Training in progress, step 5000, checkpoint
Browse files
last-checkpoint/adapter_config.json
CHANGED
|
@@ -8,16 +8,16 @@
|
|
| 8 |
"init_lora_weights": true,
|
| 9 |
"layers_pattern": null,
|
| 10 |
"layers_to_transform": null,
|
| 11 |
-
"lora_alpha":
|
| 12 |
"lora_dropout": 0.1,
|
| 13 |
"modules_to_save": null,
|
| 14 |
"peft_type": "LORA",
|
| 15 |
-
"r":
|
| 16 |
"rank_pattern": {},
|
| 17 |
"revision": null,
|
| 18 |
"target_modules": [
|
| 19 |
-
"
|
| 20 |
-
"
|
| 21 |
],
|
| 22 |
"task_type": "SEQ_2_SEQ_LM"
|
| 23 |
}
|
|
|
|
| 8 |
"init_lora_weights": true,
|
| 9 |
"layers_pattern": null,
|
| 10 |
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 32,
|
| 12 |
"lora_dropout": 0.1,
|
| 13 |
"modules_to_save": null,
|
| 14 |
"peft_type": "LORA",
|
| 15 |
+
"r": 8,
|
| 16 |
"rank_pattern": {},
|
| 17 |
"revision": null,
|
| 18 |
"target_modules": [
|
| 19 |
+
"v",
|
| 20 |
+
"q"
|
| 21 |
],
|
| 22 |
"task_type": "SEQ_2_SEQ_LM"
|
| 23 |
}
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99bb7dd9f9d339c1b6ea8b7c4913f5e5a59ddcf1d651ea8af246503e39076873
|
| 3 |
+
size 18915040
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce6d1c7f359ee86eddcfca26e584b2e996e19c9d2397f3e83bb6b92e27e4a9a8
|
| 3 |
+
size 37990394
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82fcab858db719cf57493411f22ed07e25f5e61522b33c35190b4108d341449e
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bafc5641cb904ce836a25edd0b40d466bd94a372c40e8a43b0b01499a11c7b0c
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,860 +1,79 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint":
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.01,
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 4.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
-
{
|
| 18 |
-
"epoch": 0.01,
|
| 19 |
-
"learning_rate": 9.999998551451928e-05,
|
| 20 |
-
"loss": 3.7798,
|
| 21 |
-
"step": 1000
|
| 22 |
-
},
|
| 23 |
-
{
|
| 24 |
-
"epoch": 0.02,
|
| 25 |
-
"learning_rate": 9.999994205808551e-05,
|
| 26 |
-
"loss": 3.6154,
|
| 27 |
-
"step": 1500
|
| 28 |
-
},
|
| 29 |
{
|
| 30 |
"epoch": 0.03,
|
| 31 |
-
"learning_rate":
|
| 32 |
-
"loss": 3.
|
| 33 |
-
"step":
|
| 34 |
-
},
|
| 35 |
-
{
|
| 36 |
-
"epoch": 0.04,
|
| 37 |
-
"learning_rate": 9.999976823247632e-05,
|
| 38 |
-
"loss": 3.5877,
|
| 39 |
-
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.04,
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 3.
|
| 45 |
-
"step":
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.05,
|
| 49 |
-
"learning_rate": 9.999947852357531e-05,
|
| 50 |
-
"loss": 3.5804,
|
| 51 |
-
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.06,
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 3.
|
| 57 |
-
"step":
|
| 58 |
-
},
|
| 59 |
-
{
|
| 60 |
-
"epoch": 0.07,
|
| 61 |
-
"learning_rate": 9.999907293205393e-05,
|
| 62 |
-
"loss": 3.4906,
|
| 63 |
-
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.07,
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 3.
|
| 69 |
-
"step":
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.08,
|
| 73 |
-
"learning_rate": 9.99985514588522e-05,
|
| 74 |
-
"loss": 3.4256,
|
| 75 |
-
"step": 5500
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.09,
|
| 79 |
-
"learning_rate": 9.99982472669884e-05,
|
| 80 |
-
"loss": 3.4774,
|
| 81 |
-
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.09,
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 3.
