-
"grad_norm":
|
| 988 |
-
"learning_rate": 8.
|
| 989 |
-
"loss": 0.
|
| 990 |
"step": 1400
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.5632115038945477,
|
| 994 |
-
"grad_norm":
|
| 995 |
-
"learning_rate": 8.
|
| 996 |
-
"loss": 0.
|
| 997 |
"step": 1410
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 0.567205911723587,
|
| 1001 |
-
"grad_norm":
|
| 1002 |
-
"learning_rate": 8.
|
| 1003 |
-
"loss": 0.
|
| 1004 |
"step": 1420
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 0.5712003195526263,
|
| 1008 |
-
"grad_norm": 2.
|
| 1009 |
-
"learning_rate": 8.
|
| 1010 |
-
"loss": 0.
|
| 1011 |
"step": 1430
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 0.5751947273816657,
|
| 1015 |
-
"grad_norm":
|
| 1016 |
-
"learning_rate": 8.
|
| 1017 |
-
"loss": 0.
|
| 1018 |
"step": 1440
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 0.579189135210705,
|
| 1022 |
-
"grad_norm":
|
| 1023 |
-
"learning_rate": 8.
|
| 1024 |
-
"loss": 0.
|
| 1025 |
"step": 1450
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 0.5831835430397444,
|
| 1029 |
-
"grad_norm": 3.
|
| 1030 |
-
"learning_rate": 8.
|
| 1031 |
-
"loss": 0.
|
| 1032 |
"step": 1460
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 0.5871779508687837,
|
| 1036 |
-
"grad_norm":
|
| 1037 |
-
"learning_rate": 8.
|
| 1038 |
-
"loss": 0.
|
| 1039 |
"step": 1470
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
"epoch": 0.591172358697823,
|
| 1043 |
-
"grad_norm":
|
| 1044 |
-
"learning_rate": 8.
|
| 1045 |
-
"loss": 0.
|
| 1046 |
"step": 1480
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 0.5951667665268624,
|
| 1050 |
-
"grad_norm":
|
| 1051 |
-
"learning_rate": 8.
|
| 1052 |
-
"loss": 0.
|
| 1053 |
"step": 1490
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
"epoch": 0.5991611743559018,
|
| 1057 |
-
"grad_norm":
|
| 1058 |
-
"learning_rate": 8.
|
| 1059 |
-
"loss": 0.
|
| 1060 |
"step": 1500
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 0.603155582184941,
|
| 1064 |
-
"grad_norm":
|
| 1065 |
-
"learning_rate": 8.
|
| 1066 |
-
"loss": 0.
|
| 1067 |
"step": 1510
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 0.6071499900139804,
|
| 1071 |
-
"grad_norm":
|
| 1072 |
-
"learning_rate": 8.
|
| 1073 |
-
"loss": 0.
|
| 1074 |
"step": 1520
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 0.6111443978430198,
|
| 1078 |
-
"grad_norm":
|
| 1079 |
-
"learning_rate": 8.
|
| 1080 |
-
"loss": 0.
|
| 1081 |
"step": 1530
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.6151388056720591,
|
| 1085 |
-
"grad_norm":
|
| 1086 |
-
"learning_rate": 8.
|
| 1087 |
-
"loss": 0.
|
| 1088 |
"step": 1540
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 0.6191332135010985,
|
| 1092 |
-
"grad_norm": 2.
|
| 1093 |
-
"learning_rate": 8.
|
| 1094 |
-
"loss": 0.
|
| 1095 |
"step": 1550
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 0.6231276213301378,
|
| 1099 |
-
"grad_norm": 2.
|
| 1100 |
-
"learning_rate": 8.
|
| 1101 |
-
"loss": 0.
|
| 1102 |
"step": 1560
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 0.6271220291591771,
|
| 1106 |
-
"grad_norm":
|
| 1107 |
-
"learning_rate": 8.
|
| 1108 |
-
"loss": 0.
|
| 1109 |
"step": 1570
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 0.6311164369882165,
|
| 1113 |
-
"grad_norm": 2.
|
| 1114 |
-
"learning_rate": 8.
|
| 1115 |
-
"loss": 0.
|
| 1116 |
"step": 1580
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 0.6351108448172559,
|
| 1120 |
-
"grad_norm":
|
| 1121 |
-
"learning_rate": 8.
|
| 1122 |
-
"loss": 0.
|
| 1123 |
"step": 1590
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 0.6391052526462951,
|
| 1127 |
-
"grad_norm":
|
| 1128 |
-
"learning_rate": 8.
|
| 1129 |
-
"loss": 0.
|
| 1130 |
"step": 1600
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 0.6430996604753345,
|
| 1134 |
-
"grad_norm":
|
| 1135 |
-
"learning_rate":
|
| 1136 |
-
"loss": 0.
|
| 1137 |
"step": 1610
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 0.6470940683043739,
|
| 1141 |
-
"grad_norm":
|
| 1142 |
-
"learning_rate":
|
| 1143 |
-
"loss": 0.
|
| 1144 |
"step": 1620
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 0.6510884761334133,
|
| 1148 |
-
"grad_norm": 2.
|
| 1149 |
-
"learning_rate":
|
| 1150 |
-
"loss": 0.
|
| 1151 |
"step": 1630
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 0.6550828839624525,
|
| 1155 |
-
"grad_norm": 3.
|
| 1156 |
-
"learning_rate":
|
| 1157 |
-
"loss": 0.
|
| 1158 |
"step": 1640
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 0.6590772917914919,
|
| 1162 |
-
"grad_norm":
|
| 1163 |
-
"learning_rate":
|
| 1164 |
-
"loss": 0.
|
| 1165 |
"step": 1650
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 0.6630716996205313,
|
| 1169 |
-
"grad_norm": 3.
|
| 1170 |
-
"learning_rate":
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 1660
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 0.6670661074495706,
|
| 1176 |
-
"grad_norm":
|
| 1177 |
-
"learning_rate":
|
| 1178 |
-
"loss": 0.
|
| 1179 |
"step": 1670
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.67106051527861,
|
| 1183 |
-
"grad_norm":
|
| 1184 |
-
"learning_rate":
|
| 1185 |
-
"loss": 0.
|
| 1186 |
"step": 1680
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 0.6750549231076493,
|
| 1190 |
-
"grad_norm":
|
| 1191 |
-
"learning_rate":
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 1690
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.6790493309366886,
|
| 1197 |
-
"grad_norm":
|
| 1198 |
-
"learning_rate":
|
| 1199 |
-
"loss": 0.
|
| 1200 |
"step": 1700
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.683043738765728,
|
| 1204 |
-
"grad_norm": 3.
|
| 1205 |
-
"learning_rate":
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 1710
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.6870381465947674,
|
| 1211 |
-
"grad_norm":
|
| 1212 |
-
"learning_rate":
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 1720
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.6910325544238066,
|
| 1218 |
-
"grad_norm":
|
| 1219 |
-
"learning_rate":
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 1730
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.695026962252846,
|
| 1225 |
-
"grad_norm": 2.
|
| 1226 |
-
"learning_rate":
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 1740
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.6990213700818854,
|
| 1232 |
-
"grad_norm":
|
| 1233 |
-
"learning_rate":
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 1750
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.7030157779109247,
|
| 1239 |
-
"grad_norm":
|
| 1240 |
-
"learning_rate":
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 1760
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.7070101857399641,
|
| 1246 |
-
"grad_norm": 2.
|
| 1247 |
-
"learning_rate":
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 1770
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.7110045935690034,
|
| 1253 |
-
"grad_norm":
|
| 1254 |
-
"learning_rate":
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 1780
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.7149990013980427,
|
| 1260 |
-
"grad_norm":
|
| 1261 |
-
"learning_rate":
|
| 1262 |
-
"loss": 0.
|
| 1263 |
"step": 1790
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.7189934092270821,
|
| 1267 |
-
"grad_norm": 3.
|
| 1268 |
-
"learning_rate":
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 1800
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.7229878170561215,
|
| 1274 |
-
"grad_norm": 3.
|
| 1275 |
-
"learning_rate":
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 1810
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.7269822248851607,
|
| 1281 |
-
"grad_norm": 3.
