Upload folder using huggingface_hub
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- special_tokens_map.json +21 -3
- tokenizer_config.json +4 -0
- trainer_state.json +976 -965
- training_args.bin +1 -1
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 891644712
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4816f95cb2200a6a12c5566f44048186ef83f37262177ca271572145592055f3
|
| 3 |
size 891644712
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1783448843
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfa5ef50e01fb9c41de5447c677b8e7ba6aa164eda59e5fa63128b03ff9c11a8
|
| 3 |
size 1783448843
|
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:b2c995dc901ebaef9aeedc15588f6bdbaeaaa7c0ed92315746a3ee0c7c2b9838
|
| 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 sha256:9df96ea9c16c43b5a2c22317a8049d99d80cf01a2376bd602a8fe6faaf27bad1
|
| 3 |
size 1465
|
special_tokens_map.json
CHANGED
|
@@ -101,7 +101,25 @@
|
|
| 101 |
"<extra_id_98>",
|
| 102 |
"<extra_id_99>"
|
| 103 |
],
|
| 104 |
-
"eos_token":
|
| 105 |
-
|
| 106 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 107 |
}
|
|
|
|
| 101 |
"<extra_id_98>",
|
| 102 |
"<extra_id_99>"
|
| 103 |
],
|
| 104 |
+
"eos_token": {
|
| 105 |
+
"content": "</s>",
|
| 106 |
+
"lstrip": false,
|
| 107 |
+
"normalized": false,
|
| 108 |
+
"rstrip": false,
|
| 109 |
+
"single_word": false
|
| 110 |
+
},
|
| 111 |
+
"pad_token": {
|
| 112 |
+
"content": "<pad>",
|
| 113 |
+
"lstrip": false,
|
| 114 |
+
"normalized": false,
|
| 115 |
+
"rstrip": false,
|
| 116 |
+
"single_word": false
|
| 117 |
+
},
|
| 118 |
+
"unk_token": {
|
| 119 |
+
"content": "<unk>",
|
| 120 |
+
"lstrip": false,
|
| 121 |
+
"normalized": false,
|
| 122 |
+
"rstrip": false,
|
| 123 |
+
"single_word": false
|
| 124 |
+
}
|
| 125 |
}
|
tokenizer_config.json
CHANGED
|
@@ -932,8 +932,12 @@
|
|
| 932 |
"eos_token": "</s>",
|
| 933 |
"extra_ids": 100,
|
| 934 |
"extra_special_tokens": {},
|
|
|
|
| 935 |
"model_max_length": 1000000000000000019884624838656,
|
| 936 |
"pad_token": "<pad>",
|
|
|
|
| 937 |
"tokenizer_class": "T5Tokenizer",
|
|
|
|
|
|
|
| 938 |
"unk_token": "<unk>"
|
| 939 |
}
|
|
|
|
| 932 |
"eos_token": "</s>",
|
| 933 |
"extra_ids": 100,
|
| 934 |
"extra_special_tokens": {},
|
| 935 |
+
"max_length": 64,
|
| 936 |
"model_max_length": 1000000000000000019884624838656,
|
| 937 |
"pad_token": "<pad>",
|
| 938 |
+
"stride": 0,
|
| 939 |
"tokenizer_class": "T5Tokenizer",
|
| 940 |
+
"truncation_side": "right",
|
| 941 |
+
"truncation_strategy": "longest_first",
|
| 942 |
"unk_token": "<unk>"
|
| 943 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1687 +1,1698 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint": "runs/t5_base/checkpoint-
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 2.
|
| 16 |
-
"loss":
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate": 2.
|
| 23 |
-
"loss":
|
| 24 |
"step": 100
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 2.
|
| 30 |
-
"loss":
|
| 31 |
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate": 2.
|
| 37 |
-
"loss":
|
| 38 |
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 2.
|
| 44 |
-
"loss":
|
| 45 |
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm": 1.
|
| 50 |
-
"learning_rate": 2.
|
| 51 |
-
"loss":
|
| 52 |
"step": 300
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate": 2.
|
| 58 |
-
"loss":
|
| 59 |
"step": 350
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm": 1.
|
| 64 |
-
"learning_rate": 2.
|
| 65 |
-
"loss":
|
| 66 |
"step": 400
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 2.
|
| 72 |
-
"loss":
|
| 73 |
"step": 450
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate": 2.
|
| 79 |
-
"loss":
|
| 80 |
"step": 500
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate": 2.
|
| 86 |
-
"loss":
|
| 87 |
"step": 550
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm": 1.
|
| 92 |
-
"learning_rate": 2.
|
| 93 |
-
"loss":
|
| 94 |
"step": 600
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"grad_norm": 1.
|
| 99 |
-
"learning_rate": 2.
|
| 100 |
-
"loss":
|
| 101 |
"step": 650
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch": 0.
|
| 105 |
-
"grad_norm": 1.
|
| 106 |
-
"learning_rate": 2.
|
| 107 |
-
"loss":
|
| 108 |
"step": 700
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch": 0.
|
| 112 |
-
"grad_norm": 1.
|
| 113 |
-
"learning_rate": 2.
|
| 114 |
-
"loss": 1.
|
| 115 |
"step": 750
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch": 0.
|
| 119 |
-
"grad_norm":
|
| 120 |
-
"learning_rate": 2.
|
| 121 |
-
"loss": 1.
|
| 122 |
"step": 800
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
-
"grad_norm":
|
| 127 |
-
"learning_rate": 2.
|
| 128 |
-
"loss":
|
| 129 |
"step": 850
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate": 2.
|
| 135 |
-
"loss": 1.
|
| 136 |
"step": 900
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch": 0.
|
| 140 |
-
"grad_norm": 1.
|
| 141 |
-
"learning_rate": 2.
|
| 142 |
-
"loss": 1.
|
| 143 |
"step": 950
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch": 0.
|
| 147 |
-
"grad_norm":
|
| 148 |
-
"learning_rate": 2.
|
| 149 |
-
"loss": 1.
|
| 150 |
"step": 1000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch": 0.
|
| 154 |
-
"grad_norm": 1.
|
| 155 |
-
"learning_rate": 2.
|
| 156 |
-
"loss": 1.
|
| 157 |
"step": 1050
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch": 0.
|
| 161 |
-
"grad_norm": 1.
|
| 162 |
-
"learning_rate": 2.
|
| 163 |
-
"loss":
|
| 164 |
"step": 1100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate": 2.
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 1150
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate": 2.
|
| 177 |
-
"loss": 1.
|
| 178 |
"step": 1200
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch": 0.
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate": 2.
|
| 184 |
-
"loss": 1.
|
| 185 |
"step": 1250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch": 0.
|
| 189 |
-
"grad_norm": 1.
|
| 190 |
-
"learning_rate": 2.
|
| 191 |
-
"loss":
|
| 192 |
"step": 1300
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch": 0.
|
| 196 |
-
"grad_norm": 1.
|
| 197 |
-
"learning_rate": 2.
|
| 198 |
-
"loss": 1.
|
| 199 |
"step": 1350
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch": 0.
|
| 203 |
-
"grad_norm": 1.
|
| 204 |
-
"learning_rate": 2.
|
| 205 |
-
"loss": 1.
|
| 206 |
"step": 1400
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate": 2.
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 1450
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 0.
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate": 2.
|
| 219 |
-
"loss": 1.
|
| 220 |
"step": 1500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch": 0.
|
| 224 |
-
"grad_norm":
|
| 225 |
-
"learning_rate": 2.
|
| 226 |
-
"loss": 1.
|
| 227 |
"step": 1550
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch": 0.
|
| 231 |
-
"grad_norm":
|
| 232 |
-
"learning_rate": 2.
|
| 233 |
-
"loss": 1.
|
| 234 |
"step": 1600
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch": 0.
|
| 238 |
-
"grad_norm":
|
| 239 |
-
"learning_rate": 2.
|
| 240 |
-
"loss": 1.
|
| 241 |
"step": 1650
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch": 0.
|
| 245 |
-
"grad_norm": 1.
|
| 246 |
-
"learning_rate": 2.
|
| 247 |
-
"loss": 1.
|
| 248 |
"step": 1700
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
-
"grad_norm":
|
| 253 |
-
"learning_rate": 2.
|
| 254 |
-
"loss": 1.
|
| 255 |
"step": 1750
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 0.
|
| 259 |
-
"grad_norm": 1.
|
| 260 |
-
"learning_rate": 2.
|
| 261 |
-
"loss": 1.
|
| 262 |
"step": 1800
|
| 263 |
},
|
| 264 |
{
|
| 265 |
-
"epoch": 0.
|
| 266 |
-
"grad_norm": 1.
|
| 267 |
-
"learning_rate": 2.
|
| 268 |
-
"loss": 1.
|
| 269 |
"step": 1850
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch": 0.
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate": 2.
|
| 275 |
-
"loss": 1.
|
| 276 |
"step": 1900
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch": 0.
|
| 280 |
-
"grad_norm": 1.
|
| 281 |
-
"learning_rate": 2.
|
| 282 |
-
"loss": 1.
|
| 283 |
"step": 1950
|
| 284 |
},
|
| 285 |
{
|
| 286 |
-
"epoch": 0.
|
| 287 |
-
"grad_norm": 1.
|
| 288 |
-
"learning_rate": 2.
|
| 289 |
-
"loss": 1.
|
| 290 |
"step": 2000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 0.
|
| 294 |
-
"grad_norm": 1.
|
| 295 |
-
"learning_rate": 2.
|
| 296 |
-
"loss": 1.
|
| 297 |
"step": 2050
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 0.
|
| 301 |
-
"grad_norm":
|
| 302 |
-
"learning_rate": 2.
|
| 303 |
-
"loss": 1.
|
| 304 |
"step": 2100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
-
"epoch": 0.
|
| 308 |
-
"grad_norm":
|
| 309 |
-
"learning_rate": 2.
|
| 310 |
-
"loss": 1.
|
| 311 |
"step": 2150
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch": 0.
|
| 315 |
-
"grad_norm": 1.
|
| 316 |
-
"learning_rate": 2.
|
| 317 |
-
"loss": 1.
|
| 318 |
"step": 2200
|
| 319 |
},
|
| 320 |
{
|
| 321 |
-
"epoch": 0.
|
| 322 |
-
"grad_norm": 1.
|
| 323 |
-
"learning_rate": 2.
|
| 324 |
-
"loss": 1.
|
| 325 |
"step": 2250
|
| 326 |
},
|
| 327 |
{
|
| 328 |
-
"epoch": 0.
|
| 329 |
-
"grad_norm": 1.
|
| 330 |
-
"learning_rate": 2.
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 2300
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
-
"grad_norm": 1.
|
| 337 |
-
"learning_rate": 2.
|
| 338 |
-
"loss": 1.
|
| 339 |
"step": 2350
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 0.
|
| 343 |
-
"grad_norm": 1.
|
| 344 |
-
"learning_rate": 2.
|
| 345 |
-
"loss": 1.
|
| 346 |
"step": 2400
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch": 0.
|
| 350 |
-
"grad_norm":
|
| 351 |
-
"learning_rate": 2.
|
| 352 |
-
"loss": 1.
|
| 353 |
"step": 2450
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch": 0.
|
| 357 |
-
"grad_norm": 1.
|
| 358 |
-
"learning_rate": 2.
|
| 359 |
-
"loss": 1.
|
| 360 |
"step": 2500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch": 0.
|
| 364 |
-
"grad_norm": 1.
|
| 365 |
-
"learning_rate": 2.
|
| 366 |
-
"loss": 1.
|
| 367 |
"step": 2550
|
| 368 |
},
|
| 369 |
{
|
| 370 |
-
"epoch": 0.
|
| 371 |
-
"grad_norm": 1.
|
| 372 |
-
"learning_rate": 2.
|
| 373 |
-
"loss": 1.
|
| 374 |
"step": 2600
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch": 0.
|
| 378 |
-
"grad_norm": 1.
|
| 379 |
-
"learning_rate": 2.
|
| 380 |
-
"loss": 1.
|
| 381 |
"step": 2650
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 0.
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate": 2.
|
| 387 |
-
"loss": 1.
|
| 388 |
"step": 2700
|
| 389 |
},
|
| 390 |
{
|
| 391 |
-
"epoch": 0.
|
| 392 |
-
"grad_norm": 1.
|
| 393 |
-
"learning_rate": 2.
|
| 394 |
-
"loss": 1.
|
| 395 |
"step": 2750
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch": 0.
|
| 399 |
-
"grad_norm": 1.
|
| 400 |
-
"learning_rate": 2.
|
| 401 |
-
"loss": 1.
|
| 402 |
"step": 2800
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch": 0.
|
| 406 |
-
"grad_norm":
|
| 407 |
-
"learning_rate": 2.
|
| 408 |
-
"loss": 1.
|
| 409 |
"step": 2850
|
| 410 |
},
|
| 411 |
{
|
| 412 |
-
"epoch": 0.
|
| 413 |
-
"grad_norm": 1.
|
| 414 |
-
"learning_rate": 2.
|
| 415 |
-
"loss": 1.
|
| 416 |
"step": 2900
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch": 0.
|
| 420 |
-
"grad_norm": 1.
|
| 421 |
-
"learning_rate": 2.
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 2950
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 0.
|
| 427 |
-
"grad_norm": 1.
|
| 428 |
-
"learning_rate": 2.
|
| 429 |
-
"loss": 1.
|
| 430 |
"step": 3000
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch": 0.
|
| 434 |
-
"grad_norm": 1.
|
| 435 |
-
"learning_rate": 2.
|
| 436 |
-
"loss": 1.
|
| 437 |
"step": 3050
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch": 0.
|
| 441 |
-
"grad_norm": 1.
|
| 442 |
-
"learning_rate": 2.
|
| 443 |
-
"loss": 1.
|
| 444 |
"step": 3100
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch": 0.
|
| 448 |
-
"grad_norm":
|
| 449 |
-
"learning_rate": 2.
|
| 450 |
-
"loss": 1.
|
| 451 |
"step": 3150
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch": 0.
|
| 455 |
-
"grad_norm": 1.
|
| 456 |
-
"learning_rate": 2.
|
| 457 |
-
"loss": 1.
|
| 458 |
"step": 3200
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 0.
|
| 462 |
-
"grad_norm": 1.
|
| 463 |
-
"learning_rate": 2.
|
| 464 |
-
"loss": 1.
|
| 465 |
"step": 3250
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch": 0.
|
| 469 |
-
"grad_norm": 1.
|
| 470 |
-
"learning_rate": 2.
|
| 471 |
-
"loss": 1.
|
| 472 |
"step": 3300
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch": 0.
|
| 476 |
-
"grad_norm": 1.
|
| 477 |
-
"learning_rate": 2.
|
| 478 |
-
"loss": 1.
|
| 479 |
"step": 3350
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch": 0.
|
| 483 |
-
"grad_norm":
|
| 484 |
-
"learning_rate": 2.
|
| 485 |
-
"loss": 1.
|
| 486 |
"step": 3400
|
| 487 |
},
|
| 488 |
{
|
| 489 |
-
"epoch": 0.
|
| 490 |
-
"grad_norm": 1.
|
| 491 |
-
"learning_rate": 2.
|
| 492 |
-
"loss": 1.
|
| 493 |
"step": 3450
|
| 494 |
},
|
| 495 |
{
|
| 496 |
-
"epoch": 0.
|
| 497 |
-
"grad_norm": 1.
|
| 498 |
-
"learning_rate": 2.
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 3500
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch": 0.
|
| 504 |
-
"grad_norm": 1.
|
| 505 |
-
"learning_rate": 2.
