Upload folder using huggingface_hub
Browse files- adapter_config.json +8 -8
- adapter_model.safetensors +2 -2
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +298 -298
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -12,24 +12,24 @@
|
|
| 12 |
"layers_pattern": null,
|
| 13 |
"layers_to_transform": null,
|
| 14 |
"loftq_config": {},
|
| 15 |
-
"lora_alpha":
|
| 16 |
"lora_bias": false,
|
| 17 |
-
"lora_dropout": 0,
|
| 18 |
"megatron_config": null,
|
| 19 |
"megatron_core": "megatron.core",
|
| 20 |
"modules_to_save": null,
|
| 21 |
"peft_type": "LORA",
|
| 22 |
-
"r":
|
| 23 |
"rank_pattern": {},
|
| 24 |
"revision": null,
|
| 25 |
"target_modules": [
|
| 26 |
-
"o_proj",
|
| 27 |
-
"down_proj",
|
| 28 |
"k_proj",
|
| 29 |
-
"q_proj",
|
| 30 |
"gate_proj",
|
| 31 |
-
"
|
| 32 |
-
"
|
|
|
|
|
|
|
|
|
|
| 33 |
],
|
| 34 |
"task_type": "CAUSAL_LM",
|
| 35 |
"use_dora": false,
|
|
|
|
| 12 |
"layers_pattern": null,
|
| 13 |
"layers_to_transform": null,
|
| 14 |
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 96,
|
| 16 |
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.05,
|
| 18 |
"megatron_config": null,
|
| 19 |
"megatron_core": "megatron.core",
|
| 20 |
"modules_to_save": null,
|
| 21 |
"peft_type": "LORA",
|
| 22 |
+
"r": 48,
|
| 23 |
"rank_pattern": {},
|
| 24 |
"revision": null,
|
| 25 |
"target_modules": [
|
|
|
|
|
|
|
| 26 |
"k_proj",
|
|
|
|
| 27 |
"gate_proj",
|
| 28 |
+
"up_proj",
|
| 29 |
+
"o_proj",
|
| 30 |
+
"q_proj",
|
| 31 |
+
"down_proj",
|
| 32 |
+
"v_proj"
|
| 33 |
],
|
| 34 |
"task_type": "CAUSAL_LM",
|
| 35 |
"use_dora": false,
|
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:cba0b76173da997b59cb7a5e1fcd715c9862b2f1bbf3b91d8ff2fc9c798e757d
|
| 3 |
+
size 359270696
|
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:cbb604a23b99a11cbf1db6886d978776df8e140f11a8a277fc8b32f968eb15d7
|
| 3 |
+
size 718831691
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3d312928d0bb60518eb9856d5ab0ae1674bcb745294bf27f615cb6d07b0463e
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5418e0fc9a3a6a50ea3a7b440dfb8b2fa26686b28c8f28256150a09922035962
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -11,702 +11,702 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.05063291139240506,
|
| 14 |
-
"grad_norm":
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
"loss": 3.0474,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.10126582278481013,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate": 3.
|
| 23 |
"loss": 3.2925,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.1518987341772152,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 6.666666666666667e-
|
| 30 |
-
"loss": 3.
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.20253164556962025,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 2.
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.25316455696202533,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.3037974683544304,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 2.
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.35443037974683544,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 2.
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.4050632911392405,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 2.
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.45569620253164556,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 2.
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.5063291139240507,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 2.
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.5569620253164557,
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 2.
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.6075949367088608,
|
| 91 |
-
"grad_norm":
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss": 2.
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.6582278481012658,
|
| 98 |
-
"grad_norm":
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss": 2.
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.7088607594936709,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss": 2.
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.759493670886076,
|
| 112 |
-
"grad_norm":
|
| 113 |
-
"learning_rate":
|
| 114 |
-
"loss": 2.
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.810126582278481,
|
| 119 |
-
"grad_norm":
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss": 2.
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.8607594936708861,
|
| 126 |
-
"grad_norm":
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 2.
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.9113924050632911,
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"loss": 2.
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.9620253164556962,
|
| 140 |
-
"grad_norm":
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss": 2.
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 1.0,
|
| 147 |
-
"grad_norm":
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss": 2.
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 1.0506329113924051,
|
| 154 |
-
"grad_norm":
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss": 1.
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 1.1012658227848102,
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss": 1.
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 1.1518987341772151,
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 1.2025316455696202,
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 1.
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 1.2531645569620253,
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss": 1.
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 1.3037974683544304,
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss":
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 1.3544303797468356,
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss":
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 1.4050632911392404,
|
| 203 |
-
"grad_norm":
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 1.
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 1.4556962025316456,
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 1.5063291139240507,
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 1.