|
| 1282 |
-
"learning_rate":
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 1820
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.7309766327142001,
|
| 1288 |
-
"grad_norm":
|
| 1289 |
-
"learning_rate":
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 1830
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.7349710405432395,
|
| 1295 |
-
"grad_norm":
|
| 1296 |
-
"learning_rate":
|
| 1297 |
-
"loss": 0.
|
| 1298 |
"step": 1840
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.7389654483722788,
|
| 1302 |
-
"grad_norm": 2.
|
| 1303 |
-
"learning_rate":
|
| 1304 |
-
"loss": 0.
|
| 1305 |
"step": 1850
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.7429598562013181,
|
| 1309 |
-
"grad_norm":
|
| 1310 |
-
"learning_rate":
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 1860
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.7469542640303575,
|
| 1316 |
-
"grad_norm":
|
| 1317 |
-
"learning_rate":
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 1870
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.7509486718593968,
|
| 1323 |
-
"grad_norm": 2.
|
| 1324 |
-
"learning_rate":
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 1880
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.7549430796884362,
|
| 1330 |
-
"grad_norm": 2.
|
| 1331 |
-
"learning_rate":
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 1890
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.7589374875174756,
|
| 1337 |
-
"grad_norm": 2.
|
| 1338 |
-
"learning_rate":
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 1900
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.7629318953465148,
|
| 1344 |
-
"grad_norm": 3.
|
| 1345 |
-
"learning_rate": 7.
|
| 1346 |
-
"loss": 0.
|
| 1347 |
"step": 1910
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 0.7669263031755542,
|
| 1351 |
-
"grad_norm":
|
| 1352 |
-
"learning_rate": 7.
|
| 1353 |
-
"loss": 0.
|
| 1354 |
"step": 1920
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 0.7709207110045936,
|
| 1358 |
-
"grad_norm":
|
| 1359 |
-
"learning_rate": 7.
|
| 1360 |
-
"loss": 0.
|
| 1361 |
"step": 1930
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 0.774915118833633,
|
| 1365 |
-
"grad_norm":
|
| 1366 |
-
"learning_rate": 7.
|
| 1367 |
-
"loss": 0.
|
| 1368 |
"step": 1940
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 0.7789095266626722,
|
| 1372 |
-
"grad_norm": 2.
|
| 1373 |
-
"learning_rate": 7.
|
| 1374 |
-
"loss": 0.
|
| 1375 |
"step": 1950
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 0.7829039344917116,
|
| 1379 |
-
"grad_norm":
|
| 1380 |
-
"learning_rate": 7.
|
| 1381 |
-
"loss": 0.
|
| 1382 |
"step": 1960
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 0.786898342320751,
|
| 1386 |
-
"grad_norm":
|
| 1387 |
-
"learning_rate": 7.
|
| 1388 |
-
"loss": 0.
|
| 1389 |
"step": 1970
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 0.7908927501497903,
|
| 1393 |
-
"grad_norm": 2.
|
| 1394 |
-
"learning_rate": 7.
|
| 1395 |
-
"loss": 0.
|
| 1396 |
"step": 1980
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 0.7948871579788296,
|
| 1400 |
-
"grad_norm":
|
| 1401 |
-
"learning_rate": 7.
|
| 1402 |
-
"loss": 0.
|
| 1403 |
"step": 1990
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 0.798881565807869,
|
| 1407 |
-
"grad_norm": 2.
|
| 1408 |
-
"learning_rate": 7.
|
| 1409 |
-
"loss": 0.
|
| 1410 |
"step": 2000
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 0.798881565807869,
|
| 1414 |
-
"eval_loss": 0.
|
| 1415 |
-
"eval_runtime":
|
| 1416 |
-
"eval_samples_per_second": 2.
|
| 1417 |
-
"eval_steps_per_second": 0.
|
| 1418 |
-
"eval_wer": 1.
|
| 1419 |
"step": 2000
|
| 1420 |
}
|
| 1421 |
],
|
| 1422 |
"logging_steps": 10,
|
| 1423 |
-
"max_steps":
|
| 1424 |
"num_input_tokens_seen": 0,
|
| 1425 |
"num_train_epochs": 3,
|
| 1426 |
"save_steps": 2000,
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 2000,
|
| 3 |
+
"best_metric": 1.5431796145591616,
|
| 4 |
"best_model_checkpoint": "./SALAMA_NEW8/checkpoint-2000",
|
| 5 |
"epoch": 0.798881565807869,
|
| 6 |
"eval_steps": 2000,
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.003994407829039345,
|
| 14 |
+
"grad_norm": 2.2177462577819824,
|
| 15 |
"learning_rate": 1.8e-07,
|
| 16 |
+
"loss": 0.022,
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.00798881565807869,
|
| 21 |
+
"grad_norm": 2.724672794342041,
|
| 22 |
"learning_rate": 3.8e-07,
|
| 23 |
+
"loss": 0.0249,
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.011983223487118035,
|
| 28 |
+
"grad_norm": 2.5243964195251465,
|
| 29 |
"learning_rate": 5.800000000000001e-07,
|
| 30 |
+
"loss": 0.0215,
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.01597763131615738,
|
| 35 |
+
"grad_norm": 2.4517173767089844,
|
| 36 |
"learning_rate": 7.8e-07,
|
| 37 |
+
"loss": 0.0229,
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.019972039145196723,
|
| 42 |
+
"grad_norm": 1.1104662418365479,
|
| 43 |
"learning_rate": 9.800000000000001e-07,
|
| 44 |
+
"loss": 0.0212,
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.02396644697423607,
|
| 49 |
+
"grad_norm": 3.2835190296173096,
|
| 50 |
"learning_rate": 1.1800000000000001e-06,
|
| 51 |
+
"loss": 0.0177,
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.027960854803275415,
|
| 56 |
+
"grad_norm": 2.933971405029297,
|
| 57 |
"learning_rate": 1.3800000000000001e-06,
|
| 58 |
+
"loss": 0.0181,
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.03195526263231476,
|
| 63 |
+
"grad_norm": 2.667775869369507,
|
| 64 |
"learning_rate": 1.5800000000000001e-06,
|
| 65 |
+
"loss": 0.0243,
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.035949670461354104,
|
| 70 |
+
"grad_norm": 2.8503150939941406,
|
| 71 |
"learning_rate": 1.7800000000000001e-06,
|
| 72 |
+
"loss": 0.0158,
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.039944078290393446,
|
| 77 |
+
"grad_norm": 2.6173582077026367,
|
| 78 |
"learning_rate": 1.98e-06,
|
| 79 |
+
"loss": 0.0217,
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.043938486119432796,
|
| 84 |
+
"grad_norm": 2.148167371749878,
|
| 85 |
"learning_rate": 2.1800000000000003e-06,
|
| 86 |
+
"loss": 0.0216,
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.04793289394847214,
|
| 91 |
+
"grad_norm": 1.6248950958251953,
|
| 92 |
"learning_rate": 2.38e-06,
|
| 93 |
+
"loss": 0.0211,
|
| 94 |
"step": 120
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.05192730177751148,
|
| 98 |
+
"grad_norm": 2.5831820964813232,
|
| 99 |
"learning_rate": 2.5800000000000003e-06,
|
| 100 |
+
"loss": 0.0282,
|
| 101 |
"step": 130
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.05592170960655083,
|
| 105 |
+
"grad_norm": 2.1337149143218994,
|
| 106 |
"learning_rate": 2.7800000000000005e-06,
|
| 107 |
+
"loss": 0.0215,
|
| 108 |
"step": 140
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.05991611743559017,
|
| 112 |
+
"grad_norm": 2.95695161819458,
|
| 113 |
"learning_rate": 2.9800000000000003e-06,
|
| 114 |
+
"loss": 0.0177,
|
| 115 |
"step": 150
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.06391052526462952,
|
| 119 |
+
"grad_norm": 2.277665853500366,
|
| 120 |
"learning_rate": 3.1800000000000005e-06,
|
| 121 |
+
