Delete checkpoint-471/trainer_state.json
Browse files
checkpoint-471/trainer_state.json
DELETED
|
@@ -1,738 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": null,
|
| 3 |
-
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 3.0,
|
| 5 |
-
"eval_steps": 10,
|
| 6 |
-
"global_step": 471,
|
| 7 |
-
"is_hyper_param_search": false,
|
| 8 |
-
"is_local_process_zero": true,
|
| 9 |
-
"is_world_process_zero": true,
|
| 10 |
-
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.06369426751592357,
|
| 13 |
-
"grad_norm": 5.655877590179443,
|
| 14 |
-
"learning_rate": 4.893842887473461e-05,
|
| 15 |
-
"loss": 0.9888,
|
| 16 |
-
"step": 10
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.06369426751592357,
|
| 20 |
-
"eval_loss": 0.6204623579978943,
|
| 21 |
-
"eval_runtime": 260.4945,
|
| 22 |
-
"eval_samples_per_second": 1.332,
|
| 23 |
-
"eval_steps_per_second": 0.334,
|
| 24 |
-
"step": 10
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.12738853503184713,
|
| 28 |
-
"grad_norm": 5.2134904861450195,
|
| 29 |
-
"learning_rate": 4.787685774946922e-05,
|
| 30 |
-
"loss": 0.6134,
|
| 31 |
-
"step": 20
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.12738853503184713,
|
| 35 |
-
"eval_loss": 0.5212348699569702,
|
| 36 |
-
"eval_runtime": 261.046,
|
| 37 |
-
"eval_samples_per_second": 1.329,
|
| 38 |
-
"eval_steps_per_second": 0.333,
|
| 39 |
-
"step": 20
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"epoch": 0.1910828025477707,
|
| 43 |
-
"grad_norm": 6.658833026885986,
|
| 44 |
-
"learning_rate": 4.681528662420383e-05,
|
| 45 |
-
"loss": 0.5704,
|
| 46 |
-
"step": 30
|
| 47 |
-
},
|
| 48 |
-
{
|
| 49 |
-
"epoch": 0.1910828025477707,
|
| 50 |
-
"eval_loss": 0.5708295106887817,
|
| 51 |
-
"eval_runtime": 266.4556,
|
| 52 |
-
"eval_samples_per_second": 1.302,
|
| 53 |
-
"eval_steps_per_second": 0.327,
|
| 54 |
-
"step": 30
|
| 55 |
-
},
|
| 56 |
-
{
|
| 57 |
-
"epoch": 0.25477707006369427,
|
| 58 |
-
"grad_norm": 4.93190336227417,
|
| 59 |
-
"learning_rate": 4.575371549893843e-05,
|
| 60 |
-
"loss": 0.5118,
|
| 61 |
-
"step": 40
|
| 62 |
-
},
|
| 63 |
-
{
|
| 64 |
-
"epoch": 0.25477707006369427,
|
| 65 |
-
"eval_loss": 0.45317262411117554,
|
| 66 |
-
"eval_runtime": 263.8344,
|
| 67 |
-
"eval_samples_per_second": 1.315,
|
| 68 |
-
"eval_steps_per_second": 0.33,
|
| 69 |
-
"step": 40
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.3184713375796178,
|
| 73 |
-
"grad_norm": 6.612313747406006,
|
| 74 |
-
"learning_rate": 4.469214437367304e-05,
|
| 75 |
-
"loss": 0.5768,
|
| 76 |
-
"step": 50
|
| 77 |
-
},
|
| 78 |
-
{
|
| 79 |
-
"epoch": 0.3184713375796178,
|
| 80 |
-
"eval_loss": 0.5027761459350586,
|
| 81 |
-
"eval_runtime": 317.3646,
|
| 82 |
-
"eval_samples_per_second": 1.093,
|
| 83 |
-
"eval_steps_per_second": 0.274,
|
| 84 |
-
"step": 50
|
| 85 |
-
},
|
| 86 |
-
{
|
| 87 |
-
"epoch": 0.3821656050955414,
|
| 88 |
-
"grad_norm": 6.43605375289917,
|
| 89 |
-
"learning_rate": 4.3630573248407646e-05,
|
| 90 |
-
"loss": 0.4945,
|
| 91 |
-
"step": 60
