Commit
•
f7f71db
1
Parent(s):
b1d6a31
Upload folder using huggingface_hub
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +2 -2
- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +689 -1886
- training_args.bin +2 -2
adapter_config.json
CHANGED
|
@@ -11,17 +11,17 @@
|
|
| 11 |
"layers_to_transform": null,
|
| 12 |
"loftq_config": {},
|
| 13 |
"lora_alpha": 16,
|
| 14 |
-
"lora_dropout": 0.
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
| 17 |
"modules_to_save": null,
|
| 18 |
"peft_type": "LORA",
|
| 19 |
-
"r":
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"
|
| 24 |
-
"
|
| 25 |
],
|
| 26 |
"task_type": "CAUSAL_LM",
|
| 27 |
"use_dora": false,
|
|
|
|
| 11 |
"layers_to_transform": null,
|
| 12 |
"loftq_config": {},
|
| 13 |
"lora_alpha": 16,
|
| 14 |
+
"lora_dropout": 0.15,
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
| 17 |
"modules_to_save": null,
|
| 18 |
"peft_type": "LORA",
|
| 19 |
+
"r": 256,
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"v_proj",
|
| 24 |
+
"q_proj"
|
| 25 |
],
|
| 26 |
"task_type": "CAUSAL_LM",
|
| 27 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bead1aedc8910fcc9b2bd6be5c43e6f3f09b2eb2adfd053c047dcbd3112bce7f
|
| 3 |
+
size 204486560
|
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4988025488
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:913606a6440e60476970cd6cd86be1b695d3311ede7dd8cc86416552ee2ac500
|
| 3 |
size 4988025488
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 240691712
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a0556278ab291851893954c2da87ed3fe2e20fe2c813ae91ac11062678129ec
|
| 3 |
size 240691712
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a58c1393e0480d870912037bcdf402732a8a7aa79ebdfce9c548518490a2546
|
| 3 |
+
size 409009786
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:457fde7689fecd18a1e7e772622122b6415ae5f684c1c653fda8583567535a48
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd2103bfcabaab6163aebcdf8618dc42afda820d0d3cde332bbcda47ae31fee6
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,2412 +1,1215 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 4.
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 4.
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 4.
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 4.
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss": 3.
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 0.
|
| 57 |
-
"loss": 3.
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate": 0.
|
| 64 |
-
"loss": 3.
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate": 0.
|
| 71 |
-
"loss": 3.
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate": 0.
|
| 78 |
-
"loss": 3.
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate": 0.
|
| 85 |
-
"loss": 3.
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 0.
|
| 92 |
-
"loss": 3.
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate": 0.
|
| 99 |
-
"loss": 3.
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate": 0.
|
| 106 |
-
"loss": 3.
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate": 0.
|
| 113 |
-
"loss": 3.
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate": 0.
|
| 120 |
-
"loss": 3.
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 0.
|
| 127 |
-
"loss": 3.
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate": 0.
|
| 134 |
-
"loss":
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate": 0.
|
| 141 |
-
"loss": 2.
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate": 0.
|
| 148 |
-
"loss": 2.
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate": 0.
|
| 155 |
-
"loss": 2.
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate": 0.
|
| 162 |
-
"loss": 2.
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate": 0.
|
| 169 |
-
"loss": 2.
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm":
|
| 175 |
-
"learning_rate": 0.
|
| 176 |
-
"loss": 2.
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm":
|
| 182 |
-
"learning_rate": 0.
|
| 183 |
-
"loss": 2.
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm":
|
| 189 |
-
"learning_rate": 0.
|
| 190 |
-
"loss": 2.
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch": 0.
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate": 0.
|
| 197 |
-
"loss": 2.
|
| 198 |
"step": 2700
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 0.
|
| 202 |
-
"grad_norm":
|
| 203 |
-
"learning_rate": 0.
|
| 204 |
-
"loss": 2.
|
| 205 |
"step": 2800
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"grad_norm":
|
| 210 |
-
"learning_rate": 0.
|
| 211 |
-
"loss": 2.
|
| 212 |
"step": 2900
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"grad_norm":
|
| 217 |
-
"learning_rate": 0.
|
| 218 |
-
"loss": 2.
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate": 0.
|
| 225 |
-
"loss": 2.
|
| 226 |
"step": 3100
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch": 0.