"loss": 0.0217,
|
| 122 |
"step": 160
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.06790493309366886,
|
| 126 |
+
"grad_norm": 2.531909704208374,
|
| 127 |
"learning_rate": 3.3800000000000007e-06,
|
| 128 |
+
"loss": 0.0375,
|
| 129 |
"step": 170
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.07189934092270821,
|
| 133 |
+
"grad_norm": 2.5161631107330322,
|
| 134 |
"learning_rate": 3.58e-06,
|
| 135 |
+
"loss": 0.0262,
|
| 136 |
"step": 180
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.07589374875174755,
|
| 140 |
+
"grad_norm": 2.9737093448638916,
|
| 141 |
"learning_rate": 3.7800000000000002e-06,
|
| 142 |
+
"loss": 0.0237,
|
| 143 |
"step": 190
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.07988815658078689,
|
| 147 |
+
"grad_norm": 2.2957046031951904,
|
| 148 |
"learning_rate": 3.980000000000001e-06,
|
| 149 |
+
"loss": 0.0211,
|
| 150 |
"step": 200
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.08388256440982625,
|
| 154 |
+
"grad_norm": 1.9400784969329834,
|
| 155 |
"learning_rate": 4.18e-06,
|
| 156 |
+
"loss": 0.0241,
|
| 157 |
"step": 210
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.08787697223886559,
|
| 161 |
+
"grad_norm": 3.7293314933776855,
|
| 162 |
"learning_rate": 4.38e-06,
|
| 163 |
+
"loss": 0.0263,
|
| 164 |
"step": 220
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.09187138006790493,
|
| 168 |
+
"grad_norm": 4.266306400299072,
|
| 169 |
"learning_rate": 4.58e-06,
|
| 170 |
+
"loss": 0.0387,
|
| 171 |
"step": 230
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.09586578789694428,
|
| 175 |
+
"grad_norm": 2.510282278060913,
|
| 176 |
"learning_rate": 4.78e-06,
|
| 177 |
+
"loss": 0.029,
|
| 178 |
"step": 240
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.09986019572598362,
|
| 182 |
+
"grad_norm": 1.4938842058181763,
|
| 183 |
"learning_rate": 4.980000000000001e-06,
|
| 184 |
+
"loss": 0.0154,
|
| 185 |
"step": 250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.10385460355502296,
|
| 189 |
+
"grad_norm": 2.0076608657836914,
|
| 190 |
"learning_rate": 5.18e-06,
|
| 191 |
+
"loss": 0.0263,
|
| 192 |
"step": 260
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.10784901138406232,
|
| 196 |
+
"grad_norm": 2.252916097640991,
|
| 197 |
"learning_rate": 5.380000000000001e-06,
|
| 198 |
+
"loss": 0.0273,
|
| 199 |
"step": 270
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.11184341921310166,
|
| 203 |
+
"grad_norm": 0.9760869145393372,
|
| 204 |
"learning_rate": 5.580000000000001e-06,
|
| 205 |
+
"loss": 0.0227,
|
| 206 |
"step": 280
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.115837827042141,
|
| 210 |
+
"grad_norm": 1.7485207319259644,
|
| 211 |
"learning_rate": 5.78e-06,
|
| 212 |
+
"loss": 0.0234,
|
| 213 |
"step": 290
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.11983223487118035,
|
| 217 |
+
"grad_norm": 2.540903329849243,
|
| 218 |
"learning_rate": 5.98e-06,
|
| 219 |
+
"loss": 0.0216,
|
| 220 |
"step": 300
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.12382664270021969,
|
| 224 |
+
"grad_norm": 2.970639228820801,
|
| 225 |
"learning_rate": 6.18e-06,
|
| 226 |
+
"loss": 0.0244,
|
| 227 |
"step": 310
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.12782105052925904,
|
| 231 |
+
"grad_norm": 4.563722610473633,
|
| 232 |
"learning_rate": 6.380000000000001e-06,
|
| 233 |
+
"loss": 0.0332,
|
| 234 |
"step": 320
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.1318154583582984,
|
| 238 |
+
"grad_norm": 3.670563220977783,
|
| 239 |
"learning_rate": 6.5800000000000005e-06,
|
| 240 |
+
"loss": 0.0259,
|
| 241 |
"step": 330
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.13580986618733773,
|
| 245 |
+
"grad_norm": 3.6714515686035156,
|
| 246 |
"learning_rate": 6.780000000000001e-06,
|
| 247 |
+
"loss": 0.032,
|
| 248 |
"step": 340
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.13980427401637707,
|
| 252 |
+
"grad_norm": 2.029430389404297,
|
| 253 |
"learning_rate": 6.98e-06,
|
| 254 |
+
"loss": 0.0264,
|
| 255 |
"step": 350
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.14379868184541642,
|
| 259 |
+
"grad_norm": 2.609748363494873,
|
| 260 |
"learning_rate": 7.180000000000001e-06,
|
| 261 |
+
"loss": 0.0369,
|
| 262 |
"step": 360
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.14779308967445576,
|
| 266 |
+
"grad_norm": 1.438359260559082,
|
| 267 |
"learning_rate": 7.3800000000000005e-06,
|
| 268 |
+
"loss": 0.0242,
|
| 269 |
"step": 370
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.1517874975034951,
|
| 273 |
+
"grad_norm": 2.1897497177124023,
|
| 274 |
"learning_rate": 7.58e-06,
|
| 275 |
+
"loss": 0.02,
|
| 276 |
"step": 380
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.15578190533253444,
|
| 280 |
+
"grad_norm": 1.9179186820983887,
|
| 281 |
"learning_rate": 7.78e-06,
|
| 282 |
+
"loss": 0.0366,
|
| 283 |
"step": 390
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.15977631316157379,
|
| 287 |
+
"grad_norm": 2.6540606021881104,
|
| 288 |
"learning_rate": 7.980000000000002e-06,
|
| 289 |
+
"loss": 0.0253,
|
| 290 |
"step": 400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.16377072099061313,
|
| 294 |
+
"grad_norm": 2.5292742252349854,
|
| 295 |
"learning_rate": 8.18e-06,
|
| 296 |
+
"loss": 0.0356,
|
| 297 |
"step": 410
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.1677651288196525,
|
| 301 |
+
"grad_norm": 2.4729959964752197,
|
| 302 |
"learning_rate": 8.380000000000001e-06,
|
| 303 |
+
"loss": 0.0327,
|
| 304 |
"step": 420
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.17175953664869184,
|
| 308 |
+
"grad_norm": 3.639272689819336,
|
| 309 |
"learning_rate": 8.580000000000001e-06,
|
| 310 |
+
"loss": 0.032,
|
| 311 |
"step": 430
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.17575394447773118,
|
| 315 |
+
"grad_norm": 3.2385575771331787,
|
| 316 |
"learning_rate": 8.78e-06,
|
| 317 |
+
"loss": 0.0359,
|
| 318 |
"step": 440
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.17974835230677053,
|
| 322 |
+
"grad_norm": 3.476755142211914,
|
| 323 |
"learning_rate": 8.98e-06,
|
| 324 |
+
"loss": 0.0286,
|
| 325 |
"step": 450
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.18374276013580987,
|
| 329 |
+
"grad_norm": 2.612086772918701,
|
| 330 |
"learning_rate": 9.180000000000002e-06,
|
| 331 |
+
"loss": 0.0228,
|
| 332 |
"step": 460
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.1877371679648492,
|
| 336 |
+
"grad_norm": 3.5403027534484863,
|
| 337 |
"learning_rate": 9.38e-06,
|
| 338 |
+
"loss": 0.0243,
|
| 339 |
"step": 470
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.19173157579388855,
|
| 343 |
+
"grad_norm": 2.7765207290649414,
|
| 344 |
"learning_rate": 9.58e-06,
|
| 345 |
+
"loss": 0.0348,
|
| 346 |
"step": 480
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.1957259836229279,
|
| 350 |
+
"grad_norm": 3.589587450027466,
|
| 351 |
"learning_rate": 9.780000000000001e-06,
|
| 352 |
+
"loss": 0.0323,
|
| 353 |
"step": 490
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.19972039145196724,
|
| 357 |
+
"grad_norm": 1.8719650506973267,
|
| 358 |
"learning_rate": 9.980000000000001e-06,
|
| 359 |
+
"loss": 0.0275,
|
| 360 |