|
| 92 |
-
},
|
| 93 |
-
{
|
| 94 |
-
"epoch": 0.3821656050955414,
|
| 95 |
-
"eval_loss": 0.45588889718055725,
|
| 96 |
-
"eval_runtime": 269.9949,
|
| 97 |
-
"eval_samples_per_second": 1.285,
|
| 98 |
-
"eval_steps_per_second": 0.322,
|
| 99 |
-
"step": 60
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.445859872611465,
|
| 103 |
-
"grad_norm": 3.744384527206421,
|
| 104 |
-
"learning_rate": 4.256900212314226e-05,
|
| 105 |
-
"loss": 0.4173,
|
| 106 |
-
"step": 70
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 0.445859872611465,
|
| 110 |
-
"eval_loss": 0.44179391860961914,
|
| 111 |
-
"eval_runtime": 263.2415,
|
| 112 |
-
"eval_samples_per_second": 1.318,
|
| 113 |
-
"eval_steps_per_second": 0.33,
|
| 114 |
-
"step": 70
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.5095541401273885,
|
| 118 |
-
"grad_norm": 4.705707550048828,
|
| 119 |
-
"learning_rate": 4.150743099787686e-05,
|
| 120 |
-
"loss": 0.4277,
|
| 121 |
-
"step": 80
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.5095541401273885,
|
| 125 |
-
"eval_loss": 0.37640380859375,
|
| 126 |
-
"eval_runtime": 264.4226,
|
| 127 |
-
"eval_samples_per_second": 1.312,
|
| 128 |
-
"eval_steps_per_second": 0.329,
|
| 129 |
-
"step": 80
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.5732484076433121,
|
| 133 |
-
"grad_norm": 2.7133395671844482,
|
| 134 |
-
"learning_rate": 4.044585987261147e-05,
|
| 135 |
-
"loss": 0.4374,
|
| 136 |
-
"step": 90
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.5732484076433121,
|
| 140 |
-
"eval_loss": 0.35762831568717957,
|
| 141 |
-
"eval_runtime": 264.9493,
|
| 142 |
-
"eval_samples_per_second": 1.31,
|
| 143 |
-
"eval_steps_per_second": 0.328,
|
| 144 |
-
"step": 90
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"epoch": 0.6369426751592356,
|
| 148 |
-
"grad_norm": 4.4099650382995605,
|
| 149 |
-
"learning_rate": 3.9384288747346076e-05,
|
| 150 |
-
"loss": 0.3717,
|
| 151 |
-
"step": 100
|
| 152 |
-
},
|
| 153 |
-
{
|
| 154 |
-
"epoch": 0.6369426751592356,
|
| 155 |
-
"eval_loss": 0.35183703899383545,
|
| 156 |
-
"eval_runtime": 278.8216,
|
| 157 |
-
"eval_samples_per_second": 1.245,
|
| 158 |
-
"eval_steps_per_second": 0.312,
|
| 159 |
-
"step": 100
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.7006369426751592,
|
| 163 |
-
"grad_norm": 15.860236167907715,
|
| 164 |
-
"learning_rate": 3.8322717622080686e-05,
|
| 165 |
-
"loss": 0.4415,
|
| 166 |
-
"step": 110
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 0.7006369426751592,
|
| 170 |
-
"eval_loss": 0.6824913620948792,
|
| 171 |
-
"eval_runtime": 281.0438,
|
| 172 |
-
"eval_samples_per_second": 1.235,
|
| 173 |
-
"eval_steps_per_second": 0.31,
|
| 174 |
-
"step": 110
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 0.7643312101910829,
|
| 178 |
-
"grad_norm": 5.011375427246094,
|
| 179 |
-
"learning_rate": 3.7261146496815283e-05,
|
| 180 |
-
"loss": 0.3639,
|
| 181 |
-
"step": 120
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 0.7643312101910829,
|
| 185 |
-