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 3550
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch": 0.
|
| 511 |
-
"grad_norm": 1.
|
| 512 |
-
"learning_rate": 2.
|
| 513 |
-
"loss": 1.
|
| 514 |
"step": 3600
|
| 515 |
},
|
| 516 |
{
|
| 517 |
-
"epoch": 0.
|
| 518 |
-
"grad_norm": 1.
|
| 519 |
-
"learning_rate": 2.
|
| 520 |
-
"loss": 1.
|
| 521 |
"step": 3650
|
| 522 |
},
|
| 523 |
{
|
| 524 |
-
"epoch": 0.
|
| 525 |
-
"grad_norm": 1.
|
| 526 |
-
"learning_rate": 2.
|
| 527 |
-
"loss": 1.
|
| 528 |
"step": 3700
|
| 529 |
},
|
| 530 |
{
|
| 531 |
-
"epoch": 0.
|
| 532 |
-
"grad_norm": 1.
|
| 533 |
-
"learning_rate": 2.
|
| 534 |
-
"loss": 1.
|
| 535 |
"step": 3750
|
| 536 |
},
|
| 537 |
{
|
| 538 |
-
"epoch": 0.
|
| 539 |
-
"grad_norm": 1.
|
| 540 |
-
"learning_rate": 2.
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 3800
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch": 0.
|
| 546 |
-
"grad_norm":
|
| 547 |
-
"learning_rate": 2.
|
| 548 |
-
"loss": 1.
|
| 549 |
"step": 3850
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch": 0.
|
| 553 |
-
"grad_norm": 1.
|
| 554 |
-
"learning_rate": 2.
|
| 555 |
-
"loss": 1.
|
| 556 |
"step": 3900
|
| 557 |
},
|
| 558 |
{
|
| 559 |
-
"epoch": 0.
|
| 560 |
-
"grad_norm": 1.
|
| 561 |
-
"learning_rate": 2.
|
| 562 |
-
"loss": 1.
|
| 563 |
"step": 3950
|
| 564 |
},
|
| 565 |
{
|
| 566 |
-
"epoch": 0.
|
| 567 |
-
"grad_norm": 1.
|
| 568 |
-
"learning_rate": 2.
|
| 569 |
-
"loss": 1.
|
| 570 |
"step": 4000
|
| 571 |
},
|
| 572 |
{
|
| 573 |
-
"epoch": 0.
|
| 574 |
-
"grad_norm": 1.
|
| 575 |
-
"learning_rate": 2.
|
| 576 |
-
"loss": 1.
|
| 577 |
"step": 4050
|
| 578 |
},
|
| 579 |
{
|
| 580 |
-
"epoch": 0.
|
| 581 |
-
"grad_norm": 1.
|
| 582 |
-
"learning_rate": 2.
|
| 583 |
-
"loss": 1.
|
| 584 |
"step": 4100
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch": 0.
|
| 588 |
-
"grad_norm": 1.
|
| 589 |
-
"learning_rate": 2.
|
| 590 |
-
"loss": 1.
|
| 591 |
"step": 4150
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch": 0.
|
| 595 |
-
"grad_norm": 1.
|
| 596 |
-
"learning_rate": 2.
|
| 597 |
-
"loss": 1.
|
| 598 |
"step": 4200
|
| 599 |
},
|
| 600 |
{
|
| 601 |
-
"epoch": 0.
|
| 602 |
-
"grad_norm":
|
| 603 |
-
"learning_rate": 2.
|
| 604 |
-
"loss": 1.
|
| 605 |
"step": 4250
|
| 606 |
},
|
| 607 |
{
|
| 608 |
-
"epoch": 0.
|
| 609 |
-
"grad_norm": 1.
|
| 610 |
-
"learning_rate": 2.
|
| 611 |
-
"loss": 1.
|
| 612 |
"step": 4300
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch": 0.
|
| 616 |
-
"grad_norm": 1.
|
| 617 |
-
"learning_rate": 2.
|
| 618 |
-
"loss": 1.
|
| 619 |
"step": 4350
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch": 0.
|
| 623 |
-
"grad_norm": 1.
|
| 624 |
-
"learning_rate": 2.
|
| 625 |
-
"loss": 1.
|
| 626 |
"step": 4400
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch": 0.
|
| 630 |
-
"grad_norm": 1.
|
| 631 |
-
"learning_rate": 2.
|
| 632 |
-
"loss": 1.
|
| 633 |
"step": 4450
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch": 0.
|
| 637 |
-
"grad_norm": 1.
|
| 638 |
-
"learning_rate": 2.
|
| 639 |
-
"loss": 1.
|
| 640 |
"step": 4500
|
| 641 |
},
|
| 642 |
{
|
| 643 |
-
"epoch": 0.
|
| 644 |
-
"grad_norm": 1.
|
| 645 |
-
"learning_rate": 2.
|
| 646 |
-
"loss": 1.
|
| 647 |
"step": 4550
|
| 648 |
},
|
| 649 |
{
|
| 650 |
-
"epoch": 0.
|
| 651 |
-
"grad_norm": 1.
|
| 652 |
-
"learning_rate": 2.
|
| 653 |
-
"loss": 1.
|
| 654 |
"step": 4600
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch": 0.
|
| 658 |
-
"grad_norm": 1.
|
| 659 |
-
"learning_rate": 2.
|
| 660 |
-
"loss": 1.
|
| 661 |
"step": 4650
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch": 0.
|
| 665 |
-
"grad_norm": 1.
|
| 666 |
-
"learning_rate": 2.
|
| 667 |
-
"loss": 1.
|
| 668 |
"step": 4700
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch": 0.
|
| 672 |
-
"grad_norm":
|
| 673 |
-
"learning_rate": 2.
|
| 674 |
-
"loss": 1.
|
| 675 |
"step": 4750
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch": 0.
|
| 679 |
-
"grad_norm": 1.
|
| 680 |
-
"learning_rate": 2.
|
| 681 |
-
"loss": 1.
|
| 682 |
"step": 4800
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"epoch": 0.
|
| 686 |
-
"grad_norm": 1.
|
| 687 |
-
"learning_rate": 2.
|
| 688 |
-
"loss": 1.
|
| 689 |
"step": 4850
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch": 0.
|
| 693 |
-
"grad_norm": 1.
|
| 694 |
-
"learning_rate": 2.
|
| 695 |
-
"loss": 1.
|
| 696 |
"step": 4900
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch": 0.
|
| 700 |
-
"grad_norm": 1.
|
| 701 |
-
"learning_rate": 2.
|
| 702 |
-
"loss": 1.
|
| 703 |
"step": 4950
|
| 704 |
},
|
| 705 |
{
|
| 706 |
-
"epoch": 0.
|
| 707 |
-
"grad_norm": 1.
|
| 708 |
-
"learning_rate": 2.
|
| 709 |
-
"loss": 1.
|
| 710 |
"step": 5000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch": 0.
|
| 714 |
-
"grad_norm":
|
| 715 |
-
"learning_rate": 2.
|
| 716 |
-
"loss": 1.
|
| 717 |
"step": 5050
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch": 0.
|
| 721 |
-
"grad_norm":
|
| 722 |
-
"learning_rate": 2.
|
| 723 |
-
"loss": 1.
|
| 724 |
"step": 5100
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"epoch": 0.
|
| 728 |
-
"grad_norm": 1.
|
| 729 |
-
"learning_rate": 2.
|
| 730 |
-
"loss": 1.
|
| 731 |
"step": 5150
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch": 0.
|
| 735 |
-
"grad_norm":
|
| 736 |
-
"learning_rate": 2.
|
| 737 |
-
"loss": 1.
|
| 738 |
"step": 5200
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch": 0.
|
| 742 |
-
"grad_norm": 1.
|
| 743 |
-
"learning_rate": 2.
|
| 744 |
-
"loss": 1.
|
| 745 |
"step": 5250
|
| 746 |
},
|
| 747 |
{
|
| 748 |
-
"epoch": 0.
|
| 749 |
-
"grad_norm":
|
| 750 |
-
"learning_rate": 2.
|
| 751 |
-
"loss": 1.
|
| 752 |
"step": 5300
|
| 753 |
},
|
| 754 |
{
|
| 755 |
-
"epoch": 0.
|
| 756 |
-
"grad_norm":
|
| 757 |
-
"learning_rate": 2.
|
| 758 |
-
"loss": 1.
|
| 759 |
"step": 5350
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch": 0.
|
| 763 |
-
"grad_norm": 1.
|
| 764 |
-
"learning_rate": 2.
|
| 765 |
-
"loss": 1.
|
| 766 |
"step": 5400
|
| 767 |
},
|
| 768 |
{
|
| 769 |
-
"epoch": 0.
|
| 770 |
-
"grad_norm": 1.
|
| 771 |
-
"learning_rate": 2.
|
| 772 |
-
"loss": 1.
|
| 773 |
"step": 5450
|
| 774 |
},
|
| 775 |
{
|
| 776 |
-
"epoch": 0.
|
| 777 |
-
"grad_norm": 1.
|
| 778 |
-
"learning_rate": 2.
|
| 779 |
-
"loss": 1.
|
| 780 |
"step": 5500
|
| 781 |
},
|
| 782 |
{
|
| 783 |
-
"epoch": 0.
|
| 784 |
-
"grad_norm": 1.
|
| 785 |
-
"learning_rate": 2.
|
| 786 |
-
"loss": 1.
|
| 787 |
"step": 5550
|
| 788 |
},
|
| 789 |
{
|
| 790 |
-
"epoch": 0.
|
| 791 |
-
"grad_norm": 1.
|
| 792 |
-
"learning_rate": 2.
|
| 793 |
-
"loss": 1.
|
| 794 |
"step": 5600
|
| 795 |
},
|
| 796 |
{
|
| 797 |
-
"epoch": 0.
|
| 798 |
-
"grad_norm": 1.
|
| 799 |
-
"learning_rate": 2.
|
| 800 |
-
"loss": 1.
|
| 801 |
"step": 5650
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch": 0.
|
| 805 |
-
"grad_norm": 1.
|
| 806 |
-
"learning_rate": 2.
|
| 807 |
-
"loss": 1.
|
| 808 |
"step": 5700
|
| 809 |
},
|
| 810 |
{
|
| 811 |
-
"epoch": 0.
|
| 812 |
-
"grad_norm": 1.
|
| 813 |
-
"learning_rate": 2.
|
| 814 |
-
"loss": 1.
|
| 815 |
"step": 5750
|
| 816 |
},
|
| 817 |
{
|
| 818 |
-
"epoch": 0.
|
| 819 |
-
"grad_norm": 1.
|
| 820 |
-
"learning_rate": 2.
|
| 821 |
-
"loss": 1.
|
| 822 |
"step": 5800
|
| 823 |
},
|
| 824 |
{
|
| 825 |
-
"epoch": 0.
|
| 826 |
-
"grad_norm": 1.
|
| 827 |
-
"learning_rate": 2.
|
| 828 |
-
"loss": 1.
|
| 829 |
"step": 5850
|
| 830 |
},
|
| 831 |
{
|
| 832 |
-
"epoch": 0.
|
| 833 |
-
"grad_norm":
|
| 834 |
-
"learning_rate": 2.
|
| 835 |
-
"loss": 1.
|
| 836 |
"step": 5900
|
| 837 |
},
|
| 838 |
{
|
| 839 |
-
"epoch": 0
|
| 840 |
-
"
|
| 841 |
-
"
|
| 842 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 843 |
"step": 5950
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch":
|
| 847 |
-
"grad_norm": 1.
|
| 848 |
-
"learning_rate": 2.
|
| 849 |
-
"loss": 1.
|
| 850 |
"step": 6000
|
| 851 |
},
|
| 852 |
{
|
| 853 |
-
"epoch":
|
| 854 |
-
"grad_norm": 1.
|
| 855 |
-
"learning_rate": 2.
|
| 856 |
-
"loss": 1.
|
| 857 |
"step": 6050
|
| 858 |
},
|
| 859 |
{
|
| 860 |
-
"epoch":
|
| 861 |
-
"grad_norm": 1.
|
| 862 |
-
"learning_rate": 2.
|
| 863 |
-
"loss": 1.
|
| 864 |
"step": 6100
|
| 865 |
},
|
| 866 |
{
|
| 867 |
-
"epoch":
|
| 868 |
-
"grad_norm": 1.
|
| 869 |
-
"learning_rate": 2.
|
| 870 |
-
"loss": 1.
|
| 871 |
"step": 6150
|
| 872 |
},
|
| 873 |
{
|
| 874 |
-
"epoch":
|
| 875 |
-
"grad_norm": 1.
|
| 876 |
-
"learning_rate": 2.
|
| 877 |
-
"loss": 1.
|
| 878 |
"step": 6200
|
| 879 |
},
|
| 880 |
{
|
| 881 |
-
"epoch":
|
| 882 |
-
"grad_norm": 1.
|
| 883 |
-
"learning_rate": 2.
|
| 884 |
-
"loss": 1.
|
| 885 |
"step": 6250
|
| 886 |
},
|
| 887 |
{
|
| 888 |
-
"epoch":
|
| 889 |
-
"grad_norm": 1.
|
| 890 |
-
"learning_rate": 2.
|
| 891 |
-
"loss": 1.
|
| 892 |
"step": 6300
|
| 893 |
},
|
| 894 |
{
|
| 895 |
-
"epoch":
|
| 896 |
-
"grad_norm": 1.
|
| 897 |
-
"learning_rate": 2.
|
| 898 |
-
"loss": 1.
|
| 899 |
"step": 6350
|
| 900 |
},
|
| 901 |
{
|
| 902 |
-
"epoch":
|
| 903 |
-
"grad_norm":
|
| 904 |
-
"learning_rate": 2.
|
| 905 |
-
"loss": 1.
|
| 906 |
"step": 6400
|
| 907 |
},
|
| 908 |
{
|
| 909 |
-
"epoch":
|
| 910 |
-
"grad_norm": 1.
|
| 911 |
-
"learning_rate": 2.
|
| 912 |
-
"loss": 1.
|
| 913 |
"step": 6450
|
| 914 |
},
|
| 915 |
{
|
| 916 |
-
"epoch":
|
| 917 |
-
"grad_norm": 1.
|
| 918 |
-
"learning_rate": 2.
|
| 919 |
-
"loss": 1.
|
| 920 |
"step": 6500
|
| 921 |
},
|
| 922 |
{
|
| 923 |
-
"epoch":
|
| 924 |
-
"grad_norm": 1.
|
| 925 |
-
"learning_rate": 2.
|
| 926 |
-
"loss": 1.
|
| 927 |
"step": 6550
|
| 928 |
},
|
| 929 |
{
|
| 930 |
-
"epoch":
|
| 931 |
-
"grad_norm": 1.
|
| 932 |
-
"learning_rate": 2.
|
| 933 |
-
"loss": 1.
|
| 934 |
"step": 6600
|
| 935 |
},
|
| 936 |
{
|
| 937 |
-
"epoch":
|
| 938 |
-
"grad_norm": 1.
|
| 939 |
-
"learning_rate": 2.
|
| 940 |
-
"loss": 1.
|
| 941 |
"step": 6650
|
| 942 |
},
|
| 943 |
{
|
| 944 |
-
"epoch":
|
| 945 |
-
"grad_norm": 1.
|
| 946 |
-
"learning_rate": 2.
|
| 947 |
-
"loss": 1.
|
| 948 |
"step": 6700
|
| 949 |
},
|
| 950 |
{
|
| 951 |
-
"epoch":
|
| 952 |
-
"grad_norm": 1.
|
| 953 |
-
"learning_rate": 2.
|
| 954 |
-
"loss": 1.
|
| 955 |
"step": 6750
|
| 956 |
},
|
| 957 |
{
|
| 958 |
-
"epoch":
|
| 959 |
-
"grad_norm": 1.
|
| 960 |
-
"learning_rate": 2.
|
| 961 |
-
"loss": 1.
|
| 962 |
"step": 6800
|
| 963 |
},
|
| 964 |
{
|
| 965 |
-
"epoch":
|
| 966 |
-
"grad_norm":
|
| 967 |
-
"learning_rate": 2.
|
| 968 |
-
"loss": 1.
|
| 969 |
"step": 6850
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch":
|
| 973 |
-
"grad_norm": 1.
|
| 974 |
-
"learning_rate": 2.
|
| 975 |
-
"loss": 1.
|
| 976 |
"step": 6900
|
| 977 |
},
|
| 978 |
{
|
| 979 |
-
"epoch":
|
| 980 |
-
"grad_norm":
|
| 981 |
-
"learning_rate": 2.
|
| 982 |
-
"loss": 1.
|
| 983 |
"step": 6950
|
| 984 |
},
|
| 985 |
{
|
| 986 |
-
"epoch":
|
| 987 |
-
"grad_norm": 1.
|
| 988 |
-
"learning_rate": 2.
