Commit ·
4492e48
1
Parent(s): 40db8d8
Training in progress, step 1000, checkpoint
Browse files
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 995654149
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dfc0123cbba4f38d5cdfff68f2a0e41ff0ec9e6991bb31ff96c249568367da7b
|
| 3 |
size 995654149
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 497813341
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0dacf0a88a737dae9876a70e69ef790c67063ab7f11fc9897f2c2b60fd4134c6
|
| 3 |
size 497813341
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a207e92cacead3ab8860e639b0b71073ac9b7006ea3c6bb5a174842a663984e
|
| 3 |
size 14575
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:889074ce496d0433f5d3aaa744ef53c746b1953cf83f080c44d924a85d012261
|
| 3 |
size 627
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1319 +1,71 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.04,
|
| 13 |
-
"eval_loss": 0.
|
| 14 |
-
"eval_runtime":
|
| 15 |
-
"eval_samples_per_second":
|
| 16 |
-
"eval_steps_per_second": 3.
|
| 17 |
"step": 200
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.08,
|
| 21 |
-
"eval_loss":
|
| 22 |
-
"eval_runtime": 151.
|
| 23 |
-
"eval_samples_per_second":
|
| 24 |
-
"eval_steps_per_second": 3.
|
| 25 |
"step": 400
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.1,
|
| 29 |
-
"learning_rate": 4.
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.12,
|
| 35 |
-
"eval_loss": 0.
|
| 36 |
-
"eval_runtime":
|
| 37 |
-
"eval_samples_per_second":
|
| 38 |
-
"eval_steps_per_second": 3.
|
| 39 |
"step": 600
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.16,
|
| 43 |
-
"eval_loss": 0.
|
| 44 |
-
"eval_runtime": 151.
|
| 45 |
-
"eval_samples_per_second":
|
| 46 |
-
"eval_steps_per_second": 3.
|
| 47 |
"step": 800
|
| 48 |
},
|
| 49 |
{
|
| 50 |
"epoch": 0.2,
|
| 51 |
-
"learning_rate": 4.
|
| 52 |
-
"loss": 0.
|
| 53 |
"step": 1000
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.2,
|
| 57 |
-
"eval_loss": 0.
|
| 58 |
-
"eval_runtime":
|
| 59 |
-
"eval_samples_per_second":
|
| 60 |
-
"eval_steps_per_second": 3.706,
|
| 61 |
-
"step": 1000
|
| 62 |
-
},
|
| 63 |
-
{
|
| 64 |
-
"epoch": 0.24,
|
| 65 |
-
"eval_loss": 0.6123775243759155,
|
| 66 |
-
"eval_runtime": 151.1133,
|
| 67 |
-
"eval_samples_per_second": 948.533,
|
| 68 |
-
"eval_steps_per_second": 3.706,
|
| 69 |
-
"step": 1200
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.28,
|
| 73 |
-
"eval_loss": 0.6223951578140259,
|
| 74 |
-
"eval_runtime": 151.141,
|
| 75 |
-
"eval_samples_per_second": 948.36,
|
| 76 |
-
"eval_steps_per_second": 3.705,
|
| 77 |
-
"step": 1400
|
| 78 |
-
},
|
| 79 |
-
{
|
| 80 |
-
"epoch": 0.3,
|
| 81 |
-
"learning_rate": 4.7023218892637435e-05,
|
| 82 |
-
"loss": 0.7901,
|
| 83 |
-
"step": 1500
|
| 84 |
-
},
|
| 85 |
-
{
|
| 86 |
-
"epoch": 0.32,
|
| 87 |
-
"eval_loss": 0.5156298279762268,
|
| 88 |
-
"eval_runtime": 151.1176,
|
| 89 |
-
"eval_samples_per_second": 948.506,
|
| 90 |
-
"eval_steps_per_second": 3.706,
|
| 91 |
-
"step": 1600
|
| 92 |
-
},
|
| 93 |
-
{
|
| 94 |
-
"epoch": 0.36,
|
| 95 |
-
"eval_loss": 0.7702064514160156,
|
| 96 |
-
"eval_runtime": 151.1572,
|
| 97 |
-
"eval_samples_per_second": 948.258,
|
| 98 |
-
"eval_steps_per_second": 3.705,
|
| 99 |
-
"step": 1800
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.4,
|
| 103 |
-
"learning_rate": 4.603095852351657e-05,
|
| 104 |
-
"loss": 0.7633,
|
| 105 |
-
"step": 2000
|
| 106 |
-
},
|
| 107 |
-
{
|
| 108 |
-
"epoch": 0.4,
|
| 109 |
-
"eval_loss": 0.8383969664573669,
|
| 110 |
-
"eval_runtime": 151.2044,
|
| 111 |
-
"eval_samples_per_second": 947.962,
|
| 112 |
-
"eval_steps_per_second": 3.704,
|
| 113 |
-
"step": 2000
|
| 114 |
-
},
|
| 115 |
-
{
|
| 116 |
-
"epoch": 0.44,
|
| 117 |
-
"eval_loss": 0.7715063691139221,
|
| 118 |
-
"eval_runtime": 151.1515,
|
| 119 |
-
"eval_samples_per_second": 948.294,
|
| 120 |
-
"eval_steps_per_second": 3.705,
|
| 121 |
-
"step": 2200
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.48,
|
| 125 |
-
"eval_loss": 0.8304912447929382,
|
| 126 |
-
"eval_runtime": 151.1644,
|
| 127 |
-
"eval_samples_per_second": 948.213,
|
| 128 |
-
"eval_steps_per_second": 3.705,
|
| 129 |
-
"step": 2400
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.5,
|
| 133 |
-
"learning_rate": 4.5038698154395716e-05,
|
| 134 |
-
"loss": 0.829,
|
| 135 |
-
"step": 2500
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.52,
|
| 139 |
-
"eval_loss": 0.8085972666740417,
|
| 140 |
-
"eval_runtime": 151.1877,
|
| 141 |
-
"eval_samples_per_second": 948.066,
|
| 142 |
-
"eval_steps_per_second": 3.704,
|
| 143 |
-
"step": 2600
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.56,
|
| 147 |
-
"eval_loss": 0.9329496026039124,
|
| 148 |
-
"eval_runtime": 151.1826,
|
| 149 |
-
"eval_samples_per_second": 948.098,
|
| 150 |
-
"eval_steps_per_second": 3.704,
|
| 151 |
-
"step": 2800
|
| 152 |
-
},
|
| 153 |
-
{
|
| 154 |
-
"epoch": 0.6,
|
| 155 |
-
"learning_rate": 4.404643778527486e-05,
|
| 156 |
-
"loss": 0.829,
|
| 157 |
-
"step": 3000
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.6,
|
| 161 |
-
"eval_loss": 0.8683816194534302,
|
| 162 |
-
"eval_runtime": 151.1074,
|
| 163 |
-
"eval_samples_per_second": 948.571,
|
| 164 |
-
"eval_steps_per_second": 3.706,
|
| 165 |
-
"step": 3000
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 0.64,
|
| 169 |
-
"eval_loss": 0.9232978820800781,
|
| 170 |
-
"eval_runtime": 151.3223,
|
| 171 |
-
"eval_samples_per_second": 947.223,