"eval_loss": 0.3454379737377167,
|
| 186 |
-
"eval_runtime": 281.28,
|
| 187 |
-
"eval_samples_per_second": 1.234,
|
| 188 |
-
"eval_steps_per_second": 0.309,
|
| 189 |
-
"step": 120
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.8280254777070064,
|
| 193 |
-
"grad_norm": 3.8196499347686768,
|
| 194 |
-
"learning_rate": 3.6199575371549894e-05,
|
| 195 |
-
"loss": 0.3604,
|
| 196 |
-
"step": 130
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 0.8280254777070064,
|
| 200 |
-
"eval_loss": 0.33031439781188965,
|
| 201 |
-
"eval_runtime": 266.1981,
|
| 202 |
-
"eval_samples_per_second": 1.304,
|
| 203 |
-
"eval_steps_per_second": 0.327,
|
| 204 |
-
"step": 130
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 0.89171974522293,
|
| 208 |
-
"grad_norm": 8.475601196289062,
|
| 209 |
-
"learning_rate": 3.51380042462845e-05,
|
| 210 |
-
"loss": 0.3782,
|
| 211 |
-
"step": 140
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.89171974522293,
|
| 215 |
-
"eval_loss": 0.33289575576782227,
|
| 216 |
-
"eval_runtime": 265.8273,
|
| 217 |
-
"eval_samples_per_second": 1.305,
|
| 218 |
-
"eval_steps_per_second": 0.327,
|
| 219 |
-
"step": 140
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.9554140127388535,
|
| 223 |
-
"grad_norm": 4.658049583435059,
|
| 224 |
-
"learning_rate": 3.407643312101911e-05,
|
| 225 |
-
"loss": 0.36,
|
| 226 |
-
"step": 150
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.9554140127388535,
|
| 230 |
-
"eval_loss": 0.33607012033462524,
|
| 231 |
-
"eval_runtime": 277.0995,
|
| 232 |
-
"eval_samples_per_second": 1.252,
|
| 233 |
-
"eval_steps_per_second": 0.314,
|
| 234 |
-
"step": 150
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 1.019108280254777,
|
| 238 |
-
"grad_norm": 2.7650649547576904,
|
| 239 |
-
"learning_rate": 3.301486199575371e-05,
|
| 240 |
-
"loss": 0.3244,
|
| 241 |
-
"step": 160
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 1.019108280254777,
|
| 245 |
-
"eval_loss": 0.3019406497478485,
|
| 246 |
-
"eval_runtime": 281.1357,
|
| 247 |
-
"eval_samples_per_second": 1.234,
|
| 248 |
-
"eval_steps_per_second": 0.309,
|
| 249 |
-
"step": 160
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 1.0828025477707006,
|
| 253 |
-
"grad_norm": 7.399433612823486,
|
| 254 |
-
"learning_rate": 3.1953290870488323e-05,
|
| 255 |
-
"loss": 0.3337,
|
| 256 |
-
"step": 170
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 1.0828025477707006,
|
| 260 |
-
"eval_loss": 0.33017274737358093,
|
| 261 |
-
"eval_runtime": 279.943,
|
| 262 |
-
"eval_samples_per_second": 1.24,
|
| 263 |
-
"eval_steps_per_second": 0.311,
|
| 264 |
-
"step": 170
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"epoch": 1.1464968152866242,
|
| 268 |
-
"grad_norm": 6.1573567390441895,
|
| 269 |
-
"learning_rate": 3.089171974522293e-05,
|
| 270 |
-
"loss": 0.2967,
|
| 271 |
-
"step": 180
|
| 272 |
-
},
|
| 273 |
-
{
|
| 274 |
-
"epoch": 1.1464968152866242,
|
| 275 |
-
"eval_loss": 0.286271870136261,
|
| 276 |
-
"eval_runtime": 279.1484,
|
| 277 |