|
| 87 |
-
"step":
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.1,
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 3.
|
| 93 |
-
"step":
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.11,
|
| 97 |
-
"learning_rate": 9.999716087251072e-05,
|
| 98 |
-
"loss": 3.4648,
|
| 99 |
-
"step": 7500
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.12,
|
| 103 |
-
"learning_rate": 9.99967408020888e-05,
|
| 104 |
-
"loss": 3.4705,
|
| 105 |
-
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.12,
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 3.
|
| 111 |
-
"step":
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"epoch": 0.13,
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 3.
|
| 117 |
-
"step":
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"epoch": 0.14,
|
| 121 |
-
"learning_rate": 9.999530677744258e-05,
|
| 122 |
-
"loss": 3.4732,
|
| 123 |
-
"step": 9500
|
| 124 |
-
},
|
| 125 |
-
{
|
| 126 |
-
"epoch": 0.15,
|
| 127 |
-
"learning_rate": 9.999477083235691e-05,
|
| 128 |
-
"loss": 3.3216,
|
| 129 |
-
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.15,
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 3.
|
| 135 |
-
"step":
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.16,
|
| 139 |
-
"learning_rate": 9.999361203871817e-05,
|
| 140 |
-
"loss": 3.3641,
|
| 141 |
-
"step": 11000
|
| 142 |
-
},
|
| 143 |
-
{
|
| 144 |
-
"epoch": 0.17,
|
| 145 |
-
"learning_rate": 9.999298919083656e-05,
|
| 146 |
-
"loss": 3.4407,
|
| 147 |
-
"step": 11500
|
| 148 |
-
},
|
| 149 |
-
{
|
| 150 |
-
"epoch": 0.17,
|
| 151 |
-
"learning_rate": 9.99923373760557e-05,
|
| 152 |
-
"loss": 3.3962,
|
| 153 |
-
"step": 12000
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.18,
|
| 157 |
-
"learning_rate": 9.999165659475324e-05,
|
| 158 |
-
"loss": 3.3776,
|
| 159 |
-
"step": 12500
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.19,
|
| 163 |
-
"learning_rate": 9.999094684732369e-05,
|
| 164 |
-
"loss": 3.3157,
|
| 165 |
-
"step": 13000
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 0.2,
|
| 169 |
-
"learning_rate": 9.999020813417826e-05,
|
| 170 |
-
"loss": 3.2517,
|
| 171 |
-
"step": 13500
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.2,
|
| 175 |
-
"learning_rate": 9.998944045574499e-05,
|
| 176 |
-
"loss": 3.4232,
|
| 177 |
-
"step": 14000
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.21,
|
| 181 |
-
"learning_rate": 9.998864381246869e-05,
|
| 182 |
-
"loss": 3.3539,
|
| 183 |
-
"step": 14500
|
| 184 |
-
},
|
| 185 |
-
{
|
| 186 |
-
"epoch": 0.22,
|
| 187 |
-
"learning_rate": 9.998781820481091e-05,
|
| 188 |
-
"loss": 3.2431,
|
| 189 |
-
"step": 15000
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.23,
|
| 193 |
-
"learning_rate": 9.998696363325009e-05,
|
| 194 |
-
"loss": 3.2796,
|
| 195 |
-
"step": 15500
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.23,
|
| 199 |
-
"learning_rate": 9.998608009828132e-05,
|
| 200 |
-
"loss": 3.3468,
|
| 201 |
-
"step": 16000
|
| 202 |
-
},
|
| 203 |
-
{
|
| 204 |
-
"epoch": 0.24,
|
| 205 |
-
"learning_rate": 9.998516760041659e-05,
|
| 206 |
-
"loss": 3.3159,
|
| 207 |
-
"step": 16500
|
| 208 |
-
},
|
| 209 |
-
{
|
| 210 |
-
"epoch": 0.25,
|
| 211 |
-
"learning_rate": 9.998422614018456e-05,
|
| 212 |
-
"loss": 3.3635,
|
| 213 |
-