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 1.5569620253164556,
|
| 224 |
-
"grad_norm":
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss": 1.
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 1.6075949367088609,
|
| 231 |
-
"grad_norm":
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss": 1.
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 1.6582278481012658,
|
| 238 |
-
"grad_norm":
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 1.
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 1.7088607594936709,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 1.
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 1.759493670886076,
|
| 252 |
-
"grad_norm":
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 1.
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.810126582278481,
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 1.
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 1.8607594936708862,
|
| 266 |
-
"grad_norm":
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 1.
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 1.9113924050632911,
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss":
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 1.9620253164556962,
|
| 280 |
-
"grad_norm":
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 1.
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 2.0,
|
| 287 |
-
"grad_norm":
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss":
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 2.050632911392405,
|
| 294 |
-
"grad_norm":
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 1.
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 2.1012658227848102,
|
| 301 |
-
"grad_norm":
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss": 1.
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 2.151898734177215,
|
| 308 |
-
"grad_norm":
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss": 1.
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 2.2025316455696204,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss": 1.
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 2.2531645569620253,
|
| 322 |
-
"grad_norm":
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss": 1.
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 2.3037974683544302,
|
| 329 |
-
"grad_norm":
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 2.3544303797468356,
|
| 336 |
-
"grad_norm":
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 1.
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 2.4050632911392404,
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 1.
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 2.4556962025316453,
|
| 350 |
-
"grad_norm":
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 1.
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 2.5063291139240507,
|
| 357 |
-
"grad_norm":
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 2.5569620253164556,
|
| 364 |
-
"grad_norm":
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss":
|
| 367 |
"step": 51
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 2.607594936708861,
|
| 371 |
-
"grad_norm":
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 52
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 2.6582278481012658,
|
| 378 |
-
"grad_norm":
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss": 1.
|
| 381 |
"step": 53
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 2.708860759493671,
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 1.
|
| 388 |
"step": 54
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 2.759493670886076,
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss":
|
| 395 |
"step": 55
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 2.810126582278481,
|
| 399 |
-
"grad_norm":
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss":
|
| 402 |
"step": 56
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 2.8607594936708862,
|
| 406 |
-
"grad_norm":
|
| 407 |
-
"learning_rate":
|
| 408 |
-
"loss":
|
| 409 |
"step": 57
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 2.911392405063291,
|
| 413 |
-
"grad_norm":
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss": 1.
|
| 416 |
"step": 58
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 2.962025316455696,
|
| 420 |
-
"grad_norm":
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 59
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 3.0,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 60
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 3.050632911392405,
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate":
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 61
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 3.1012658227848102,
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate":
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 62
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 3.151898734177215,
|
| 448 |
-
"grad_norm":
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 63
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 3.2025316455696204,
|
| 455 |
-
"grad_norm":
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 64
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 3.2531645569620253,
|
| 462 |
-
"grad_norm":
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 65
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 3.3037974683544302,
|
| 469 |
-
"grad_norm":
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 66
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 3.3544303797468356,
|
| 476 |
-
"grad_norm":
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 67
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 3.4050632911392404,
|
| 483 |
-
"grad_norm":
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss": 0.
|
| 486 |
"step": 68
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 3.4556962025316453,
|
| 490 |
-
"grad_norm":
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 69
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 3.5063291139240507,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss": 0.
|
| 500 |
"step": 70
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 3.5569620253164556,
|
| 504 |
-
"grad_norm":
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 71
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 3.607594936708861,
|
| 511 |
-
"grad_norm":
|
| 512 |
-
"learning_rate":
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 72
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 3.6582278481012658,
|
| 518 |
-
"grad_norm":
|
| 519 |
-
"learning_rate":
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 73
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 3.708860759493671,
|
| 525 |
-
"grad_norm":
|
| 526 |
-
"learning_rate":
|
| 527 |
-
"loss": 0.
|
| 528 |
"step": 74
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 3.759493670886076,
|
| 532 |
-
"grad_norm":
|
| 533 |
-
"learning_rate":
|
| 534 |
-
"loss": 0.
|
| 535 |
"step": 75
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 3.810126582278481,
|
| 539 |
-
"grad_norm":
|
| 540 |
-
"learning_rate": 9.934134090518593e-
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 76
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 3.8607594936708862,
|
| 546 |
-
"grad_norm":
|
| 547 |
-
"learning_rate": 9.