-
"eval_samples_per_second": 1.243,
|
| 278 |
-
"eval_steps_per_second": 0.312,
|
| 279 |
-
"step": 180
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 1.2101910828025477,
|
| 283 |
-
"grad_norm": 5.677030086517334,
|
| 284 |
-
"learning_rate": 2.9830148619957538e-05,
|
| 285 |
-
"loss": 0.3236,
|
| 286 |
-
"step": 190
|
| 287 |
-
},
|
| 288 |
-
{
|
| 289 |
-
"epoch": 1.2101910828025477,
|
| 290 |
-
"eval_loss": 0.3013075590133667,
|
| 291 |
-
"eval_runtime": 280.1463,
|
| 292 |
-
"eval_samples_per_second": 1.239,
|
| 293 |
-
"eval_steps_per_second": 0.311,
|
| 294 |
-
"step": 190
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 1.2738853503184713,
|
| 298 |
-
"grad_norm": 3.6576144695281982,
|
| 299 |
-
"learning_rate": 2.8768577494692145e-05,
|
| 300 |
-
"loss": 0.2848,
|
| 301 |
-
"step": 200
|
| 302 |
-
},
|
| 303 |
-
{
|
| 304 |
-
"epoch": 1.2738853503184713,
|
| 305 |
-
"eval_loss": 0.28381872177124023,
|
| 306 |
-
"eval_runtime": 278.4374,
|
| 307 |
-
"eval_samples_per_second": 1.246,
|
| 308 |
-
"eval_steps_per_second": 0.312,
|
| 309 |
-
"step": 200
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 1.3375796178343948,
|
| 313 |
-
"grad_norm": 3.5609078407287598,
|
| 314 |
-
"learning_rate": 2.7707006369426753e-05,
|
| 315 |
-
"loss": 0.2138,
|
| 316 |
-
"step": 210
|
| 317 |
-
},
|
| 318 |
-
{
|
| 319 |
-
"epoch": 1.3375796178343948,
|
| 320 |
-
"eval_loss": 0.27076515555381775,
|
| 321 |
-
"eval_runtime": 269.6079,
|
| 322 |
-
"eval_samples_per_second": 1.287,
|
| 323 |
-
"eval_steps_per_second": 0.323,
|
| 324 |
-
"step": 210
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 1.4012738853503186,
|
| 328 |
-
"grad_norm": 5.47592306137085,
|
| 329 |
-
"learning_rate": 2.664543524416136e-05,
|
| 330 |
-
"loss": 0.2646,
|
| 331 |
-
"step": 220
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 1.4012738853503186,
|
| 335 |
-
"eval_loss": 0.2966860830783844,
|
| 336 |
-
"eval_runtime": 267.6504,
|
| 337 |
-
"eval_samples_per_second": 1.296,
|
| 338 |
-
"eval_steps_per_second": 0.325,
|
| 339 |
-
"step": 220
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.4649681528662422,
|
| 343 |
-
"grad_norm": 5.770505428314209,
|
| 344 |
-
"learning_rate": 2.5583864118895967e-05,
|
| 345 |
-
"loss": 0.2784,
|
| 346 |
-
"step": 230
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 1.4649681528662422,
|
| 350 |
-
"eval_loss": 0.25654563307762146,
|
| 351 |
-
"eval_runtime": 270.188,
|
| 352 |
-
"eval_samples_per_second": 1.284,
|
| 353 |
-
"eval_steps_per_second": 0.322,
|
| 354 |
-
"step": 230
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 1.5286624203821657,
|
| 358 |
-
"grad_norm": 9.245145797729492,
|
| 359 |
-
"learning_rate": 2.4522292993630575e-05,
|
| 360 |
-
"loss": 0.3004,
|
| 361 |
-
"step": 240
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 1.5286624203821657,
|
| 365 |
-
"eval_loss": 0.2588840425014496,
|
| 366 |
-
"eval_runtime": 270.9119,
|
| 367 |
-
"eval_samples_per_second": 1.281,
|
| 368 |
-
"eval_steps_per_second": 0.321,