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate": 0.
|
| 232 |
-
"loss": 2.
|
| 233 |
"step": 3200
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch": 0.
|
| 237 |
-
"grad_norm":
|
| 238 |
-
"learning_rate": 0.
|
| 239 |
-
"loss": 2.
|
| 240 |
"step": 3300
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 0.
|
| 244 |
-
"grad_norm":
|
| 245 |
-
"learning_rate": 0.
|
| 246 |
-
"loss":
|
| 247 |
"step": 3400
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch":
|
| 251 |
-
"grad_norm":
|
| 252 |
-
"learning_rate": 0.
|
| 253 |
-
"loss":
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch":
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate": 0.
|
| 260 |
-
"loss":
|
| 261 |
"step": 3600
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
-
"grad_norm":
|
| 266 |
-
"learning_rate": 0.
|
| 267 |
-
"loss":
|
| 268 |
"step": 3700
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch":
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate": 0.
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 3800
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch":
|
| 279 |
-
"grad_norm":
|
| 280 |
-
"learning_rate": 0.
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 3900
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch":
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate": 0.
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch":
|
| 293 |
-
"grad_norm":
|
| 294 |
-
"learning_rate": 0.
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 4100
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch":
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate": 0.
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 4200
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
-
"grad_norm":
|
| 308 |
-
"learning_rate": 0.
|
| 309 |
-
"loss": 1.
|
| 310 |
"step": 4300
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"epoch":
|
| 314 |
-
"grad_norm":
|
| 315 |
-
"learning_rate": 0.
|
| 316 |
-
"loss": 1.
|
| 317 |
"step": 4400
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch":
|
| 321 |
-
"grad_norm":
|
| 322 |
-
"learning_rate": 0.
|
| 323 |
-
"loss": 1.
|
| 324 |
"step": 4500
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch":
|
| 328 |
-
"grad_norm":
|
| 329 |
-
"learning_rate": 0.
|
| 330 |
-
"loss": 1.
|
| 331 |
"step": 4600
|
| 332 |
},
|
| 333 |
{
|
| 334 |
-
"epoch":
|
| 335 |
-
"grad_norm":
|
| 336 |
-
"learning_rate": 0.
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 4700
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch":
|
| 342 |
-
"grad_norm":
|
| 343 |
-
"learning_rate": 0.
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 4800
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch":
|
| 349 |
-
"grad_norm":
|
| 350 |
-
"learning_rate": 0.
|
| 351 |
-
"loss": 1.
|
| 352 |
"step": 4900
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"epoch":
|
| 356 |
-
"grad_norm":
|
| 357 |
-
"learning_rate": 0.
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 5000
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch":
|
| 363 |
-
"grad_norm":
|
| 364 |
-
"learning_rate": 0.
|
| 365 |
-
"loss": 1.
|
| 366 |
"step": 5100
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch":
|
| 370 |
-
"grad_norm":
|
| 371 |
-
"learning_rate": 0.
|
| 372 |
-
"loss": 1.
|
| 373 |
"step": 5200
|
| 374 |
},
|
| 375 |
{
|
| 376 |
-
"epoch":
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 0.
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 5300
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch":
|
| 384 |
-
"grad_norm":
|
| 385 |
-
"learning_rate": 0.
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 5400
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch":
|
| 391 |
-
"grad_norm":
|
| 392 |
-
"learning_rate": 0.
|
| 393 |
-
"loss": 1.
|
| 394 |
"step": 5500
|
| 395 |
},
|
| 396 |
{
|
| 397 |
-
"epoch":
|
| 398 |
-
"grad_norm":
|
| 399 |
-
"learning_rate": 0.
|
| 400 |
-
"loss": 1.
|
| 401 |
"step": 5600
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch":
|
| 405 |
-
"grad_norm":
|
| 406 |
-
"learning_rate": 0.
|
| 407 |
-
"loss": 1.
|
| 408 |
"step": 5700
|
| 409 |
},
|
| 410 |
{
|
| 411 |
-
"epoch":
|
| 412 |
-
"grad_norm":
|
| 413 |
-
"learning_rate": 0.
|
| 414 |
-
"loss": 1.
|
| 415 |
"step": 5800
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch":
|
| 419 |
-
"grad_norm":
|
| 420 |
-
"learning_rate": 0.
|
| 421 |
-
"loss": 1.
|
| 422 |
"step": 5900
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch":
|
| 426 |
-
"grad_norm":
|
| 427 |
-
"learning_rate": 0.