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 77
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 3.911392405063291,
|
| 553 |
-
"grad_norm":
|
| 554 |
-
"learning_rate": 8.50188662803194e-
|
| 555 |
-
"loss": 0.
|
| 556 |
"step": 78
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 3.962025316455696,
|
| 560 |
-
"grad_norm":
|
| 561 |
-
"learning_rate": 7.
|
| 562 |
-
"loss": 0.
|
| 563 |
"step": 79
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 4.0,
|
| 567 |
-
"grad_norm":
|
| 568 |
-
"learning_rate": 7.159746139706194e-
|
| 569 |
-
"loss": 0.
|
| 570 |
"step": 80
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 4.050632911392405,
|
| 574 |
-
"grad_norm":
|
| 575 |
-
"learning_rate": 6.524777069483526e-
|
| 576 |
-
"loss": 0.
|
| 577 |
"step": 81
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 4.10126582278481,
|
| 581 |
-
"grad_norm":
|
| 582 |
-
"learning_rate": 5.
|
| 583 |
-
"loss": 0.
|
| 584 |
"step": 82
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 4.151898734177215,
|
| 588 |
-
"grad_norm":
|
| 589 |
-
"learning_rate": 5.
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 83
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 4.2025316455696204,
|
| 595 |
-
"grad_norm":
|
| 596 |
-
"learning_rate": 4.
|
| 597 |
-
"loss": 0.
|
| 598 |
"step": 84
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 4.253164556962025,
|
| 602 |
-
"grad_norm":
|
| 603 |
-
"learning_rate": 4.245399229611238e-
|
| 604 |
-
"loss": 0.
|
| 605 |
"step": 85
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 4.30379746835443,
|
| 609 |
-
"grad_norm":
|
| 610 |
-
"learning_rate": 3.
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 86
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 4.3544303797468356,
|
| 616 |
-
"grad_norm":
|
| 617 |
-
"learning_rate": 3.
|
| 618 |
-
"loss": 0.
|
| 619 |
"step": 87
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 4.405063291139241,
|
| 623 |
-
"grad_norm":
|
| 624 |
-
"learning_rate": 2.
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 88
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 4.455696202531645,
|
| 630 |
-
"grad_norm":
|
| 631 |
-
"learning_rate": 2.
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 89
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 4.506329113924051,
|
| 637 |
-
"grad_norm":
|
| 638 |
-
"learning_rate": 2.
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 90
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 4.556962025316456,
|
| 644 |
-
"grad_norm":
|
| 645 |
-
"learning_rate": 1.
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 91
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 4.6075949367088604,
|
| 651 |
-
"grad_norm":
|
| 652 |
-
"learning_rate": 1.
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 92
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 4.658227848101266,
|
| 658 |
-
"grad_norm":
|
| 659 |
-
"learning_rate": 1.0848895124889818e-
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 93
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 4.708860759493671,
|
| 665 |
-
"grad_norm":
|
| 666 |
-
"learning_rate": 8.320398932703144e-
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 94
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 4.759493670886076,
|
| 672 |
-
"grad_norm":
|
| 673 |
-
"learning_rate": 6.
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 95
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 4.810126582278481,
|
| 679 |
-
"grad_norm":
|
| 680 |
-
"learning_rate": 4.256725079024554e-
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 96
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 4.860759493670886,
|
| 686 |
-
"grad_norm":
|
| 687 |
-
"learning_rate": 2.
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 97
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 4.911392405063291,
|
| 693 |
-
"grad_norm":
|
| 694 |
-
"learning_rate": 1.5352157761815977e-
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 98
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 4.962025316455696,
|
| 700 |
-
"grad_norm":
|
| 701 |
-
"learning_rate": 6.