|
| 989 |
-
"loss": 1.
|
| 990 |
"step": 7000
|
| 991 |
},
|
| 992 |
{
|
| 993 |
-
"epoch":
|
| 994 |
-
"grad_norm": 1.
|
| 995 |
-
"learning_rate": 2.
|
| 996 |
-
"loss": 1.
|
| 997 |
"step": 7050
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
-
"epoch":
|
| 1001 |
-
"grad_norm": 1.
|
| 1002 |
-
"learning_rate": 2.
|
| 1003 |
-
"loss": 1.
|
| 1004 |
"step": 7100
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"epoch":
|
| 1008 |
-
"grad_norm": 1.
|
| 1009 |
-
"learning_rate": 2.
|
| 1010 |
-
"loss": 1.
|
| 1011 |
"step": 7150
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
-
"epoch":
|
| 1015 |
-
"grad_norm": 1.
|
| 1016 |
-
"learning_rate": 2.
|
| 1017 |
-
"loss": 1.
|
| 1018 |
"step": 7200
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
-
"epoch":
|
| 1022 |
-
"grad_norm": 1.
|
| 1023 |
-
"learning_rate": 2.
|
| 1024 |
-
"loss": 1.
|
| 1025 |
"step": 7250
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
-
"epoch":
|
| 1029 |
-
"grad_norm": 1.
|
| 1030 |
-
"learning_rate": 2.
|
| 1031 |
-
"loss": 1.
|
| 1032 |
"step": 7300
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
-
"epoch":
|
| 1036 |
-
"grad_norm": 1.
|
| 1037 |
-
"learning_rate": 2.
|
| 1038 |
-
"loss": 1.
|
| 1039 |
"step": 7350
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
-
"epoch":
|
| 1043 |
-
"grad_norm": 1.
|
| 1044 |
-
"learning_rate": 2.
|
| 1045 |
-
"loss": 1.
|
| 1046 |
"step": 7400
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
-
"epoch":
|
| 1050 |
-
"grad_norm": 1.
|
| 1051 |
-
"learning_rate": 2.
|
| 1052 |
-
"loss": 1.
|
| 1053 |
"step": 7450
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
-
"epoch":
|
| 1057 |
-
"grad_norm": 1.
|
| 1058 |
-
"learning_rate": 2.
|
| 1059 |
-
"loss": 1.
|
| 1060 |
"step": 7500
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
-
"epoch":
|
| 1064 |
-
"grad_norm": 1.
|
| 1065 |
-
"learning_rate": 2.
|
| 1066 |
-
"loss": 1.
|
| 1067 |
"step": 7550
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
-
"epoch":
|
| 1071 |
-
"grad_norm": 1.
|
| 1072 |
-
"learning_rate": 2.
|
| 1073 |
-
"loss": 1.
|
| 1074 |
"step": 7600
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
-
"epoch":
|
| 1078 |
-
"grad_norm": 1.
|
| 1079 |
-
"learning_rate": 2.
|
| 1080 |
-
"loss": 1.
|
| 1081 |
"step": 7650
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
-
"epoch":
|
| 1085 |
-
"grad_norm": 1.
|
| 1086 |
-
"learning_rate": 2.
|
| 1087 |
-
"loss": 1.
|
| 1088 |
"step": 7700
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
-
"epoch":
|
| 1092 |
-
"grad_norm": 1.
|
| 1093 |
-
"learning_rate": 2.
|
| 1094 |
-
"loss": 1.
|
| 1095 |
"step": 7750
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
-
"epoch":
|
| 1099 |
-
"grad_norm": 1.
|
| 1100 |
-
"learning_rate": 2.
|
| 1101 |
-
"loss": 1.
|
| 1102 |
"step": 7800
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
-
"epoch":
|
| 1106 |
-
"grad_norm": 1.
|
| 1107 |
-
"learning_rate": 2.
|
| 1108 |
-
"loss": 1.
|
| 1109 |
"step": 7850
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
-
"epoch":
|
| 1113 |
-
"grad_norm":
|
| 1114 |
-
"learning_rate": 2.
|
| 1115 |
-
"loss": 1.
|
| 1116 |
"step": 7900
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
-
"epoch":
|
| 1120 |
-
"grad_norm": 1.
|
| 1121 |
-
"learning_rate": 2.
|
| 1122 |
-
"loss": 1.
|
| 1123 |
"step": 7950
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
-
"epoch":
|
| 1127 |
-
"grad_norm":
|
| 1128 |
-
"learning_rate": 2.
|
| 1129 |
-
"loss": 1.
|
| 1130 |
"step": 8000
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
-
"epoch":
|
| 1134 |
-
"grad_norm": 1.
|
| 1135 |
-
"learning_rate": 2.
|
| 1136 |
-
"loss": 1.
|
| 1137 |
"step": 8050
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
-
"epoch":
|
| 1141 |
-
"grad_norm": 1.
|
| 1142 |
-
"learning_rate": 2.
|
| 1143 |
-
"loss": 1.
|
| 1144 |
"step": 8100
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
-
"epoch":
|
| 1148 |
-
"grad_norm": 1.
|
| 1149 |
-
"learning_rate": 2.
|
| 1150 |
-
"loss": 1.
|
| 1151 |
"step": 8150
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
-
"epoch":
|
| 1155 |
-
"grad_norm": 1.
|
| 1156 |
-
"learning_rate": 2.
|
| 1157 |
-
"loss": 1.
|
| 1158 |
"step": 8200
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
-
"epoch":
|
| 1162 |
-
"grad_norm": 1.
|
| 1163 |
-
"learning_rate": 2.
|
| 1164 |
-
"loss": 1.
|
| 1165 |
"step": 8250
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
-
"epoch":
|
| 1169 |
-
"grad_norm": 1.
|
| 1170 |
-
"learning_rate": 2.
|
| 1171 |
-
"loss": 1.
|
| 1172 |
"step": 8300
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
-
"epoch":
|
| 1176 |
-
"grad_norm": 1.
|
| 1177 |
-
"learning_rate": 2.
|
| 1178 |
-
"loss": 1.
|
| 1179 |
"step": 8350
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
-
"epoch":
|
| 1183 |
-
"grad_norm":
|
| 1184 |
-
"learning_rate": 2.
|
| 1185 |
-
"loss": 1.
|
| 1186 |
"step": 8400
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
-
"epoch":
|
| 1190 |
-
"grad_norm":
|
| 1191 |
-
"learning_rate": 2.
|
| 1192 |
-
"loss": 1.
|
| 1193 |
"step": 8450
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
-
"epoch":
|
| 1197 |
-
"grad_norm": 1.
|
| 1198 |
-
"learning_rate": 2.
|
| 1199 |
-
"loss": 1.
|
| 1200 |
"step": 8500
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
-
"epoch":
|
| 1204 |
-
"grad_norm": 1.
|
| 1205 |
-
"learning_rate": 2.
|
| 1206 |
-
"loss": 1.
|
| 1207 |
"step": 8550
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
-
"epoch":
|
| 1211 |
-
"grad_norm": 1.
|
| 1212 |
-
"learning_rate": 2.
|
| 1213 |
-
"loss": 1.
|
| 1214 |
"step": 8600
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
-
"epoch":
|
| 1218 |
-
"grad_norm": 1.
|
| 1219 |
-
"learning_rate": 2.
|
| 1220 |
-
"loss": 1.
|
| 1221 |
"step": 8650
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
-
"epoch":
|
| 1225 |
-
"grad_norm": 1.
|
| 1226 |
-
"learning_rate": 2.
|
| 1227 |
-
"loss": 1.
|
| 1228 |
"step": 8700
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
-
"epoch":
|
| 1232 |
-
"grad_norm": 1.
|
| 1233 |
-
"learning_rate": 2.
|
| 1234 |
-
"loss": 1.
|
| 1235 |
"step": 8750
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
-
"epoch":
|
| 1239 |
-
"grad_norm": 1.
|
| 1240 |
-
"learning_rate": 2.
|
| 1241 |
-
"loss": 1.
|
| 1242 |
"step": 8800
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
-
"epoch":
|
| 1246 |
-
"grad_norm": 1.
|
| 1247 |
-
"learning_rate": 2.
|
| 1248 |
-
"loss": 1.
|
| 1249 |
"step": 8850
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
-
"epoch":
|
| 1253 |
-
"grad_norm": 1.
|
| 1254 |
-
"learning_rate": 2.
|
| 1255 |
-
"loss": 1.
|
| 1256 |
"step": 8900
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
-
"epoch":
|
| 1260 |
-
"grad_norm": 1.
|
| 1261 |
-
"learning_rate": 2.
|
| 1262 |
-
"loss": 1.
|
| 1263 |
"step": 8950
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
-
"epoch":
|
| 1267 |
-
"grad_norm": 1.
|
| 1268 |
-
"learning_rate": 2.
|
| 1269 |
-
"loss": 1.
|
| 1270 |
"step": 9000
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
-
"epoch":
|
| 1274 |
-
"grad_norm":
|
| 1275 |
-
"learning_rate": 2.
|
| 1276 |
-
"loss": 1.
|
| 1277 |
"step": 9050
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
-
"epoch":
|
| 1281 |
-
"grad_norm": 1.
|
| 1282 |
-
"learning_rate": 2.
|
| 1283 |
-
"loss": 1.
|
| 1284 |
"step": 9100
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
-
"epoch":
|
| 1288 |
-
"grad_norm": 1.
|
| 1289 |
-
"learning_rate": 2.
|
| 1290 |
-
"loss": 1.
|
| 1291 |
"step": 9150
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
-
"epoch":
|
| 1295 |
-
"grad_norm":
|
| 1296 |
-
"learning_rate": 2.
|
| 1297 |
-
"loss": 1.
|
| 1298 |
"step": 9200
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
-
"epoch":
|
| 1302 |
-
"grad_norm": 1.
|
| 1303 |
-
"learning_rate": 2.
|
| 1304 |
-
"loss": 1.
|
| 1305 |
"step": 9250
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
-
"epoch":
|
| 1309 |
-
"grad_norm": 1.
|
| 1310 |
-
"learning_rate": 2.
|
| 1311 |
-
"loss": 1.
|
| 1312 |
"step": 9300
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
-
"epoch":
|
| 1316 |
-
"grad_norm": 1.
|
| 1317 |
-
"learning_rate": 2.
|
| 1318 |
-
"loss": 1.
|
| 1319 |
"step": 9350
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
-
"epoch":
|
| 1323 |
-
"grad_norm": 1.
|
| 1324 |
-
"learning_rate": 2.
|
| 1325 |
-
"loss": 1.
|
| 1326 |
"step": 9400
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
-
"epoch":
|
| 1330 |
-
"grad_norm": 1.
|
| 1331 |
-
"learning_rate": 2.
|
| 1332 |
-
"loss": 1.
|
| 1333 |
"step": 9450
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
-
"epoch":
|
| 1337 |
-
"grad_norm": 1.
|
| 1338 |
-
"learning_rate": 2.
|
| 1339 |
-
"loss": 1.
|
| 1340 |
"step": 9500
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
-
"epoch":
|
| 1344 |
-
"grad_norm":
|
| 1345 |
-
"learning_rate": 2.
|
| 1346 |
-
"loss": 1.
|
| 1347 |
"step": 9550
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
-
"epoch":
|
| 1351 |
-
"grad_norm": 1.
|
| 1352 |
-
"learning_rate": 2.
|
| 1353 |
-
"loss": 1.
|
| 1354 |
"step": 9600
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
-
"epoch":
|
| 1358 |
-
"grad_norm": 1.
|
| 1359 |
-
"learning_rate": 2.
|
| 1360 |
-
"loss": 1.
|
| 1361 |
"step": 9650
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
-
"epoch":
|
| 1365 |
-
"grad_norm": 1.
|
| 1366 |
-
"learning_rate": 2.
|
| 1367 |
-
"loss": 1.
|
| 1368 |
"step": 9700
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
-
"epoch":
|
| 1372 |
-
"grad_norm": 1.
|
| 1373 |
-
"learning_rate": 2.
|
| 1374 |
-
"loss": 1.
|
| 1375 |
"step": 9750
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
-
"epoch":
|
| 1379 |
-
"grad_norm": 1.
|
| 1380 |
-
"learning_rate": 2.
|
| 1381 |
-
"loss": 1.
|
| 1382 |
"step": 9800
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
-
"epoch":
|
| 1386 |
-
"grad_norm": 1.
|
| 1387 |
-
"learning_rate": 2.
|
| 1388 |
-
"loss": 1.
|
| 1389 |
"step": 9850
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
-
"epoch":
|
| 1393 |
-
"grad_norm": 1.
|
| 1394 |
-
"learning_rate": 2.
|
| 1395 |
-
"loss": 1.
|
| 1396 |
"step": 9900
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
-
"epoch":
|
| 1400 |
-
"grad_norm": 1.
|
| 1401 |
-
"learning_rate": 2.
|
| 1402 |
-
"loss": 1.
|
| 1403 |
"step": 9950
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
-
"epoch":
|
| 1407 |
-
"grad_norm": 1.
|
| 1408 |
-
"learning_rate": 2.
|
| 1409 |
-
"loss": 1.
|
| 1410 |
"step": 10000
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
-
"epoch":
|
| 1414 |
-
"grad_norm": 1.
|
| 1415 |
-
"learning_rate": 2.
|
| 1416 |
-
"loss": 1.
|
| 1417 |
"step": 10050
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
-
"epoch":
|
| 1421 |
-
"grad_norm": 1.
|
| 1422 |
-
"learning_rate": 2.
|
| 1423 |
-
"loss": 1.
|
| 1424 |
"step": 10100
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
-
"epoch":
|
| 1428 |
-
"grad_norm": 1.
|
| 1429 |
-
"learning_rate": 2.
|
| 1430 |
-
"loss": 1.
|
| 1431 |
"step": 10150
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
-
"epoch":
|
| 1435 |
-
"grad_norm": 1.
|
| 1436 |
-
"learning_rate": 2.
|
| 1437 |
-
"loss": 1.
|
| 1438 |
"step": 10200
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
-
"epoch":
|
| 1442 |
-
"grad_norm": 1.
|
| 1443 |
-
"learning_rate": 2.
|
| 1444 |
-
"loss": 1.
|
| 1445 |
"step": 10250
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
-
"epoch":
|
| 1449 |
-
"grad_norm": 1.
|
| 1450 |
-
"learning_rate": 2.
|
| 1451 |
-
"loss": 1.
|
| 1452 |
"step": 10300
|
| 1453 |
},
|
| 1454 |
{
|
| 1455 |
-
"epoch":
|
| 1456 |
-
"grad_norm": 1.
|
| 1457 |
-
"learning_rate": 2.
|
| 1458 |
-
"loss": 1.
|
| 1459 |
"step": 10350
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
-
"epoch":
|
| 1463 |
-
"grad_norm":
|
| 1464 |
-
"learning_rate": 2.
|
| 1465 |
-
"loss": 1.
|
| 1466 |
"step": 10400
|
| 1467 |
},
|
| 1468 |
{
|
| 1469 |
-
"epoch":
|
| 1470 |
-
"grad_norm": 1.
|
| 1471 |
-
"learning_rate": 2.
|
| 1472 |
-
"loss": 1.
|
| 1473 |
"step": 10450
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
-
"epoch":
|
| 1477 |
-
"grad_norm": 1.
|
| 1478 |
-
"learning_rate": 2.
|
| 1479 |
-
"loss": 1.
|
| 1480 |
"step": 10500
|
| 1481 |
},
|
| 1482 |
{
|
| 1483 |
-
"epoch":
|
| 1484 |
-
"grad_norm":
|
| 1485 |
-
"learning_rate": 2.
|
| 1486 |
-
"loss": 1.
|
| 1487 |
"step": 10550
|
| 1488 |
},
|
| 1489 |
{
|
| 1490 |
-
"epoch":
|
| 1491 |
-
"grad_norm": 1.
|
| 1492 |
-
"learning_rate": 2.
|
| 1493 |
-
"loss": 1.
|
| 1494 |
"step": 10600
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
-
"epoch":
|
| 1498 |
-
"grad_norm": 1.
|
| 1499 |
-
"learning_rate": 2.
|
| 1500 |
-
"loss": 1.