"step": 500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.20371479928100658,
|
| 364 |
+
"grad_norm": 2.750931739807129,
|
| 365 |
+
"learning_rate": 9.983636363636364e-06,
|
| 366 |
+
"loss": 0.0348,
|
| 367 |
"step": 510
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.20770920711004592,
|
| 371 |
+
"grad_norm": 1.9383965730667114,
|
| 372 |
+
"learning_rate": 9.965454545454546e-06,
|
| 373 |
+
"loss": 0.0282,
|
| 374 |
"step": 520
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.2117036149390853,
|
| 378 |
+
"grad_norm": 3.6504414081573486,
|
| 379 |
+
"learning_rate": 9.947272727272728e-06,
|
| 380 |
+
"loss": 0.0316,
|
| 381 |
"step": 530
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.21569802276812464,
|
| 385 |
+
"grad_norm": 3.26383376121521,
|
| 386 |
+
"learning_rate": 9.92909090909091e-06,
|
| 387 |
+
"loss": 0.0273,
|
| 388 |
"step": 540
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.21969243059716398,
|
| 392 |
+
"grad_norm": 4.769904613494873,
|
| 393 |
+
"learning_rate": 9.910909090909092e-06,
|
| 394 |
+
"loss": 0.0295,
|
| 395 |
"step": 550
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.22368683842620332,
|
| 399 |
+
"grad_norm": 1.143069863319397,
|
| 400 |
+
"learning_rate": 9.892727272727273e-06,
|
| 401 |
+
"loss": 0.027,
|
| 402 |
"step": 560
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.22768124625524266,
|
| 406 |
+
"grad_norm": 2.988460063934326,
|
| 407 |
+
"learning_rate": 9.874545454545455e-06,
|
| 408 |
+
"loss": 0.0349,
|
| 409 |
"step": 570
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.231675654084282,
|
| 413 |
+
"grad_norm": 2.6448323726654053,
|
| 414 |
+
"learning_rate": 9.856363636363637e-06,
|
| 415 |
+
"loss": 0.036,
|
| 416 |
"step": 580
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.23567006191332135,
|
| 420 |
+
"grad_norm": 3.077092409133911,
|
| 421 |
+
"learning_rate": 9.838181818181819e-06,
|
| 422 |
+
"loss": 0.0491,
|
| 423 |
"step": 590
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.2396644697423607,
|
| 427 |
+
"grad_norm": 6.214341163635254,
|
| 428 |
+
"learning_rate": 9.820000000000001e-06,
|
| 429 |
+
"loss": 0.0313,
|
| 430 |
"step": 600
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.24365887757140003,
|
| 434 |
+
"grad_norm": 3.1780426502227783,
|
| 435 |
+
"learning_rate": 9.801818181818183e-06,
|
| 436 |
+
"loss": 0.0251,
|
| 437 |
"step": 610
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.24765328540043938,
|
| 441 |
+
"grad_norm": 1.7672063112258911,
|
| 442 |
+
"learning_rate": 9.783636363636365e-06,
|
| 443 |
+
"loss": 0.0259,
|
| 444 |
"step": 620
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.2516476932294787,
|
| 448 |
+
"grad_norm": 3.9317915439605713,
|
| 449 |
+
"learning_rate": 9.765454545454546e-06,
|
| 450 |
+
"loss": 0.0263,
|
| 451 |
"step": 630
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.2556421010585181,
|
| 455 |
+
"grad_norm": 2.23710298538208,
|
| 456 |
+
"learning_rate": 9.747272727272728e-06,
|
| 457 |
+
"loss": 0.0335,
|
| 458 |
"step": 640
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.2596365088875574,
|
| 462 |
+
"grad_norm": 3.24867844581604,
|
| 463 |
+
"learning_rate": 9.72909090909091e-06,
|
| 464 |
+
"loss": 0.0383,
|
| 465 |
"step": 650
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.2636309167165968,
|
| 469 |
+
"grad_norm": 3.0995874404907227,
|
| 470 |
+
"learning_rate": 9.710909090909092e-06,
|
| 471 |
+
"loss": 0.0296,
|
| 472 |
"step": 660
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.2676253245456361,
|
| 476 |
+
"grad_norm": 1.4767210483551025,
|
| 477 |
+
"learning_rate": 9.692727272727274e-06,
|
| 478 |
+
"loss": 0.04,
|
| 479 |
"step": 670
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.27161973237467546,
|
| 483 |
+
"grad_norm": 1.7182132005691528,
|
| 484 |
+
"learning_rate": 9.674545454545456e-06,
|
| 485 |
+
"loss": 0.0294,
|
| 486 |
"step": 680
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.2756141402037148,
|
| 490 |
+
"grad_norm": 3.0770034790039062,
|
| 491 |
+
"learning_rate": 9.656363636363637e-06,
|
| 492 |
+
"loss": 0.0393,
|
| 493 |
"step": 690
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.27960854803275415,
|
| 497 |
+
"grad_norm": 2.452723503112793,
|
| 498 |
+
"learning_rate": 9.63818181818182e-06,
|
| 499 |
+
"loss": 0.0321,
|
| 500 |
"step": 700
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.2836029558617935,
|
| 504 |
+
"grad_norm": 2.9560225009918213,
|
| 505 |
+
"learning_rate": 9.620000000000001e-06,
|
| 506 |
"loss": 0.0292,
|
| 507 |
"step": 710
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.28759736369083283,
|
| 511 |
+
"grad_norm": 2.8095853328704834,
|
| 512 |
+
"learning_rate": 9.601818181818183e-06,
|
| 513 |
+
"loss": 0.028,
|
| 514 |
"step": 720
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.2915917715198722,
|
| 518 |
+
"grad_norm": 3.4851746559143066,
|
| 519 |
+
"learning_rate": 9.583636363636365e-06,
|
| 520 |
+
"loss": 0.0296,
|
| 521 |
"step": 730
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.2955861793489115,
|
| 525 |
+
"grad_norm": 4.2011027336120605,
|
| 526 |
+
"learning_rate": 9.565454545454547e-06,
|
| 527 |
+
"loss": 0.0353,
|
| 528 |
"step": 740
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.2995805871779509,
|
| 532 |
+
"grad_norm": 5.23146390914917,
|
| 533 |
+
"learning_rate": 9.547272727272728e-06,
|
| 534 |
+
"loss": 0.0393,
|
| 535 |
"step": 750
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.3035749950069902,
|
| 539 |
+
"grad_norm": 2.8136441707611084,
|
| 540 |
+
"learning_rate": 9.52909090909091e-06,
|
| 541 |
+
"loss": 0.0339,
|
| 542 |
"step": 760
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.30756940283602957,
|
| 546 |
+
"grad_norm": 3.0712337493896484,
|
| 547 |
+
"learning_rate": 9.510909090909092e-06,
|
| 548 |
+
"loss": 0.0414,
|
| 549 |
"step": 770
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.3115638106650689,
|
| 553 |
+
"grad_norm": 2.1798360347747803,
|
| 554 |
+
"learning_rate": 9.492727272727274e-06,
|
| 555 |
+
"loss": 0.03,
|
| 556 |
"step": 780
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.31555821849410826,
|
| 560 |
+
"grad_norm": 3.7893567085266113,
|
| 561 |
+
"learning_rate": 9.474545454545456e-06,
|
| 562 |
+
"loss": 0.0325,
|
| 563 |
"step": 790
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.31955262632314757,
|
| 567 |
+
"grad_norm": 2.0599608421325684,
|
| 568 |
+
"learning_rate": 9.456363636363638e-06,
|
| 569 |
+
"loss": 0.0389,
|
| 570 |
"step": 800
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.32354703415218694,
|
| 574 |
+
"grad_norm": 2.7613537311553955,
|
| 575 |
+
"learning_rate": 9.438181818181818e-06,
|
| 576 |
+
"loss": 0.0284,
|
| 577 |
"step": 810
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.32754144198122626,
|
| 581 |
+
"grad_norm": 2.630587577819824,
|
| 582 |
+
"learning_rate": 9.42e-06,
|
| 583 |
+
"loss": 0.0308,
|
| 584 |
"step": 820
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.3315358498102656,
|
| 588 |
+
"grad_norm": 2.8372304439544678,
|
| 589 |
+
"learning_rate": 9.401818181818183e-06,
|
| 590 |
+
"loss": 0.029,
|
| 591 |