|
| 172 |
-
"eval_steps_per_second": 3.701,
|
| 173 |
-
"step": 3200
|
| 174 |
-
},
|
| 175 |
-
{
|
| 176 |
-
"epoch": 0.67,
|
| 177 |
-
"eval_loss": 0.8544695973396301,
|
| 178 |
-
"eval_runtime": 151.1372,
|
| 179 |
-
"eval_samples_per_second": 948.383,
|
| 180 |
-
"eval_steps_per_second": 3.705,
|
| 181 |
-
"step": 3400
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 0.69,
|
| 185 |
-
"learning_rate": 4.3054177416154005e-05,
|
| 186 |
-
"loss": 0.9221,
|
| 187 |
-
"step": 3500
|
| 188 |
-
},
|
| 189 |
-
{
|
| 190 |
-
"epoch": 0.71,
|
| 191 |
-
"eval_loss": 0.9146988987922668,
|
| 192 |
-
"eval_runtime": 151.1788,
|
| 193 |
-
"eval_samples_per_second": 948.122,
|
| 194 |
-
"eval_steps_per_second": 3.704,
|
| 195 |
-
"step": 3600
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.75,
|
| 199 |
-
"eval_loss": 0.8903905153274536,
|
| 200 |
-
"eval_runtime": 151.1333,
|
| 201 |
-
"eval_samples_per_second": 948.408,
|
| 202 |
-
"eval_steps_per_second": 3.705,
|
| 203 |
-
"step": 3800
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 0.79,
|
| 207 |
-
"learning_rate": 4.206191704703314e-05,
|
| 208 |
-
"loss": 0.9301,
|
| 209 |
-
"step": 4000
|
| 210 |
-
},
|
| 211 |
-
{
|
| 212 |
-
"epoch": 0.79,
|
| 213 |
-
"eval_loss": 0.7672401666641235,
|
| 214 |
-
"eval_runtime": 151.0676,
|
| 215 |
-
"eval_samples_per_second": 948.821,
|
| 216 |
-
"eval_steps_per_second": 3.707,
|
| 217 |
-
"step": 4000
|
| 218 |
-
},
|
| 219 |
-
{
|
| 220 |
-
"epoch": 0.83,
|
| 221 |
-
"eval_loss": 0.9532115459442139,
|
| 222 |
-
"eval_runtime": 151.0924,
|
| 223 |
-
"eval_samples_per_second": 948.665,
|
| 224 |
-
"eval_steps_per_second": 3.706,
|
| 225 |
-
"step": 4200
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.87,
|
| 229 |
-
"eval_loss": 0.25937971472740173,
|
| 230 |
-
"eval_runtime": 151.2151,
|
| 231 |
-
"eval_samples_per_second": 947.895,
|
| 232 |
-
"eval_steps_per_second": 3.703,
|
| 233 |
-
"step": 4400
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.89,
|
| 237 |
-
"learning_rate": 4.1069656677912286e-05,
|
| 238 |
-
"loss": 1.5584,
|
| 239 |
-
"step": 4500
|
| 240 |
-
},
|
| 241 |
-
{
|
| 242 |
-
"epoch": 0.91,
|
| 243 |
-
"eval_loss": 0.13394613564014435,
|
| 244 |
-
"eval_runtime": 151.3943,
|
| 245 |
-
"eval_samples_per_second": 946.773,
|
| 246 |
-
"eval_steps_per_second": 3.699,
|
| 247 |
-
"step": 4600
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.95,
|
| 251 |
-
"eval_loss": 0.1273290514945984,
|
| 252 |
-
"eval_runtime": 150.9105,
|
| 253 |
-
"eval_samples_per_second": 949.808,
|
| 254 |
-
"eval_steps_per_second": 3.711,
|
| 255 |
-
"step": 4800
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.99,
|
| 259 |
-
"learning_rate": 4.0077396308791423e-05,
|
| 260 |
-
"loss": 0.135,
|
| 261 |
-
"step": 5000
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.99,
|
| 265 |
-
"eval_loss": 0.11838869005441666,
|
| 266 |
-
"eval_runtime": 150.9534,
|
| 267 |
-
"eval_samples_per_second": 949.538,
|
| 268 |
-
"eval_steps_per_second": 3.71,
|
| 269 |
-
"step": 5000
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 1.03,
|
| 273 |
-
"eval_loss": 0.11631189286708832,
|
| 274 |
-
"eval_runtime": 150.9499,
|
| 275 |
-
"eval_samples_per_second": 949.56,
|
| 276 |
-
"eval_steps_per_second": 3.71,
|
| 277 |
-
"step": 5200
|
| 278 |
-
},
|
| 279 |
-
{
|
| 280 |
-
"epoch": 1.07,
|
| 281 |
-
"eval_loss": 0.11566050350666046,
|
| 282 |
-
"eval_runtime": 151.0546,
|
| 283 |
-
"eval_samples_per_second": 948.902,
|
| 284 |
-
"eval_steps_per_second": 3.707,
|
| 285 |
-
"step": 5400
|
| 286 |
-
},
|
| 287 |
-
{
|
| 288 |
-
"epoch": 1.09,
|
| 289 |
-
"learning_rate": 3.908513593967057e-05,
|
| 290 |
-
"loss": 0.1189,
|
| 291 |
-
"step": 5500
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 1.11,
|
| 295 |
-
"eval_loss": 0.1144029051065445,
|
| 296 |
-
"eval_runtime": 151.022,
|
| 297 |
-
"eval_samples_per_second": 949.107,
|
| 298 |
-
"eval_steps_per_second": 3.708,
|
| 299 |
-
"step": 5600
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 1.15,
|
| 303 |
-
"eval_loss": 0.11364156752824783,
|
| 304 |
-
"eval_runtime": 151.0116,
|
| 305 |
-
"eval_samples_per_second": 949.172,
|
| 306 |
-
"eval_steps_per_second": 3.708,
|
| 307 |
-
"step": 5800
|
| 308 |
-
},
|
| 309 |
-
{
|
| 310 |
-
"epoch": 1.19,
|
| 311 |
-
"learning_rate": 3.809287557054971e-05,
|
| 312 |
-
"loss": 0.1162,
|
| 313 |
-
"step": 6000
|
| 314 |
-
},
|
| 315 |
-
{
|
| 316 |
-
"epoch": 1.19,
|
| 317 |
-
"eval_loss": 0.113379567861557,
|
| 318 |
-
"eval_runtime": 151.0775,
|
| 319 |
-
"eval_samples_per_second": 948.758,
|
| 320 |
-
"eval_steps_per_second": 3.707,
|
| 321 |
-
"step": 6000
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 1.23,
|
| 325 |
-
"eval_loss": 0.11304181814193726,
|
| 326 |
-
"eval_runtime": 151.1881,
|
| 327 |
-
"eval_samples_per_second": 948.064,
|
| 328 |
-
"eval_steps_per_second": 3.704,
|
| 329 |
-
"step": 6200
|
| 330 |
-
},
|
| 331 |
-
{
|
| 332 |
-
"epoch": 1.27,
|
| 333 |
-
"eval_loss": 0.11305490881204605,
|
| 334 |
-
"eval_runtime": 151.0078,
|
| 335 |
-
"eval_samples_per_second": 949.196,
|
| 336 |
-
"eval_steps_per_second": 3.708,
|
| 337 |
-
"step": 6400
|
| 338 |
-
},
|
| 339 |
-
{
|
| 340 |
-
"epoch": 1.29,
|
| 341 |
-
"learning_rate": 3.7100615201428856e-05,
|
| 342 |
-
"loss": 0.1165,
|
| 343 |
-
"step": 6500
|
| 344 |
-
},
|
| 345 |
-
{
|
| 346 |
-
"epoch": 1.31,
|
| 347 |
-
"eval_loss": 0.11227019131183624,
|
| 348 |
-