"step": 17000
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.25,
|
| 217 |
-
"learning_rate": 9.998325571813079e-05,
|
| 218 |
-
"loss": 3.3708,
|
| 219 |
-
"step": 17500
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.26,
|
| 223 |
-
"learning_rate": 9.998225633481753e-05,
|
| 224 |
-
"loss": 3.3435,
|
| 225 |
-
"step": 18000
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.27,
|
| 229 |
-
"learning_rate": 9.998122799082386e-05,
|
| 230 |
-
"loss": 3.3649,
|
| 231 |
-
"step": 18500
|
| 232 |
-
},
|
| 233 |
-
{
|
| 234 |
-
"epoch": 0.28,
|
| 235 |
-
"learning_rate": 9.998017068674558e-05,
|
| 236 |
-
"loss": 3.3519,
|
| 237 |
-
"step": 19000
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 0.28,
|
| 241 |
-
"learning_rate": 9.997908442319536e-05,
|
| 242 |
-
"loss": 3.3408,
|
| 243 |
-
"step": 19500
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 0.29,
|
| 247 |
-
"learning_rate": 9.99779692008026e-05,
|
| 248 |
-
"loss": 3.2678,
|
| 249 |
-
"step": 20000
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.3,
|
| 253 |
-
"learning_rate": 9.997682502021345e-05,
|
| 254 |
-
"loss": 3.2453,
|
| 255 |
-
"step": 20500
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.31,
|
| 259 |
-
"learning_rate": 9.997565188209089e-05,
|
| 260 |
-
"loss": 3.2948,
|
| 261 |
-
"step": 21000
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.31,
|
| 265 |
-
"learning_rate": 9.997444978711465e-05,
|
| 266 |
-
"loss": 3.2849,
|
| 267 |
-
"step": 21500
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 0.32,
|
| 271 |
-
"learning_rate": 9.997321873598125e-05,
|
| 272 |
-
"loss": 3.3274,
|
| 273 |
-
"step": 22000
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 0.33,
|
| 277 |
-
"learning_rate": 9.9971958729404e-05,
|
| 278 |
-
"loss": 3.3084,
|
| 279 |
-
"step": 22500
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 0.33,
|
| 283 |
-
"learning_rate": 9.997066976811294e-05,
|
| 284 |
-
"loss": 3.3019,
|
| 285 |
-
"step": 23000
|
| 286 |
-
},
|
| 287 |
-
{
|
| 288 |
-
"epoch": 0.34,
|
| 289 |
-
"learning_rate": 9.996935185285495e-05,
|
| 290 |
-
"loss": 3.2998,
|
| 291 |
-
"step": 23500
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.35,
|
| 295 |
-
"learning_rate": 9.996800498439362e-05,
|
| 296 |
-
"loss": 3.268,
|
| 297 |
-
"step": 24000
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.36,
|
| 301 |
-
"learning_rate": 9.99666291635094e-05,
|
| 302 |
-
"loss": 3.248,
|
| 303 |
-
"step": 24500
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.36,
|
| 307 |
-
"learning_rate": 9.996522439099943e-05,
|
| 308 |
-
"loss": 3.2204,
|
| 309 |
-
"step": 25000
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 0.37,
|
| 313 |
-
"learning_rate": 9.99637906676777e-05,
|
| 314 |
-
"loss": 3.2922,
|
| 315 |
-
"step": 25500
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 0.38,
|
| 319 |
-
"learning_rate": 9.996232799437487e-05,
|
| 320 |
-
"loss": 3.2716,
|
| 321 |
-
"step": 26000
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 0.39,
|
| 325 |
-
"learning_rate": 9.996083637193849e-05,
|
| 326 |
-
"loss": 3.3181,
|
| 327 |
-
"step": 26500
|
| 328 |
-
},
|
| 329 |
-
{
|
| 330 |
-
"epoch": 0.39,
|
| 331 |
-
"learning_rate": 9.995931580123284e-05,
|
| 332 |
-
"loss": 3.2899,
|
| 333 |
-
"step": 27000