"step": 830
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.335530257639305,
|
| 595 |
+
"grad_norm": 2.21663236618042,
|
| 596 |
+
"learning_rate": 9.383636363636365e-06,
|
| 597 |
+
"loss": 0.0367,
|
| 598 |
"step": 840
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.3395246654683443,
|
| 602 |
+
"grad_norm": 3.325446367263794,
|
| 603 |
+
"learning_rate": 9.365454545454547e-06,
|
| 604 |
+
"loss": 0.0256,
|
| 605 |
"step": 850
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.3435190732973837,
|
| 609 |
+
"grad_norm": 2.7013704776763916,
|
| 610 |
+
"learning_rate": 9.347272727272729e-06,
|
| 611 |
+
"loss": 0.0279,
|
| 612 |
"step": 860
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.347513481126423,
|
| 616 |
+
"grad_norm": 1.6712983846664429,
|
| 617 |
+
"learning_rate": 9.32909090909091e-06,
|
| 618 |
+
"loss": 0.0256,
|
| 619 |
"step": 870
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.35150788895546237,
|
| 623 |
+
"grad_norm": 3.0498671531677246,
|
| 624 |
+
"learning_rate": 9.310909090909092e-06,
|
| 625 |
+
"loss": 0.0351,
|
| 626 |
"step": 880
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.3555022967845017,
|
| 630 |
+
"grad_norm": 2.060898542404175,
|
| 631 |
+
"learning_rate": 9.292727272727272e-06,
|
| 632 |
+
"loss": 0.0349,
|
| 633 |
"step": 890
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.35949670461354105,
|
| 637 |
+
"grad_norm": 2.191037178039551,
|
| 638 |
+
"learning_rate": 9.274545454545454e-06,
|
| 639 |
+
"loss": 0.0236,
|
| 640 |
"step": 900
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.36349111244258037,
|
| 644 |
+
"grad_norm": 2.7442283630371094,
|
| 645 |
+
"learning_rate": 9.256363636363636e-06,
|
| 646 |
+
"loss": 0.0339,
|
| 647 |
"step": 910
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.36748552027161974,
|
| 651 |
+
"grad_norm": 2.151813268661499,
|
| 652 |
+
"learning_rate": 9.23818181818182e-06,
|
| 653 |
+
"loss": 0.0272,
|
| 654 |
"step": 920
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.37147992810065905,
|
| 658 |
+
"grad_norm": 3.0903069972991943,
|
| 659 |
+
"learning_rate": 9.220000000000002e-06,
|
| 660 |
+
"loss": 0.0343,
|
| 661 |
"step": 930
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.3754743359296984,
|
| 665 |
+
"grad_norm": 3.2857322692871094,
|
| 666 |
+
"learning_rate": 9.201818181818183e-06,
|
| 667 |
+
"loss": 0.028,
|
| 668 |
"step": 940
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.3794687437587378,
|
| 672 |
+
"grad_norm": 2.7144224643707275,
|
| 673 |
+
"learning_rate": 9.183636363636365e-06,
|
| 674 |
+
"loss": 0.0224,
|
| 675 |
"step": 950
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.3834631515877771,
|
| 679 |
+
"grad_norm": 2.985253095626831,
|
| 680 |
+
"learning_rate": 9.165454545454547e-06,
|
| 681 |
+
"loss": 0.0297,
|
| 682 |
"step": 960
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.3874575594168165,
|
| 686 |
+
"grad_norm": 2.6801540851593018,
|
| 687 |
+
"learning_rate": 9.147272727272727e-06,
|
| 688 |
+
"loss": 0.0306,
|
| 689 |
"step": 970
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.3914519672458558,
|
| 693 |
+
"grad_norm": 2.944761276245117,
|
| 694 |
+
"learning_rate": 9.129090909090909e-06,
|
| 695 |
+
"loss": 0.0324,
|
| 696 |
"step": 980
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.39544637507489516,
|
| 700 |
+
"grad_norm": 2.6776340007781982,
|
| 701 |
+
"learning_rate": 9.11090909090909e-06,
|
| 702 |
+
"loss": 0.04,
|
| 703 |
"step": 990
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.3994407829039345,
|
| 707 |
+
"grad_norm": 4.007734775543213,
|
| 708 |
+
"learning_rate": 9.092727272727273e-06,
|
| 709 |
+
"loss": 0.0286,
|
| 710 |
"step": 1000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.40343519073297385,
|
| 714 |
+
"grad_norm": 3.652127265930176,
|
| 715 |
+
"learning_rate": 9.074545454545455e-06,
|
| 716 |
+
"loss": 0.0373,
|
| 717 |
"step": 1010
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.40742959856201316,
|
| 721 |
+
"grad_norm": 1.8009936809539795,
|
| 722 |
+
"learning_rate": 9.056363636363638e-06,
|
| 723 |
+
"loss": 0.024,
|
| 724 |
"step": 1020
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.41142400639105253,
|
| 728 |
+
"grad_norm": 2.0933914184570312,
|
| 729 |
+
"learning_rate": 9.03818181818182e-06,
|
| 730 |
+
"loss": 0.0341,
|
| 731 |
"step": 1030
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.41541841422009185,
|
| 735 |
+
"grad_norm": 2.683140277862549,
|
| 736 |
+
"learning_rate": 9.020000000000002e-06,
|
| 737 |
+
"loss": 0.0473,
|
| 738 |
"step": 1040
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.4194128220491312,
|
| 742 |
+
"grad_norm": 2.7582523822784424,
|
| 743 |
+
"learning_rate": 9.001818181818182e-06,
|
| 744 |
+
"loss": 0.0322,
|
| 745 |
"step": 1050
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.4234072298781706,
|
| 749 |
+
"grad_norm": 3.6831445693969727,
|
| 750 |
+
"learning_rate": 8.983636363636364e-06,
|
| 751 |
+
"loss": 0.0382,
|
| 752 |
"step": 1060
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.4274016377072099,
|
| 756 |
+
"grad_norm": 2.053619623184204,
|
| 757 |
+
"learning_rate": 8.965454545454546e-06,
|
| 758 |
+
"loss": 0.0304,
|
| 759 |
"step": 1070
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.4313960455362493,
|
| 763 |
+
"grad_norm": 2.5573840141296387,
|
| 764 |
+
"learning_rate": 8.947272727272727e-06,
|
| 765 |
+
"loss": 0.0567,
|
| 766 |
"step": 1080
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.4353904533652886,
|
| 770 |
+
"grad_norm": 3.5455336570739746,
|
| 771 |
+
"learning_rate": 8.92909090909091e-06,
|
| 772 |
+
"loss": 0.0338,
|
| 773 |
"step": 1090
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.43938486119432796,
|
| 777 |
+
"grad_norm": 2.8842337131500244,
|
| 778 |
+
"learning_rate": 8.910909090909091e-06,
|
| 779 |
+
"loss": 0.0314,
|
| 780 |
"step": 1100
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.4433792690233673,
|
| 784 |
+
"grad_norm": 2.779266834259033,
|
| 785 |
+
"learning_rate": 8.892727272727275e-06,
|
| 786 |
+
"loss": 0.0486,
|
| 787 |
"step": 1110
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.44737367685240664,
|
| 791 |
+
"grad_norm": 3.959299325942993,
|
| 792 |
+
"learning_rate": 8.874545454545456e-06,
|
| 793 |
+
"loss": 0.0396,
|
| 794 |
"step": 1120
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.45136808468144596,
|
| 798 |
+
"grad_norm": 2.053091526031494,
|
| 799 |
+
"learning_rate": 8.856363636363637e-06,
|
| 800 |
+
"loss": 0.0329,
|
| 801 |
"step": 1130
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.45536249251048533,
|
| 805 |
+
"grad_norm": 1.1432011127471924,
|
| 806 |
+
"learning_rate": 8.838181818181818e-06,
|
| 807 |
+
"loss": 0.0268,
|
| 808 |
"step": 1140
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.45935690033952464,
|
| 812 |
+
"grad_norm": 2.738510847091675,
|
| 813 |
+
"learning_rate": 8.82e-06,
|
| 814 |
+
"loss": 0.0388,
|
| 815 |
"step": 1150
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.463351308168564,
|
| 819 |
+
"grad_norm": 1.9435960054397583,
|
| 820 |
+
"learning_rate": 8.801818181818182e-06,
|
| 821 |
+