|
| 369 |
-
"step": 240
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 1.5923566878980893,
|
| 373 |
-
"grad_norm": 6.825544834136963,
|
| 374 |
-
"learning_rate": 2.3460721868365182e-05,
|
| 375 |
-
"loss": 0.2585,
|
| 376 |
-
"step": 250
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 1.5923566878980893,
|
| 380 |
-
"eval_loss": 0.2777319848537445,
|
| 381 |
-
"eval_runtime": 271.4967,
|
| 382 |
-
"eval_samples_per_second": 1.278,
|
| 383 |
-
"eval_steps_per_second": 0.32,
|
| 384 |
-
"step": 250
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 1.6560509554140128,
|
| 388 |
-
"grad_norm": 5.504091262817383,
|
| 389 |
-
"learning_rate": 2.239915074309979e-05,
|
| 390 |
-
"loss": 0.239,
|
| 391 |
-
"step": 260
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 1.6560509554140128,
|
| 395 |
-
"eval_loss": 0.2845991551876068,
|
| 396 |
-
"eval_runtime": 275.4719,
|
| 397 |
-
"eval_samples_per_second": 1.26,
|
| 398 |
-
"eval_steps_per_second": 0.316,
|
| 399 |
-
"step": 260
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 1.7197452229299364,
|
| 403 |
-
"grad_norm": 1.7976171970367432,
|
| 404 |
-
"learning_rate": 2.1337579617834397e-05,
|
| 405 |
-
"loss": 0.324,
|
| 406 |
-
"step": 270
|
| 407 |
-
},
|
| 408 |
-
{
|
| 409 |
-
"epoch": 1.7197452229299364,
|
| 410 |
-
"eval_loss": 0.2418244183063507,
|
| 411 |
-
"eval_runtime": 273.3976,
|
| 412 |
-
"eval_samples_per_second": 1.269,
|
| 413 |
-
"eval_steps_per_second": 0.318,
|
| 414 |
-
"step": 270
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 1.78343949044586,
|
| 418 |
-
"grad_norm": 7.518586158752441,
|
| 419 |
-
"learning_rate": 2.0276008492569004e-05,
|
| 420 |
-
"loss": 0.2661,
|
| 421 |
-
"step": 280
|
| 422 |
-
},
|
| 423 |
-
{
|
| 424 |
-
"epoch": 1.78343949044586,
|
| 425 |
-
"eval_loss": 0.2677413821220398,
|
| 426 |
-
"eval_runtime": 273.2159,
|
| 427 |
-
"eval_samples_per_second": 1.27,
|
| 428 |
-
"eval_steps_per_second": 0.318,
|
| 429 |
-
"step": 280
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 1.8471337579617835,
|
| 433 |
-
"grad_norm": 3.743008613586426,
|
| 434 |
-
"learning_rate": 1.921443736730361e-05,
|
| 435 |
-
"loss": 0.2763,
|
| 436 |
-
"step": 290
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 1.8471337579617835,
|
| 440 |
-
"eval_loss": 0.2228710651397705,
|
| 441 |
-
"eval_runtime": 272.9055,
|
| 442 |
-
"eval_samples_per_second": 1.272,
|
| 443 |
-
"eval_steps_per_second": 0.319,
|
| 444 |
-
"step": 290
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 1.910828025477707,
|
| 448 |
-
"grad_norm": 7.198996543884277,
|
| 449 |
-
"learning_rate": 1.8152866242038215e-05,
|
| 450 |
-
"loss": 0.1957,
|
| 451 |
-
"step": 300
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 1.910828025477707,
|
| 455 |
-
"eval_loss": 0.25310221314430237,
|
| 456 |
-
"eval_runtime": 272.8452,
|
| 457 |
-
"eval_samples_per_second": 1.272,
|
| 458 |
-
"eval_steps_per_second": 0.319,
|
| 459 |
-
"step": 300
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 1.9745222929936306,