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 0.4,
|
| 337 |
-
"learning_rate": 9.995776628313896e-05,
|
| 338 |
-
"loss": 3.266,
|
| 339 |
-
"step": 27500
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.41,
|
| 343 |
-
"learning_rate": 9.995618781855464e-05,
|
| 344 |
-
"loss": 3.3644,
|
| 345 |
-
"step": 28000
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.41,
|
| 349 |
-
"learning_rate": 9.995458040839452e-05,
|
| 350 |
-
"loss": 3.2413,
|
| 351 |
-
"step": 28500
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 0.42,
|
| 355 |
-
"learning_rate": 9.995294405358993e-05,
|
| 356 |
-
"loss": 3.2759,
|
| 357 |
-
"step": 29000
|
| 358 |
-
},
|
| 359 |
-
{
|
| 360 |
-
"epoch": 0.43,
|
| 361 |
-
"learning_rate": 9.995127875508903e-05,
|
| 362 |
-
"loss": 3.3065,
|
| 363 |
-
"step": 29500
|
| 364 |
-
},
|
| 365 |
-
{
|
| 366 |
-
"epoch": 0.44,
|
| 367 |
-
"learning_rate": 9.99495845138567e-05,
|
| 368 |
-
"loss": 3.3028,
|
| 369 |
-
"step": 30000
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.44,
|
| 373 |
-
"learning_rate": 9.994786133087464e-05,
|
| 374 |
-
"loss": 3.2613,
|
| 375 |
-
"step": 30500
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 0.45,
|
| 379 |
-
"learning_rate": 9.994610920714126e-05,
|
| 380 |
-
"loss": 3.2137,
|
| 381 |
-
"step": 31000
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.46,
|
| 385 |
-
"learning_rate": 9.994432814367183e-05,
|
| 386 |
-
"loss": 3.2899,
|
| 387 |
-
"step": 31500
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.47,
|
| 391 |
-
"learning_rate": 9.99425181414983e-05,
|
| 392 |
-
"loss": 3.2877,
|
| 393 |
-
"step": 32000
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 0.47,
|
| 397 |
-
"learning_rate": 9.994067920166939e-05,
|
| 398 |
-
"loss": 3.2139,
|
| 399 |
-
"step": 32500
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 0.48,
|
| 403 |
-
"learning_rate": 9.993881132525067e-05,
|
| 404 |
-
"loss": 3.2407,
|
| 405 |
-
"step": 33000
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 0.49,
|
| 409 |
-
"learning_rate": 9.993691451332439e-05,
|
| 410 |
-
"loss": 3.3226,
|
| 411 |
-
"step": 33500
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 0.49,
|
| 415 |
-
"learning_rate": 9.993498876698963e-05,
|
| 416 |
-
"loss": 3.2004,
|
| 417 |
-
"step": 34000
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 0.5,
|
| 421 |
-
"learning_rate": 9.993303408736217e-05,
|
| 422 |
-
"loss": 3.2773,
|
| 423 |
-
"step": 34500
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.51,
|
| 427 |
-
"learning_rate": 9.993105047557461e-05,
|
| 428 |
-
"loss": 3.2183,
|
| 429 |
-
"step": 35000
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.52,
|
| 433 |
-
"learning_rate": 9.992903793277628e-05,
|
| 434 |
-
"loss": 3.2648,
|
| 435 |
-
"step": 35500
|
| 436 |
-
},
|
| 437 |
-
{
|
| 438 |
-
"epoch": 0.52,
|
| 439 |
-
"learning_rate": 9.99269964601333e-05,
|
| 440 |
-
"loss": 3.1963,
|
| 441 |
-
"step": 36000
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"epoch": 0.53,
|
| 445 |
-
"learning_rate": 9.992492605882853e-05,
|
| 446 |
-
"loss": 3.2825,
|
| 447 |
-
"step": 36500
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 0.54,
|
| 451 |
-
"learning_rate": 9.99228267300616e-05,
|
| 452 |
-
"loss": 3.2658,
|
| 453 |
-
"step": 37000