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 6000
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
-
"grad_norm":
|
| 434 |
-
"learning_rate": 0.
|
| 435 |
-
"loss": 1.
|
| 436 |
"step": 6100
|
| 437 |
},
|
| 438 |
{
|
| 439 |
-
"epoch":
|
| 440 |
-
"grad_norm":
|
| 441 |
-
"learning_rate": 0.
|
| 442 |
-
"loss": 1.
|
| 443 |
"step": 6200
|
| 444 |
},
|
| 445 |
{
|
| 446 |
-
"epoch":
|
| 447 |
-
"grad_norm":
|
| 448 |
-
"learning_rate": 0.
|
| 449 |
-
"loss": 1.
|
| 450 |
"step": 6300
|
| 451 |
},
|
| 452 |
{
|
| 453 |
-
"epoch":
|
| 454 |
-
"grad_norm":
|
| 455 |
-
"learning_rate": 0.
|
| 456 |
-
"loss":
|
| 457 |
"step": 6400
|
| 458 |
},
|
| 459 |
{
|
| 460 |
-
"epoch":
|
| 461 |
-
"grad_norm":
|
| 462 |
-
"learning_rate": 0.
|
| 463 |
-
"loss": 1.
|
| 464 |
"step": 6500
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch":
|
| 468 |
-
"grad_norm":
|
| 469 |
-
"learning_rate": 0.
|
| 470 |
-
"loss": 1.
|
| 471 |
"step": 6600
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
-
"grad_norm":
|
| 476 |
-
"learning_rate": 0.
|
| 477 |
-
"loss": 1.
|
| 478 |
"step": 6700
|
| 479 |
},
|
| 480 |
{
|
| 481 |
-
"epoch":
|
| 482 |
-
"grad_norm":
|
| 483 |
-
"learning_rate": 0.
|
| 484 |
-
"loss":
|
| 485 |
"step": 6800
|
| 486 |
},
|
| 487 |
{
|
| 488 |
-
"epoch":
|
| 489 |
-
"grad_norm":
|
| 490 |
-
"learning_rate": 0.
|
| 491 |
-
"loss":
|
| 492 |
"step": 6900
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"epoch":
|
| 496 |
-
"grad_norm":
|
| 497 |
-
"learning_rate": 0.
|
| 498 |
-
"loss":
|
| 499 |
"step": 7000
|
| 500 |
},
|
| 501 |
{
|
| 502 |
-
"epoch":
|
| 503 |
-
"grad_norm":
|
| 504 |
-
"learning_rate": 0.
|
| 505 |
-
"loss":
|
| 506 |
"step": 7100
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch":
|
| 510 |
-
"grad_norm":
|
| 511 |
-
"learning_rate": 0.
|
| 512 |
-
"loss":
|
| 513 |
"step": 7200
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch":
|
| 517 |
-
"grad_norm":
|
| 518 |
-
"learning_rate": 0.
|
| 519 |
-
"loss":
|
| 520 |
"step": 7300
|
| 521 |
},
|
| 522 |
{
|
| 523 |
-
"epoch":
|
| 524 |
-
"grad_norm":
|
| 525 |
-
"learning_rate": 0.
|
| 526 |
-
"loss":
|
| 527 |
"step": 7400
|
| 528 |
},
|
| 529 |
{
|
| 530 |
-
"epoch":
|
| 531 |
-
"grad_norm":
|
| 532 |
-
"learning_rate": 0.
|
| 533 |
-
"loss":
|
| 534 |
"step": 7500
|
| 535 |
},
|
| 536 |
{
|
| 537 |
-
"epoch":
|
| 538 |
-
"grad_norm":
|
| 539 |
-
"learning_rate": 0.
|
| 540 |
-
"loss":
|
| 541 |
"step": 7600
|
| 542 |
},
|
| 543 |
{
|
| 544 |
-
"epoch":
|
| 545 |
-
"grad_norm":
|
| 546 |
-
"learning_rate": 0.
|
| 547 |
-
"loss":
|
| 548 |
"step": 7700
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch":
|
| 552 |
-
"grad_norm":
|
| 553 |
-
"learning_rate": 0.
|
| 554 |
-
"loss":
|
| 555 |
"step": 7800
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch":
|
| 559 |
-
"grad_norm":
|
| 560 |
-
"learning_rate": 0.
|
| 561 |
-
"loss":
|
| 562 |
"step": 7900
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch":
|
| 566 |
-
"grad_norm":
|
| 567 |
-
"learning_rate": 0.
|
| 568 |
-
"loss":
|
| 569 |
"step": 8000
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch":
|
| 573 |
-
"grad_norm":
|
| 574 |
-
"learning_rate": 0.
|
| 575 |
-
"loss":
|
| 576 |
"step": 8100
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch":
|
| 580 |
-
"grad_norm":
|
| 581 |
-
"learning_rate": 0.
|
| 582 |
-
"loss":
|
| 583 |
"step": 8200
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch":
|
| 587 |
-
"grad_norm":
|
| 588 |
-
"learning_rate": 0.
|
| 589 |
-
"loss":
|
| 590 |
"step": 8300
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch":
|
| 594 |
-
"grad_norm":
|
| 595 |
-
"learning_rate": 0.