|
| 463 |
-
"grad_norm": 7.609084606170654,
|
| 464 |
-
"learning_rate": 1.7091295116772823e-05,
|
| 465 |
-
"loss": 0.251,
|
| 466 |
-
"step": 310
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 1.9745222929936306,
|
| 470 |
-
"eval_loss": 0.250274658203125,
|
| 471 |
-
"eval_runtime": 274.2439,
|
| 472 |
-
"eval_samples_per_second": 1.265,
|
| 473 |
-
"eval_steps_per_second": 0.317,
|
| 474 |
-
"step": 310
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 2.038216560509554,
|
| 478 |
-
"grad_norm": 5.631246566772461,
|
| 479 |
-
"learning_rate": 1.602972399150743e-05,
|
| 480 |
-
"loss": 0.2124,
|
| 481 |
-
"step": 320
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 2.038216560509554,
|
| 485 |
-
"eval_loss": 0.23672077059745789,
|
| 486 |
-
"eval_runtime": 273.0475,
|
| 487 |
-
"eval_samples_per_second": 1.271,
|
| 488 |
-
"eval_steps_per_second": 0.319,
|
| 489 |
-
"step": 320
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 2.1019108280254777,
|
| 493 |
-
"grad_norm": 8.399256706237793,
|
| 494 |
-
"learning_rate": 1.4968152866242039e-05,
|
| 495 |
-
"loss": 0.1832,
|
| 496 |
-
"step": 330
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 2.1019108280254777,
|
| 500 |
-
"eval_loss": 0.2170046865940094,
|
| 501 |
-
"eval_runtime": 274.911,
|
| 502 |
-
"eval_samples_per_second": 1.262,
|
| 503 |
-
"eval_steps_per_second": 0.316,
|
| 504 |
-
"step": 330
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 2.1656050955414012,
|
| 508 |
-
"grad_norm": 6.787008285522461,
|
| 509 |
-
"learning_rate": 1.3906581740976646e-05,
|
| 510 |
-
"loss": 0.1968,
|
| 511 |
-
"step": 340
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 2.1656050955414012,
|
| 515 |
-
"eval_loss": 0.25750404596328735,
|
| 516 |
-
"eval_runtime": 274.5796,
|
| 517 |
-
"eval_samples_per_second": 1.264,
|
| 518 |
-
"eval_steps_per_second": 0.317,
|
| 519 |
-
"step": 340
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 2.229299363057325,
|
| 523 |
-
"grad_norm": 4.561034202575684,
|
| 524 |
-
"learning_rate": 1.2845010615711253e-05,
|
| 525 |
-
"loss": 0.2169,
|
| 526 |
-
"step": 350
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 2.229299363057325,
|
| 530 |
-
"eval_loss": 0.21437929570674896,
|
| 531 |
-
"eval_runtime": 274.1716,
|
| 532 |
-
"eval_samples_per_second": 1.266,
|
| 533 |
-
"eval_steps_per_second": 0.317,
|
| 534 |
-
"step": 350
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 2.2929936305732483,
|
| 538 |
-
"grad_norm": 2.4580085277557373,
|
| 539 |
-
"learning_rate": 1.178343949044586e-05,
|
| 540 |
-
"loss": 0.1645,
|
| 541 |
-
"step": 360
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 2.2929936305732483,
|
| 545 |
-
"eval_loss": 0.21549557149410248,
|
| 546 |
-
"eval_runtime": 274.4038,
|
| 547 |
-
"eval_samples_per_second": 1.265,
|
| 548 |
-
"eval_steps_per_second": 0.317,
|
| 549 |
-
"step": 360
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 2.356687898089172,
|
| 553 |
-
"grad_norm": 5.2749786376953125,