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 0.55,
|
| 457 |
-
"learning_rate": 9.992069847504891e-05,
|
| 458 |
-
"loss": 3.1058,
|
| 459 |
-
"step": 37500
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 0.55,
|
| 463 |
-
"learning_rate": 9.99185412950236e-05,
|
| 464 |
-
"loss": 3.1375,
|
| 465 |
-
"step": 38000
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.56,
|
| 469 |
-
"learning_rate": 9.991635519123559e-05,
|
| 470 |
-
"loss": 3.2075,
|
| 471 |
-
"step": 38500
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.57,
|
| 475 |
-
"learning_rate": 9.991414016495155e-05,
|
| 476 |
-
"loss": 3.2228,
|
| 477 |
-
"step": 39000
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 0.57,
|
| 481 |
-
"learning_rate": 9.99118962174549e-05,
|
| 482 |
-
"loss": 3.2004,
|
| 483 |
-
"step": 39500
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 0.58,
|
| 487 |
-
"learning_rate": 9.990962335004584e-05,
|
| 488 |
-
"loss": 3.1924,
|
| 489 |
-
"step": 40000
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 0.59,
|
| 493 |
-
"learning_rate": 9.99073215640413e-05,
|
| 494 |
-
"loss": 3.2956,
|
| 495 |
-
"step": 40500
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 0.6,
|
| 499 |
-
"learning_rate": 9.990499086077498e-05,
|
| 500 |
-
"loss": 3.2698,
|
| 501 |
-
"step": 41000
|
| 502 |
-
},
|
| 503 |
-
{
|
| 504 |
-
"epoch": 0.6,
|
| 505 |
-
"learning_rate": 9.990263124159736e-05,
|
| 506 |
-
"loss": 3.2863,
|
| 507 |
-
"step": 41500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.61,
|
| 511 |
-
"learning_rate": 9.990024270787561e-05,
|
| 512 |
-
"loss": 3.3611,
|
| 513 |
-
"step": 42000
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.62,
|
| 517 |
-
"learning_rate": 9.989782526099372e-05,
|
| 518 |
-
"loss": 3.1691,
|
| 519 |
-
"step": 42500
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 0.62,
|
| 523 |
-
"learning_rate": 9.989537890235238e-05,
|
| 524 |
-
"loss": 3.1085,
|
| 525 |
-
"step": 43000
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 0.63,
|
| 529 |
-
"learning_rate": 9.989290363336908e-05,
|
| 530 |
-
"loss": 3.1825,
|
| 531 |
-
"step": 43500
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 0.64,
|
| 535 |
-
"learning_rate": 9.989039945547803e-05,
|
| 536 |
-
"loss": 3.2333,
|
| 537 |
-
"step": 44000
|
| 538 |
-
},
|
| 539 |
-
{
|
| 540 |
-
"epoch": 0.65,
|
| 541 |
-
"learning_rate": 9.98878663701302e-05,
|
| 542 |
-
"loss": 3.2635,
|
| 543 |
-
"step": 44500
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 0.65,
|
| 547 |
-
"learning_rate": 9.988530437879333e-05,
|
| 548 |
-
"loss": 3.2907,
|
| 549 |
-
"step": 45000
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.66,
|
| 553 |
-
"learning_rate": 9.988271348295184e-05,
|
| 554 |
-
"loss": 3.1334,
|
| 555 |
-
"step": 45500
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.67,
|
| 559 |
-
"learning_rate": 9.988009368410698e-05,
|
| 560 |
-
"loss": 3.2239,
|
| 561 |
-
"step": 46000
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 0.68,
|
| 565 |
-
"learning_rate": 9.98774449837767e-05,
|
| 566 |
-
"loss": 3.1904,
|
| 567 |
-
"step": 46500
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"epoch": 0.68,
|
| 571 |
-
"learning_rate": 9.987476738349571e-05,
|
| 572 |
-
"loss": 3.2781,
|
| 573 |
-
"step": 47000
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 0.69,