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 99
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 5.0,
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate": 1.7073496424427348e-
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 100
|
| 711 |
}
|
| 712 |
],
|
|
@@ -727,7 +727,7 @@
|
|
| 727 |
"attributes": {}
|
| 728 |
}
|
| 729 |
},
|
| 730 |
-
"total_flos":
|
| 731 |
"train_batch_size": 2,
|
| 732 |
"trial_name": null,
|
| 733 |
"trial_params": null
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.05063291139240506,
|
| 14 |
+
"grad_norm": 22.15937042236328,
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
"loss": 3.0474,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.10126582278481013,
|
| 21 |
+
"grad_norm": 36.183231353759766,
|
| 22 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 23 |
"loss": 3.2925,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.1518987341772152,
|
| 28 |
+
"grad_norm": 5.320329189300537,
|
| 29 |
+
"learning_rate": 6.666666666666667e-05,
|
| 30 |
+
"loss": 3.0096,
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.20253164556962025,
|
| 35 |
+
"grad_norm": 4.796537399291992,
|
| 36 |
+
"learning_rate": 0.0001,
|
| 37 |
+
"loss": 2.6045,
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.25316455696202533,
|
| 42 |
+
"grad_norm": 2.9281771183013916,
|
| 43 |
+
"learning_rate": 0.00013333333333333334,
|
| 44 |
+
"loss": 2.1625,
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.3037974683544304,
|
| 49 |
+
"grad_norm": 7.840775012969971,
|
| 50 |
+
"learning_rate": 0.00016666666666666666,
|
| 51 |
+
"loss": 2.7606,
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.35443037974683544,
|
| 56 |
+
"grad_norm": 3.6150004863739014,
|
| 57 |
+
"learning_rate": 0.0002,
|
| 58 |
+
"loss": 2.5675,
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.4050632911392405,
|
| 63 |
+
"grad_norm": 3.3033154010772705,
|
| 64 |
+
"learning_rate": 0.00023333333333333333,
|
| 65 |
+
"loss": 2.3897,
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.45569620253164556,
|
| 70 |
+
"grad_norm": 4.086965560913086,
|
| 71 |
+
"learning_rate": 0.0002666666666666667,
|
| 72 |
+
"loss": 2.755,
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.5063291139240507,
|
| 77 |
+
"grad_norm": 3.734769105911255,
|
| 78 |
+
"learning_rate": 0.0003,
|
| 79 |
+
"loss": 2.5837,
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.5569620253164557,
|
| 84 |
+
"grad_norm": 3.2234697341918945,
|
| 85 |
+
"learning_rate": 0.0003333333333333333,
|
| 86 |
+
"loss": 2.6353,
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.6075949367088608,
|
| 91 |
+
"grad_norm": 2.629314422607422,
|
| 92 |
+
"learning_rate": 0.00036666666666666667,
|
| 93 |
+
"loss": 2.732,
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.6582278481012658,
|
| 98 |
+
"grad_norm": 3.710653066635132,
|
| 99 |
+
"learning_rate": 0.0004,
|
| 100 |
+
"loss": 2.4126,
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.7088607594936709,
|
| 105 |
+
"grad_norm": 3.1855616569519043,
|
| 106 |
+
"learning_rate": 0.00043333333333333337,
|
| 107 |
+
"loss": 2.0725,
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.759493670886076,
|
| 112 |
+
"grad_norm": 3.34596848487854,
|
| 113 |
+
"learning_rate": 0.00046666666666666666,
|
| 114 |
+
"loss": 2.3351,
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.810126582278481,
|
| 119 |
+
"grad_norm": 3.2423255443573,
|
| 120 |
+
"learning_rate": 0.0005,
|
| 121 |
+
"loss": 2.7354,
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.8607594936708861,
|
| 126 |
+
"grad_norm": 3.6997056007385254,
|
| 127 |
+
"learning_rate": 0.0004998292650357557,
|
| 128 |
+
"loss": 2.6154,
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.9113924050632911,
|
| 133 |
+
"grad_norm": 3.3852779865264893,
|
| 134 |
+
"learning_rate": 0.0004993172933464471,
|
| 135 |
+
"loss": 2.9623,
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.9620253164556962,
|
| 140 |
+
"grad_norm": 5.298639297485352,
|
| 141 |
+
"learning_rate": 0.0004984647842238185,
|
| 142 |
+
"loss": 2.7019,
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 1.0,
|
| 147 |
+
"grad_norm": 4.402466773986816,
|
| 148 |
+
"learning_rate": 0.0004972729020927865,
|
| 149 |
+
"loss": 2.4495,