"eval_runtime": 151.0464,
|
| 349 |
-
"eval_samples_per_second": 948.953,
|
| 350 |
-
"eval_steps_per_second": 3.707,
|
| 351 |
-
"step": 6600
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 1.35,
|
| 355 |
-
"eval_loss": 0.11202485859394073,
|
| 356 |
-
"eval_runtime": 150.9714,
|
| 357 |
-
"eval_samples_per_second": 949.425,
|
| 358 |
"eval_steps_per_second": 3.709,
|
| 359 |
-
"step":
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 1.39,
|
| 363 |
-
"learning_rate": 3.6108354832308e-05,
|
| 364 |
-
"loss": 0.1136,
|
| 365 |
-
"step": 7000
|
| 366 |
-
},
|
| 367 |
-
{
|
| 368 |
-
"epoch": 1.39,
|
| 369 |
-
"eval_loss": 0.11163973808288574,
|
| 370 |
-
"eval_runtime": 151.0638,
|
| 371 |
-
"eval_samples_per_second": 948.844,
|
| 372 |
-
"eval_steps_per_second": 3.707,
|
| 373 |
-
"step": 7000
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 1.43,
|
| 377 |
-
"eval_loss": 0.11151301115751266,
|
| 378 |
-
"eval_runtime": 151.052,
|
| 379 |
-
"eval_samples_per_second": 948.918,
|
| 380 |
-
"eval_steps_per_second": 3.707,
|
| 381 |
-
"step": 7200
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.47,
|
| 385 |
-
"eval_loss": 0.11108649522066116,
|
| 386 |
-
"eval_runtime": 151.035,
|
| 387 |
-
"eval_samples_per_second": 949.025,
|
| 388 |
-
"eval_steps_per_second": 3.708,
|
| 389 |
-
"step": 7400
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 1.49,
|
| 393 |
-
"learning_rate": 3.5116094463187144e-05,
|
| 394 |
-
"loss": 0.1124,
|
| 395 |
-
"step": 7500
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 1.51,
|
| 399 |
-
"eval_loss": 0.11082353442907333,
|
| 400 |
-
"eval_runtime": 150.9886,
|
| 401 |
-
"eval_samples_per_second": 949.317,
|
| 402 |
-
"eval_steps_per_second": 3.709,
|
| 403 |
-
"step": 7600
|
| 404 |
-
},
|
| 405 |
-
{
|
| 406 |
-
"epoch": 1.55,
|
| 407 |
-
"eval_loss": 0.11104652285575867,
|
| 408 |
-
"eval_runtime": 150.9566,
|
| 409 |
-
"eval_samples_per_second": 949.518,
|
| 410 |
-
"eval_steps_per_second": 3.71,
|
| 411 |
-
"step": 7800
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 1.59,
|
| 415 |
-
"learning_rate": 3.412383409406629e-05,
|
| 416 |
-
"loss": 0.1128,
|
| 417 |
-
"step": 8000
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 1.59,
|
| 421 |
-
"eval_loss": 0.11066918820142746,
|
| 422 |
-
"eval_runtime": 150.949,
|
| 423 |
-
"eval_samples_per_second": 949.566,
|
| 424 |
-
"eval_steps_per_second": 3.71,
|
| 425 |
-
"step": 8000
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 1.63,
|
| 429 |
-
"eval_loss": 0.11038321256637573,
|
| 430 |
-
"eval_runtime": 151.0411,
|
| 431 |
-
"eval_samples_per_second": 948.987,
|
| 432 |
-
"eval_steps_per_second": 3.708,
|
| 433 |
-
"step": 8200
|
| 434 |
-
},
|
| 435 |
-
{
|
| 436 |
-
"epoch": 1.67,
|
| 437 |
-
"eval_loss": 0.11011925339698792,
|
| 438 |
-
"eval_runtime": 151.052,
|
| 439 |
-
"eval_samples_per_second": 948.918,
|
| 440 |
-
"eval_steps_per_second": 3.707,
|
| 441 |
-
"step": 8400
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"epoch": 1.69,
|
| 445 |
-
"learning_rate": 3.3131573724945426e-05,
|
| 446 |
-
"loss": 0.1114,
|
| 447 |
-
"step": 8500
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 1.71,
|
| 451 |
-
"eval_loss": 0.11026579141616821,
|
| 452 |
-
"eval_runtime": 151.2041,
|
| 453 |
-
"eval_samples_per_second": 947.963,
|
| 454 |
-
"eval_steps_per_second": 3.704,
|
| 455 |
-
"step": 8600
|
| 456 |
-
},
|
| 457 |
-
{
|
| 458 |
-
"epoch": 1.75,
|
| 459 |
-
"eval_loss": 0.10981647670269012,
|
| 460 |
-
"eval_runtime": 151.0892,
|
| 461 |
-
"eval_samples_per_second": 948.685,
|
| 462 |
-
"eval_steps_per_second": 3.706,
|
| 463 |
-
"step": 8800
|
| 464 |
-
},
|
| 465 |
-
{
|
| 466 |
-
"epoch": 1.79,
|
| 467 |
-
"learning_rate": 3.213931335582457e-05,
|
| 468 |
-
"loss": 0.1094,
|
| 469 |
-
"step": 9000
|
| 470 |
-
},
|
| 471 |
-
{
|
| 472 |
-
"epoch": 1.79,
|
| 473 |
-
"eval_loss": 0.11008855700492859,
|
| 474 |
-
"eval_runtime": 150.9863,
|
| 475 |
-
"eval_samples_per_second": 949.331,
|
| 476 |
-
"eval_steps_per_second": 3.709,
|
| 477 |
-
"step": 9000
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 1.83,
|
| 481 |
-
"eval_loss": 0.10983282327651978,
|
| 482 |
-
"eval_runtime": 150.9699,
|
| 483 |
-
"eval_samples_per_second": 949.434,
|
| 484 |
-
"eval_steps_per_second": 3.709,
|
| 485 |
-
"step": 9200
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 1.87,
|
| 489 |
-
"eval_loss": 0.10982002317905426,
|
| 490 |
-
"eval_runtime": 151.0614,
|
| 491 |
-
"eval_samples_per_second": 948.859,
|
| 492 |
-
"eval_steps_per_second": 3.707,
|
| 493 |
-
"step": 9400
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 1.89,
|
| 497 |
-
"learning_rate": 3.1147052986703714e-05,
|
| 498 |
-
"loss": 0.111,
|
| 499 |
-
"step": 9500
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 1.91,
|
| 503 |
-
"eval_loss": 0.11063655465841293,
|
| 504 |
-
"eval_runtime": 151.0974,
|
| 505 |
-
"eval_samples_per_second": 948.633,
|
| 506 |
-
"eval_steps_per_second": 3.706,
|
| 507 |
-
"step": 9600
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.94,
|
| 511 |
-
"eval_loss": 0.10998239368200302,
|
| 512 |
-
"eval_runtime": 150.9779,
|
| 513 |
-
"eval_samples_per_second": 949.384,
|
| 514 |
-
"eval_steps_per_second": 3.709,
|
| 515 |
-
"step": 9800
|
| 516 |
-
},
|
| 517 |
-
{
|
| 518 |
-
"epoch": 1.98,
|
| 519 |
-
"learning_rate": 3.0154792617582855e-05,
|
| 520 |
-
"loss": 0.1118,
|
| 521 |
-
"step": 10000
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 1.98,