|
| 596 |
-
"loss":
|
| 597 |
"step": 8400
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"grad_norm":
|
| 602 |
-
"learning_rate": 0.
|
| 603 |
-
"loss":
|
| 604 |
"step": 8500
|
| 605 |
},
|
| 606 |
{
|
| 607 |
-
"epoch":
|
| 608 |
-
"grad_norm":
|
| 609 |
-
"learning_rate": 0.
|
| 610 |
-
"loss":
|
| 611 |
"step": 8600
|
| 612 |
},
|
| 613 |
{
|
| 614 |
-
"epoch":
|
| 615 |
-
"grad_norm":
|
| 616 |
-
"learning_rate": 0.
|
| 617 |
-
"loss":
|
| 618 |
"step": 8700
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch":
|
| 622 |
-
"grad_norm":
|
| 623 |
-
"learning_rate": 0.
|
| 624 |
-
"loss":
|
| 625 |
"step": 8800
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch":
|
| 629 |
-
"grad_norm":
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss":
|
| 632 |
"step": 8900
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch":
|
| 636 |
-
"grad_norm":
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss":
|
| 639 |
"step": 9000
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch":
|
| 643 |
-
"grad_norm": 3.
|
| 644 |
-
"learning_rate":
|
| 645 |
-
"loss":
|
| 646 |
"step": 9100
|
| 647 |
},
|
| 648 |
{
|
| 649 |
-
"epoch":
|
| 650 |
-
"grad_norm":
|
| 651 |
-
"learning_rate":
|
| 652 |
-
"loss":
|
| 653 |
"step": 9200
|
| 654 |
},
|
| 655 |
{
|
| 656 |
-
"epoch":
|
| 657 |
-
"grad_norm":
|
| 658 |
-
"learning_rate":
|
| 659 |
-
"loss":
|
| 660 |
"step": 9300
|
| 661 |
},
|
| 662 |
{
|
| 663 |
-
"epoch":
|
| 664 |
-
"grad_norm":
|
| 665 |
-
"learning_rate":
|
| 666 |
-
"loss":
|
| 667 |
"step": 9400
|
| 668 |
},
|
| 669 |
{
|
| 670 |
-
"epoch":
|
| 671 |
-
"grad_norm":
|
| 672 |
-
"learning_rate":
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 9500
|
| 675 |
},
|
| 676 |
{
|
| 677 |
-
"epoch":
|
| 678 |
-
"grad_norm":
|
| 679 |
-
"learning_rate":
|
| 680 |
-
"loss":
|
| 681 |
"step": 9600
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch":
|
| 685 |
-
"grad_norm":
|
| 686 |
-
"learning_rate":
|
| 687 |
-
"loss":
|
| 688 |
"step": 9700
|
| 689 |
},
|
| 690 |
{
|
| 691 |
-
"epoch":
|
| 692 |
-
"grad_norm":
|
| 693 |
-
"learning_rate":
|
| 694 |
-
"loss":
|
| 695 |
"step": 9800
|
| 696 |
},
|
| 697 |
{
|
| 698 |
-
"epoch":
|
| 699 |
-
"grad_norm":
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss":
|
| 702 |
"step": 9900
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"epoch":
|
| 706 |
-
"grad_norm":
|
| 707 |
-
"learning_rate":
|
| 708 |
-
"loss":
|
| 709 |
"step": 10000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
-
"epoch":
|
| 713 |
-
"grad_norm":
|
| 714 |
-
"learning_rate":
|
| 715 |
-
"loss":
|
| 716 |
"step": 10100
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch":
|
| 720 |
-
"grad_norm":
|
| 721 |
-
"learning_rate":
|
| 722 |
-
"loss":
|
| 723 |
"step": 10200
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch":
|
| 727 |
-
"grad_norm":
|
| 728 |
-
"learning_rate":
|
| 729 |
-
"loss":
|
| 730 |
"step": 10300
|
| 731 |
},
|
| 732 |
{
|
| 733 |
-
"epoch":
|
| 734 |
-
"grad_norm":
|
| 735 |
-
"learning_rate":
|
| 736 |
-
"loss":
|
| 737 |
"step": 10400
|
| 738 |
},
|
| 739 |
{
|
| 740 |
-
"epoch":
|
| 741 |
-
"grad_norm":
|
| 742 |
-
"learning_rate":
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 10500
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch":
|
| 748 |
-
"grad_norm":
|
| 749 |
-
"learning_rate":
|
| 750 |
-
"loss":
|
| 751 |
"step": 10600
|
| 752 |
},
|
| 753 |
{
|
| 754 |
-
"epoch":
|
| 755 |
-
"grad_norm":
|
| 756 |
-
"learning_rate":
|
| 757 |
-
"loss":
|
| 758 |
"step": 10700
|
| 759 |
},
|
| 760 |
{
|
| 761 |
-
"epoch":
|
| 762 |
-
"grad_norm":
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 10800
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch":
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 10900
|
| 773 |
},
|
| 774 |
{
|
| 775 |
-
"epoch":
|
| 776 |
-
"grad_norm":
|
| 777 |
-