|
| 577 |
-
"learning_rate": 9.987206088481545e-05,
|
| 578 |
-
"loss": 3.19,
|
| 579 |
-
"step": 47500
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 0.7,
|
| 583 |
-
"learning_rate": 9.986932548930414e-05,
|
| 584 |
-
"loss": 3.2235,
|
| 585 |
-
"step": 48000
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 0.7,
|
| 589 |
-
"learning_rate": 9.986656119854672e-05,
|
| 590 |
-
"loss": 3.2302,
|
| 591 |
-
"step": 48500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.71,
|
| 595 |
-
"learning_rate": 9.986376801414485e-05,
|
| 596 |
-
"loss": 3.289,
|
| 597 |
-
"step": 49000
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.72,
|
| 601 |
-
"learning_rate": 9.986094593771699e-05,
|
| 602 |
-
"loss": 3.2874,
|
| 603 |
-
"step": 49500
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 0.73,
|
| 607 |
-
"learning_rate": 9.985809497089827e-05,
|
| 608 |
-
"loss": 3.144,
|
| 609 |
-
"step": 50000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 0.73,
|
| 613 |
-
"learning_rate": 9.985521511534062e-05,
|
| 614 |
-
"loss": 3.1967,
|
| 615 |
-
"step": 50500
|
| 616 |
-
},
|
| 617 |
-
{
|
| 618 |
-
"epoch": 0.74,
|
| 619 |
-
"learning_rate": 9.985230637271266e-05,
|
| 620 |
-
"loss": 3.248,
|
| 621 |
-
"step": 51000
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 0.75,
|
| 625 |
-
"learning_rate": 9.984936874469979e-05,
|
| 626 |
-
"loss": 3.1933,
|
| 627 |
-
"step": 51500
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 0.76,
|
| 631 |
-
"learning_rate": 9.984640223300413e-05,
|
| 632 |
-
"loss": 3.2108,
|
| 633 |
-
"step": 52000
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.76,
|
| 637 |
-
"learning_rate": 9.98434068393445e-05,
|
| 638 |
-
"loss": 3.0986,
|
| 639 |
-
"step": 52500
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.77,
|
| 643 |
-
"learning_rate": 9.984038256545653e-05,
|
| 644 |
-
"loss": 3.2016,
|
| 645 |
-
"step": 53000
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 0.78,
|
| 649 |
-
"learning_rate": 9.983732941309253e-05,
|
| 650 |
-
"loss": 3.1967,
|
| 651 |
-
"step": 53500
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 0.78,
|
| 655 |
-
"learning_rate": 9.983424738402156e-05,
|
| 656 |
-
"loss": 3.2803,
|
| 657 |
-
"step": 54000
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"epoch": 0.79,
|
| 661 |
-
"learning_rate": 9.98311364800294e-05,
|
| 662 |
-
"loss": 3.2952,
|
| 663 |
-
"step": 54500
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 0.8,
|
| 667 |
-
"learning_rate": 9.982799670291857e-05,
|
| 668 |
-
"loss": 3.18,
|
| 669 |
-
"step": 55000
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.81,
|
| 673 |
-
"learning_rate": 9.98248280545083e-05,
|
| 674 |
-
"loss": 3.1768,
|
| 675 |
-
"step": 55500
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 0.81,
|
| 679 |
-
"learning_rate": 9.982163053663459e-05,
|
| 680 |
-
"loss": 3.1986,
|
| 681 |
-
"step": 56000
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 0.82,
|
| 685 |
-
"learning_rate": 9.981840415115014e-05,
|
| 686 |
-
"loss": 3.2241,
|
| 687 |
-
"step": 56500
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 0.83,
|
| 691 |
-
"learning_rate": 9.981514889992436e-05,
|
| 692 |
-
"loss": 3.2754,
|
| 693 |
-
"step": 57000
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"epoch": 0.84,