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 1.0506329113924051,
|
| 154 |
+
"grad_norm": 3.915959358215332,
|
| 155 |
+
"learning_rate": 0.0004957432749209755,
|
| 156 |
+
"loss": 1.9027,
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 1.1012658227848102,
|
| 161 |
+
"grad_norm": 2.6463258266448975,
|
| 162 |
+
"learning_rate": 0.0004938779919951092,
|
| 163 |
+
"loss": 1.6182,
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 1.1518987341772151,
|
| 168 |
+
"grad_norm": 3.179638385772705,
|
| 169 |
+
"learning_rate": 0.0004916796010672969,
|
| 170 |
+
"loss": 1.6585,
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 1.2025316455696202,
|
| 175 |
+
"grad_norm": 5.375019550323486,
|
| 176 |
+
"learning_rate": 0.0004891511048751102,
|
| 177 |
+
"loss": 1.9379,
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 1.2531645569620253,
|
| 182 |
+
"grad_norm": 3.4863626956939697,
|
| 183 |
+
"learning_rate": 0.00048629595704020493,
|
| 184 |
+
"loss": 1.5123,
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 1.3037974683544304,
|
| 189 |
+
"grad_norm": 2.945317506790161,
|
| 190 |
+
"learning_rate": 0.00048311805735108893,
|
| 191 |
+
"loss": 1.8359,
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 1.3544303797468356,
|
| 196 |
+
"grad_norm": 4.178781986236572,
|
| 197 |
+
"learning_rate": 0.0004796217464364808,
|
| 198 |
+
"loss": 1.8808,
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 1.4050632911392404,
|
| 203 |
+
"grad_norm": 3.691697597503662,
|
| 204 |
+
"learning_rate": 0.0004758117998365322,
|
| 205 |
+
"loss": 1.7988,
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 1.4556962025316456,
|
| 210 |
+
"grad_norm": 3.262970447540283,
|
| 211 |
+
"learning_rate": 0.00047169342148001547,
|
| 212 |
+
"loss": 1.8361,
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 1.5063291139240507,
|
| 217 |
+
"grad_norm": 3.550689220428467,
|
| 218 |
+
"learning_rate": 0.0004672722365763821,
|
| 219 |
+
"loss": 1.609,
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 1.5569620253164556,
|
| 224 |
+
"grad_norm": 3.615360736846924,
|
| 225 |
+
"learning_rate": 0.0004625542839324036,
|
| 226 |
+
"loss": 1.5563,
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 1.6075949367088609,
|
| 231 |
+
"grad_norm": 2.8636770248413086,
|
| 232 |
+
"learning_rate": 0.00045754600770388763,
|
| 233 |
+
"loss": 1.5551,
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 1.6582278481012658,
|
| 238 |
+
"grad_norm": 4.557415962219238,
|
| 239 |
+
"learning_rate": 0.0004522542485937369,
|
| 240 |
+
"loss": 1.7539,
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 1.7088607594936709,
|
| 245 |
+
"grad_norm": 3.1868419647216797,
|
| 246 |
+
"learning_rate": 0.0004466862345083708,
|
| 247 |
+
"loss": 1.7852,
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 1.759493670886076,
|
| 252 |
+
"grad_norm": 3.380448818206787,
|
| 253 |
+
"learning_rate": 0.0004408495706852758,
|
| 254 |
+
"loss": 1.7807,
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.810126582278481,
|
| 259 |
+
"grad_norm": 4.706411838531494,
|
| 260 |
+
"learning_rate": 0.00043475222930516476,
|
| 261 |
+
"loss": 1.9616,
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 1.8607594936708862,
|
| 266 |
+
"grad_norm": 2.568176031112671,
|
| 267 |
+
"learning_rate": 0.0004284025386029381,
|
| 268 |
+
"loss": 1.5545,
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 1.9113924050632911,
|
| 273 |
+
"grad_norm": 3.263295888900757,
|
| 274 |
+
"learning_rate": 0.00042180917149231567,
|
| 275 |
+
"loss": 1.7313,
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 1.9620253164556962,
|
| 280 |
+
"grad_norm": 5.326169967651367,
|
| 281 |
+
"learning_rate": 0.0004149811337196807,
|
| 282 |
+
"loss": 1.6959,
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 2.0,
|
| 287 |
+
"grad_norm": 4.056178569793701,
|
| 288 |
+
"learning_rate": 0.00040792775156331276,
|
| 289 |
+
"loss": 1.9885,
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 2.050632911392405,
|
| 294 |
+
"grad_norm": 2.816775321960449,
|
| 295 |
+
"learning_rate": 0.0004006586590948141,
|
| 296 |
+
"loss": 1.079,
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 2.1012658227848102,
|
| 301 |
+
"grad_norm": 3.266395330429077,
|
| 302 |
+
"learning_rate": 0.0003931837850201263,
|
| 303 |
+