|
| 525 |
-
"eval_loss": 0.10957028716802597,
|
| 526 |
-
"eval_runtime": 151.0796,
|
| 527 |
-
"eval_samples_per_second": 948.745,
|
| 528 |
-
"eval_steps_per_second": 3.707,
|
| 529 |
-
"step": 10000
|
| 530 |
-
},
|
| 531 |
-
{
|
| 532 |
-
"epoch": 2.02,
|
| 533 |
-
"eval_loss": 0.10959649085998535,
|
| 534 |
-
"eval_runtime": 150.9465,
|
| 535 |
-
"eval_samples_per_second": 949.582,
|
| 536 |
-
"eval_steps_per_second": 3.71,
|
| 537 |
-
"step": 10200
|
| 538 |
-
},
|
| 539 |
-
{
|
| 540 |
-
"epoch": 2.06,
|
| 541 |
-
"eval_loss": 0.10964351147413254,
|
| 542 |
-
"eval_runtime": 151.0871,
|
| 543 |
-
"eval_samples_per_second": 948.698,
|
| 544 |
-
"eval_steps_per_second": 3.706,
|
| 545 |
-
"step": 10400
|
| 546 |
-
},
|
| 547 |
-
{
|
| 548 |
-
"epoch": 2.08,
|
| 549 |
-
"learning_rate": 2.9162532248462e-05,
|
| 550 |
-
"loss": 0.1102,
|
| 551 |
-
"step": 10500
|
| 552 |
-
},
|
| 553 |
-
{
|
| 554 |
-
"epoch": 2.1,
|
| 555 |
-
"eval_loss": 0.1093754768371582,
|
| 556 |
-
"eval_runtime": 150.92,
|
| 557 |
-
"eval_samples_per_second": 949.748,
|
| 558 |
-
"eval_steps_per_second": 3.711,
|
| 559 |
-
"step": 10600
|
| 560 |
-
},
|
| 561 |
-
{
|
| 562 |
-
"epoch": 2.14,
|
| 563 |
-
"eval_loss": 0.10980819910764694,
|
| 564 |
-
"eval_runtime": 151.0665,
|
| 565 |
-
"eval_samples_per_second": 948.827,
|
| 566 |
-
"eval_steps_per_second": 3.707,
|
| 567 |
-
"step": 10800
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"epoch": 2.18,
|
| 571 |
-
"learning_rate": 2.8170271879341143e-05,
|
| 572 |
-
"loss": 0.1083,
|
| 573 |
-
"step": 11000
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 2.18,
|
| 577 |
-
"eval_loss": 0.10928498953580856,
|
| 578 |
-
"eval_runtime": 151.0595,
|
| 579 |
-
"eval_samples_per_second": 948.871,
|
| 580 |
-
"eval_steps_per_second": 3.707,
|
| 581 |
-
"step": 11000
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 2.22,
|
| 585 |
-
"eval_loss": 0.10924935340881348,
|
| 586 |
-
"eval_runtime": 151.0906,
|
| 587 |
-
"eval_samples_per_second": 948.676,
|
| 588 |
-
"eval_steps_per_second": 3.706,
|
| 589 |
-
"step": 11200
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 2.26,
|
| 593 |
-
"eval_loss": 0.10928213596343994,
|
| 594 |
-
"eval_runtime": 151.0496,
|
| 595 |
-
"eval_samples_per_second": 948.933,
|
| 596 |
-
"eval_steps_per_second": 3.707,
|
| 597 |
-
"step": 11400
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 2.28,
|
| 601 |
-
"learning_rate": 2.717801151022028e-05,
|
| 602 |
-
"loss": 0.1113,
|
| 603 |
-
"step": 11500
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 2.3,
|
| 607 |
-
"eval_loss": 0.10921794176101685,
|
| 608 |
-
"eval_runtime": 151.0887,
|
| 609 |
-
"eval_samples_per_second": 948.688,
|
| 610 |
-
"eval_steps_per_second": 3.706,
|
| 611 |
-
"step": 11600
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 2.34,
|
| 615 |
-
"eval_loss": 0.1091688945889473,
|
| 616 |
-
"eval_runtime": 151.1277,
|
| 617 |
-
"eval_samples_per_second": 948.443,
|
| 618 |
-
"eval_steps_per_second": 3.705,
|
| 619 |
-
"step": 11800
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 2.38,
|
| 623 |
-
"learning_rate": 2.6185751141099424e-05,
|
| 624 |
-
"loss": 0.1102,
|
| 625 |
-
"step": 12000
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 2.38,
|
| 629 |
-
"eval_loss": 0.10950697958469391,
|
| 630 |
-
"eval_runtime": 151.1121,
|
| 631 |
-
"eval_samples_per_second": 948.541,
|
| 632 |
-
"eval_steps_per_second": 3.706,
|
| 633 |
-
"step": 12000
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 2.42,
|
| 637 |
-
"eval_loss": 0.10911945253610611,
|
| 638 |
-
"eval_runtime": 151.0456,
|
| 639 |
-
"eval_samples_per_second": 948.958,
|
| 640 |
-
"eval_steps_per_second": 3.707,
|
| 641 |
-
"step": 12200
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch": 2.46,
|
| 645 |
-
"eval_loss": 0.1094910129904747,
|
| 646 |
-
"eval_runtime": 150.9865,
|
| 647 |
-
"eval_samples_per_second": 949.33,
|
| 648 |
-
"eval_steps_per_second": 3.709,
|
| 649 |
-
"step": 12400
|
| 650 |
-
},
|
| 651 |
-
{
|
| 652 |
-
"epoch": 2.48,
|
| 653 |
-
"learning_rate": 2.519349077197857e-05,
|
| 654 |
-
"loss": 0.1093,
|
| 655 |
-
"step": 12500
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 2.5,
|
| 659 |
-
"eval_loss": 0.10911854356527328,
|
| 660 |
-
"eval_runtime": 150.8255,
|
| 661 |
-
"eval_samples_per_second": 950.343,
|
| 662 |
-
"eval_steps_per_second": 3.713,
|
| 663 |
-
"step": 12600
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 2.54,
|
| 667 |
-
"eval_loss": 0.1091616079211235,
|
| 668 |
-
"eval_runtime": 150.8284,
|
| 669 |
-
"eval_samples_per_second": 950.325,
|
| 670 |
-
"eval_steps_per_second": 3.713,
|
| 671 |
-
"step": 12800
|
| 672 |
-
},
|
| 673 |
-
{
|
| 674 |
-
"epoch": 2.58,
|
| 675 |
-
"learning_rate": 2.420123040285771e-05,
|
| 676 |
-
"loss": 0.1105,
|
| 677 |
-
"step": 13000
|
| 678 |
-
},
|
| 679 |
-
{
|
| 680 |
-
"epoch": 2.58,
|
| 681 |
-
"eval_loss": 0.10907712578773499,
|
| 682 |
-
"eval_runtime": 151.0685,
|
| 683 |
-
"eval_samples_per_second": 948.815,
|
| 684 |
-
"eval_steps_per_second": 3.707,
|
| 685 |
-
"step": 13000
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 2.62,
|
| 689 |
-
"eval_loss": 0.10917963832616806,
|
| 690 |
-
"eval_runtime": 150.9953,
|
| 691 |
-
"eval_samples_per_second": 949.274,
|
| 692 |
-
"eval_steps_per_second": 3.709,
|
| 693 |
-
"step": 13200
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"epoch": 2.66,
|
| 697 |
-