|
| 554 |
-
"learning_rate": 1.0721868365180468e-05,
|
| 555 |
-
"loss": 0.1292,
|
| 556 |
-
"step": 370
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 2.356687898089172,
|
| 560 |
-
"eval_loss": 0.24459926784038544,
|
| 561 |
-
"eval_runtime": 275.4779,
|
| 562 |
-
"eval_samples_per_second": 1.26,
|
| 563 |
-
"eval_steps_per_second": 0.316,
|
| 564 |
-
"step": 370
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 2.4203821656050954,
|
| 568 |
-
"grad_norm": 11.909132957458496,
|
| 569 |
-
"learning_rate": 9.660297239915075e-06,
|
| 570 |
-
"loss": 0.2064,
|
| 571 |
-
"step": 380
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 2.4203821656050954,
|
| 575 |
-
"eval_loss": 0.2202015370130539,
|
| 576 |
-
"eval_runtime": 276.3076,
|
| 577 |
-
"eval_samples_per_second": 1.256,
|
| 578 |
-
"eval_steps_per_second": 0.315,
|
| 579 |
-
"step": 380
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 2.484076433121019,
|
| 583 |
-
"grad_norm": 8.857806205749512,
|
| 584 |
-
"learning_rate": 8.598726114649681e-06,
|
| 585 |
-
"loss": 0.2967,
|
| 586 |
-
"step": 390
|
| 587 |
-
},
|
| 588 |
-
{
|
| 589 |
-
"epoch": 2.484076433121019,
|
| 590 |
-
"eval_loss": 0.2075081467628479,
|
| 591 |
-
"eval_runtime": 275.5237,
|
| 592 |
-
"eval_samples_per_second": 1.259,
|
| 593 |
-
"eval_steps_per_second": 0.316,
|
| 594 |
-
"step": 390
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 2.5477707006369426,
|
| 598 |
-
"grad_norm": 8.795026779174805,
|
| 599 |
-
"learning_rate": 7.537154989384289e-06,
|
| 600 |
-
"loss": 0.2349,
|
| 601 |
-
"step": 400
|
| 602 |
-
},
|
| 603 |
-
{
|
| 604 |
-
"epoch": 2.5477707006369426,
|
| 605 |
-
"eval_loss": 0.2141169160604477,
|
| 606 |
-
"eval_runtime": 274.8429,
|
| 607 |
-
"eval_samples_per_second": 1.263,
|
| 608 |
-
"eval_steps_per_second": 0.317,
|
| 609 |
-
"step": 400
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 2.611464968152866,
|
| 613 |
-
"grad_norm": 4.121735572814941,
|
| 614 |
-
"learning_rate": 6.4755838641188965e-06,
|
| 615 |
-
"loss": 0.161,
|
| 616 |
-
"step": 410
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 2.611464968152866,
|
| 620 |
-
"eval_loss": 0.2091439962387085,
|
| 621 |
-
"eval_runtime": 275.865,
|
| 622 |
-
"eval_samples_per_second": 1.258,
|
| 623 |
-
"eval_steps_per_second": 0.315,
|
| 624 |
-
"step": 410
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 2.6751592356687897,
|
| 628 |
-
"grad_norm": 7.995586395263672,
|
| 629 |
-
"learning_rate": 5.414012738853504e-06,
|
| 630 |
-
"loss": 0.1786,
|
| 631 |
-
"step": 420
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 2.6751592356687897,
|
| 635 |
-
"eval_loss": 0.21283170580863953,
|
| 636 |
-
"eval_runtime": 274.9723,
|
| 637 |
-
"eval_samples_per_second": 1.262,
|
| 638 |
-
"eval_steps_per_second": 0.316,
|
| 639 |
-
"step": 420
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 2.738853503184713,
|
| 643 |
-
"grad_norm": 6.569371700286865,
|
| 644 |
-
"learning_rate": 4.35244161358811e-06,
|
| 645 |
-
"loss": 0.2665,