"learning_rate":
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 11000
|
| 780 |
},
|
| 781 |
{
|
| 782 |
-
"epoch":
|
| 783 |
-
"grad_norm":
|
| 784 |
-
"learning_rate":
|
| 785 |
-
"loss":
|
| 786 |
"step": 11100
|
| 787 |
},
|
| 788 |
{
|
| 789 |
-
"epoch":
|
| 790 |
-
"grad_norm":
|
| 791 |
-
"learning_rate":
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 11200
|
| 794 |
},
|
| 795 |
{
|
| 796 |
-
"epoch":
|
| 797 |
-
"grad_norm":
|
| 798 |
-
"learning_rate":
|
| 799 |
-
"loss":
|
| 800 |
"step": 11300
|
| 801 |
},
|
| 802 |
{
|
| 803 |
-
"epoch":
|
| 804 |
-
"grad_norm":
|
| 805 |
-
"learning_rate":
|
| 806 |
-
"loss":
|
| 807 |
"step": 11400
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch":
|
| 811 |
-
"grad_norm":
|
| 812 |
-
"learning_rate":
|
| 813 |
-
"loss":
|
| 814 |
"step": 11500
|
| 815 |
},
|
| 816 |
{
|
| 817 |
-
"epoch":
|
| 818 |
-
"grad_norm":
|
| 819 |
-
"learning_rate":
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 11600
|
| 822 |
},
|
| 823 |
{
|
| 824 |
-
"epoch":
|
| 825 |
-
"grad_norm":
|
| 826 |
-
"learning_rate":
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 11700
|
| 829 |
},
|
| 830 |
{
|
| 831 |
-
"epoch":
|
| 832 |
-
"grad_norm":
|
| 833 |
-
"learning_rate":
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 11800
|
| 836 |
},
|
| 837 |
{
|
| 838 |
-
"epoch":
|
| 839 |
-
"grad_norm":
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss":
|
| 842 |
"step": 11900
|
| 843 |
},
|
| 844 |
{
|
| 845 |
-
"epoch":
|
| 846 |
-
"grad_norm":
|
| 847 |
-
"learning_rate":
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 12000
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch":
|
| 853 |
-
"grad_norm":
|
| 854 |
-
"learning_rate":
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 12100
|
| 857 |
},
|
| 858 |
{
|
| 859 |
-
"epoch":
|
| 860 |
-
"grad_norm":
|
| 861 |
-
"learning_rate":
|
| 862 |
-
"loss":
|
| 863 |
"step": 12200
|
| 864 |
},
|
| 865 |
{
|
| 866 |
-
"epoch":
|
| 867 |
-
"grad_norm":
|
| 868 |
-
"learning_rate":
|
| 869 |
-
"loss":
|
| 870 |
"step": 12300
|
| 871 |
},
|
| 872 |
{
|
| 873 |
-
"epoch":
|
| 874 |
-
"grad_norm":
|
| 875 |
-
"learning_rate":
|
| 876 |
-
"loss":
|
| 877 |
"step": 12400
|
| 878 |
},
|
| 879 |
{
|
| 880 |
-
"epoch":
|
| 881 |
-
"grad_norm":
|
| 882 |
-
"learning_rate":
|
| 883 |
-
"loss":
|
| 884 |
"step": 12500
|
| 885 |
},
|
| 886 |
{
|
| 887 |
-
"epoch":
|
| 888 |
-
"grad_norm":
|
| 889 |
-
"learning_rate":
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 12600
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch":
|
| 895 |
-
"grad_norm":
|
| 896 |
-
"learning_rate":
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 12700
|
| 899 |
},
|
| 900 |
{
|
| 901 |
-
"epoch":
|
| 902 |
-
"grad_norm":
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss":
|
| 905 |
"step": 12800
|
| 906 |
},
|
| 907 |
{
|
| 908 |
-
"epoch":
|
| 909 |
-
"grad_norm":
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 12900
|
| 913 |
},
|
| 914 |
{
|
| 915 |
-
"epoch":
|
| 916 |
-
"grad_norm":
|
| 917 |
-
"learning_rate":
|
| 918 |
-
"loss":
|
| 919 |
"step": 13000
|
| 920 |
},
|
| 921 |
{
|
| 922 |
-
"epoch":
|
| 923 |
-
"grad_norm":
|
| 924 |
-
"learning_rate":
|
| 925 |
-
"loss":
|
| 926 |
"step": 13100
|
| 927 |
},
|
| 928 |
{
|
| 929 |
-
"epoch":
|
| 930 |
-
"grad_norm":
|
| 931 |
-
"learning_rate":
|
| 932 |
-
"loss":
|
| 933 |
"step": 13200
|
| 934 |
},
|
| 935 |
{
|
| 936 |
-
"epoch":
|
| 937 |
-
"grad_norm":
|
| 938 |
-
"learning_rate":
|
| 939 |
-
"loss":
|
| 940 |
"step": 13300
|
| 941 |
},
|
| 942 |
{
|
| 943 |
-
"epoch":
|
| 944 |
-
"grad_norm":
|
| 945 |
-
"learning_rate":
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 13400
|
| 948 |
},
|
| 949 |
{
|
| 950 |
-
"epoch":
|
| 951 |
-
"grad_norm":
|
| 952 |
-
"learning_rate":
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 13500