"eval_loss": 0.10913284868001938,
|
| 698 |
-
"eval_runtime": 150.8911,
|
| 699 |
-
"eval_samples_per_second": 949.93,
|
| 700 |
-
"eval_steps_per_second": 3.711,
|
| 701 |
-
"step": 13400
|
| 702 |
-
},
|
| 703 |
-
{
|
| 704 |
-
"epoch": 2.68,
|
| 705 |
-
"learning_rate": 2.3208970033736854e-05,
|
| 706 |
-
"loss": 0.1094,
|
| 707 |
-
"step": 13500
|
| 708 |
-
},
|
| 709 |
-
{
|
| 710 |
-
"epoch": 2.7,
|
| 711 |
-
"eval_loss": 0.10903245210647583,
|
| 712 |
-
"eval_runtime": 150.9858,
|
| 713 |
-
"eval_samples_per_second": 949.334,
|
| 714 |
-
"eval_steps_per_second": 3.709,
|
| 715 |
-
"step": 13600
|
| 716 |
-
},
|
| 717 |
-
{
|
| 718 |
-
"epoch": 2.74,
|
| 719 |
-
"eval_loss": 0.10889765620231628,
|
| 720 |
-
"eval_runtime": 150.9473,
|
| 721 |
-
"eval_samples_per_second": 949.577,
|
| 722 |
-
"eval_steps_per_second": 3.71,
|
| 723 |
-
"step": 13800
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 2.78,
|
| 727 |
-
"learning_rate": 2.2216709664615994e-05,
|
| 728 |
-
"loss": 0.1104,
|
| 729 |
-
"step": 14000
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 2.78,
|
| 733 |
-
"eval_loss": 0.1090923622250557,
|
| 734 |
-
"eval_runtime": 151.128,
|
| 735 |
-
"eval_samples_per_second": 948.441,
|
| 736 |
-
"eval_steps_per_second": 3.705,
|
| 737 |
-
"step": 14000
|
| 738 |
-
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 2.82,
|
| 741 |
-
"eval_loss": 0.10899552702903748,
|
| 742 |
-
"eval_runtime": 151.5342,
|
| 743 |
-
"eval_samples_per_second": 945.899,
|
| 744 |
-
"eval_steps_per_second": 3.696,
|
| 745 |
-
"step": 14200
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 2.86,
|
| 749 |
-
"eval_loss": 0.10895191133022308,
|
| 750 |
-
"eval_runtime": 150.9811,
|
| 751 |
-
"eval_samples_per_second": 949.364,
|
| 752 |
-
"eval_steps_per_second": 3.709,
|
| 753 |
-
"step": 14400
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 2.88,
|
| 757 |
-
"learning_rate": 2.122444929549514e-05,
|
| 758 |
-
"loss": 0.1117,
|
| 759 |
-
"step": 14500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 2.9,
|
| 763 |
-
"eval_loss": 0.10904989391565323,
|
| 764 |
-
"eval_runtime": 150.9339,
|
| 765 |
-
"eval_samples_per_second": 949.661,
|
| 766 |
-
"eval_steps_per_second": 3.71,
|
| 767 |
-
"step": 14600
|
| 768 |
-
},
|
| 769 |
-
{
|
| 770 |
-
"epoch": 2.94,
|
| 771 |
-
"eval_loss": 0.10929518938064575,
|
| 772 |
-
"eval_runtime": 150.8669,
|
| 773 |
-
"eval_samples_per_second": 950.082,
|
| 774 |
-
"eval_steps_per_second": 3.712,
|
| 775 |
-
"step": 14800
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 2.98,
|
| 779 |
-
"learning_rate": 2.0232188926374283e-05,
|
| 780 |
-
"loss": 0.1093,
|
| 781 |
-
"step": 15000
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 2.98,
|
| 785 |
-
"eval_loss": 0.10902168601751328,
|
| 786 |
-
"eval_runtime": 151.0426,
|
| 787 |
-
"eval_samples_per_second": 948.977,
|
| 788 |
-
"eval_steps_per_second": 3.708,
|
| 789 |
-
"step": 15000
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 3.02,
|
| 793 |
-
"eval_loss": 0.10888814926147461,
|
| 794 |
-
"eval_runtime": 150.8829,
|
| 795 |
-
"eval_samples_per_second": 949.982,
|
| 796 |
-
"eval_steps_per_second": 3.711,
|
| 797 |
-
"step": 15200
|
| 798 |
-
},
|
| 799 |
-
{
|
| 800 |
-
"epoch": 3.06,
|
| 801 |
-
"eval_loss": 0.1088201105594635,
|
| 802 |
-
"eval_runtime": 150.8765,
|
| 803 |
-
"eval_samples_per_second": 950.022,
|
| 804 |
-
"eval_steps_per_second": 3.712,
|
| 805 |
-
"step": 15400
|
| 806 |
-
},
|
| 807 |
-
{
|
| 808 |
-
"epoch": 3.08,
|
| 809 |
-
"learning_rate": 1.9239928557253427e-05,
|
| 810 |
-
"loss": 0.1098,
|
| 811 |
-
"step": 15500
|
| 812 |
-
},
|
| 813 |
-
{
|
| 814 |
-
"epoch": 3.1,
|
| 815 |
-
"eval_loss": 0.10886531323194504,
|
| 816 |
-
"eval_runtime": 150.9989,
|
| 817 |
-
"eval_samples_per_second": 949.252,
|
| 818 |
-
"eval_steps_per_second": 3.709,
|
| 819 |
-
"step": 15600
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 3.14,
|
| 823 |
-
"eval_loss": 0.1089038997888565,
|
| 824 |
-
"eval_runtime": 150.8728,
|
| 825 |
-
"eval_samples_per_second": 950.045,
|
| 826 |
-
"eval_steps_per_second": 3.712,
|
| 827 |
-
"step": 15800
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"epoch": 3.18,
|
| 831 |
-
"learning_rate": 1.8247668188132567e-05,
|
| 832 |
-
"loss": 0.1102,
|
| 833 |
-
"step": 16000
|
| 834 |
-
},
|
| 835 |
-
{
|
| 836 |
-
"epoch": 3.18,
|
| 837 |
-
"eval_loss": 0.10882215201854706,
|
| 838 |
-
"eval_runtime": 151.4626,
|
| 839 |
-
"eval_samples_per_second": 946.346,
|
| 840 |
-
"eval_steps_per_second": 3.697,
|
| 841 |
-
"step": 16000
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 3.21,
|
| 845 |
-
"eval_loss": 0.10881613940000534,
|
| 846 |
-
"eval_runtime": 151.4856,
|
| 847 |
-
"eval_samples_per_second": 946.202,
|
| 848 |
-
"eval_steps_per_second": 3.697,
|
| 849 |
-
"step": 16200
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 3.25,
|
| 853 |
-
"eval_loss": 0.10882557183504105,
|
| 854 |
-
"eval_runtime": 150.9956,
|
| 855 |
-
"eval_samples_per_second": 949.273,
|
| 856 |
-
"eval_steps_per_second": 3.709,
|
| 857 |
-
"step": 16400
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 3.27,
|
| 861 |
-
"learning_rate": 1.7255407819011708e-05,
|
| 862 |
-
"loss": 0.1087,
|
| 863 |
-
"step": 16500
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 3.29,
|
| 867 |
-
"eval_loss": 0.10881206393241882,
|
| 868 |
-
"eval_runtime": 150.911,
|
| 869 |
-
"eval_samples_per_second": 949.805,
|
| 870 |
-