"loss": 0.0321,
|
| 822 |
"step": 1160
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 0.46734571599760333,
|
| 826 |
+
"grad_norm": 2.462301015853882,
|
| 827 |
+
"learning_rate": 8.783636363636364e-06,
|
| 828 |
+
"loss": 0.0344,
|
| 829 |
"step": 1170
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 0.4713401238266427,
|
| 833 |
+
"grad_norm": 2.8463451862335205,
|
| 834 |
+
"learning_rate": 8.765454545454546e-06,
|
| 835 |
+
"loss": 0.0372,
|
| 836 |
"step": 1180
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.47533453165568207,
|
| 840 |
+
"grad_norm": 4.407367706298828,
|
| 841 |
+
"learning_rate": 8.747272727272728e-06,
|
| 842 |
+
"loss": 0.0399,
|
| 843 |
"step": 1190
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 0.4793289394847214,
|
| 847 |
+
"grad_norm": 2.5179121494293213,
|
| 848 |
+
"learning_rate": 8.72909090909091e-06,
|
| 849 |
+
"loss": 0.0331,
|
| 850 |
"step": 1200
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 0.48332334731376075,
|
| 854 |
+
"grad_norm": 3.2438509464263916,
|
| 855 |
+
"learning_rate": 8.710909090909091e-06,
|
| 856 |
+
"loss": 0.0348,
|
| 857 |
"step": 1210
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.48731775514280007,
|
| 861 |
+
"grad_norm": 2.54004168510437,
|
| 862 |
+
"learning_rate": 8.692727272727273e-06,
|
| 863 |
+
"loss": 0.0317,
|
| 864 |
"step": 1220
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.49131216297183944,
|
| 868 |
+
"grad_norm": 2.8079185485839844,
|
| 869 |
+
"learning_rate": 8.674545454545455e-06,
|
| 870 |
+
"loss": 0.0386,
|
| 871 |
"step": 1230
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 0.49530657080087875,
|
| 875 |
+
"grad_norm": 4.827033519744873,
|
| 876 |
+
"learning_rate": 8.656363636363637e-06,
|
| 877 |
+
"loss": 0.0338,
|
| 878 |
"step": 1240
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.4993009786299181,
|
| 882 |
+
"grad_norm": 2.919968843460083,
|
| 883 |
+
"learning_rate": 8.638181818181819e-06,
|
| 884 |
+
"loss": 0.0265,
|
| 885 |
"step": 1250
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 0.5032953864589574,
|
| 889 |
+
"grad_norm": 3.6935982704162598,
|
| 890 |
+
"learning_rate": 8.62e-06,
|
| 891 |
+
"loss": 0.0327,
|
| 892 |
"step": 1260
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 0.5072897942879968,
|
| 896 |
+
"grad_norm": 4.243749141693115,
|
| 897 |
+
"learning_rate": 8.601818181818182e-06,
|
| 898 |
+
"loss": 0.0282,
|
| 899 |
"step": 1270
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 0.5112842021170362,
|
| 903 |
+
"grad_norm": 3.979485273361206,
|
| 904 |
+
"learning_rate": 8.583636363636364e-06,
|
| 905 |
+
"loss": 0.0242,
|
| 906 |
"step": 1280
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 0.5152786099460755,
|
| 910 |
+
"grad_norm": 3.32183575630188,
|
| 911 |
+
"learning_rate": 8.565454545454546e-06,
|
| 912 |
+
"loss": 0.0356,
|
| 913 |
"step": 1290
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 0.5192730177751148,
|
| 917 |
+
"grad_norm": 2.474187135696411,
|
| 918 |
+
"learning_rate": 8.547272727272728e-06,
|
| 919 |
+
"loss": 0.0201,
|
| 920 |
"step": 1300
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 0.5232674256041542,
|
| 924 |
+
"grad_norm": 3.150151252746582,
|
| 925 |
+
"learning_rate": 8.52909090909091e-06,
|
| 926 |
+
"loss": 0.0304,
|
| 927 |
"step": 1310
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.5272618334331935,
|
| 931 |
+
"grad_norm": 2.039504289627075,
|
| 932 |
+
"learning_rate": 8.510909090909092e-06,
|
| 933 |
+
"loss": 0.045,
|
| 934 |
"step": 1320
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 0.5312562412622329,
|
| 938 |
+
"grad_norm": 2.7388851642608643,
|
| 939 |
+
"learning_rate": 8.492727272727273e-06,
|
| 940 |
+
"loss": 0.0268,
|
| 941 |
"step": 1330
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.5352506490912722,
|
| 945 |
+
"grad_norm": 2.9932122230529785,
|
| 946 |
+
"learning_rate": 8.474545454545455e-06,
|
| 947 |
+
"loss": 0.0331,
|
| 948 |
"step": 1340
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 0.5392450569203115,
|
| 952 |
+
"grad_norm": 3.3818037509918213,
|
| 953 |
+
"learning_rate": 8.456363636363637e-06,
|
| 954 |
+
"loss": 0.0365,
|
| 955 |
"step": 1350
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 0.5432394647493509,
|
| 959 |
+
"grad_norm": 3.251274585723877,
|
| 960 |
+
"learning_rate": 8.438181818181819e-06,
|
| 961 |
+
"loss": 0.0395,
|
| 962 |
"step": 1360
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.5472338725783903,
|
| 966 |
+
"grad_norm": 3.46567964553833,
|
| 967 |
+
"learning_rate": 8.42e-06,
|
| 968 |
+
"loss": 0.0287,
|
| 969 |
"step": 1370
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.5512282804074295,
|
| 973 |
+
"grad_norm": 2.459820032119751,
|
| 974 |
+
"learning_rate": 8.401818181818183e-06,
|
| 975 |
+
"loss": 0.0329,
|
| 976 |
"step": 1380
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 0.5552226882364689,
|
| 980 |
+
"grad_norm": 1.735729694366455,
|
| 981 |
+
"learning_rate": 8.383636363636364e-06,
|
| 982 |
+
"loss": 0.0313,
|
| 983 |
"step": 1390
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 0.5592170960655083,
|
| 987 |
+
"grad_norm": 2.2887370586395264,
|
| 988 |
+
"learning_rate": 8.365454545454546e-06,
|
| 989 |
+
"loss": 0.0215,
|
| 990 |
"step": 1400
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.5632115038945477,
|
| 994 |
+
"grad_norm": 3.110576868057251,
|
| 995 |
+
"learning_rate": 8.347272727272728e-06,
|
| 996 |
+
"loss": 0.0213,
|
| 997 |
"step": 1410
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 0.567205911723587,
|
| 1001 |
+
"grad_norm": 3.1144895553588867,
|
| 1002 |
+
"learning_rate": 8.32909090909091e-06,
|
| 1003 |
+
"loss": 0.0312,
|
| 1004 |
"step": 1420
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 0.5712003195526263,
|
| 1008 |
+
"grad_norm": 2.9777989387512207,
|
| 1009 |
+
"learning_rate": 8.310909090909092e-06,
|
| 1010 |
+
"loss": 0.0322,
|
| 1011 |
"step": 1430
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 0.5751947273816657,
|
| 1015 |
+
"grad_norm": 2.2770936489105225,
|
| 1016 |
+
"learning_rate": 8.292727272727274e-06,
|
| 1017 |
+
"loss": 0.0304,
|
| 1018 |
"step": 1440
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 0.579189135210705,
|
| 1022 |
+
"grad_norm": 3.5356180667877197,
|
| 1023 |
+
"learning_rate": 8.274545454545455e-06,
|
| 1024 |
+
"loss": 0.0347,
|
| 1025 |
"step": 1450
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 0.5831835430397444,
|
| 1029 |
+
"grad_norm": 3.6384565830230713,
|
| 1030 |
+
"learning_rate": 8.256363636363637e-06,
|
| 1031 |
+
"loss": 0.0289,
|
| 1032 |
"step": 1460
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 0.5871779508687837,
|
| 1036 |
+
"grad_norm": 3.053424119949341,
|
| 1037 |
+
"learning_rate": 8.238181818181819e-06,
|
| 1038 |
+
"loss": 0.0356,
|
| 1039 |
"step": 1470
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
"epoch": 0.591172358697823,
|
| 1043 |
+
"grad_norm": 4.9105963706970215,
|
| 1044 |
+
"learning_rate": 8.220000000000001e-06,
|
| 1045 |
+
"loss": 0.0315,
|
| 1046 |
"step": 1480
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 0.5951667665268624,
|