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"epoch":
|
| 958 |
-
"grad_norm":
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"loss":
|
| 961 |
"step": 13600
|
| 962 |
},
|
| 963 |
{
|
| 964 |
-
"epoch": 0
|
| 965 |
-
"grad_norm":
|
| 966 |
-
"learning_rate":
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 13700
|
| 969 |
},
|
| 970 |
{
|
| 971 |
-
"epoch":
|
| 972 |
-
"grad_norm":
|
| 973 |
-
"learning_rate":
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 13800
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch":
|
| 979 |
-
"grad_norm": 2.
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 13900
|
| 983 |
},
|
| 984 |
{
|
| 985 |
-
"epoch":
|
| 986 |
-
"grad_norm":
|
| 987 |
-
"learning_rate":
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 14000
|
| 990 |
},
|
| 991 |
{
|
| 992 |
-
"epoch":
|
| 993 |
-
"grad_norm":
|
| 994 |
-
"learning_rate":
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 14100
|
| 997 |
},
|
| 998 |
{
|
| 999 |
-
"epoch":
|
| 1000 |
-
"grad_norm":
|
| 1001 |
-
"learning_rate":
|
| 1002 |
-
"loss":
|
| 1003 |
"step": 14200
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
-
"epoch":
|
| 1007 |
-
"grad_norm": 3.
|
| 1008 |
-
"learning_rate":
|
| 1009 |
-
"loss":
|
| 1010 |
"step": 14300
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
-
"epoch":
|
| 1014 |
-
"grad_norm":
|
| 1015 |
-
"learning_rate":
|
| 1016 |
-
"loss":
|
| 1017 |
"step": 14400
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
-
"epoch":
|
| 1021 |
-
"grad_norm":
|
| 1022 |
-
"learning_rate":
|
| 1023 |
-
"loss":
|
| 1024 |
"step": 14500
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
-
"epoch":
|
| 1028 |
-
"grad_norm":
|
| 1029 |
-
"learning_rate":
|
| 1030 |
-
"loss":
|
| 1031 |
"step": 14600
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
-
"epoch":
|
| 1035 |
-
"grad_norm":
|
| 1036 |
-
"learning_rate":
|
| 1037 |
-
"loss":
|
| 1038 |
"step": 14700
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
-
"epoch":
|
| 1042 |
-
"grad_norm":
|
| 1043 |
-
"learning_rate":
|
| 1044 |
-
"loss":
|
| 1045 |
"step": 14800
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
-
"epoch":
|
| 1049 |
-
"grad_norm":
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 14900
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
-
"epoch":
|
| 1056 |
-
"grad_norm":
|
| 1057 |
-
"learning_rate":
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 15000
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch":
|
| 1063 |
-
"grad_norm":
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 15100
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
-
"epoch":
|
| 1070 |
-
"grad_norm":
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss":
|
| 1073 |
"step": 15200
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch":
|
| 1077 |
-
"grad_norm":
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss":
|
| 1080 |
"step": 15300
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
-
"epoch":
|
| 1084 |
-
"grad_norm":
|
| 1085 |
-
"learning_rate":
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 15400
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
-
"epoch":
|
| 1091 |
-
"grad_norm":
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss":
|
| 1094 |
"step": 15500
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch":
|
| 1098 |
-
"grad_norm":
|
| 1099 |
-
"learning_rate":
|
| 1100 |
-
"loss":
|
| 1101 |
"step": 15600
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch":
|
| 1105 |
-
"grad_norm":
|
| 1106 |
-
"learning_rate":
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 15700
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
-
"epoch":
|
| 1112 |
-
"grad_norm":
|
| 1113 |
-
"learning_rate":
|
| 1114 |
-
"loss":
|
| 1115 |
"step": 15800
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
-
"epoch":
|
| 1119 |
-
"grad_norm": 1.