|
| 1501 |
"step": 10650
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
-
"epoch":
|
| 1505 |
-
"grad_norm": 1.
|
| 1506 |
-
"learning_rate": 2.
|
| 1507 |
-
"loss": 1.
|
| 1508 |
"step": 10700
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
-
"epoch":
|
| 1512 |
-
"grad_norm":
|
| 1513 |
-
"learning_rate": 2.
|
| 1514 |
-
"loss": 1.
|
| 1515 |
"step": 10750
|
| 1516 |
},
|
| 1517 |
{
|
| 1518 |
-
"epoch":
|
| 1519 |
-
"grad_norm":
|
| 1520 |
-
"learning_rate": 2.
|
| 1521 |
-
"loss": 1.
|
| 1522 |
"step": 10800
|
| 1523 |
},
|
| 1524 |
{
|
| 1525 |
-
"epoch":
|
| 1526 |
-
"grad_norm": 1.
|
| 1527 |
-
"learning_rate": 2.
|
| 1528 |
-
"loss": 1.
|
| 1529 |
"step": 10850
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
-
"epoch":
|
| 1533 |
-
"grad_norm": 1.
|
| 1534 |
-
"learning_rate": 2.
|
| 1535 |
-
"loss": 1.
|
| 1536 |
"step": 10900
|
| 1537 |
},
|
| 1538 |
{
|
| 1539 |
-
"epoch":
|
| 1540 |
-
"grad_norm": 1.
|
| 1541 |
-
"learning_rate": 2.
|
| 1542 |
-
"loss": 1.
|
| 1543 |
"step": 10950
|
| 1544 |
},
|
| 1545 |
{
|
| 1546 |
-
"epoch":
|
| 1547 |
-
"grad_norm": 1.
|
| 1548 |
-
"learning_rate": 2.
|
| 1549 |
-
"loss": 1.
|
| 1550 |
"step": 11000
|
| 1551 |
},
|
| 1552 |
{
|
| 1553 |
-
"epoch":
|
| 1554 |
-
"grad_norm": 1.
|
| 1555 |
-
"learning_rate": 2.
|
| 1556 |
-
"loss": 1.
|
| 1557 |
"step": 11050
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
-
"epoch":
|
| 1561 |
-
"grad_norm":
|
| 1562 |
-
"learning_rate": 2.
|
| 1563 |
-
"loss": 1.
|
| 1564 |
"step": 11100
|
| 1565 |
},
|
| 1566 |
{
|
| 1567 |
-
"epoch":
|
| 1568 |
-
"grad_norm": 1.
|
| 1569 |
-
"learning_rate": 2.
|
| 1570 |
-
"loss": 1.
|
| 1571 |
"step": 11150
|
| 1572 |
},
|
| 1573 |
{
|
| 1574 |
-
"epoch":
|
| 1575 |
-
"grad_norm": 1.
|
| 1576 |
-
"learning_rate": 2.
|
| 1577 |
-
"loss": 1.
|
| 1578 |
"step": 11200
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
-
"epoch":
|
| 1582 |
-
"grad_norm": 1.
|
| 1583 |
-
"learning_rate": 2.
|
| 1584 |
-
"loss": 1.
|
| 1585 |
"step": 11250
|
| 1586 |
},
|
| 1587 |
{
|
| 1588 |
-
"epoch":
|
| 1589 |
-
"grad_norm": 1.
|
| 1590 |
-
"learning_rate": 2.
|
| 1591 |
-
"loss": 1.
|
| 1592 |
"step": 11300
|
| 1593 |
},
|
| 1594 |
{
|
| 1595 |
-
"epoch":
|
| 1596 |
-
"grad_norm": 1.
|
| 1597 |
-
"learning_rate": 2.
|
| 1598 |
-
"loss": 1.
|
| 1599 |
"step": 11350
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
-
"epoch":
|
| 1603 |
-
"grad_norm": 1.
|
| 1604 |
-
"learning_rate": 2.
|
| 1605 |
-
"loss": 1.
|
| 1606 |
"step": 11400
|
| 1607 |
},
|
| 1608 |
{
|
| 1609 |
-
"epoch":
|
| 1610 |
-
"grad_norm": 1.
|
| 1611 |
-
"learning_rate": 2.
|
| 1612 |
-
"loss": 1.
|
| 1613 |
"step": 11450
|
| 1614 |
},
|
| 1615 |
{
|
| 1616 |
-
"epoch":
|
| 1617 |
-
"grad_norm": 1.
|
| 1618 |
-
"learning_rate": 2.
|
| 1619 |
-
"loss": 1.
|
| 1620 |
"step": 11500
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
-
"epoch":
|
| 1624 |
-
"grad_norm": 1.
|
| 1625 |
-
"learning_rate": 2.
|
| 1626 |
-
"loss": 1.
|
| 1627 |
"step": 11550
|
| 1628 |
},
|
| 1629 |
{
|
| 1630 |
-
"epoch":
|
| 1631 |
-
"grad_norm": 1.
|
| 1632 |
-
"learning_rate": 2.
|
| 1633 |
-
"loss": 1.
|
| 1634 |
"step": 11600
|
| 1635 |
},
|
| 1636 |
{
|
| 1637 |
-
"epoch":
|
| 1638 |
-
"grad_norm": 1.
|
| 1639 |
-
"learning_rate": 2.
|
| 1640 |
-
"loss": 1.
|
| 1641 |
"step": 11650
|
| 1642 |
},
|
| 1643 |
{
|
| 1644 |
-
"epoch":
|
| 1645 |
-
"grad_norm": 1.
|
| 1646 |
-
"learning_rate": 2.
|
| 1647 |
-
"loss": 1.
|
| 1648 |
"step": 11700
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
-
"epoch":
|
| 1652 |
-
"grad_norm": 1.
|
| 1653 |
-
"learning_rate": 2.
|
| 1654 |
-
"loss": 1.
|
| 1655 |
"step": 11750
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
-
"epoch":
|
| 1659 |
-
"grad_norm":
|
| 1660 |
-
"learning_rate": 2.
|
| 1661 |
-
"loss": 1.
|
| 1662 |
"step": 11800
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
-
"epoch":
|
| 1666 |
-
"grad_norm": 1.
|
| 1667 |
-
"learning_rate": 2.
|
| 1668 |
-
"loss": 1.
|
| 1669 |
"step": 11850
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
-
"epoch":
|
| 1673 |
-
"eval_loss": 1.
|
| 1674 |
-
"eval_rouge1": 0.
|
| 1675 |
-
"eval_rouge2": 0.
|
| 1676 |
-
"eval_rougeL": 0.
|
| 1677 |
-
"eval_runtime":
|
| 1678 |
-
"eval_samples_per_second":
|
| 1679 |
-
"eval_steps_per_second": 1.
|
| 1680 |
-
"step":
|
| 1681 |
}
|
| 1682 |
],
|
| 1683 |
"logging_steps": 50,
|
| 1684 |
-
"max_steps":
|
| 1685 |
"num_input_tokens_seen": 0,
|
| 1686 |
"num_train_epochs": 50,
|
| 1687 |
"save_steps": 500,
|
|
@@ -1706,8 +1717,8 @@
|
|
| 1706 |
"attributes": {}
|
| 1707 |
}
|
| 1708 |
},
|
| 1709 |
-
"total_flos":
|
| 1710 |
-
"train_batch_size":
|
| 1711 |
"trial_name": null,
|
| 1712 |
"trial_params": null
|
| 1713 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 11852,
|
| 3 |
+
"best_metric": 0.3635343246206635,
|
| 4 |
+
"best_model_checkpoint": "runs/t5_base/checkpoint-11852",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 11852,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.00843810648890389,
|
| 14 |
+
"grad_norm": 1.1823029518127441,
|
| 15 |
+
"learning_rate": 2.999503881201485e-05,
|
| 16 |
+
"loss": 1.7915,
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.01687621297780778,
|
| 21 |
+
"grad_norm": 1.172362208366394,
|
| 22 |
+
"learning_rate": 2.9989976375295308e-05,
|
| 23 |
+
"loss": 1.8189,
|
| 24 |
"step": 100
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.02531431946671167,
|
| 28 |
+
"grad_norm": 1.6581450700759888,
|
| 29 |
+
"learning_rate": 2.998491393857577e-05,
|
| 30 |
+
"loss": 1.8142,
|
| 31 |
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.03375242595561556,
|
| 35 |
+
"grad_norm": 1.1387466192245483,
|
| 36 |
+
"learning_rate": 2.9979851501856227e-05,
|
| 37 |
+
"loss": 1.8261,
|
| 38 |
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.04219053244451945,
|
| 42 |
+
"grad_norm": 1.1661549806594849,
|
| 43 |
+
"learning_rate": 2.9974789065136688e-05,
|
| 44 |
+
"loss": 1.8097,
|
| 45 |
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.05062863893342334,
|
| 49 |
+
"grad_norm": 1.2071425914764404,
|
| 50 |
+
"learning_rate": 2.9969726628417145e-05,
|
| 51 |
+
"loss": 1.791,
|
| 52 |
"step": 300
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.05906674542232723,
|
| 56 |
+
"grad_norm": 1.1316887140274048,
|
| 57 |
+
"learning_rate": 2.9964664191697606e-05,
|
| 58 |
+
"loss": 1.8034,
|
| 59 |
"step": 350
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.06750485191123112,
|
| 63 |
+
"grad_norm": 1.1885789632797241,
|
| 64 |
+
"learning_rate": 2.9959601754978064e-05,
|
| 65 |
+
"loss": 1.7741,
|
| 66 |
"step": 400
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.07594295840013501,
|
| 70 |
+
"grad_norm": 1.2099308967590332,
|
| 71 |
+
"learning_rate": 2.995453931825852e-05,
|
| 72 |
+
"loss": 1.8011,
|
| 73 |
"step": 450
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.0843810648890389,
|
| 77 |
+
"grad_norm": 1.2386366128921509,
|
| 78 |
+
"learning_rate": 2.9949476881538982e-05,
|
| 79 |
+
"loss": 1.8063,
|
| 80 |
"step": 500
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.0928191713779428,
|
| 84 |
+
"grad_norm": 1.2590100765228271,
|
| 85 |
+
"learning_rate": 2.994441444481944e-05,
|
| 86 |
+
"loss": 1.7954,
|
| 87 |
"step": 550
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.10125727786684668,
|
| 91 |
+
"grad_norm": 1.7030580043792725,
|
| 92 |
+
"learning_rate": 2.99393520080999e-05,
|
| 93 |
+
"loss": 1.7486,
|
| 94 |
"step": 600
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 0.10969538435575057,
|
| 98 |
+
"grad_norm": 1.1653929948806763,
|
| 99 |
+
"learning_rate": 2.9934289571380358e-05,
|
| 100 |
+
"loss": 1.8254,
|
| 101 |
"step": 650
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 0.11813349084465447,
|
| 105 |
+
"grad_norm": 1.2317183017730713,
|
| 106 |
+
"learning_rate": 2.992922713466082e-05,
|
| 107 |
+
"loss": 1.7634,
|
| 108 |
"step": 700
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 0.12657159733355836,
|
| 112 |
+
"grad_norm": 1.2524069547653198,
|
| 113 |
+
"learning_rate": 2.9924164697941276e-05,
|
| 114 |
+
"loss": 1.758,
|
| 115 |
"step": 750
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 0.13500970382246225,
|
| 119 |
+
"grad_norm": 1.220339298248291,
|
| 120 |
+
"learning_rate": 2.9919102261221737e-05,
|
| 121 |
+
"loss": 1.7594,
|
| 122 |
"step": 800
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.14344781031136614,
|
| 126 |
+
"grad_norm": 1.3024489879608154,
|
| 127 |
+
"learning_rate": 2.991403982450219e-05,
|
| 128 |
+
"loss": 1.7682,
|
| 129 |
"step": 850
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.15188591680027003,
|
| 133 |
+
"grad_norm": 1.2505040168762207,
|
| 134 |
+
"learning_rate": 2.9908977387782652e-05,
|
| 135 |
+
"loss": 1.7908,
|
| 136 |
"step": 900
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 0.16032402328917392,
|
| 140 |
+
"grad_norm": 1.3853548765182495,
|
| 141 |
+
"learning_rate": 2.9903914951063113e-05,
|
| 142 |
+
"loss": 1.7203,
|
| 143 |
"step": 950
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 0.1687621297780778,
|
| 147 |
+
"grad_norm": 1.197028636932373,
|
| 148 |
+
"learning_rate": 2.989885251434357e-05,
|
| 149 |
+
"loss": 1.826,
|
| 150 |
"step": 1000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 0.1772002362669817,
|
| 154 |
+
"grad_norm": 1.2575501203536987,
|
| 155 |
+
"learning_rate": 2.989379007762403e-05,
|
| 156 |
+
"loss": 1.7429,
|
| 157 |
"step": 1050
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 0.1856383427558856,
|
| 161 |
+
"grad_norm": 1.6052565574645996,
|
| 162 |
+
"learning_rate": 2.988872764090449e-05,
|
| 163 |
+
"loss": 1.7866,
|
| 164 |
"step": 1100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.19407644924478948,
|
| 168 |
+
"grad_norm": 1.1177057027816772,
|
| 169 |
+
"learning_rate": 2.988366520418495e-05,
|
| 170 |
+
"loss": 1.7606,
|
| 171 |
"step": 1150
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.20251455573369337,
|
| 175 |
+
"grad_norm": 1.1401762962341309,
|
| 176 |
+
"learning_rate": 2.9878602767465407e-05,
|
| 177 |
+
"loss": 1.8021,
|
| 178 |
"step": 1200
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 0.21095266222259726,
|
| 182 |
+
"grad_norm": 1.583141565322876,
|
| 183 |
+
"learning_rate": 2.987354033074587e-05,
|
| 184 |
+
"loss": 1.775,
|
| 185 |
"step": 1250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 0.21939076871150115,
|
| 189 |
+
"grad_norm": 1.1614867448806763,
|
| 190 |
+
"learning_rate": 2.9868477894026326e-05,
|
| 191 |
+
"loss": 1.7617,
|
| 192 |
"step": 1300
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 0.22782887520040504,
|
| 196 |
+
"grad_norm": 1.340226411819458,
|
| 197 |
+
"learning_rate": 2.9863415457306783e-05,
|
| 198 |
+
"loss": 1.7679,
|
| 199 |
"step": 1350
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 0.23626698168930893,
|
| 203 |
+
"grad_norm": 1.1907130479812622,
|
| 204 |
+
"learning_rate": 2.985835302058724e-05,
|
| 205 |
+
"loss": 1.7642,
|
| 206 |
"step": 1400
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.24470508817821282,
|
| 210 |
+
"grad_norm": 1.1323164701461792,
|
| 211 |
+
"learning_rate": 2.9853290583867702e-05,
|
| 212 |
+
"loss": 1.7563,
|
| 213 |
"step": 1450
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 0.2531431946671167,
|
| 217 |
+
"grad_norm": 1.2015341520309448,
|
| 218 |
+
"learning_rate": 2.9848228147148163e-05,
|
| 219 |
+
"loss": 1.7832,
|
| 220 |
"step": 1500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 0.2615813011560206,
|
| 224 |
+
"grad_norm": 1.1307079792022705,
|
| 225 |
+
"learning_rate": 2.984316571042862e-05,
|
| 226 |
+
"loss": 1.7815,
|
| 227 |
"step": 1550
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 0.2700194076449245,
|
| 231 |
+
"grad_norm": 1.2381619215011597,
|
| 232 |
+