"loss": 1.021,
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 2.151898734177215,
|
| 308 |
+
"grad_norm": 4.753467559814453,
|
| 309 |
+
"learning_rate": 0.00038551333911811237,
|
| 310 |
+
"loss": 1.0574,
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 2.2025316455696204,
|
| 315 |
+
"grad_norm": 4.696927547454834,
|
| 316 |
+
"learning_rate": 0.00037765779829522674,
|
| 317 |
+
"loss": 1.1594,
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 2.2531645569620253,
|
| 322 |
+
"grad_norm": 3.3051199913024902,
|
| 323 |
+
"learning_rate": 0.00036962789227532164,
|
| 324 |
+
"loss": 1.0123,
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 2.3037974683544302,
|
| 329 |
+
"grad_norm": 3.196387767791748,
|
| 330 |
+
"learning_rate": 0.0003614345889441346,
|
| 331 |
+
"loss": 1.2621,
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 2.3544303797468356,
|
| 336 |
+
"grad_norm": 3.5468530654907227,
|
| 337 |
+
"learning_rate": 0.0003530890793684759,
|
| 338 |
+
"loss": 1.0165,
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 2.4050632911392404,
|
| 343 |
+
"grad_norm": 2.5015392303466797,
|
| 344 |
+
"learning_rate": 0.0003446027625105776,
|
| 345 |
+
"loss": 1.216,
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 2.4556962025316453,
|
| 350 |
+
"grad_norm": 2.7021989822387695,
|
| 351 |
+
"learning_rate": 0.00033598722965848206,
|
| 352 |
+
"loss": 1.2118,
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 2.5063291139240507,
|
| 357 |
+
"grad_norm": 3.0691046714782715,
|
| 358 |
+
"learning_rate": 0.00032725424859373687,
|
| 359 |
+
"loss": 0.9781,
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 2.5569620253164556,
|
| 364 |
+
"grad_norm": 2.657027244567871,
|
| 365 |
+
"learning_rate": 0.0003184157475180208,
|
| 366 |
+
"loss": 0.9974,
|
| 367 |
"step": 51
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 2.607594936708861,
|
| 371 |
+
"grad_norm": 3.9573814868927,
|
| 372 |
+
"learning_rate": 0.00030948379876065467,
|
| 373 |
+
"loss": 0.8675,
|
| 374 |
"step": 52
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 2.6582278481012658,
|
| 378 |
+
"grad_norm": 2.825610876083374,
|
| 379 |
+
"learning_rate": 0.00030047060228925254,
|
| 380 |
+
"loss": 1.0503,
|
| 381 |
"step": 53
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 2.708860759493671,
|
| 385 |
+
"grad_norm": 4.903167724609375,
|
| 386 |
+
"learning_rate": 0.0002913884690460325,
|
| 387 |
+
"loss": 1.4037,
|
| 388 |
"step": 54
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 2.759493670886076,
|
| 392 |
+
"grad_norm": 2.883190393447876,
|
| 393 |
+
"learning_rate": 0.00028224980413255084,
|
| 394 |
+
"loss": 0.9242,
|
| 395 |
"step": 55
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 2.810126582278481,
|
| 399 |
+
"grad_norm": 2.112745523452759,
|
| 400 |
+
"learning_rate": 0.0002730670898658255,
|
| 401 |
+
"loss": 0.8504,
|
| 402 |
"step": 56
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 2.8607594936708862,
|
| 406 |
+
"grad_norm": 2.8920631408691406,
|
| 407 |
+
"learning_rate": 0.0002638528687289925,
|
| 408 |
+
"loss": 0.8863,
|
| 409 |
"step": 57
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 2.911392405063291,
|
| 413 |
+
"grad_norm": 2.817871570587158,
|
| 414 |
+
"learning_rate": 0.0002546197262397825,
|
| 415 |
+
"loss": 1.0101,
|
| 416 |
"step": 58
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 2.962025316455696,
|
| 420 |
+
"grad_norm": 3.5625603199005127,
|
| 421 |
+
"learning_rate": 0.0002453802737602176,
|
| 422 |
+
"loss": 1.138,
|
| 423 |
"step": 59
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 3.0,
|
| 427 |
+
"grad_norm": 2.3167858123779297,
|
| 428 |
+
"learning_rate": 0.00023614713127100752,
|
| 429 |
+
"loss": 0.5741,
|
| 430 |
"step": 60
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 3.050632911392405,
|
| 434 |
+
"grad_norm": 2.5570881366729736,
|
| 435 |
+
"learning_rate": 0.00022693291013417452,
|
| 436 |
+
"loss": 0.6103,
|
| 437 |
"step": 61
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 3.1012658227848102,
|
| 441 |
+
"grad_norm": 1.9336360692977905,
|
| 442 |
+
"learning_rate": 0.00021775019586744925,
|
| 443 |
+
"loss": 0.4277,
|
| 444 |
"step": 62
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 3.151898734177215,
|
| 448 |
+