|
| 1120 |
-
"learning_rate":
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 15900
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
-
"epoch":
|
| 1126 |
-
"grad_norm":
|
| 1127 |
-
"learning_rate":
|
| 1128 |
-
"loss":
|
| 1129 |
"step": 16000
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
-
"epoch":
|
| 1133 |
-
"grad_norm":
|
| 1134 |
-
"learning_rate":
|
| 1135 |
-
"loss":
|
| 1136 |
"step": 16100
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"epoch":
|
| 1140 |
-
"grad_norm":
|
| 1141 |
-
"learning_rate":
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 16200
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch":
|
| 1147 |
-
"grad_norm": 3.
|
| 1148 |
-
"learning_rate":
|
| 1149 |
-
"loss":
|
| 1150 |
"step": 16300
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
-
"epoch":
|
| 1154 |
-
"grad_norm":
|
| 1155 |
-
"learning_rate":
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 16400
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
-
"epoch":
|
| 1161 |
-
"grad_norm": 3.
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss":
|
| 1164 |
"step": 16500
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
-
"epoch":
|
| 1168 |
-
"grad_norm":
|
| 1169 |
-
"learning_rate":
|
| 1170 |
-
"loss":
|
| 1171 |
"step": 16600
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
-
"epoch":
|
| 1175 |
-
"grad_norm":
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss":
|
| 1178 |
"step": 16700
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
-
"epoch":
|
| 1182 |
-
"grad_norm":
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 16800
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch":
|
| 1189 |
-
"grad_norm":
|
| 1190 |
-
"learning_rate":
|
| 1191 |
-
"loss":
|
| 1192 |
"step": 16900
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
-
"epoch":
|
| 1196 |
-
"grad_norm":
|
| 1197 |
-
"learning_rate":
|
| 1198 |
-
"loss":
|
| 1199 |
"step": 17000
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
-
"epoch":
|
| 1203 |
-
"grad_norm":
|
| 1204 |
-
"learning_rate":
|
| 1205 |
-
"loss":
|
| 1206 |
"step": 17100
|
| 1207 |
-
},
|
| 1208 |
-
{
|
| 1209 |
-
"epoch": 0.5022337723012235,
|
| 1210 |
-
"grad_norm": 1.6674872636795044,
|
| 1211 |
-
"learning_rate": 0.00010422139226811105,
|
| 1212 |
-
"loss": 0.5571,
|
| 1213 |
-
"step": 17200
|
| 1214 |
-
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 0.5051537360936724,
|
| 1217 |
-
"grad_norm": 5.221986293792725,
|
| 1218 |
-
"learning_rate": 0.0001032763389755395,
|
| 1219 |
-
"loss": 0.6719,
|
| 1220 |
-
"step": 17300
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 0.5080736998861214,
|
| 1224 |
-
"grad_norm": 5.639619827270508,
|
| 1225 |
-
"learning_rate": 0.00010233099265287182,
|
| 1226 |
-
"loss": 0.5489,
|
| 1227 |
-
"step": 17400
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 0.5109936636785704,
|
| 1231 |
-
"grad_norm": 5.451557636260986,
|
| 1232 |
-
"learning_rate": 0.0001013854378502496,
|
| 1233 |
-
"loss": 0.6173,
|
| 1234 |
-
"step": 17500
|
| 1235 |
-
},
|
| 1236 |
-
{
|
| 1237 |
-
"epoch": 0.5139136274710193,
|
| 1238 |
-
"grad_norm": 5.069969177246094,
|
| 1239 |
-
"learning_rate": 0.00010043975913646017,
|
| 1240 |
-
"loss": 0.5701,
|
| 1241 |
-
"step": 17600
|
| 1242 |
-
},
|
| 1243 |
-
{
|
| 1244 |
-
"epoch": 0.5168335912634683,
|
| 1245 |
-
"grad_norm": 1.3468672037124634,
|
| 1246 |
-
"learning_rate": 9.949404109137361e-05,
|
| 1247 |
-
"loss": 0.5582,
|
| 1248 |
-
"step": 17700
|
| 1249 |
-
},
|
| 1250 |
-
{
|
| 1251 |
-
"epoch": 0.5197535550559174,
|
| 1252 |
-
"grad_norm": 7.0001420974731445,
|
| 1253 |
-
"learning_rate": 9.854836829837741e-05,
|
| 1254 |
-
"loss": 0.6168,
|
| 1255 |
-
"step": 17800
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 0.5226735188483663,