"eval_steps_per_second": 3.711,
|
| 871 |
-
"step": 16600
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 3.33,
|
| 875 |
-
"eval_loss": 0.10894536972045898,
|
| 876 |
-
"eval_runtime": 150.8614,
|
| 877 |
-
"eval_samples_per_second": 950.117,
|
| 878 |
-
"eval_steps_per_second": 3.712,
|
| 879 |
-
"step": 16800
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 3.37,
|
| 883 |
-
"learning_rate": 1.6263147449890852e-05,
|
| 884 |
-
"loss": 0.1082,
|
| 885 |
-
"step": 17000
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 3.37,
|
| 889 |
-
"eval_loss": 0.10886503010988235,
|
| 890 |
-
"eval_runtime": 151.0504,
|
| 891 |
-
"eval_samples_per_second": 948.928,
|
| 892 |
-
"eval_steps_per_second": 3.707,
|
| 893 |
-
"step": 17000
|
| 894 |
-
},
|
| 895 |
-
{
|
| 896 |
-
"epoch": 3.41,
|
| 897 |
-
"eval_loss": 0.10877899825572968,
|
| 898 |
-
"eval_runtime": 150.8388,
|
| 899 |
-
"eval_samples_per_second": 950.259,
|
| 900 |
-
"eval_steps_per_second": 3.713,
|
| 901 |
-
"step": 17200
|
| 902 |
-
},
|
| 903 |
-
{
|
| 904 |
-
"epoch": 3.45,
|
| 905 |
-
"eval_loss": 0.10876929759979248,
|
| 906 |
-
"eval_runtime": 150.9938,
|
| 907 |
-
"eval_samples_per_second": 949.284,
|
| 908 |
-
"eval_steps_per_second": 3.709,
|
| 909 |
-
"step": 17400
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 3.47,
|
| 913 |
-
"learning_rate": 1.5270887080769993e-05,
|
| 914 |
-
"loss": 0.1097,
|
| 915 |
-
"step": 17500
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 3.49,
|
| 919 |
-
"eval_loss": 0.1089547798037529,
|
| 920 |
-
"eval_runtime": 151.0328,
|
| 921 |
-
"eval_samples_per_second": 949.039,
|
| 922 |
-
"eval_steps_per_second": 3.708,
|
| 923 |
-
"step": 17600
|
| 924 |
-
},
|
| 925 |
-
{
|
| 926 |
-
"epoch": 3.53,
|
| 927 |
-
"eval_loss": 0.10884077101945877,
|
| 928 |
-
"eval_runtime": 151.0214,
|
| 929 |
-
"eval_samples_per_second": 949.111,
|
| 930 |
-
"eval_steps_per_second": 3.708,
|
| 931 |
-
"step": 17800
|
| 932 |
-
},
|
| 933 |
-
{
|
| 934 |
-
"epoch": 3.57,
|
| 935 |
-
"learning_rate": 1.4278626711649137e-05,
|
| 936 |
-
"loss": 0.1105,
|
| 937 |
-
"step": 18000
|
| 938 |
-
},
|
| 939 |
-
{
|
| 940 |
-
"epoch": 3.57,
|
| 941 |
-
"eval_loss": 0.1087304875254631,
|
| 942 |
-
"eval_runtime": 150.7922,
|
| 943 |
-
"eval_samples_per_second": 950.553,
|
| 944 |
-
"eval_steps_per_second": 3.714,
|
| 945 |
-
"step": 18000
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 3.61,
|
| 949 |
-
"eval_loss": 0.10879357904195786,
|
| 950 |
-
"eval_runtime": 150.8914,
|
| 951 |
-
"eval_samples_per_second": 949.929,
|
| 952 |
-
"eval_steps_per_second": 3.711,
|
| 953 |
-
"step": 18200
|
| 954 |
-
},
|
| 955 |
-
{
|
| 956 |
-
"epoch": 3.65,
|
| 957 |
-
"eval_loss": 0.10872866213321686,
|
| 958 |
-
"eval_runtime": 150.992,
|
| 959 |
-
"eval_samples_per_second": 949.295,
|
| 960 |
-
"eval_steps_per_second": 3.709,
|
| 961 |
-
"step": 18400
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 3.67,
|
| 965 |
-
"learning_rate": 1.3286366342528281e-05,
|
| 966 |
-
"loss": 0.1089,
|
| 967 |
-
"step": 18500
|
| 968 |
-
},
|
| 969 |
-
{
|
| 970 |
-
"epoch": 3.69,
|
| 971 |
-
"eval_loss": 0.10874517261981964,
|
| 972 |
-
"eval_runtime": 151.2717,
|
| 973 |
-
"eval_samples_per_second": 947.54,
|
| 974 |
-
"eval_steps_per_second": 3.702,
|
| 975 |
-
"step": 18600
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 3.73,
|
| 979 |
-
"eval_loss": 0.10880085825920105,
|
| 980 |
-
"eval_runtime": 150.8035,
|
| 981 |
-
"eval_samples_per_second": 950.482,
|
| 982 |
-
"eval_steps_per_second": 3.713,
|
| 983 |
-
"step": 18800
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 3.77,
|
| 987 |
-
"learning_rate": 1.2294105973407422e-05,
|
| 988 |
-
"loss": 0.1101,
|
| 989 |
-
"step": 19000
|
| 990 |
-
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 3.77,
|
| 993 |
-
"eval_loss": 0.10869602859020233,
|
| 994 |
-
"eval_runtime": 150.9258,
|
| 995 |
-
"eval_samples_per_second": 949.712,
|
| 996 |
-
"eval_steps_per_second": 3.71,
|
| 997 |
-
"step": 19000
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 3.81,
|
| 1001 |
-
"eval_loss": 0.10874003916978836,
|
| 1002 |
-
"eval_runtime": 150.8809,
|
| 1003 |
-
"eval_samples_per_second": 949.994,
|
| 1004 |
-
"eval_steps_per_second": 3.712,
|
| 1005 |
-
"step": 19200
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 3.85,
|
| 1009 |
-
"eval_loss": 0.10871053487062454,
|
| 1010 |
-
"eval_runtime": 151.4705,
|
| 1011 |
-
"eval_samples_per_second": 946.296,
|
| 1012 |
-
"eval_steps_per_second": 3.697,
|
| 1013 |
-
"step": 19400
|
| 1014 |
-
},
|
| 1015 |
-
{
|
| 1016 |
-
"epoch": 3.87,
|
| 1017 |
-
"learning_rate": 1.1301845604286565e-05,
|
| 1018 |
-
"loss": 0.1095,
|
| 1019 |
-
"step": 19500
|
| 1020 |
-
},
|
| 1021 |
-
{
|
| 1022 |
-
"epoch": 3.89,
|
| 1023 |
-
"eval_loss": 0.10870121419429779,
|
| 1024 |
-
"eval_runtime": 151.2796,
|
| 1025 |
-
"eval_samples_per_second": 947.491,
|
| 1026 |
-
"eval_steps_per_second": 3.702,
|
| 1027 |
-
"step": 19600
|
| 1028 |
-
},
|
| 1029 |
-
{
|
| 1030 |
-
"epoch": 3.93,
|
| 1031 |
-
"eval_loss": 0.1087031215429306,
|
| 1032 |
-
"eval_runtime": 151.3836,
|
| 1033 |
-
"eval_samples_per_second": 946.839,
|
| 1034 |
-
"eval_steps_per_second": 3.699,
|
| 1035 |
-
"step": 19800
|
| 1036 |
-
},
|
| 1037 |
-
{
|
| 1038 |
-
"epoch": 3.97,
|
| 1039 |
-
"learning_rate": 1.0309585235165709e-05,
|
| 1040 |
-
"loss": 0.1089,
|
| 1041 |
-
"step": 20000
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 3.97,