|
| 646 |
-
"step": 430
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 2.738853503184713,
|
| 650 |
-
"eval_loss": 0.19985607266426086,
|
| 651 |
-
"eval_runtime": 275.6861,
|
| 652 |
-
"eval_samples_per_second": 1.259,
|
| 653 |
-
"eval_steps_per_second": 0.316,
|
| 654 |
-
"step": 430
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 2.802547770700637,
|
| 658 |
-
"grad_norm": 2.53202223777771,
|
| 659 |
-
"learning_rate": 3.2908704883227177e-06,
|
| 660 |
-
"loss": 0.1377,
|
| 661 |
-
"step": 440
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 2.802547770700637,
|
| 665 |
-
"eval_loss": 0.2025289386510849,
|
| 666 |
-
"eval_runtime": 275.6519,
|
| 667 |
-
"eval_samples_per_second": 1.259,
|
| 668 |
-
"eval_steps_per_second": 0.316,
|
| 669 |
-
"step": 440
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 2.8662420382165603,
|
| 673 |
-
"grad_norm": 2.821432590484619,
|
| 674 |
-
"learning_rate": 2.229299363057325e-06,
|
| 675 |
-
"loss": 0.2082,
|
| 676 |
-
"step": 450
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 2.8662420382165603,
|
| 680 |
-
"eval_loss": 0.20033259689807892,
|
| 681 |
-
"eval_runtime": 276.9112,
|
| 682 |
-
"eval_samples_per_second": 1.253,
|
| 683 |
-
"eval_steps_per_second": 0.314,
|
| 684 |
-
"step": 450
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 2.9299363057324843,
|
| 688 |
-
"grad_norm": 3.187981605529785,
|
| 689 |
-
"learning_rate": 1.167728237791932e-06,
|
| 690 |
-
"loss": 0.1408,
|
| 691 |
-
"step": 460
|
| 692 |
-
},
|
| 693 |
-
{
|
| 694 |
-
"epoch": 2.9299363057324843,
|
| 695 |
-
"eval_loss": 0.19696064293384552,
|
| 696 |
-
"eval_runtime": 276.0427,
|
| 697 |
-
"eval_samples_per_second": 1.257,
|
| 698 |
-
"eval_steps_per_second": 0.315,
|
| 699 |
-
"step": 460
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 2.9936305732484074,
|
| 703 |
-
"grad_norm": 3.8218815326690674,
|
| 704 |
-
"learning_rate": 1.0615711252653928e-07,
|
| 705 |
-
"loss": 0.1506,
|
| 706 |
-
"step": 470
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 2.9936305732484074,
|
| 710 |
-
"eval_loss": 0.19708757102489471,
|
| 711 |
-
"eval_runtime": 274.6555,
|
| 712 |
-
"eval_samples_per_second": 1.263,
|
| 713 |
-
"eval_steps_per_second": 0.317,
|
| 714 |
-
"step": 470
|
| 715 |
-
}
|
| 716 |
-
],
|
| 717 |
-
"logging_steps": 10,
|
| 718 |
-
"max_steps": 471,
|
| 719 |
-
"num_input_tokens_seen": 0,
|
| 720 |
-
"num_train_epochs": 3,
|
| 721 |
-
"save_steps": 500,
|
| 722 |
-
"stateful_callbacks": {
|
| 723 |
-
"TrainerControl": {
|
| 724 |
-
"args": {
|
| 725 |
-
"should_epoch_stop": false,
|
| 726 |
-
"should_evaluate": false,
|
| 727 |
-
"should_log": false,
|
| 728 |
-
"should_save": true,
|
| 729 |
-
"should_training_stop": true
|
| 730 |
-
},
|
| 731 |
-
"attributes": {}
|
| 732 |
-
}
|
| 733 |
-
},
|
| 734 |
-
"total_flos": 495041961535488.0,
|
| 735 |
-
"train_batch_size": 4,
|
| 736 |
-
"trial_name": null,
|
| 737 |
-
"trial_params": null
|
| 738 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|