|
| 697 |
-
"learning_rate": 9.981186478484344e-05,
|
| 698 |
-
"loss": 3.2162,
|
| 699 |
-
"step": 57500
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 0.84,
|
| 703 |
-
"learning_rate": 9.980855180781021e-05,
|
| 704 |
-
"loss": 3.2659,
|
| 705 |
-
"step": 58000
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 0.85,
|
| 709 |
-
"learning_rate": 9.980520997074432e-05,
|
| 710 |
-
"loss": 3.172,
|
| 711 |
-
"step": 58500
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 0.86,
|
| 715 |
-
"learning_rate": 9.980183927558207e-05,
|
| 716 |
-
"loss": 3.3035,
|
| 717 |
-
"step": 59000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 0.86,
|
| 721 |
-
"learning_rate": 9.97984397242765e-05,
|
| 722 |
-
"loss": 3.1746,
|
| 723 |
-
"step": 59500
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 0.87,
|
| 727 |
-
"learning_rate": 9.979501131879741e-05,
|
| 728 |
-
"loss": 3.1678,
|
| 729 |
-
"step": 60000
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 0.88,
|
| 733 |
-
"learning_rate": 9.979155406113124e-05,
|
| 734 |
-
"loss": 3.162,
|
| 735 |
-
"step": 60500
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 0.89,
|
| 739 |
-
"learning_rate": 9.978806795328121e-05,
|
| 740 |
-
"loss": 3.3041,
|
| 741 |
-
"step": 61000
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 0.89,
|
| 745 |
-
"learning_rate": 9.978455299726726e-05,
|
| 746 |
-
"loss": 3.2086,
|
| 747 |
-
"step": 61500
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 0.9,
|
| 751 |
-
"learning_rate": 9.978100919512598e-05,
|
| 752 |
-
"loss": 3.1832,
|
| 753 |
-
"step": 62000
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 0.91,
|
| 757 |
-
"learning_rate": 9.977743654891077e-05,
|
| 758 |
-
"loss": 3.1839,
|
| 759 |
-
"step": 62500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.92,
|
| 763 |
-
"learning_rate": 9.977383506069164e-05,
|
| 764 |
-
"loss": 3.123,
|
| 765 |
-
"step": 63000
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 0.92,
|
| 769 |
-
"learning_rate": 9.977020473255539e-05,
|
| 770 |
-
"loss": 3.2868,
|
| 771 |
-
"step": 63500
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 0.93,
|
| 775 |
-
"learning_rate": 9.976654556660548e-05,
|
| 776 |
-
"loss": 3.1488,
|
| 777 |
-
"step": 64000
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 0.94,
|
| 781 |
-
"learning_rate": 9.976285756496211e-05,
|
| 782 |
-
"loss": 3.2116,
|
| 783 |
-
"step": 64500
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"epoch": 0.94,
|
| 787 |
-
"learning_rate": 9.97591407297622e-05,
|
| 788 |
-
"loss": 3.2233,
|
| 789 |
-
"step": 65000
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 0.95,
|
| 793 |
-
"learning_rate": 9.975539506315933e-05,
|
| 794 |
-
"loss": 3.1018,
|
| 795 |
-
"step": 65500
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 0.96,
|
| 799 |
-
"learning_rate": 9.975162056732385e-05,
|
| 800 |
-
"loss": 3.1404,
|
| 801 |
-
"step": 66000
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 0.97,
|
| 805 |
-
"learning_rate": 9.974781724444272e-05,
|
| 806 |
-
"loss": 3.1984,
|
| 807 |
-
"step": 66500
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 0.97,
|
| 811 |
-
"learning_rate": 9.974398509671969e-05,
|
| 812 |
-
"loss": 3.2421,
|
| 813 |
-
"step": 67000