"learning_rate": 2.983810327370908e-05,
|
| 233 |
+
"loss": 1.7602,
|
| 234 |
"step": 1600
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 0.27845751413382835,
|
| 238 |
+
"grad_norm": 1.1170424222946167,
|
| 239 |
+
"learning_rate": 2.983304083698954e-05,
|
| 240 |
+
"loss": 1.7683,
|
| 241 |
"step": 1650
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 0.2868956206227323,
|
| 245 |
+
"grad_norm": 1.3253265619277954,
|
| 246 |
+
"learning_rate": 2.982797840027e-05,
|
| 247 |
+
"loss": 1.7387,
|
| 248 |
"step": 1700
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.29533372711163614,
|
| 252 |
+
"grad_norm": 1.391984224319458,
|
| 253 |
+
"learning_rate": 2.9822915963550457e-05,
|
| 254 |
+
"loss": 1.7655,
|
| 255 |
"step": 1750
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.30377183360054005,
|
| 259 |
+
"grad_norm": 1.2357732057571411,
|
| 260 |
+
"learning_rate": 2.9817853526830915e-05,
|
| 261 |
+
"loss": 1.7711,
|
| 262 |
"step": 1800
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 0.3122099400894439,
|
| 266 |
+
"grad_norm": 1.8593379259109497,
|
| 267 |
+
"learning_rate": 2.9812791090111372e-05,
|
| 268 |
+
"loss": 1.7707,
|
| 269 |
"step": 1850
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 0.32064804657834783,
|
| 273 |
+
"grad_norm": 1.1849147081375122,
|
| 274 |
+
"learning_rate": 2.9807728653391833e-05,
|
| 275 |
+
"loss": 1.7511,
|
| 276 |
"step": 1900
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 0.3290861530672517,
|
| 280 |
+
"grad_norm": 1.3520323038101196,
|
| 281 |
+
"learning_rate": 2.980266621667229e-05,
|
| 282 |
+
"loss": 1.7641,
|
| 283 |
"step": 1950
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 0.3375242595561556,
|
| 287 |
+
"grad_norm": 1.1148391962051392,
|
| 288 |
+
"learning_rate": 2.979760377995275e-05,
|
| 289 |
+
"loss": 1.7849,
|
| 290 |
"step": 2000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.3459623660450595,
|
| 294 |
+
"grad_norm": 1.2033883333206177,
|
| 295 |
+
"learning_rate": 2.979254134323321e-05,
|
| 296 |
+
"loss": 1.7861,
|
| 297 |
"step": 2050
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 0.3544004725339634,
|
| 301 |
+
"grad_norm": 1.2050273418426514,
|
| 302 |
+
"learning_rate": 2.978747890651367e-05,
|
| 303 |
+
"loss": 1.7531,
|
| 304 |
"step": 2100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
+
"epoch": 0.36283857902286726,
|
| 308 |
+
"grad_norm": 1.1736046075820923,
|
| 309 |
+
"learning_rate": 2.978241646979413e-05,
|
| 310 |
+
"loss": 1.7748,
|
| 311 |
"step": 2150
|
| 312 |
},
|
| 313 |
{
|
| 314 |
+
"epoch": 0.3712766855117712,
|
| 315 |
+
"grad_norm": 1.2393854856491089,
|
| 316 |
+
"learning_rate": 2.9777354033074588e-05,
|
| 317 |
+
"loss": 1.7972,
|
| 318 |
"step": 2200
|
| 319 |
},
|
| 320 |
{
|
| 321 |
+
"epoch": 0.37971479200067504,
|
| 322 |
+
"grad_norm": 1.3532406091690063,
|
| 323 |
+
"learning_rate": 2.977229159635505e-05,
|
| 324 |
+
"loss": 1.7565,
|
| 325 |
"step": 2250
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 0.38815289848957896,
|
| 329 |
+
"grad_norm": 1.2599409818649292,
|
| 330 |
+
"learning_rate": 2.9767229159635503e-05,
|
| 331 |
+
"loss": 1.7591,
|
| 332 |
"step": 2300
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.3965910049784828,
|
| 336 |
+
"grad_norm": 1.1961002349853516,
|
| 337 |
+
"learning_rate": 2.9762166722915964e-05,
|
| 338 |
+
"loss": 1.7725,
|
| 339 |
"step": 2350
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 0.40502911146738674,
|
| 343 |
+
"grad_norm": 1.3700023889541626,
|
| 344 |
+
"learning_rate": 2.975710428619642e-05,
|
| 345 |
+
"loss": 1.7616,
|
| 346 |
"step": 2400
|
| 347 |
},
|
| 348 |
{
|
| 349 |
+
"epoch": 0.4134672179562906,
|
| 350 |
+
"grad_norm": 2.6831870079040527,
|
| 351 |
+
"learning_rate": 2.9752041849476883e-05,
|
| 352 |
+
"loss": 1.7372,
|
| 353 |
"step": 2450
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 0.4219053244451945,
|
| 357 |
+
"grad_norm": 1.1522475481033325,
|
| 358 |
+
"learning_rate": 2.974697941275734e-05,
|
| 359 |
+
"loss": 1.7758,
|
| 360 |
"step": 2500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 0.4303434309340984,
|
| 364 |
+
"grad_norm": 1.4393203258514404,
|
| 365 |
+
"learning_rate": 2.97419169760378e-05,
|
| 366 |
+
"loss": 1.7663,
|
| 367 |
"step": 2550
|
| 368 |
},
|
| 369 |
{
|
| 370 |
+
"epoch": 0.4387815374230023,
|
| 371 |
+
"grad_norm": 1.3723366260528564,
|
| 372 |
+
"learning_rate": 2.973685453931826e-05,
|
| 373 |
+
"loss": 1.7618,
|
| 374 |
"step": 2600
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 0.44721964391190616,
|
| 378 |
+
"grad_norm": 1.1089024543762207,
|
| 379 |
+
"learning_rate": 2.973179210259872e-05,
|
| 380 |
+
"loss": 1.7842,
|
| 381 |
"step": 2650
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 0.4556577504008101,
|
| 385 |
+
"grad_norm": 1.2044650316238403,
|
| 386 |
+
"learning_rate": 2.972672966587918e-05,
|
| 387 |
+
"loss": 1.7793,
|
| 388 |
"step": 2700
|
| 389 |
},
|
| 390 |
{
|
| 391 |
+
"epoch": 0.46409585688971394,
|
| 392 |
+
"grad_norm": 1.2730334997177124,
|
| 393 |
+
"learning_rate": 2.9721667229159634e-05,
|
| 394 |
+
"loss": 1.757,
|
| 395 |
"step": 2750
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 0.47253396337861786,
|
| 399 |
+
"grad_norm": 1.1678314208984375,
|
| 400 |
+
"learning_rate": 2.9716604792440095e-05,
|
| 401 |
+
"loss": 1.7724,
|
| 402 |
"step": 2800
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"epoch": 0.4809720698675217,
|
| 406 |
+
"grad_norm": 1.220035195350647,
|
| 407 |
+
"learning_rate": 2.9711542355720553e-05,
|
| 408 |
+
"loss": 1.7381,
|
| 409 |
"step": 2850
|
| 410 |
},
|
| 411 |
{
|
| 412 |
+
"epoch": 0.48941017635642564,
|
| 413 |
+
"grad_norm": 1.1194424629211426,
|
| 414 |
+
"learning_rate": 2.9706479919001014e-05,
|
| 415 |
+
"loss": 1.783,
|
| 416 |
"step": 2900
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.4978482828453295,
|
| 420 |
+
"grad_norm": 1.2477567195892334,
|
| 421 |
+
"learning_rate": 2.970141748228147e-05,
|
| 422 |
+
"loss": 1.7778,
|
| 423 |
"step": 2950
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 0.5062863893342334,
|
| 427 |
+
"grad_norm": 1.2932846546173096,
|
| 428 |
+
"learning_rate": 2.9696355045561932e-05,
|
| 429 |
+
"loss": 1.7495,
|
| 430 |
"step": 3000
|
| 431 |
},
|
| 432 |
{
|
| 433 |
+
"epoch": 0.5147244958231373,
|
| 434 |
+
"grad_norm": 1.265342354774475,
|
| 435 |
+
"learning_rate": 2.969129260884239e-05,
|
| 436 |
+
"loss": 1.7731,
|
| 437 |
"step": 3050
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 0.5231626023120411,
|
| 441 |
+
"grad_norm": 1.1403651237487793,
|
| 442 |
+
"learning_rate": 2.968623017212285e-05,
|
| 443 |
+
"loss": 1.775,
|
| 444 |
"step": 3100
|
| 445 |
},
|
| 446 |
{
|
| 447 |
+
"epoch": 0.5316007088009451,
|
| 448 |
+
"grad_norm": 1.2105836868286133,
|
| 449 |
+
"learning_rate": 2.9681167735403308e-05,
|
| 450 |
+
"loss": 1.7812,
|
| 451 |
"step": 3150
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 0.540038815289849,
|
| 455 |
+
"grad_norm": 1.1452655792236328,
|
| 456 |
+
"learning_rate": 2.9676105298683766e-05,
|
| 457 |
+
"loss": 1.7555,
|
| 458 |
"step": 3200
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.5484769217787528,
|
| 462 |
+
"grad_norm": 1.0883064270019531,
|
| 463 |
+
"learning_rate": 2.9671042861964226e-05,
|
| 464 |
+
"loss": 1.7783,
|
| 465 |
"step": 3250
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 0.5569150282676567,
|
| 469 |
+
"grad_norm": 1.279936671257019,
|
| 470 |
+
"learning_rate": 2.9665980425244684e-05,
|
| 471 |
+
"loss": 1.7481,
|
| 472 |
"step": 3300
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"epoch": 0.5653531347565607,
|
| 476 |
+
"grad_norm": 1.2121195793151855,
|
| 477 |
+
"learning_rate": 2.9660917988525145e-05,
|
| 478 |
+
"loss": 1.774,
|
| 479 |
"step": 3350
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 0.5737912412454645,
|
| 483 |
+
"grad_norm": 1.2857924699783325,
|
| 484 |
+
"learning_rate": 2.9655855551805602e-05,
|
| 485 |
+
"loss": 1.7754,
|
| 486 |
"step": 3400
|
| 487 |
},
|
| 488 |
{
|
| 489 |
+
"epoch": 0.5822293477343684,
|
| 490 |
+
"grad_norm": 1.172126293182373,
|
| 491 |
+
"learning_rate": 2.9650793115086063e-05,
|
| 492 |
+
"loss": 1.757,
|
| 493 |
"step": 3450
|
| 494 |
},
|
| 495 |
{
|
| 496 |
+
"epoch": 0.5906674542232723,
|
| 497 |
+
"grad_norm": 1.261569857597351,
|
| 498 |
+
"learning_rate": 2.964573067836652e-05,
|
| 499 |
+
"loss": 1.7542,
|
| 500 |
"step": 3500
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 0.5991055607121761,
|
| 504 |
+
"grad_norm": 1.223888635635376,
|
| 505 |
+
"learning_rate": 2.964066824164698e-05,
|
| 506 |
+
"loss": 1.7315,
|
| 507 |
"step": 3550
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 0.6075436672010801,
|
| 511 |
+
"grad_norm": 1.1595600843429565,
|
| 512 |
+
"learning_rate": 2.963560580492744e-05,
|
| 513 |
+
"loss": 1.7447,
|
| 514 |
"step": 3600
|
| 515 |
},
|
| 516 |
{
|
| 517 |
+
"epoch": 0.615981773689984,
|
| 518 |
+
"grad_norm": 1.255895733833313,
|
| 519 |
+
"learning_rate": 2.96305433682079e-05,
|
| 520 |
+
"loss": 1.798,
|
| 521 |
"step": 3650
|
| 522 |
},
|
| 523 |
{
|
| 524 |
+
"epoch": 0.6244198801788878,
|
| 525 |
+
"grad_norm": 1.321952223777771,
|
| 526 |
+
"learning_rate": 2.9625480931488354e-05,
|
| 527 |
+
"loss": 1.7578,
|
| 528 |
"step": 3700
|
| 529 |
},
|
| 530 |
{
|
| 531 |
+
"epoch": 0.6328579866677917,
|
| 532 |
+
"grad_norm": 1.2312285900115967,
|
| 533 |
+
"learning_rate": 2.9620418494768815e-05,
|
| 534 |
+
"loss": 1.7962,
|
| 535 |
"step": 3750
|
| 536 |
},
|
| 537 |
{
|
| 538 |
+
"epoch": 0.6412960931566957,
|
| 539 |
+
"grad_norm": 1.4945430755615234,
|
| 540 |
+
"learning_rate": 2.9615356058049276e-05,
|
| 541 |
+
"loss": 1.7392,
|
| 542 |
"step": 3800
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 0.6497341996455995,
|
| 546 |
+
"grad_norm": 1.5319666862487793,
|
| 547 |
+
"learning_rate": 2.9610293621329734e-05,
|
| 548 |
+
"loss": 1.7249,
|
| 549 |
"step": 3850
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 0.6581723061345034,
|
| 553 |
+
"grad_norm": 1.3336591720581055,
|
| 554 |
+
"learning_rate": 2.9605231184610194e-05,
|
| 555 |
+
"loss": 1.7569,
|
| 556 |
"step": 3900
|
| 557 |
},
|
| 558 |
{
|
| 559 |
+
"epoch": 0.6666104126234073,
|
| 560 |
+
"grad_norm": 1.366384744644165,
|
| 561 |
+
"learning_rate": 2.9600168747890652e-05,
|
| 562 |
+
"loss": 1.7298,
|
| 563 |
"step": 3950
|
| 564 |
},
|
| 565 |
{
|
| 566 |
+
"epoch": 0.6750485191123112,
|
| 567 |
+
"grad_norm": 1.5491745471954346,
|
| 568 |
+
"learning_rate": 2.9595106311171113e-05,
|
| 569 |
+
"loss": 1.7618,
|
| 570 |
"step": 4000
|
| 571 |
},
|
| 572 |
{
|
| 573 |
+
"epoch": 0.6834866256012151,
|
| 574 |
+
"grad_norm": 1.3339734077453613,
|
| 575 |
+
"learning_rate": 2.959004387445157e-05,
|
| 576 |
+
"loss": 1.7208,
|
| 577 |
"step": 4050
|
| 578 |
},
|
| 579 |
{
|
| 580 |
+
"epoch": 0.691924732090119,
|
| 581 |
+
"grad_norm": 1.2187597751617432,
|
| 582 |
+
"learning_rate": 2.958498143773203e-05,
|
| 583 |
+
"loss": 1.7822,
|
| 584 |
"step": 4100
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 0.7003628385790228,
|
| 588 |
+
"grad_norm": 1.428877353668213,
|
| 589 |
+
"learning_rate": 2.9579919001012485e-05,
|
| 590 |
+
"loss": 1.7534,
|
| 591 |
"step": 4150
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 0.7088009450679268,
|
| 595 |
+
"grad_norm": 1.2236136198043823,
|
| 596 |
+
"learning_rate": 2.9574856564292946e-05,
|
| 597 |
+
"loss": 1.7512,
|
| 598 |
"step": 4200
|
| 599 |
},
|
| 600 |
{
|
| 601 |
+
"epoch": 0.7172390515568307,
|
| 602 |
+
"grad_norm": 1.5085322856903076,
|
| 603 |
+
"learning_rate": 2.9569794127573404e-05,
|
| 604 |
+
"loss": 1.7556,
|
| 605 |
"step": 4250
|
| 606 |
},
|
| 607 |
{
|
| 608 |
+
"epoch": 0.7256771580457345,
|
| 609 |
+
"grad_norm": 1.3578091859817505,
|
| 610 |
+
"learning_rate": 2.9564731690853865e-05,