"grad_norm": 2.514486789703369,
|
| 449 |
+
"learning_rate": 0.0002086115309539675,
|
| 450 |
+
"loss": 0.6049,
|
| 451 |
"step": 63
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 3.2025316455696204,
|
| 455 |
+
"grad_norm": 1.7973552942276,
|
| 456 |
+
"learning_rate": 0.0001995293977107475,
|
| 457 |
+
"loss": 0.3149,
|
| 458 |
"step": 64
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 3.2531645569620253,
|
| 462 |
+
"grad_norm": 1.9369994401931763,
|
| 463 |
+
"learning_rate": 0.00019051620123934537,
|
| 464 |
+
"loss": 0.4559,
|
| 465 |
"step": 65
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 3.3037974683544302,
|
| 469 |
+
"grad_norm": 2.178471803665161,
|
| 470 |
+
"learning_rate": 0.0001815842524819793,
|
| 471 |
+
"loss": 0.5129,
|
| 472 |
"step": 66
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 3.3544303797468356,
|
| 476 |
+
"grad_norm": 2.5989177227020264,
|
| 477 |
+
"learning_rate": 0.00017274575140626317,
|
| 478 |
+
"loss": 0.5667,
|
| 479 |
"step": 67
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 3.4050632911392404,
|
| 483 |
+
"grad_norm": 2.144813299179077,
|
| 484 |
+
"learning_rate": 0.00016401277034151795,
|
| 485 |
+
"loss": 0.575,
|
| 486 |
"step": 68
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 3.4556962025316453,
|
| 490 |
+
"grad_norm": 2.2235898971557617,
|
| 491 |
+
"learning_rate": 0.00015539723748942243,
|
| 492 |
+
"loss": 0.5374,
|
| 493 |
"step": 69
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 3.5063291139240507,
|
| 497 |
+
"grad_norm": 1.9977900981903076,
|
| 498 |
+
"learning_rate": 0.00014691092063152418,
|
| 499 |
+
"loss": 0.3938,
|
| 500 |
"step": 70
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 3.5569620253164556,
|
| 504 |
+
"grad_norm": 1.750430941581726,
|
| 505 |
+
"learning_rate": 0.00013856541105586545,
|
| 506 |
+
"loss": 0.3894,
|
| 507 |
"step": 71
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 3.607594936708861,
|
| 511 |
+
"grad_norm": 2.250420570373535,
|
| 512 |
+
"learning_rate": 0.0001303721077246784,
|
| 513 |
+
"loss": 0.4814,
|
| 514 |
"step": 72
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 3.6582278481012658,
|
| 518 |
+
"grad_norm": 2.961397409439087,
|
| 519 |
+
"learning_rate": 0.0001223422017047733,
|
| 520 |
+
"loss": 0.4686,
|
| 521 |
"step": 73
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 3.708860759493671,
|
| 525 |
+
"grad_norm": 2.334899425506592,
|
| 526 |
+
"learning_rate": 0.00011448666088188764,
|
| 527 |
+
"loss": 0.5233,
|
| 528 |
"step": 74
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 3.759493670886076,
|
| 532 |
+
"grad_norm": 2.6683523654937744,
|
| 533 |
+
"learning_rate": 0.00010681621497987371,
|
| 534 |
+
"loss": 0.4644,
|
| 535 |
"step": 75
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 3.810126582278481,
|
| 539 |
+
"grad_norm": 2.351604461669922,
|
| 540 |
+
"learning_rate": 9.934134090518593e-05,
|
| 541 |
+
"loss": 0.4128,
|
| 542 |
"step": 76
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 3.8607594936708862,
|
| 546 |
+
"grad_norm": 2.5051443576812744,
|
| 547 |
+
"learning_rate": 9.207224843668733e-05,
|
| 548 |
+
"loss": 0.3683,
|
| 549 |
"step": 77
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 3.911392405063291,
|
| 553 |
+
"grad_norm": 2.8908872604370117,
|
| 554 |
+
"learning_rate": 8.50188662803194e-05,
|
| 555 |
+
"loss": 0.3682,
|
| 556 |
"step": 78
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 3.962025316455696,
|
| 560 |
+
"grad_norm": 2.4416487216949463,
|
| 561 |
+
"learning_rate": 7.819082850768433e-05,
|
| 562 |
+
"loss": 0.5503,
|
| 563 |
"step": 79
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 4.0,
|
| 567 |
+
"grad_norm": 2.0517449378967285,
|
| 568 |
+
"learning_rate": 7.159746139706194e-05,
|
| 569 |
+
"loss": 0.3605,
|
| 570 |
"step": 80
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 4.050632911392405,
|
| 574 |
+
"grad_norm": 1.0866498947143555,
|
| 575 |
+
"learning_rate": 6.524777069483526e-05,
|
| 576 |
+
"loss": 0.1385,
|
| 577 |
"step": 81
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 4.10126582278481,
|
| 581 |
+
"grad_norm": 1.568613052368164,
|
| 582 |
+
"learning_rate": 5.9150429314724254e-05,
|
| 583 |
+
"loss": 0.2764,
|
| 584 |
"step": 82