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 0.98,
|
| 817 |
-
"learning_rate": 9.974012412637517e-05,
|
| 818 |
-
"loss": 3.1499,
|
| 819 |
-
"step": 67500
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 0.99,
|
| 823 |
-
"learning_rate": 9.97362343356463e-05,
|
| 824 |
-
"loss": 3.174,
|
| 825 |
-
"step": 68000
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 1.0,
|
| 829 |
-
"learning_rate": 9.973231572678686e-05,
|
| 830 |
-
"loss": 3.1816,
|
| 831 |
-
"step": 68500
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 1.0,
|
| 835 |
-
"eval_bleu": 1.0,
|
| 836 |
-
"eval_brevity_penalty": 1.0,
|
| 837 |
-
"eval_length_ratio": 1.0,
|
| 838 |
-
"eval_loss": 3.016300916671753,
|
| 839 |
-
"eval_precisions": [
|
| 840 |
-
1.0,
|
| 841 |
-
1.0,
|
| 842 |
-
1.0,
|
| 843 |
-
1.0
|
| 844 |
-
],
|
| 845 |
-
"eval_reference_length": 7828480,
|
| 846 |
-
"eval_runtime": 20103.3013,
|
| 847 |
-
"eval_samples_per_second": 0.761,
|
| 848 |
-
"eval_steps_per_second": 0.38,
|
| 849 |
-
"eval_translation_length": 7828480,
|
| 850 |
-
"step": 68803
|
| 851 |
}
|
| 852 |
],
|
| 853 |
"logging_steps": 500,
|
| 854 |
-
"max_steps":
|
| 855 |
-
"num_train_epochs":
|
| 856 |
-
"save_steps":
|
| 857 |
-
"total_flos": 1.
|
| 858 |
"trial_name": null,
|
| 859 |
"trial_params": null
|
| 860 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.14534249960030812,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 4e-05,
|
| 14 |
+
"loss": 4.4951,
|
| 15 |
"step": 500
|
| 16 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
{
|
| 18 |
"epoch": 0.03,
|
| 19 |
+
"learning_rate": 8e-05,
|
| 20 |
+
"loss": 3.777,
|
| 21 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.04,
|
| 25 |
+
"learning_rate": 7.999831247941866e-05,
|
| 26 |
+
"loss": 3.6246,
|
| 27 |
+
"step": 1500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.06,
|
| 31 |
+
"learning_rate": 7.99932500600609e-05,
|
| 32 |
+
"loss": 3.5067,
|
| 33 |
+
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.07,
|
| 37 |
+
"learning_rate": 7.998481316907362e-05,
|
| 38 |
+
"loss": 3.4947,
|
| 39 |
+
"step": 2500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.09,
|
| 43 |
+
"learning_rate": 7.99730025183281e-05,
|
| 44 |
+
"loss": 3.4452,
|
| 45 |
+
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.1,
|
| 49 |
+
"learning_rate": 7.995781910436019e-05,
|
| 50 |
+
"loss": 3.3696,
|
| 51 |
+
"step": 3500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.12,
|
| 55 |
+
"learning_rate": 7.993926420828609e-05,
|
| 56 |
+
"loss": 3.4226,
|
| 57 |
+
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.13,
|
| 61 |
+
"learning_rate": 7.991733939569422e-05,
|
| 62 |
+
"loss": 3.3765,
|
| 63 |
+
"step": 4500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.15,
|
| 67 |
+
"learning_rate": 7.989204651651322e-05,
|
| 68 |
+
"loss": 3.4237,
|
| 69 |
+
"step": 5000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 70 |
}
|
| 71 |
],
|
| 72 |
"logging_steps": 500,
|
| 73 |
+
"max_steps": 172005,
|
| 74 |
+
"num_train_epochs": 5,
|
| 75 |
+
"save_steps": 5000,
|
| 76 |
+
"total_flos": 1.7133622788096e+17,
|
| 77 |
"trial_name": null,
|
| 78 |
"trial_params": null
|
| 79 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac93b83e20cb3266b0249c2b9cb223a898cd7a840eb97f009fc035d490eabb4d
|
| 3 |
size 4728
|