| 1050 |
+
"grad_norm": 3.0485212802886963,
|
| 1051 |
+
"learning_rate": 8.201818181818183e-06,
|
| 1052 |
+
"loss": 0.0298,
|
| 1053 |
"step": 1490
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
"epoch": 0.5991611743559018,
|
| 1057 |
+
"grad_norm": 3.3632636070251465,
|
| 1058 |
+
"learning_rate": 8.183636363636365e-06,
|
| 1059 |
+
"loss": 0.0268,
|
| 1060 |
"step": 1500
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 0.603155582184941,
|
| 1064 |
+
"grad_norm": 2.593235969543457,
|
| 1065 |
+
"learning_rate": 8.165454545454546e-06,
|
| 1066 |
+
"loss": 0.0286,
|
| 1067 |
"step": 1510
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 0.6071499900139804,
|
| 1071 |
+
"grad_norm": 2.542865753173828,
|
| 1072 |
+
"learning_rate": 8.147272727272728e-06,
|
| 1073 |
+
"loss": 0.0317,
|
| 1074 |
"step": 1520
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 0.6111443978430198,
|
| 1078 |
+
"grad_norm": 3.246321201324463,
|
| 1079 |
+
"learning_rate": 8.12909090909091e-06,
|
| 1080 |
+
"loss": 0.0358,
|
| 1081 |
"step": 1530
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.6151388056720591,
|
| 1085 |
+
"grad_norm": 4.592155456542969,
|
| 1086 |
+
"learning_rate": 8.110909090909092e-06,
|
| 1087 |
+
"loss": 0.0295,
|
| 1088 |
"step": 1540
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 0.6191332135010985,
|
| 1092 |
+
"grad_norm": 2.1040351390838623,
|
| 1093 |
+
"learning_rate": 8.092727272727274e-06,
|
| 1094 |
+
"loss": 0.044,
|
| 1095 |
"step": 1550
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 0.6231276213301378,
|
| 1099 |
+
"grad_norm": 2.832470417022705,
|
| 1100 |
+
"learning_rate": 8.074545454545456e-06,
|
| 1101 |
+
"loss": 0.0335,
|
| 1102 |
"step": 1560
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 0.6271220291591771,
|
| 1106 |
+
"grad_norm": 3.7737035751342773,
|
| 1107 |
+
"learning_rate": 8.056363636363636e-06,
|
| 1108 |
+
"loss": 0.0289,
|
| 1109 |
"step": 1570
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 0.6311164369882165,
|
| 1113 |
+
"grad_norm": 2.9322657585144043,
|
| 1114 |
+
"learning_rate": 8.038181818181818e-06,
|
| 1115 |
+
"loss": 0.0209,
|
| 1116 |
"step": 1580
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 0.6351108448172559,
|
| 1120 |
+
"grad_norm": 3.0506591796875,
|
| 1121 |
+
"learning_rate": 8.020000000000001e-06,
|
| 1122 |
+
"loss": 0.0277,
|
| 1123 |
"step": 1590
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 0.6391052526462951,
|
| 1127 |
+
"grad_norm": 1.679126501083374,
|
| 1128 |
+
"learning_rate": 8.001818181818183e-06,
|
| 1129 |
+
"loss": 0.0313,
|
| 1130 |
"step": 1600
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 0.6430996604753345,
|
| 1134 |
+
"grad_norm": 1.4025191068649292,
|
| 1135 |
+
"learning_rate": 7.983636363636365e-06,
|
| 1136 |
+
"loss": 0.0342,
|
| 1137 |
"step": 1610
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 0.6470940683043739,
|
| 1141 |
+
"grad_norm": 3.2698376178741455,
|
| 1142 |
+
"learning_rate": 7.965454545454547e-06,
|
| 1143 |
+
"loss": 0.0334,
|
| 1144 |
"step": 1620
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 0.6510884761334133,
|
| 1148 |
+
"grad_norm": 2.007560968399048,
|
| 1149 |
+
"learning_rate": 7.947272727272728e-06,
|
| 1150 |
+
"loss": 0.0361,
|
| 1151 |
"step": 1630
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 0.6550828839624525,
|
| 1155 |
+
"grad_norm": 3.021299123764038,
|
| 1156 |
+
"learning_rate": 7.92909090909091e-06,
|
| 1157 |
+
"loss": 0.0316,
|
| 1158 |
"step": 1640
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 0.6590772917914919,
|
| 1162 |
+
"grad_norm": 3.4559459686279297,
|
| 1163 |
+
"learning_rate": 7.91090909090909e-06,
|
| 1164 |
+
"loss": 0.0336,
|
| 1165 |
"step": 1650
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 0.6630716996205313,
|
| 1169 |
+
"grad_norm": 3.786959648132324,
|
| 1170 |
+
"learning_rate": 7.892727272727272e-06,
|
| 1171 |
+
"loss": 0.0337,
|
| 1172 |
"step": 1660
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 0.6670661074495706,
|
| 1176 |
+
"grad_norm": 2.5222368240356445,
|
| 1177 |
+
"learning_rate": 7.874545454545454e-06,
|
| 1178 |
+
"loss": 0.026,
|
| 1179 |
"step": 1670
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.67106051527861,
|
| 1183 |
+
"grad_norm": 2.880535364151001,
|
| 1184 |
+
"learning_rate": 7.856363636363638e-06,
|
| 1185 |
+
"loss": 0.023,
|
| 1186 |
"step": 1680
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 0.6750549231076493,
|
| 1190 |
+
"grad_norm": 3.375427007675171,
|
| 1191 |
+
"learning_rate": 7.83818181818182e-06,
|
| 1192 |
+
"loss": 0.025,
|
| 1193 |
"step": 1690
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.6790493309366886,
|
| 1197 |
+
"grad_norm": 1.8976999521255493,
|
| 1198 |
+
"learning_rate": 7.820000000000001e-06,
|
| 1199 |
+
"loss": 0.0267,
|
| 1200 |
"step": 1700
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.683043738765728,
|
| 1204 |
+
"grad_norm": 3.0969624519348145,
|
| 1205 |
+
"learning_rate": 7.801818181818183e-06,
|
| 1206 |
+
"loss": 0.0334,
|
| 1207 |
"step": 1710
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.6870381465947674,
|
| 1211 |
+
"grad_norm": 2.949564218521118,
|
| 1212 |
+
"learning_rate": 7.783636363636365e-06,
|
| 1213 |
+
"loss": 0.0317,
|
| 1214 |
"step": 1720
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.6910325544238066,
|
| 1218 |
+
"grad_norm": 3.1288681030273438,
|
| 1219 |
+
"learning_rate": 7.765454545454545e-06,
|
| 1220 |
+
"loss": 0.036,
|
| 1221 |
"step": 1730
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.695026962252846,
|
| 1225 |
+
"grad_norm": 2.337353229522705,
|
| 1226 |
+
"learning_rate": 7.747272727272727e-06,
|
| 1227 |
+
"loss": 0.0233,
|
| 1228 |
"step": 1740
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.6990213700818854,
|
| 1232 |
+
"grad_norm": 2.593578577041626,
|
| 1233 |
+
"learning_rate": 7.729090909090909e-06,
|
| 1234 |
+
"loss": 0.0338,
|
| 1235 |
"step": 1750
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.7030157779109247,
|
| 1239 |
+
"grad_norm": 2.1465232372283936,
|
| 1240 |
+
"learning_rate": 7.71090909090909e-06,
|
| 1241 |
+
"loss": 0.0289,
|
| 1242 |
"step": 1760
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.7070101857399641,
|
| 1246 |
+
"grad_norm": 2.932419538497925,
|
| 1247 |
+
"learning_rate": 7.692727272727273e-06,
|
| 1248 |
+
"loss": 0.0263,
|
| 1249 |
"step": 1770
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.7110045935690034,
|
| 1253 |
+
"grad_norm": 4.221918106079102,
|
| 1254 |
+
"learning_rate": 7.674545454545456e-06,
|
| 1255 |
+
"loss": 0.0262,
|
| 1256 |
"step": 1780
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.7149990013980427,
|
| 1260 |
+
"grad_norm": 3.3574588298797607,
|
| 1261 |
+
"learning_rate": 7.656363636363638e-06,
|
| 1262 |
+
"loss": 0.0298,
|
| 1263 |
"step": 1790
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.7189934092270821,
|
| 1267 |
+
"grad_norm": 3.61956787109375,
|
| 1268 |
+
"learning_rate": 7.63818181818182e-06,
|
| 1269 |
+
"loss": 0.0361,
|
| 1270 |
"step": 1800
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.7229878170561215,
|
| 1274 |
+