|
| 611 |
+
"loss": 1.7768,
|
| 612 |
"step": 4300
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"epoch": 0.7341152645346384,
|
| 616 |
+
"grad_norm": 1.2682510614395142,
|
| 617 |
+
"learning_rate": 2.9559669254134322e-05,
|
| 618 |
+
"loss": 1.7726,
|
| 619 |
"step": 4350
|
| 620 |
},
|
| 621 |
{
|
| 622 |
+
"epoch": 0.7425533710235424,
|
| 623 |
+
"grad_norm": 1.1269199848175049,
|
| 624 |
+
"learning_rate": 2.9554606817414783e-05,
|
| 625 |
+
"loss": 1.7732,
|
| 626 |
"step": 4400
|
| 627 |
},
|
| 628 |
{
|
| 629 |
+
"epoch": 0.7509914775124462,
|
| 630 |
+
"grad_norm": 1.368332862854004,
|
| 631 |
+
"learning_rate": 2.9549544380695244e-05,
|
| 632 |
+
"loss": 1.7475,
|
| 633 |
"step": 4450
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 0.7594295840013501,
|
| 637 |
+
"grad_norm": 1.813037395477295,
|
| 638 |
+
"learning_rate": 2.95444819439757e-05,
|
| 639 |
+
"loss": 1.7674,
|
| 640 |
"step": 4500
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 0.7678676904902539,
|
| 644 |
+
"grad_norm": 1.1015700101852417,
|
| 645 |
+
"learning_rate": 2.9539419507256162e-05,
|
| 646 |
+
"loss": 1.7419,
|
| 647 |
"step": 4550
|
| 648 |
},
|
| 649 |
{
|
| 650 |
+
"epoch": 0.7763057969791579,
|
| 651 |
+
"grad_norm": 1.2412892580032349,
|
| 652 |
+
"learning_rate": 2.953435707053662e-05,
|
| 653 |
+
"loss": 1.7903,
|
| 654 |
"step": 4600
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 0.7847439034680618,
|
| 658 |
+
"grad_norm": 1.2447668313980103,
|
| 659 |
+
"learning_rate": 2.9529294633817077e-05,
|
| 660 |
+
"loss": 1.7427,
|
| 661 |
"step": 4650
|
| 662 |
},
|
| 663 |
{
|
| 664 |
+
"epoch": 0.7931820099569656,
|
| 665 |
+
"grad_norm": 1.1194871664047241,
|
| 666 |
+
"learning_rate": 2.9524232197097535e-05,
|
| 667 |
+
"loss": 1.7353,
|
| 668 |
"step": 4700
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 0.8016201164458695,
|
| 672 |
+
"grad_norm": 1.322404146194458,
|
| 673 |
+
"learning_rate": 2.9519169760377996e-05,
|
| 674 |
+
"loss": 1.7522,
|
| 675 |
"step": 4750
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 0.8100582229347735,
|
| 679 |
+
"grad_norm": 1.2130954265594482,
|
| 680 |
+
"learning_rate": 2.9514107323658453e-05,
|
| 681 |
+
"loss": 1.7373,
|
| 682 |
"step": 4800
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 0.8184963294236773,
|
| 686 |
+
"grad_norm": 1.2840181589126587,
|
| 687 |
+
"learning_rate": 2.9509044886938914e-05,
|
| 688 |
+
"loss": 1.7588,
|
| 689 |
"step": 4850
|
| 690 |
},
|
| 691 |
{
|
| 692 |
+
"epoch": 0.8269344359125812,
|
| 693 |
+
"grad_norm": 1.1954180002212524,
|
| 694 |
+
"learning_rate": 2.9503982450219372e-05,
|
| 695 |
+
"loss": 1.7523,
|
| 696 |
"step": 4900
|
| 697 |
},
|
| 698 |
{
|
| 699 |
+
"epoch": 0.8353725424014851,
|
| 700 |
+
"grad_norm": 1.2900581359863281,
|
| 701 |
+
"learning_rate": 2.9498920013499833e-05,
|
| 702 |
+
"loss": 1.7747,
|
| 703 |
"step": 4950
|
| 704 |
},
|
| 705 |
{
|
| 706 |
+
"epoch": 0.843810648890389,
|
| 707 |
+
"grad_norm": 1.2370872497558594,
|
| 708 |
+
"learning_rate": 2.9493857576780294e-05,
|
| 709 |
+
"loss": 1.744,
|
| 710 |
"step": 5000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 0.8522487553792929,
|
| 714 |
+
"grad_norm": 2.0214121341705322,
|
| 715 |
+
"learning_rate": 2.948879514006075e-05,
|
| 716 |
+
"loss": 1.7506,
|
| 717 |
"step": 5050
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 0.8606868618681968,
|
| 721 |
+
"grad_norm": 1.257506012916565,
|
| 722 |
+
"learning_rate": 2.948373270334121e-05,
|
| 723 |
+
"loss": 1.692,
|
| 724 |
"step": 5100
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"epoch": 0.8691249683571006,
|
| 728 |
+
"grad_norm": 1.2164514064788818,
|
| 729 |
+
"learning_rate": 2.9478670266621666e-05,
|
| 730 |
+
"loss": 1.7464,
|
| 731 |
"step": 5150
|
| 732 |
},
|
| 733 |
{
|
| 734 |
+
"epoch": 0.8775630748460046,
|
| 735 |
+
"grad_norm": 1.2734804153442383,
|
| 736 |
+
"learning_rate": 2.9473607829902127e-05,
|
| 737 |
+
"loss": 1.7383,
|
| 738 |
"step": 5200
|
| 739 |
},
|
| 740 |
{
|
| 741 |
+
"epoch": 0.8860011813349085,
|
| 742 |
+
"grad_norm": 1.2429083585739136,
|
| 743 |
+
"learning_rate": 2.9468545393182585e-05,
|
| 744 |
+
"loss": 1.7666,
|
| 745 |
"step": 5250
|
| 746 |
},
|
| 747 |
{
|
| 748 |
+
"epoch": 0.8944392878238123,
|
| 749 |
+
"grad_norm": 1.2660380601882935,
|
| 750 |
+
"learning_rate": 2.9463482956463045e-05,
|
| 751 |
+
"loss": 1.7715,
|
| 752 |
"step": 5300
|
| 753 |
},
|
| 754 |
{
|
| 755 |
+
"epoch": 0.9028773943127162,
|
| 756 |
+
"grad_norm": 1.426188588142395,
|
| 757 |
+
"learning_rate": 2.9458420519743503e-05,
|
| 758 |
+
"loss": 1.7812,
|
| 759 |
"step": 5350
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 0.9113155008016202,
|
| 763 |
+
"grad_norm": 1.23947274684906,
|
| 764 |
+
"learning_rate": 2.9453358083023964e-05,
|
| 765 |
+
"loss": 1.7604,
|
| 766 |
"step": 5400
|
| 767 |
},
|
| 768 |
{
|
| 769 |
+
"epoch": 0.919753607290524,
|
| 770 |
+
"grad_norm": 1.183780312538147,
|
| 771 |
+
"learning_rate": 2.944829564630442e-05,
|
| 772 |
+
"loss": 1.7618,
|
| 773 |
"step": 5450
|
| 774 |
},
|
| 775 |
{
|
| 776 |
+
"epoch": 0.9281917137794279,
|
| 777 |
+
"grad_norm": 1.3170071840286255,
|
| 778 |
+
"learning_rate": 2.9443233209584882e-05,
|
| 779 |
+
"loss": 1.7323,
|
| 780 |
"step": 5500
|
| 781 |
},
|
| 782 |
{
|
| 783 |
+
"epoch": 0.9366298202683317,
|
| 784 |
+
"grad_norm": 1.2505546808242798,
|
| 785 |
+
"learning_rate": 2.9438170772865343e-05,
|
| 786 |
+
"loss": 1.7306,
|
| 787 |
"step": 5550
|
| 788 |
},
|
| 789 |
{
|
| 790 |
+
"epoch": 0.9450679267572357,
|
| 791 |
+
"grad_norm": 1.2599437236785889,
|
| 792 |
+
"learning_rate": 2.9433108336145797e-05,
|
| 793 |
+
"loss": 1.7135,
|
| 794 |
"step": 5600
|
| 795 |
},
|
| 796 |
{
|
| 797 |
+
"epoch": 0.9535060332461396,
|
| 798 |
+
"grad_norm": 1.3927714824676514,
|
| 799 |
+
"learning_rate": 2.9428045899426258e-05,
|
| 800 |
+
"loss": 1.7458,
|
| 801 |
"step": 5650
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 0.9619441397350434,
|
| 805 |
+
"grad_norm": 1.5024762153625488,
|
| 806 |
+
"learning_rate": 2.9422983462706716e-05,
|
| 807 |
+
"loss": 1.751,
|
| 808 |
"step": 5700
|
| 809 |
},
|
| 810 |
{
|
| 811 |
+
"epoch": 0.9703822462239473,
|
| 812 |
+
"grad_norm": 1.2158094644546509,
|
| 813 |
+
"learning_rate": 2.9417921025987177e-05,
|
| 814 |
+
"loss": 1.7217,
|
| 815 |
"step": 5750
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 0.9788203527128513,
|
| 819 |
+
"grad_norm": 1.177870750427246,
|
| 820 |
+
"learning_rate": 2.9412858589267634e-05,
|
| 821 |
+
"loss": 1.7481,
|
| 822 |
"step": 5800
|
| 823 |
},
|
| 824 |
{
|
| 825 |
+
"epoch": 0.9872584592017551,
|
| 826 |
+
"grad_norm": 1.2789208889007568,
|
| 827 |
+
"learning_rate": 2.9407796152548095e-05,
|
| 828 |
+
"loss": 1.7697,
|
| 829 |
"step": 5850
|
| 830 |
},
|
| 831 |
{
|
| 832 |
+
"epoch": 0.995696565690659,
|
| 833 |
+
"grad_norm": 1.2305048704147339,
|
| 834 |
+
"learning_rate": 2.9402733715828553e-05,
|
| 835 |
+
"loss": 1.7322,
|
| 836 |
"step": 5900
|
| 837 |
},
|
| 838 |
{
|
| 839 |
+
"epoch": 1.0,
|
| 840 |
+
"eval_loss": 1.6421738862991333,
|
| 841 |
+
"eval_rouge1": 0.4250170405747811,
|
| 842 |
+
"eval_rouge2": 0.21400043951132275,
|
| 843 |
+
"eval_rougeL": 0.35855678329236096,
|
| 844 |
+
"eval_runtime": 1163.9596,
|
| 845 |
+
"eval_samples_per_second": 20.363,
|
| 846 |
+
"eval_steps_per_second": 1.273,
|
| 847 |
+
"step": 5926
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 1.0040502911146738,
|
| 851 |
+
"grad_norm": 1.1294411420822144,
|
| 852 |
+
"learning_rate": 2.9397671279109013e-05,
|
| 853 |
+
"loss": 1.7502,
|
| 854 |
"step": 5950
|
| 855 |
},
|
| 856 |
{
|
| 857 |
+
"epoch": 1.0124883976035777,
|
| 858 |
+
"grad_norm": 1.2580525875091553,
|
| 859 |
+
"learning_rate": 2.939260884238947e-05,
|
| 860 |
+
"loss": 1.6828,
|
| 861 |
"step": 6000
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 1.0209265040924818,
|
| 865 |
+
"grad_norm": 1.9175118207931519,
|
| 866 |
+
"learning_rate": 2.938754640566993e-05,
|
| 867 |
+
"loss": 1.7069,
|
| 868 |
"step": 6050
|
| 869 |
},
|
| 870 |
{
|
| 871 |
+
"epoch": 1.0293646105813856,
|
| 872 |
+
"grad_norm": 1.4026094675064087,
|
| 873 |
+
"learning_rate": 2.938248396895039e-05,
|
| 874 |
+
"loss": 1.7445,
|
| 875 |
"step": 6100
|
| 876 |
},
|
| 877 |
{
|
| 878 |
+
"epoch": 1.0378027170702895,
|
| 879 |
+
"grad_norm": 1.1968200206756592,
|
| 880 |
+
"learning_rate": 2.9377421532230847e-05,
|
| 881 |
+
"loss": 1.6942,
|
| 882 |
"step": 6150
|
| 883 |
},
|
| 884 |
{
|
| 885 |
+
"epoch": 1.0462408235591933,
|
| 886 |
+
"grad_norm": 1.1554464101791382,
|
| 887 |
+
"learning_rate": 2.9372359095511308e-05,
|
| 888 |
+
"loss": 1.7288,
|
| 889 |
"step": 6200
|
| 890 |
},
|
| 891 |
{
|
| 892 |
+
"epoch": 1.0546789300480972,
|
| 893 |
+
"grad_norm": 1.3406728506088257,
|
| 894 |
+
"learning_rate": 2.9367296658791765e-05,
|
| 895 |
+
"loss": 1.7135,
|
| 896 |
"step": 6250
|
| 897 |
},
|
| 898 |
{
|
| 899 |
+
"epoch": 1.063117036537001,
|
| 900 |
+
"grad_norm": 1.1203522682189941,
|
| 901 |
+
"learning_rate": 2.9362234222072226e-05,
|
| 902 |
+
"loss": 1.7165,
|
| 903 |
"step": 6300
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 1.071555143025905,
|
| 907 |
+
"grad_norm": 1.1261606216430664,
|
| 908 |
+
"learning_rate": 2.9357171785352684e-05,
|
| 909 |
+
"loss": 1.7016,
|
| 910 |
"step": 6350
|
| 911 |
},
|
| 912 |
{
|
| 913 |
+
"epoch": 1.0799932495148088,
|
| 914 |
+
"grad_norm": 2.1716997623443604,
|
| 915 |
+
"learning_rate": 2.9352109348633145e-05,
|
| 916 |
+
"loss": 1.6976,
|
| 917 |
"step": 6400
|
| 918 |
},
|
| 919 |
{
|
| 920 |
+
"epoch": 1.0884313560037127,
|
| 921 |
+
"grad_norm": 1.335428237915039,
|
| 922 |
+
"learning_rate": 2.9347046911913602e-05,
|
| 923 |
+
"loss": 1.7018,
|
| 924 |
"step": 6450
|
| 925 |
},
|
| 926 |
{
|
| 927 |
+
"epoch": 1.0968694624926167,
|
| 928 |
+
"grad_norm": 1.216572642326355,
|
| 929 |
+
"learning_rate": 2.9341984475194063e-05,
|
| 930 |
+
"loss": 1.7092,
|
| 931 |
"step": 6500
|
| 932 |
},
|
| 933 |
{
|
| 934 |
+
"epoch": 1.1053075689815206,
|
| 935 |
+
"grad_norm": 1.151798129081726,
|
| 936 |
+
"learning_rate": 2.9336922038474517e-05,
|
| 937 |
+
"loss": 1.7169,
|
| 938 |
"step": 6550
|
| 939 |
},
|
| 940 |
{
|
| 941 |
+
"epoch": 1.1137456754704245,
|
| 942 |
+
"grad_norm": 1.283575415611267,
|
| 943 |
+
"learning_rate": 2.9331859601754978e-05,
|
| 944 |
+
"loss": 1.7384,
|
| 945 |
"step": 6600
|
| 946 |
},
|
| 947 |
{
|
| 948 |
+
"epoch": 1.1221837819593283,
|
| 949 |
+
"grad_norm": 1.1927815675735474,
|
| 950 |
+
"learning_rate": 2.9326797165035436e-05,
|
| 951 |
+
"loss": 1.7135,
|
| 952 |
"step": 6650
|
| 953 |
},
|
| 954 |
{
|
| 955 |
+
"epoch": 1.1306218884482322,
|
| 956 |
+
"grad_norm": 1.230566382408142,
|
| 957 |
+
"learning_rate": 2.9321734728315896e-05,
|
| 958 |
+
"loss": 1.7062,
|
| 959 |
"step": 6700
|
| 960 |
},
|
| 961 |
{
|
| 962 |
+
"epoch": 1.139059994937136,
|
| 963 |
+
"grad_norm": 1.1863012313842773,
|
| 964 |
+
"learning_rate": 2.9316672291596357e-05,
|
| 965 |
+
"loss": 1.7038,
|
| 966 |
"step": 6750
|
| 967 |
},
|
| 968 |
{
|
| 969 |
+
"epoch": 1.14749810142604,
|
| 970 |
+
"grad_norm": 1.2071080207824707,
|
| 971 |
+
"learning_rate": 2.9311609854876815e-05,
|
| 972 |
+
"loss": 1.7232,
|
| 973 |
"step": 6800
|
| 974 |
},
|
| 975 |
{
|
| 976 |
+
"epoch": 1.155936207914944,
|
| 977 |
+
"grad_norm": 1.308000922203064,
|
| 978 |
+
"learning_rate": 2.9306547418157276e-05,
|
| 979 |
+
"loss": 1.7493,
|
| 980 |
"step": 6850
|
| 981 |
},
|
| 982 |
{
|
| 983 |
+
"epoch": 1.1643743144038479,
|
| 984 |
+
"grad_norm": 1.7674635648727417,
|
| 985 |
+
"learning_rate": 2.9301484981437733e-05,
|
| 986 |
+
"loss": 1.6925,
|
| 987 |
"step": 6900
|
| 988 |
},
|
| 989 |
{
|
| 990 |
+
"epoch": 1.1728124208927517,
|
| 991 |
+
"grad_norm": 1.2214921712875366,
|