|
| 1045 |
-
"eval_loss": 0.10877121239900589,
|
| 1046 |
-
"eval_runtime": 150.9637,
|
| 1047 |
-
"eval_samples_per_second": 949.473,
|
| 1048 |
-
"eval_steps_per_second": 3.71,
|
| 1049 |
-
"step": 20000
|
| 1050 |
-
},
|
| 1051 |
-
{
|
| 1052 |
-
"epoch": 4.01,
|
| 1053 |
-
"eval_loss": 0.10867509990930557,
|
| 1054 |
-
"eval_runtime": 151.1135,
|
| 1055 |
-
"eval_samples_per_second": 948.532,
|
| 1056 |
-
"eval_steps_per_second": 3.706,
|
| 1057 |
-
"step": 20200
|
| 1058 |
-
},
|
| 1059 |
-
{
|
| 1060 |
-
"epoch": 4.05,
|
| 1061 |
-
"eval_loss": 0.10896100848913193,
|
| 1062 |
-
"eval_runtime": 151.009,
|
| 1063 |
-
"eval_samples_per_second": 949.189,
|
| 1064 |
-
"eval_steps_per_second": 3.708,
|
| 1065 |
-
"step": 20400
|
| 1066 |
-
},
|
| 1067 |
-
{
|
| 1068 |
-
"epoch": 4.07,
|
| 1069 |
-
"learning_rate": 9.317324866044851e-06,
|
| 1070 |
-
"loss": 0.1093,
|
| 1071 |
-
"step": 20500
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 4.09,
|
| 1075 |
-
"eval_loss": 0.10866597294807434,
|
| 1076 |
-
"eval_runtime": 151.1827,
|
| 1077 |
-
"eval_samples_per_second": 948.098,
|
| 1078 |
-
"eval_steps_per_second": 3.704,
|
| 1079 |
-
"step": 20600
|
| 1080 |
-
},
|
| 1081 |
-
{
|
| 1082 |
-
"epoch": 4.13,
|
| 1083 |
-
"eval_loss": 0.1086614802479744,
|
| 1084 |
-
"eval_runtime": 150.9777,
|
| 1085 |
-
"eval_samples_per_second": 949.385,
|
| 1086 |
-
"eval_steps_per_second": 3.709,
|
| 1087 |
-
"step": 20800
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 4.17,
|
| 1091 |
-
"learning_rate": 8.325064496923992e-06,
|
| 1092 |
-
"loss": 0.1091,
|
| 1093 |
-
"step": 21000
|
| 1094 |
-
},
|
| 1095 |
-
{
|
| 1096 |
-
"epoch": 4.17,
|
| 1097 |
-
"eval_loss": 0.1086558997631073,
|
| 1098 |
-
"eval_runtime": 150.9361,
|
| 1099 |
-
"eval_samples_per_second": 949.647,
|
| 1100 |
-
"eval_steps_per_second": 3.71,
|
| 1101 |
-
"step": 21000
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 4.21,
|
| 1105 |
-
"eval_loss": 0.10874707996845245,
|
| 1106 |
-
"eval_runtime": 150.8938,
|
| 1107 |
-
"eval_samples_per_second": 949.913,
|
| 1108 |
-
"eval_steps_per_second": 3.711,
|
| 1109 |
-
"step": 21200
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 4.25,
|
| 1113 |
-
"eval_loss": 0.10870933532714844,
|
| 1114 |
-
"eval_runtime": 150.9348,
|
| 1115 |
-
"eval_samples_per_second": 949.655,
|
| 1116 |
-
"eval_steps_per_second": 3.71,
|
| 1117 |
-
"step": 21400
|
| 1118 |
-
},
|
| 1119 |
-
{
|
| 1120 |
-
"epoch": 4.27,
|
| 1121 |
-
"learning_rate": 7.332804127803136e-06,
|
| 1122 |
-
"loss": 0.11,
|
| 1123 |
-
"step": 21500
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 4.29,
|
| 1127 |
-
"eval_loss": 0.10863141715526581,
|
| 1128 |
-
"eval_runtime": 151.0321,
|
| 1129 |
-
"eval_samples_per_second": 949.043,
|
| 1130 |
-
"eval_steps_per_second": 3.708,
|
| 1131 |
-
"step": 21600
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 4.33,
|
| 1135 |
-
"eval_loss": 0.1086675301194191,
|
| 1136 |
-
"eval_runtime": 150.8629,
|
| 1137 |
-
"eval_samples_per_second": 950.108,
|
| 1138 |
-
"eval_steps_per_second": 3.712,
|
| 1139 |
-
"step": 21800
|
| 1140 |
-
},
|
| 1141 |
-
{
|
| 1142 |
-
"epoch": 4.37,
|
| 1143 |
-
"learning_rate": 6.3405437586822785e-06,
|
| 1144 |
-
"loss": 0.11,
|
| 1145 |
-
"step": 22000
|
| 1146 |
-
},
|
| 1147 |
-
{
|
| 1148 |
-
"epoch": 4.37,
|
| 1149 |
-
"eval_loss": 0.10875381529331207,
|
| 1150 |
-
"eval_runtime": 150.9082,
|
| 1151 |
-
"eval_samples_per_second": 949.822,
|
| 1152 |
-
"eval_steps_per_second": 3.711,
|
| 1153 |
-
"step": 22000
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 4.41,
|
| 1157 |
-
"eval_loss": 0.10868263989686966,
|
| 1158 |
-
"eval_runtime": 150.9029,
|
| 1159 |
-
"eval_samples_per_second": 949.856,
|
| 1160 |
-
"eval_steps_per_second": 3.711,
|
| 1161 |
-
"step": 22200
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 4.45,
|
| 1165 |
-
"eval_loss": 0.10862214863300323,
|
| 1166 |
-
"eval_runtime": 150.9465,
|
| 1167 |
-
"eval_samples_per_second": 949.582,
|
| 1168 |
-
"eval_steps_per_second": 3.71,
|
| 1169 |
-
"step": 22400
|
| 1170 |
-
},
|
| 1171 |
-
{
|
| 1172 |
-
"epoch": 4.47,
|
| 1173 |
-
"learning_rate": 5.348283389561421e-06,
|
| 1174 |
-
"loss": 0.1094,
|
| 1175 |
-
"step": 22500
|
| 1176 |
-
},
|
| 1177 |
-
{
|
| 1178 |
-
"epoch": 4.49,
|
| 1179 |
-
"eval_loss": 0.10863359272480011,
|
| 1180 |
-
"eval_runtime": 150.949,
|
| 1181 |
-
"eval_samples_per_second": 949.566,
|
| 1182 |
-
"eval_steps_per_second": 3.71,
|
| 1183 |
-
"step": 22600
|
| 1184 |
-
},
|
| 1185 |
-
{
|
| 1186 |
-
"epoch": 4.52,
|
| 1187 |
-
"eval_loss": 0.10862518846988678,
|
| 1188 |
-
"eval_runtime": 150.9951,
|
| 1189 |
-
"eval_samples_per_second": 949.276,
|
| 1190 |
-
"eval_steps_per_second": 3.709,
|
| 1191 |
-
"step": 22800
|
| 1192 |
-
},
|
| 1193 |
-
{
|
| 1194 |
-
"epoch": 4.56,
|
| 1195 |
-
"learning_rate": 4.356023020440564e-06,
|
| 1196 |
-
"loss": 0.1086,
|
| 1197 |
-
"step": 23000
|
| 1198 |
-
},
|
| 1199 |
-
{
|
| 1200 |
-
"epoch": 4.56,
|
| 1201 |
-
"eval_loss": 0.10860537737607956,
|
| 1202 |
-
"eval_runtime": 150.9863,
|
| 1203 |
-
"eval_samples_per_second": 949.331,
|
| 1204 |
-
"eval_steps_per_second": 3.709,
|
| 1205 |
-
"step": 23000
|
| 1206 |
-
},
|
| 1207 |
-
{
|
| 1208 |
-
"epoch": 4.6,
|
| 1209 |
-
"eval_loss": 0.10861887782812119,
|
| 1210 |
-
"eval_runtime": 150.9451,
|
| 1211 |
-
"eval_samples_per_second": 949.59,
|
| 1212 |
-
"eval_steps_per_second": 3.71,
|
| 1213 |
-
"step": 23200
|