"grad_norm": 3.8640527725219727,
|
| 1275 |
+
"learning_rate": 7.620000000000001e-06,
|
| 1276 |
+
"loss": 0.0379,
|
| 1277 |
"step": 1810
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.7269822248851607,
|
| 1281 |
+
"grad_norm": 3.021359920501709,
|
| 1282 |
+
"learning_rate": 7.6018181818181826e-06,
|
| 1283 |
+
"loss": 0.0256,
|
| 1284 |
"step": 1820
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.7309766327142001,
|
| 1288 |
+
"grad_norm": 3.4771692752838135,
|
| 1289 |
+
"learning_rate": 7.583636363636364e-06,
|
| 1290 |
+
"loss": 0.0317,
|
| 1291 |
"step": 1830
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.7349710405432395,
|
| 1295 |
+
"grad_norm": 1.799800992012024,
|
| 1296 |
+
"learning_rate": 7.565454545454546e-06,
|
| 1297 |
+
"loss": 0.0318,
|
| 1298 |
"step": 1840
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.7389654483722788,
|
| 1302 |
+
"grad_norm": 2.8619134426116943,
|
| 1303 |
+
"learning_rate": 7.547272727272727e-06,
|
| 1304 |
+
"loss": 0.0351,
|
| 1305 |
"step": 1850
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.7429598562013181,
|
| 1309 |
+
"grad_norm": 1.9078686237335205,
|
| 1310 |
+
"learning_rate": 7.529090909090909e-06,
|
| 1311 |
+
"loss": 0.0223,
|
| 1312 |
"step": 1860
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.7469542640303575,
|
| 1316 |
+
"grad_norm": 3.3508458137512207,
|
| 1317 |
+
"learning_rate": 7.510909090909092e-06,
|
| 1318 |
+
"loss": 0.0219,
|
| 1319 |
"step": 1870
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.7509486718593968,
|
| 1323 |
+
"grad_norm": 2.5900681018829346,
|
| 1324 |
+
"learning_rate": 7.492727272727274e-06,
|
| 1325 |
+
"loss": 0.0231,
|
| 1326 |
"step": 1880
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.7549430796884362,
|
| 1330 |
+
"grad_norm": 2.878042697906494,
|
| 1331 |
+
"learning_rate": 7.4745454545454554e-06,
|
| 1332 |
+
"loss": 0.0269,
|
| 1333 |
"step": 1890
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.7589374875174756,
|
| 1337 |
+
"grad_norm": 2.814326524734497,
|
| 1338 |
+
"learning_rate": 7.456363636363637e-06,
|
| 1339 |
+
"loss": 0.027,
|
| 1340 |
"step": 1900
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.7629318953465148,
|
| 1344 |
+
"grad_norm": 3.0135231018066406,
|
| 1345 |
+
"learning_rate": 7.438181818181819e-06,
|
| 1346 |
+
"loss": 0.0363,
|
| 1347 |
"step": 1910
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 0.7669263031755542,
|
| 1351 |
+
"grad_norm": 4.630500793457031,
|
| 1352 |
+
"learning_rate": 7.420000000000001e-06,
|
| 1353 |
+
"loss": 0.0318,
|
| 1354 |
"step": 1920
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 0.7709207110045936,
|
| 1358 |
+
"grad_norm": 2.1792924404144287,
|
| 1359 |
+
"learning_rate": 7.401818181818182e-06,
|
| 1360 |
+
"loss": 0.0299,
|
| 1361 |
"step": 1930
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 0.774915118833633,
|
| 1365 |
+
"grad_norm": 2.716294527053833,
|
| 1366 |
+
"learning_rate": 7.383636363636364e-06,
|
| 1367 |
+
"loss": 0.028,
|
| 1368 |
"step": 1940
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 0.7789095266626722,
|
| 1372 |
+
"grad_norm": 2.591440439224243,
|
| 1373 |
+
"learning_rate": 7.365454545454546e-06,
|
| 1374 |
+
"loss": 0.0311,
|
| 1375 |
"step": 1950
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 0.7829039344917116,
|
| 1379 |
+
"grad_norm": 1.3216569423675537,
|
| 1380 |
+
"learning_rate": 7.3472727272727275e-06,
|
| 1381 |
+
"loss": 0.0213,
|
| 1382 |
"step": 1960
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 0.786898342320751,
|
| 1386 |
+
"grad_norm": 1.4774867296218872,
|
| 1387 |
+
"learning_rate": 7.32909090909091e-06,
|
| 1388 |
+
"loss": 0.0218,
|
| 1389 |
"step": 1970
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 0.7908927501497903,
|
| 1393 |
+
"grad_norm": 2.8911545276641846,
|
| 1394 |
+
"learning_rate": 7.310909090909092e-06,
|
| 1395 |
+
"loss": 0.0251,
|
| 1396 |
"step": 1980
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 0.7948871579788296,
|
| 1400 |
+
"grad_norm": 4.558359146118164,
|
| 1401 |
+
"learning_rate": 7.292727272727274e-06,
|
| 1402 |
+
"loss": 0.023,
|
| 1403 |
"step": 1990
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 0.798881565807869,
|
| 1407 |
+
"grad_norm": 2.78124737739563,
|
| 1408 |
+
"learning_rate": 7.274545454545456e-06,
|
| 1409 |
+
"loss": 0.0361,
|
| 1410 |
"step": 2000
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 0.798881565807869,
|
| 1414 |
+
"eval_loss": 0.017742320895195007,
|
| 1415 |
+
"eval_runtime": 7487.3562,
|
| 1416 |
+
"eval_samples_per_second": 2.675,
|
| 1417 |
+
"eval_steps_per_second": 0.334,
|
| 1418 |
+
"eval_wer": 1.5431796145591616,
|
| 1419 |
"step": 2000
|
| 1420 |
}
|
| 1421 |
],
|
| 1422 |
"logging_steps": 10,
|
| 1423 |
+
"max_steps": 6000,
|
| 1424 |
"num_input_tokens_seen": 0,
|
| 1425 |
"num_train_epochs": 3,
|
| 1426 |
"save_steps": 2000,
|
checkpoint-2000/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5905
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28019e08a9b2d84c03558928082cffce47629e113d3b3bcc740699fd5598b51a
|
| 3 |
size 5905
|
checkpoint-4000/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 966995080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d52ac480043f4af8dbb549281e01b226e2dafd2812dfcf79e84e9e58b0a873e
|
| 3 |
size 966995080
|
checkpoint-4000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1925071165
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83af9706c621501c0fb3d7b4a5db5537c327c3708a3067941a8e01a8e699f6c7
|
| 3 |
size 1925071165
|
checkpoint-4000/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:146f30cf11dcf27c81b44d556a85c036910889abbbbcf6c384f44964f5f8aeae
|
| 3 |
size 1383
|
checkpoint-4000/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:443ff1716f789088883efb7cf95512b1a8051d71d311454c17b75cb74f54c32d
|
| 3 |
size 1465
|
checkpoint-4000/trainer_state.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4000/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5905
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28019e08a9b2d84c03558928082cffce47629e113d3b3bcc740699fd5598b51a
|
| 3 |
size 5905
|
checkpoint-6000/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 966995080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c17719c258969e39f9360d0c70da5811b807ca5ab4aec4f31e40141c4c529d6
|
| 3 |
size 966995080
|
checkpoint-6000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1925071165
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa6cdf665dc69fdbfe1898c77ae59f54c35f8756893415ad793befbf261542b7
|
| 3 |
size 1925071165
|
checkpoint-6000/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6636c06535c8fca3968cc19bdf49014057579ef4a50067c1ae414ab16b925c24
|
| 3 |
size 1383
|
checkpoint-6000/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:245abc9317a42ca88784672abfa1666a766e488ecd0c89ad77b0cb3f023e808e
|
| 3 |
size 1465
|
checkpoint-6000/trainer_state.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-6000/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5905
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28019e08a9b2d84c03558928082cffce47629e113d3b3bcc740699fd5598b51a
|
| 3 |
size 5905
|