| 992 |
+
"learning_rate": 2.9296422544718194e-05,
|
| 993 |
+
"loss": 1.7767,
|
| 994 |
"step": 6950
|
| 995 |
},
|
| 996 |
{
|
| 997 |
+
"epoch": 1.1812505273816556,
|
| 998 |
+
"grad_norm": 1.303446650505066,
|
| 999 |
+
"learning_rate": 2.929136010799865e-05,
|
| 1000 |
+
"loss": 1.7432,
|
| 1001 |
"step": 7000
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
+
"epoch": 1.1896886338705595,
|
| 1005 |
+
"grad_norm": 1.3376314640045166,
|
| 1006 |
+
"learning_rate": 2.928629767127911e-05,
|
| 1007 |
+
"loss": 1.7055,
|
| 1008 |
"step": 7050
|
| 1009 |
},
|
| 1010 |
{
|
| 1011 |
+
"epoch": 1.1981267403594633,
|
| 1012 |
+
"grad_norm": 1.422675609588623,
|
| 1013 |
+
"learning_rate": 2.9281235234559567e-05,
|
| 1014 |
+
"loss": 1.7143,
|
| 1015 |
"step": 7100
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
+
"epoch": 1.2065648468483672,
|
| 1019 |
+
"grad_norm": 1.2204426527023315,
|
| 1020 |
+
"learning_rate": 2.9276172797840028e-05,
|
| 1021 |
+
"loss": 1.6867,
|
| 1022 |
"step": 7150
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
+
"epoch": 1.215002953337271,
|
| 1026 |
+
"grad_norm": 1.1497398614883423,
|
| 1027 |
+
"learning_rate": 2.9271110361120485e-05,
|
| 1028 |
+
"loss": 1.7242,
|
| 1029 |
"step": 7200
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
+
"epoch": 1.223441059826175,
|
| 1033 |
+
"grad_norm": 1.2231690883636475,
|
| 1034 |
+
"learning_rate": 2.9266047924400946e-05,
|
| 1035 |
+
"loss": 1.7091,
|
| 1036 |
"step": 7250
|
| 1037 |
},
|
| 1038 |
{
|
| 1039 |
+
"epoch": 1.231879166315079,
|
| 1040 |
+
"grad_norm": 1.1508055925369263,
|
| 1041 |
+
"learning_rate": 2.9260985487681407e-05,
|
| 1042 |
+
"loss": 1.7229,
|
| 1043 |
"step": 7300
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
+
"epoch": 1.2403172728039829,
|
| 1047 |
+
"grad_norm": 1.1513855457305908,
|
| 1048 |
+
"learning_rate": 2.9255923050961864e-05,
|
| 1049 |
+
"loss": 1.7057,
|
| 1050 |
"step": 7350
|
| 1051 |
},
|
| 1052 |
{
|
| 1053 |
+
"epoch": 1.2487553792928867,
|
| 1054 |
+
"grad_norm": 1.2278095483779907,
|
| 1055 |
+
"learning_rate": 2.9250860614242325e-05,
|
| 1056 |
+
"loss": 1.7257,
|
| 1057 |
"step": 7400
|
| 1058 |
},
|
| 1059 |
{
|
| 1060 |
+
"epoch": 1.2571934857817906,
|
| 1061 |
+
"grad_norm": 1.4047170877456665,
|
| 1062 |
+
"learning_rate": 2.924579817752278e-05,
|
| 1063 |
+
"loss": 1.7463,
|
| 1064 |
"step": 7450
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
+
"epoch": 1.2656315922706944,
|
| 1068 |
+
"grad_norm": 1.2649122476577759,
|
| 1069 |
+
"learning_rate": 2.924073574080324e-05,
|
| 1070 |
+
"loss": 1.7019,
|
| 1071 |
"step": 7500
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 1.2740696987595983,
|
| 1075 |
+
"grad_norm": 1.2084012031555176,
|
| 1076 |
+
"learning_rate": 2.9235673304083698e-05,
|
| 1077 |
+
"loss": 1.7373,
|
| 1078 |
"step": 7550
|
| 1079 |
},
|
| 1080 |
{
|
| 1081 |
+
"epoch": 1.2825078052485022,
|
| 1082 |
+
"grad_norm": 1.3185824155807495,
|
| 1083 |
+
"learning_rate": 2.923061086736416e-05,
|
| 1084 |
+
"loss": 1.7063,
|
| 1085 |
"step": 7600
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
+
"epoch": 1.2909459117374062,
|
| 1089 |
+
"grad_norm": 1.1844264268875122,
|
| 1090 |
+
"learning_rate": 2.9225548430644616e-05,
|
| 1091 |
+
"loss": 1.7037,
|
| 1092 |
"step": 7650
|
| 1093 |
},
|
| 1094 |
{
|
| 1095 |
+
"epoch": 1.2993840182263101,
|
| 1096 |
+
"grad_norm": 1.2209100723266602,
|
| 1097 |
+
"learning_rate": 2.9220485993925077e-05,
|
| 1098 |
+
"loss": 1.7318,
|
| 1099 |
"step": 7700
|
| 1100 |
},
|
| 1101 |
{
|
| 1102 |
+
"epoch": 1.307822124715214,
|
| 1103 |
+
"grad_norm": 1.3436676263809204,
|
| 1104 |
+
"learning_rate": 2.9215423557205535e-05,
|
| 1105 |
+
"loss": 1.7082,
|
| 1106 |
"step": 7750
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
+
"epoch": 1.3162602312041178,
|
| 1110 |
+
"grad_norm": 1.3716071844100952,
|
| 1111 |
+
"learning_rate": 2.9210361120485996e-05,
|
| 1112 |
+
"loss": 1.7387,
|
| 1113 |
"step": 7800
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
+
"epoch": 1.3246983376930217,
|
| 1117 |
+
"grad_norm": 1.18326735496521,
|
| 1118 |
+
"learning_rate": 2.9205298683766457e-05,
|
| 1119 |
+
"loss": 1.7277,
|
| 1120 |
"step": 7850
|
| 1121 |
},
|
| 1122 |
{
|
| 1123 |
+
"epoch": 1.3331364441819256,
|
| 1124 |
+
"grad_norm": 1.1182228326797485,
|
| 1125 |
+
"learning_rate": 2.9200236247046914e-05,
|
| 1126 |
+
"loss": 1.7437,
|
| 1127 |
"step": 7900
|
| 1128 |
},
|
| 1129 |
{
|
| 1130 |
+
"epoch": 1.3415745506708294,
|
| 1131 |
+
"grad_norm": 1.2569571733474731,
|
| 1132 |
+
"learning_rate": 2.919517381032737e-05,
|
| 1133 |
+
"loss": 1.7196,
|
| 1134 |
"step": 7950
|
| 1135 |
},
|
| 1136 |
{
|
| 1137 |
+
"epoch": 1.3500126571597333,
|
| 1138 |
+
"grad_norm": 1.2209763526916504,
|
| 1139 |
+
"learning_rate": 2.919011137360783e-05,
|
| 1140 |
+
"loss": 1.6957,
|
| 1141 |
"step": 8000
|
| 1142 |
},
|
| 1143 |
{
|
| 1144 |
+
"epoch": 1.3584507636486371,
|
| 1145 |
+
"grad_norm": 1.291174054145813,
|
| 1146 |
+
"learning_rate": 2.918504893688829e-05,
|
| 1147 |
+
"loss": 1.7115,
|
| 1148 |
"step": 8050
|
| 1149 |
},
|
| 1150 |
{
|
| 1151 |
+
"epoch": 1.366888870137541,
|
| 1152 |
+
"grad_norm": 1.2739378213882446,
|
| 1153 |
+
"learning_rate": 2.9179986500168747e-05,
|
| 1154 |
+
"loss": 1.7291,
|
| 1155 |
"step": 8100
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
+
"epoch": 1.375326976626445,
|
| 1159 |
+
"grad_norm": 1.325810194015503,
|
| 1160 |
+
"learning_rate": 2.917492406344921e-05,
|
| 1161 |
+
"loss": 1.7318,
|
| 1162 |
"step": 8150
|
| 1163 |
},
|
| 1164 |
{
|
| 1165 |
+
"epoch": 1.383765083115349,
|
| 1166 |
+
"grad_norm": 1.3240125179290771,
|
| 1167 |
+
"learning_rate": 2.9169861626729666e-05,
|
| 1168 |
+
"loss": 1.7334,
|
| 1169 |
"step": 8200
|
| 1170 |
},
|
| 1171 |
{
|
| 1172 |
+
"epoch": 1.3922031896042528,
|
| 1173 |
+
"grad_norm": 1.1416797637939453,
|
| 1174 |
+
"learning_rate": 2.9164799190010127e-05,
|
| 1175 |
+
"loss": 1.7169,
|
| 1176 |
"step": 8250
|
| 1177 |
},
|
| 1178 |
{
|
| 1179 |
+
"epoch": 1.4006412960931567,
|
| 1180 |
+
"grad_norm": 1.7354291677474976,
|
| 1181 |
+
"learning_rate": 2.9159736753290584e-05,
|
| 1182 |
+
"loss": 1.7374,
|
| 1183 |
"step": 8300
|
| 1184 |
},
|
| 1185 |
{
|
| 1186 |
+
"epoch": 1.4090794025820605,
|
| 1187 |
+
"grad_norm": 1.2505598068237305,
|
| 1188 |
+
"learning_rate": 2.9154674316571045e-05,
|
| 1189 |
+
"loss": 1.7079,
|
| 1190 |
"step": 8350
|
| 1191 |
},
|
| 1192 |
{
|
| 1193 |
+
"epoch": 1.4175175090709644,
|
| 1194 |
+
"grad_norm": 1.2037113904953003,
|
| 1195 |
+
"learning_rate": 2.91496118798515e-05,
|
| 1196 |
+
"loss": 1.7151,
|
| 1197 |
"step": 8400
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
+
"epoch": 1.4259556155598685,
|
| 1201 |
+
"grad_norm": 1.33303701877594,
|
| 1202 |
+
"learning_rate": 2.914454944313196e-05,
|
| 1203 |
+
"loss": 1.7099,
|
| 1204 |
"step": 8450
|
| 1205 |
},
|
| 1206 |
{
|
| 1207 |
+
"epoch": 1.4343937220487724,
|
| 1208 |
+
"grad_norm": 1.2704479694366455,
|
| 1209 |
+
"learning_rate": 2.913948700641242e-05,
|
| 1210 |
+
"loss": 1.6913,
|
| 1211 |
"step": 8500
|
| 1212 |
},
|
| 1213 |
{
|
| 1214 |
+
"epoch": 1.4428318285376762,
|
| 1215 |
+
"grad_norm": 1.3634933233261108,
|
| 1216 |
+
"learning_rate": 2.913442456969288e-05,
|
| 1217 |
+
"loss": 1.7029,
|
| 1218 |
"step": 8550
|
| 1219 |
},
|
| 1220 |
{
|
| 1221 |
+
"epoch": 1.45126993502658,
|
| 1222 |
+
"grad_norm": 1.1784166097640991,
|
| 1223 |
+
"learning_rate": 2.912936213297334e-05,
|
| 1224 |
+
"loss": 1.6844,
|
| 1225 |
"step": 8600
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
+
"epoch": 1.459708041515484,
|
| 1229 |
+
"grad_norm": 1.2901785373687744,
|
| 1230 |
+
"learning_rate": 2.9124299696253797e-05,
|
| 1231 |
+
"loss": 1.7226,
|
| 1232 |
"step": 8650
|
| 1233 |
},
|
| 1234 |
{
|
| 1235 |
+
"epoch": 1.4681461480043878,
|
| 1236 |
+
"grad_norm": 1.2651816606521606,
|
| 1237 |
+
"learning_rate": 2.9119237259534258e-05,
|
| 1238 |
+
"loss": 1.727,
|
| 1239 |
"step": 8700
|
| 1240 |
},
|
| 1241 |
{
|
| 1242 |
+
"epoch": 1.4765842544932917,
|
| 1243 |
+
"grad_norm": 1.3113429546356201,
|
| 1244 |
+
"learning_rate": 2.9114174822814715e-05,
|
| 1245 |
+
"loss": 1.7239,
|
| 1246 |
"step": 8750
|
| 1247 |
},
|
| 1248 |
{
|
| 1249 |
+
"epoch": 1.4850223609821955,
|
| 1250 |
+
"grad_norm": 1.322194218635559,
|
| 1251 |
+
"learning_rate": 2.9109112386095176e-05,
|
| 1252 |
+
"loss": 1.6862,
|
| 1253 |
"step": 8800
|
| 1254 |
},
|
| 1255 |
{
|
| 1256 |
+
"epoch": 1.4934604674710994,
|
| 1257 |
+
"grad_norm": 1.1685904264450073,
|
| 1258 |
+
"learning_rate": 2.9104049949375634e-05,
|
| 1259 |
+
"loss": 1.7337,
|
| 1260 |
"step": 8850
|
| 1261 |
},
|
| 1262 |
{
|
| 1263 |
+
"epoch": 1.5018985739600033,
|
| 1264 |
+
"grad_norm": 1.3465616703033447,
|
| 1265 |
+
"learning_rate": 2.909898751265609e-05,
|
| 1266 |
+
"loss": 1.7007,
|
| 1267 |
"step": 8900
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
+
"epoch": 1.5103366804489071,
|
| 1271 |
+
"grad_norm": 1.2812334299087524,
|
| 1272 |
+
"learning_rate": 2.909392507593655e-05,
|
| 1273 |
+
"loss": 1.692,
|
| 1274 |
"step": 8950
|
| 1275 |
},
|
| 1276 |
{
|
| 1277 |
+
"epoch": 1.5187747869378112,
|
| 1278 |
+
"grad_norm": 1.2780709266662598,
|
| 1279 |
+
"learning_rate": 2.908886263921701e-05,
|
| 1280 |
+
"loss": 1.7178,
|
| 1281 |
"step": 9000
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
+
"epoch": 1.527212893426715,
|
| 1285 |
+
"grad_norm": 2.6635632514953613,
|
| 1286 |
+
"learning_rate": 2.908380020249747e-05,
|
| 1287 |
+
"loss": 1.7218,
|
| 1288 |
"step": 9050
|
| 1289 |
},
|
| 1290 |
{
|
| 1291 |
+
"epoch": 1.535650999915619,
|
| 1292 |
+
"grad_norm": 1.440506935119629,
|
| 1293 |
+
"learning_rate": 2.9078737765777928e-05,
|
| 1294 |
+
"loss": 1.7053,
|
| 1295 |
"step": 9100
|
| 1296 |
},
|
| 1297 |
{
|
| 1298 |
+
"epoch": 1.5440891064045228,
|
| 1299 |
+
"grad_norm": 1.1698048114776611,
|
| 1300 |
+
"learning_rate": 2.907367532905839e-05,
|
| 1301 |
+
"loss": 1.6817,
|
| 1302 |
"step": 9150
|
| 1303 |
},
|
| 1304 |
{
|
| 1305 |
+
"epoch": 1.5525272128934269,
|
| 1306 |
+
"grad_norm": 2.538287401199341,
|
| 1307 |
+
"learning_rate": 2.9068612892338847e-05,
|
| 1308 |
+
"loss": 1.7249,
|
| 1309 |
"step": 9200
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
+
"epoch": 1.5609653193823307,
|
| 1313 |
+
"grad_norm": 1.577597737312317,
|
| 1314 |
+
"learning_rate": 2.9063550455619308e-05,
|
| 1315 |
+
"loss": 1.7239,
|
| 1316 |
"step": 9250
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
+
"epoch": 1.5694034258712346,
|
| 1320 |
+
"grad_norm": 1.2279037237167358,
|
| 1321 |
+
"learning_rate": 2.9058488018899765e-05,
|
| 1322 |
+
"loss": 1.7013,
|
| 1323 |
"step": 9300
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
+
"epoch": 1.5778415323601385,
|
| 1327 |
+
"grad_norm": 1.2282285690307617,
|
| 1328 |
+
"learning_rate": 2.9053425582180223e-05,
|
| 1329 |
+
"loss": 1.7367,
|
| 1330 |
"step": 9350
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
+
"epoch": 1.5862796388490423,
|
| 1334 |
+
"grad_norm": 1.1559499502182007,
|
| 1335 |
+
"learning_rate": 2.904836314546068e-05,
|
| 1336 |
+
"loss": 1.6994,
|
| 1337 |
"step": 9400
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
+
"epoch": 1.5947177453379462,
|
| 1341 |
+
"grad_norm": 1.1341310739517212,
|
| 1342 |
+
"learning_rate": 2.904330070874114e-05,
|
| 1343 |
+
"loss": 1.6864,
|
| 1344 |
"step": 9450
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
+
"epoch": 1.60315585182685,
|
| 1348 |
+
"grad_norm": 1.2908226251602173,
|
| 1349 |
+
"learning_rate": 2.90382382720216e-05,
|
| 1350 |
+
"loss": 1.6823,
|
| 1351 |
"step": 9500
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
+
"epoch": 1.611593958315754,
|
| 1355 |
+
"grad_norm": 1.2425497770309448,
|
| 1356 |
+
"learning_rate": 2.903317583530206e-05,