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 4.151898734177215,
|
| 588 |
+
"grad_norm": 1.359748363494873,
|
| 589 |
+
"learning_rate": 5.3313765491629194e-05,
|
| 590 |
+
"loss": 0.2143,
|
| 591 |
"step": 83
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 4.2025316455696204,
|
| 595 |
+
"grad_norm": 1.2184932231903076,
|
| 596 |
+
"learning_rate": 4.7745751406263163e-05,
|
| 597 |
+
"loss": 0.1412,
|
| 598 |
"step": 84
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 4.253164556962025,
|
| 602 |
+
"grad_norm": 1.2645926475524902,
|
| 603 |
+
"learning_rate": 4.245399229611238e-05,
|
| 604 |
+
"loss": 0.1776,
|
| 605 |
"step": 85
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 4.30379746835443,
|
| 609 |
+
"grad_norm": 1.2445096969604492,
|
| 610 |
+
"learning_rate": 3.7445716067596506e-05,
|
| 611 |
+
"loss": 0.2054,
|
| 612 |
"step": 86
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 4.3544303797468356,
|
| 616 |
+
"grad_norm": 0.9805382490158081,
|
| 617 |
+
"learning_rate": 3.2727763423617915e-05,
|
| 618 |
+
"loss": 0.0969,
|
| 619 |
"step": 87
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 4.405063291139241,
|
| 623 |
+
"grad_norm": 1.2975034713745117,
|
| 624 |
+
"learning_rate": 2.8306578519984528e-05,
|
| 625 |
+
"loss": 0.1607,
|
| 626 |
"step": 88
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 4.455696202531645,
|
| 630 |
+
"grad_norm": 1.2653566598892212,
|
| 631 |
+
"learning_rate": 2.4188200163467787e-05,
|
| 632 |
+
"loss": 0.1053,
|
| 633 |
"step": 89
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 4.506329113924051,
|
| 637 |
+
"grad_norm": 1.334803581237793,
|
| 638 |
+
"learning_rate": 2.0378253563519245e-05,
|
| 639 |
+
"loss": 0.1769,
|
| 640 |
"step": 90
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 4.556962025316456,
|
| 644 |
+
"grad_norm": 1.574942946434021,
|
| 645 |
+
"learning_rate": 1.6881942648911074e-05,
|
| 646 |
+
"loss": 0.1549,
|
| 647 |
"step": 91
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 4.6075949367088604,
|
| 651 |
+
"grad_norm": 1.321705937385559,
|
| 652 |
+
"learning_rate": 1.3704042959795133e-05,
|
| 653 |
+
"loss": 0.2005,
|
| 654 |
"step": 92
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 4.658227848101266,
|
| 658 |
+
"grad_norm": 1.0175703763961792,
|
| 659 |
+
"learning_rate": 1.0848895124889818e-05,
|
| 660 |
+
"loss": 0.1525,
|
| 661 |
"step": 93
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 4.708860759493671,
|
| 665 |
+
"grad_norm": 1.5820319652557373,
|
| 666 |
+
"learning_rate": 8.320398932703144e-06,
|
| 667 |
+
"loss": 0.1657,
|
| 668 |
"step": 94
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 4.759493670886076,
|
| 672 |
+
"grad_norm": 1.1965199708938599,
|
| 673 |
+
"learning_rate": 6.12200800489085e-06,
|
| 674 |
+
"loss": 0.1327,
|
| 675 |
"step": 95
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 4.810126582278481,
|
| 679 |
+
"grad_norm": 1.2993100881576538,
|
| 680 |
+
"learning_rate": 4.256725079024554e-06,
|
| 681 |
+
"loss": 0.1132,
|
| 682 |
"step": 96
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 4.860759493670886,
|
| 686 |
+
"grad_norm": 1.2790327072143555,
|
| 687 |
+
"learning_rate": 2.7270979072135106e-06,
|
| 688 |
+
"loss": 0.1908,
|
| 689 |
"step": 97
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 4.911392405063291,
|
| 693 |
+
"grad_norm": 1.3205540180206299,
|
| 694 |
+
"learning_rate": 1.5352157761815977e-06,
|
| 695 |
+
"loss": 0.1363,
|
| 696 |
"step": 98
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 4.962025316455696,
|
| 700 |
+
"grad_norm": 1.3380309343338013,
|
| 701 |
+
"learning_rate": 6.827066535529947e-07,
|
| 702 |
+
"loss": 0.1689,
|
| 703 |
"step": 99
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 5.0,
|
| 707 |
+
"grad_norm": 1.3308898210525513,
|
| 708 |
+
"learning_rate": 1.7073496424427348e-07,
|
| 709 |
+
"loss": 0.0823,
|
| 710 |
"step": 100
|
| 711 |
}
|
| 712 |
],
|
|
|
|
| 727 |
"attributes": {}
|
| 728 |
}
|
| 729 |
},
|
| 730 |
+
"total_flos": 2.9265650012307456e+16,
|
| 731 |
"train_batch_size": 2,
|
| 732 |
"trial_name": null,
|
| 733 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6033
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ba1f2914799e90728253ee6efdcfb4b949f06f16db0ce9229e365f91a8f35af
|
| 3 |
size 6033
|