| 1214 |
-
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 4.64,
|
| 1217 |
-
"eval_loss": 0.10863388329744339,
|
| 1218 |
-
"eval_runtime": 150.9571,
|
| 1219 |
-
"eval_samples_per_second": 949.515,
|
| 1220 |
-
"eval_steps_per_second": 3.71,
|
| 1221 |
-
"step": 23400
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 4.66,
|
| 1225 |
-
"learning_rate": 3.3637626513197067e-06,
|
| 1226 |
-
"loss": 0.1087,
|
| 1227 |
-
"step": 23500
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 4.68,
|
| 1231 |
-
"eval_loss": 0.10861673206090927,
|
| 1232 |
-
"eval_runtime": 150.8,
|
| 1233 |
-
"eval_samples_per_second": 950.504,
|
| 1234 |
-
"eval_steps_per_second": 3.714,
|
| 1235 |
-
"step": 23600
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 4.72,
|
| 1239 |
-
"eval_loss": 0.10860313475131989,
|
| 1240 |
-
"eval_runtime": 150.9239,
|
| 1241 |
-
"eval_samples_per_second": 949.724,
|
| 1242 |
-
"eval_steps_per_second": 3.71,
|
| 1243 |
-
"step": 23800
|
| 1244 |
-
},
|
| 1245 |
-
{
|
| 1246 |
-
"epoch": 4.76,
|
| 1247 |
-
"learning_rate": 2.371502282198849e-06,
|
| 1248 |
-
"loss": 0.1105,
|
| 1249 |
-
"step": 24000
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 4.76,
|
| 1253 |
-
"eval_loss": 0.10860700160264969,
|
| 1254 |
-
"eval_runtime": 151.02,
|
| 1255 |
-
"eval_samples_per_second": 949.119,
|
| 1256 |
-
"eval_steps_per_second": 3.708,
|
| 1257 |
-
"step": 24000
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 4.8,
|
| 1261 |
-
"eval_loss": 0.1085989773273468,
|
| 1262 |
-
"eval_runtime": 151.0214,
|
| 1263 |
-
"eval_samples_per_second": 949.111,
|
| 1264 |
-
"eval_steps_per_second": 3.708,
|
| 1265 |
-
"step": 24200
|
| 1266 |
-
},
|
| 1267 |
-
{
|
| 1268 |
-
"epoch": 4.84,
|
| 1269 |
-
"eval_loss": 0.10860409587621689,
|
| 1270 |
-
"eval_runtime": 151.0306,
|
| 1271 |
-
"eval_samples_per_second": 949.053,
|
| 1272 |
-
"eval_steps_per_second": 3.708,
|
| 1273 |
-
"step": 24400
|
| 1274 |
-
},
|
| 1275 |
-
{
|
| 1276 |
-
"epoch": 4.86,
|
| 1277 |
-
"learning_rate": 1.3792419130779918e-06,
|
| 1278 |
-
"loss": 0.1092,
|
| 1279 |
-
"step": 24500
|
| 1280 |
-
},
|
| 1281 |
-
{
|
| 1282 |
-
"epoch": 4.88,
|
| 1283 |
-
"eval_loss": 0.10858762264251709,
|
| 1284 |
-
"eval_runtime": 150.7929,
|
| 1285 |
-
"eval_samples_per_second": 950.549,
|
| 1286 |
-
"eval_steps_per_second": 3.714,
|
| 1287 |
-
"step": 24600
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 4.92,
|
| 1291 |
-
"eval_loss": 0.10861339420080185,
|
| 1292 |
-
"eval_runtime": 150.9395,
|
| 1293 |
-
"eval_samples_per_second": 949.626,
|
| 1294 |
-
"eval_steps_per_second": 3.71,
|
| 1295 |
-
"step": 24800
|
| 1296 |
-
},
|
| 1297 |
-
{
|
| 1298 |
-
"epoch": 4.96,
|
| 1299 |
-
"learning_rate": 3.869815439571344e-07,
|
| 1300 |
-
"loss": 0.108,
|
| 1301 |
-
"step": 25000
|
| 1302 |
-
},
|
| 1303 |
-
{
|
| 1304 |
-
"epoch": 4.96,
|
| 1305 |
-
"eval_loss": 0.10859096795320511,
|
| 1306 |
-
"eval_runtime": 150.9591,
|
| 1307 |
-
"eval_samples_per_second": 949.502,
|
| 1308 |
-
"eval_steps_per_second": 3.71,
|
| 1309 |
-
"step": 25000
|
| 1310 |
}
|
| 1311 |
],
|
| 1312 |
"logging_steps": 500,
|
| 1313 |
-
"max_steps":
|
| 1314 |
-
"num_train_epochs":
|
| 1315 |
"save_steps": 1000,
|
| 1316 |
-
"total_flos":
|
| 1317 |
"trial_name": null,
|
| 1318 |
"trial_params": null
|
| 1319 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.19845207382417146,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.04,
|
| 13 |
+
"eval_loss": 0.24929340183734894,
|
| 14 |
+
"eval_runtime": 152.0524,
|
| 15 |
+
"eval_samples_per_second": 942.675,
|
| 16 |
+
"eval_steps_per_second": 3.683,
|
| 17 |
"step": 200
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.08,
|
| 21 |
+
"eval_loss": 0.397050142288208,
|
| 22 |
+
"eval_runtime": 151.7752,
|
| 23 |
+
"eval_samples_per_second": 944.397,
|
| 24 |
+
"eval_steps_per_second": 3.69,
|
| 25 |
"step": 400
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.1,
|
| 29 |
+
"learning_rate": 4.751934907719786e-05,
|
| 30 |
+
"loss": 0.4919,
|
| 31 |
"step": 500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.12,
|
| 35 |
+
"eval_loss": 0.619749903678894,
|
| 36 |
+
"eval_runtime": 153.0348,
|
| 37 |
+
"eval_samples_per_second": 936.624,
|
| 38 |
+
"eval_steps_per_second": 3.659,
|
| 39 |
"step": 600
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.16,
|
| 43 |
+
"eval_loss": 0.5482326149940491,
|
| 44 |
+
"eval_runtime": 151.3209,
|
| 45 |
+
"eval_samples_per_second": 947.232,
|
| 46 |
+
"eval_steps_per_second": 3.701,
|
| 47 |
"step": 800
|
| 48 |
},
|
| 49 |
{
|
| 50 |
"epoch": 0.2,
|
| 51 |
+
"learning_rate": 4.5038698154395716e-05,
|
| 52 |
+
"loss": 0.9307,
|
| 53 |
"step": 1000
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.2,
|
| 57 |
+
"eval_loss": 0.861940324306488,
|
| 58 |
+
"eval_runtime": 150.9686,
|
| 59 |
+
"eval_samples_per_second": 949.442,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
"eval_steps_per_second": 3.709,
|
| 61 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
}
|
| 63 |
],
|
| 64 |
"logging_steps": 500,
|
| 65 |
+
"max_steps": 10078,
|
| 66 |
+
"num_train_epochs": 2,
|
| 67 |
"save_steps": 1000,
|
| 68 |
+
"total_flos": 1.6722690048e+16,
|
| 69 |
"trial_name": null,
|
| 70 |
"trial_params": null
|
| 71 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4027
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12d4324537606b6fc0897b8a9dc3e771514e3581e4386565dc3d3bd2f4e1e518
|
| 3 |
size 4027
|