|
| 1357 |
+
"loss": 1.7256,
|
| 1358 |
"step": 9550
|
| 1359 |
},
|
| 1360 |
{
|
| 1361 |
+
"epoch": 1.6200320648046578,
|
| 1362 |
+
"grad_norm": 1.4069569110870361,
|
| 1363 |
+
"learning_rate": 2.902811339858252e-05,
|
| 1364 |
+
"loss": 1.6999,
|
| 1365 |
"step": 9600
|
| 1366 |
},
|
| 1367 |
{
|
| 1368 |
+
"epoch": 1.6284701712935616,
|
| 1369 |
+
"grad_norm": 1.2929143905639648,
|
| 1370 |
+
"learning_rate": 2.9023050961862978e-05,
|
| 1371 |
+
"loss": 1.7009,
|
| 1372 |
"step": 9650
|
| 1373 |
},
|
| 1374 |
{
|
| 1375 |
+
"epoch": 1.6369082777824655,
|
| 1376 |
+
"grad_norm": 1.2135717868804932,
|
| 1377 |
+
"learning_rate": 2.901798852514344e-05,
|
| 1378 |
+
"loss": 1.7176,
|
| 1379 |
"step": 9700
|
| 1380 |
},
|
| 1381 |
{
|
| 1382 |
+
"epoch": 1.6453463842713694,
|
| 1383 |
+
"grad_norm": 1.3558820486068726,
|
| 1384 |
+
"learning_rate": 2.9012926088423896e-05,
|
| 1385 |
+
"loss": 1.6833,
|
| 1386 |
"step": 9750
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
+
"epoch": 1.6537844907602732,
|
| 1390 |
+
"grad_norm": 1.3430869579315186,
|
| 1391 |
+
"learning_rate": 2.9007863651704357e-05,
|
| 1392 |
+
"loss": 1.7055,
|
| 1393 |
"step": 9800
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
+
"epoch": 1.6622225972491773,
|
| 1397 |
+
"grad_norm": 1.2483643293380737,
|
| 1398 |
+
"learning_rate": 2.900280121498481e-05,
|
| 1399 |
+
"loss": 1.6712,
|
| 1400 |
"step": 9850
|
| 1401 |
},
|
| 1402 |
{
|
| 1403 |
+
"epoch": 1.6706607037380812,
|
| 1404 |
+
"grad_norm": 1.2428581714630127,
|
| 1405 |
+
"learning_rate": 2.8997738778265272e-05,
|
| 1406 |
+
"loss": 1.7177,
|
| 1407 |
"step": 9900
|
| 1408 |
},
|
| 1409 |
{
|
| 1410 |
+
"epoch": 1.679098810226985,
|
| 1411 |
+
"grad_norm": 1.265365481376648,
|
| 1412 |
+
"learning_rate": 2.899267634154573e-05,
|
| 1413 |
+
"loss": 1.6931,
|
| 1414 |
"step": 9950
|
| 1415 |
},
|
| 1416 |
{
|
| 1417 |
+
"epoch": 1.687536916715889,
|
| 1418 |
+
"grad_norm": 1.2224442958831787,
|
| 1419 |
+
"learning_rate": 2.898761390482619e-05,
|
| 1420 |
+
"loss": 1.7572,
|
| 1421 |
"step": 10000
|
| 1422 |
},
|
| 1423 |
{
|
| 1424 |
+
"epoch": 1.695975023204793,
|
| 1425 |
+
"grad_norm": 1.0398801565170288,
|
| 1426 |
+
"learning_rate": 2.8982551468106648e-05,
|
| 1427 |
+
"loss": 1.7232,
|
| 1428 |
"step": 10050
|
| 1429 |
},
|
| 1430 |
{
|
| 1431 |
+
"epoch": 1.7044131296936968,
|
| 1432 |
+
"grad_norm": 1.4071274995803833,
|
| 1433 |
+
"learning_rate": 2.897748903138711e-05,
|
| 1434 |
+
"loss": 1.7086,
|
| 1435 |
"step": 10100
|
| 1436 |
},
|
| 1437 |
{
|
| 1438 |
+
"epoch": 1.7128512361826007,
|
| 1439 |
+
"grad_norm": 1.2596487998962402,
|
| 1440 |
+
"learning_rate": 2.8972426594667566e-05,
|
| 1441 |
+
"loss": 1.7067,
|
| 1442 |
"step": 10150
|
| 1443 |
},
|
| 1444 |
{
|
| 1445 |
+
"epoch": 1.7212893426715046,
|
| 1446 |
+
"grad_norm": 1.2502026557922363,
|
| 1447 |
+
"learning_rate": 2.8967364157948027e-05,
|
| 1448 |
+
"loss": 1.7091,
|
| 1449 |
"step": 10200
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
+
"epoch": 1.7297274491604084,
|
| 1453 |
+
"grad_norm": 1.3957078456878662,
|
| 1454 |
+
"learning_rate": 2.8962301721228488e-05,
|
| 1455 |
+
"loss": 1.6856,
|
| 1456 |
"step": 10250
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
+
"epoch": 1.7381655556493123,
|
| 1460 |
+
"grad_norm": 1.3089699745178223,
|
| 1461 |
+
"learning_rate": 2.8957239284508942e-05,
|
| 1462 |
+
"loss": 1.7031,
|
| 1463 |
"step": 10300
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
+
"epoch": 1.7466036621382162,
|
| 1467 |
+
"grad_norm": 1.8227264881134033,
|
| 1468 |
+
"learning_rate": 2.8952176847789403e-05,
|
| 1469 |
+
"loss": 1.7182,
|
| 1470 |
"step": 10350
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
+
"epoch": 1.75504176862712,
|
| 1474 |
+
"grad_norm": 1.3877559900283813,
|
| 1475 |
+
"learning_rate": 2.894711441106986e-05,
|
| 1476 |
+
"loss": 1.7356,
|
| 1477 |
"step": 10400
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
+
"epoch": 1.7634798751160239,
|
| 1481 |
+
"grad_norm": 1.3717120885849,
|
| 1482 |
+
"learning_rate": 2.8942051974350322e-05,
|
| 1483 |
+
"loss": 1.724,
|
| 1484 |
"step": 10450
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
+
"epoch": 1.7719179816049277,
|
| 1488 |
+
"grad_norm": 1.173403024673462,
|
| 1489 |
+
"learning_rate": 2.893698953763078e-05,
|
| 1490 |
+
"loss": 1.7179,
|
| 1491 |
"step": 10500
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
+
"epoch": 1.7803560880938316,
|
| 1495 |
+
"grad_norm": 1.3358752727508545,
|
| 1496 |
+
"learning_rate": 2.893192710091124e-05,
|
| 1497 |
+
"loss": 1.7217,
|
| 1498 |
"step": 10550
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
+
"epoch": 1.7887941945827355,
|
| 1502 |
+
"grad_norm": 1.1707042455673218,
|
| 1503 |
+
"learning_rate": 2.8926864664191698e-05,
|
| 1504 |
+
"loss": 1.6956,
|
| 1505 |
"step": 10600
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
+
"epoch": 1.7972323010716396,
|
| 1509 |
+
"grad_norm": 1.0948277711868286,
|
| 1510 |
+
"learning_rate": 2.892180222747216e-05,
|
| 1511 |
+
"loss": 1.7153,
|
| 1512 |
"step": 10650
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
+
"epoch": 1.8056704075605434,
|
| 1516 |
+
"grad_norm": 1.251162052154541,
|
| 1517 |
+
"learning_rate": 2.8916739790752616e-05,
|
| 1518 |
+
"loss": 1.7083,
|
| 1519 |
"step": 10700
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
+
"epoch": 1.8141085140494473,
|
| 1523 |
+
"grad_norm": 1.2486658096313477,
|
| 1524 |
+
"learning_rate": 2.8911677354033074e-05,
|
| 1525 |
+
"loss": 1.6887,
|
| 1526 |
"step": 10750
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
+
"epoch": 1.8225466205383511,
|
| 1530 |
+
"grad_norm": 1.1763782501220703,
|
| 1531 |
+
"learning_rate": 2.8906614917313534e-05,
|
| 1532 |
+
"loss": 1.7169,
|
| 1533 |
"step": 10800
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
+
"epoch": 1.8309847270272552,
|
| 1537 |
+
"grad_norm": 1.2847461700439453,
|
| 1538 |
+
"learning_rate": 2.8901552480593992e-05,
|
| 1539 |
+
"loss": 1.7095,
|
| 1540 |
"step": 10850
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
+
"epoch": 1.839422833516159,
|
| 1544 |
+
"grad_norm": 1.3423278331756592,
|
| 1545 |
+
"learning_rate": 2.8896490043874453e-05,
|
| 1546 |
+
"loss": 1.7019,
|
| 1547 |
"step": 10900
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
+
"epoch": 1.847860940005063,
|
| 1551 |
+
"grad_norm": 1.1392067670822144,
|
| 1552 |
+
"learning_rate": 2.889142760715491e-05,
|
| 1553 |
+
"loss": 1.7134,
|
| 1554 |
"step": 10950
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
+
"epoch": 1.8562990464939668,
|
| 1558 |
+
"grad_norm": 1.2327152490615845,
|
| 1559 |
+
"learning_rate": 2.888636517043537e-05,
|
| 1560 |
+
"loss": 1.7111,
|
| 1561 |
"step": 11000
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
+
"epoch": 1.8647371529828707,
|
| 1565 |
+
"grad_norm": 1.2953259944915771,
|
| 1566 |
+
"learning_rate": 2.888130273371583e-05,
|
| 1567 |
+
"loss": 1.6891,
|
| 1568 |
"step": 11050
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
+
"epoch": 1.8731752594717745,
|
| 1572 |
+
"grad_norm": 1.1197056770324707,
|
| 1573 |
+
"learning_rate": 2.887624029699629e-05,
|
| 1574 |
+
"loss": 1.7304,
|
| 1575 |
"step": 11100
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
+
"epoch": 1.8816133659606784,
|
| 1579 |
+
"grad_norm": 1.2429778575897217,
|
| 1580 |
+
"learning_rate": 2.8871177860276747e-05,
|
| 1581 |
+
"loss": 1.7034,
|
| 1582 |
"step": 11150
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
+
"epoch": 1.8900514724495823,
|
| 1586 |
+
"grad_norm": 1.2462618350982666,
|
| 1587 |
+
"learning_rate": 2.8866115423557208e-05,
|
| 1588 |
+
"loss": 1.7144,
|
| 1589 |
"step": 11200
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
+
"epoch": 1.8984895789384861,
|
| 1593 |
+
"grad_norm": 1.3959741592407227,
|
| 1594 |
+
"learning_rate": 2.8861052986837662e-05,
|
| 1595 |
+
"loss": 1.7403,
|
| 1596 |
"step": 11250
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
+
"epoch": 1.90692768542739,
|
| 1600 |
+
"grad_norm": 1.173663854598999,
|
| 1601 |
+
"learning_rate": 2.8855990550118123e-05,
|
| 1602 |
+
"loss": 1.688,
|
| 1603 |
"step": 11300
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
+
"epoch": 1.9153657919162939,
|
| 1607 |
+
"grad_norm": 1.3880196809768677,
|
| 1608 |
+
"learning_rate": 2.8850928113398584e-05,
|
| 1609 |
+
"loss": 1.712,
|
| 1610 |
"step": 11350
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
+
"epoch": 1.9238038984051977,
|
| 1614 |
+
"grad_norm": 1.1173053979873657,
|
| 1615 |
+
"learning_rate": 2.884586567667904e-05,
|
| 1616 |
+
"loss": 1.6835,
|
| 1617 |
"step": 11400
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
+
"epoch": 1.9322420048941018,
|
| 1621 |
+
"grad_norm": 1.3038270473480225,
|
| 1622 |
+
"learning_rate": 2.8840803239959502e-05,
|
| 1623 |
+
"loss": 1.7065,
|
| 1624 |
"step": 11450
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
+
"epoch": 1.9406801113830057,
|
| 1628 |
+
"grad_norm": 1.2091000080108643,
|
| 1629 |
+
"learning_rate": 2.883574080323996e-05,
|
| 1630 |
+
"loss": 1.7301,
|
| 1631 |
"step": 11500
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
+
"epoch": 1.9491182178719095,
|
| 1635 |
+
"grad_norm": 1.5804047584533691,
|
| 1636 |
+
"learning_rate": 2.883067836652042e-05,
|
| 1637 |
+
"loss": 1.7425,
|
| 1638 |
"step": 11550
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
+
"epoch": 1.9575563243608134,
|
| 1642 |
+
"grad_norm": 1.3017399311065674,
|
| 1643 |
+
"learning_rate": 2.882561592980088e-05,
|
| 1644 |
+
"loss": 1.6849,
|
| 1645 |
"step": 11600
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
+
"epoch": 1.9659944308497175,
|
| 1649 |
+
"grad_norm": 1.2739709615707397,
|
| 1650 |
+
"learning_rate": 2.882055349308134e-05,
|
| 1651 |
+
"loss": 1.7163,
|
| 1652 |
"step": 11650
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
+
"epoch": 1.9744325373386213,
|
| 1656 |
+
"grad_norm": 1.9534889459609985,
|
| 1657 |
+
"learning_rate": 2.8815491056361793e-05,
|
| 1658 |
+
"loss": 1.6686,
|
| 1659 |
"step": 11700
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
+
"epoch": 1.9828706438275252,
|
| 1663 |
+
"grad_norm": 1.0748012065887451,
|
| 1664 |
+
"learning_rate": 2.8810428619642254e-05,
|
| 1665 |
+
"loss": 1.6797,
|
| 1666 |
"step": 11750
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
+
"epoch": 1.991308750316429,
|
| 1670 |
+
"grad_norm": 1.3093863725662231,
|
| 1671 |
+
"learning_rate": 2.8805366182922712e-05,
|
| 1672 |
+
"loss": 1.6907,
|
| 1673 |
"step": 11800
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
+
"epoch": 1.999746856805333,
|
| 1677 |
+
"grad_norm": 1.2243276834487915,
|
| 1678 |
+
"learning_rate": 2.8800303746203173e-05,
|
| 1679 |
+
"loss": 1.699,
|
| 1680 |
"step": 11850
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
+
"epoch": 2.0,
|
| 1684 |
+
"eval_loss": 1.623704195022583,
|
| 1685 |
+
"eval_rouge1": 0.42930090258128883,
|
| 1686 |
+
"eval_rouge2": 0.2188434687019127,
|
| 1687 |
+
"eval_rougeL": 0.3635343246206635,
|
| 1688 |
+
"eval_runtime": 1180.6849,
|
| 1689 |
+
"eval_samples_per_second": 20.075,
|
| 1690 |
+
"eval_steps_per_second": 1.255,
|
| 1691 |
+
"step": 11852
|
| 1692 |
}
|
| 1693 |
],
|
| 1694 |
"logging_steps": 50,
|
| 1695 |
+
"max_steps": 296300,
|
| 1696 |
"num_input_tokens_seen": 0,
|
| 1697 |
"num_train_epochs": 50,
|
| 1698 |
"save_steps": 500,
|
|
|
|
| 1717 |
"attributes": {}
|
| 1718 |
}
|
| 1719 |
},
|
| 1720 |
+
"total_flos": 3.459019927198925e+17,
|
| 1721 |
+
"train_batch_size": 16,
|
| 1722 |
"trial_name": null,
|
| 1723 |
"trial_params": null
|
| 1724 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5969
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e6d04f1121f054b64c0985c0ac862052750a81b18728c1138917f2f05e6e0b2
|
| 3 |
size 5969
|