|
| 1259 |
-
"grad_norm": 2.2107551097869873,
|
| 1260 |
-
"learning_rate": 9.760282533681206e-05,
|
| 1261 |
-
"loss": 0.4888,
|
| 1262 |
-
"step": 17900
|
| 1263 |
-
},
|
| 1264 |
-
{
|
| 1265 |
-
"epoch": 0.5255934826408153,
|
| 1266 |
-
"grad_norm": 5.56157922744751,
|
| 1267 |
-
"learning_rate": 9.665749677440597e-05,
|
| 1268 |
-
"loss": 0.5833,
|
| 1269 |
-
"step": 18000
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 0.5285134464332643,
|
| 1273 |
-
"grad_norm": 3.0002171993255615,
|
| 1274 |
-
"learning_rate": 9.571246715971211e-05,
|
| 1275 |
-
"loss": 0.5705,
|
| 1276 |
-
"step": 18100
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 0.5314334102257132,
|
| 1280 |
-
"grad_norm": 3.6839287281036377,
|
| 1281 |
-
"learning_rate": 9.476782101454611e-05,
|
| 1282 |
-
"loss": 0.5261,
|
| 1283 |
-
"step": 18200
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 0.5343533740181622,
|
| 1287 |
-
"grad_norm": 6.553487777709961,
|
| 1288 |
-
"learning_rate": 9.382364282642675e-05,
|
| 1289 |
-
"loss": 0.5936,
|
| 1290 |
-
"step": 18300
|
| 1291 |
-
},
|
| 1292 |
-
{
|
| 1293 |
-
"epoch": 0.5372733378106112,
|
| 1294 |
-
"grad_norm": 6.64651346206665,
|
| 1295 |
-
"learning_rate": 9.28800170410195e-05,
|
| 1296 |
-
"loss": 0.5576,
|
| 1297 |
-
"step": 18400
|
| 1298 |
-
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 0.5401933016030601,
|
| 1301 |
-
"grad_norm": 4.049524307250977,
|
| 1302 |
-
"learning_rate": 9.193702805458388e-05,
|
| 1303 |
-
"loss": 0.5326,
|
| 1304 |
-
"step": 18500
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 0.5431132653955091,
|
| 1308 |
-
"grad_norm": 3.995262861251831,
|
| 1309 |
-
"learning_rate": 9.09947602064253e-05,
|
| 1310 |
-
"loss": 0.577,
|
| 1311 |
-
"step": 18600
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 0.546033229187958,
|
| 1315 |
-
"grad_norm": 5.921045303344727,
|
| 1316 |
-
"learning_rate": 9.005329777135172e-05,
|
| 1317 |
-
"loss": 0.5901,
|
| 1318 |
-
"step": 18700
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 0.548953192980407,
|
| 1322 |
-
"grad_norm": 1.7855268716812134,
|
| 1323 |
-
"learning_rate": 8.911272495213639e-05,
|
| 1324 |
-
"loss": 0.5026,
|
| 1325 |
-
"step": 18800
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 0.551873156772856,
|
| 1329 |
-
"grad_norm": 2.2023262977600098,
|
| 1330 |
-
"learning_rate": 8.818251676677135e-05,
|
| 1331 |
-
"loss": 0.526,
|
| 1332 |
-
"step": 18900
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 0.5547931205653049,
|
| 1336 |
-
"grad_norm": 2.3260550498962402,
|
| 1337 |
-
"learning_rate": 8.724396446837928e-05,
|
| 1338 |
-
"loss": 0.5715,
|
| 1339 |
-
"step": 19000
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 0.557713084357754,
|
| 1343 |
-
"grad_norm": 4.143796920776367,
|
| 1344 |
-
"learning_rate": 8.63065530477612e-05,
|
| 1345 |
-
"loss": 0.5346,
|
| 1346 |
-
"step": 19100
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 0.560633048150203,
|
| 1350 |
-
"grad_norm": 3.3329811096191406,
|
| 1351 |
-
"learning_rate": 8.537036634537457e-05,
|
| 1352 |
-
"loss": 0.525,
|
| 1353 |
-
"step": 19200
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 0.563553011942652,
|
| 1357 |
-
"grad_norm": 2.1307923793792725,
|
| 1358 |
-
"learning_rate": 8.443548809214026e-05,
|
| 1359 |
-
"loss": 0.6003,
|
| 1360 |
-
"step": 19300
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 0.5664729757351009,
|
| 1364 |
-
"grad_norm": 9.149153709411621,
|
| 1365 |
-
"learning_rate": 8.350200190195363e-05,
|
| 1366 |
-
"loss": 0.5832,
|
| 1367 |
-
"step": 19400
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 0.5693929395275499,
|
| 1371 |
-
"grad_norm": 2.979999303817749,
|
| 1372 |
-
"learning_rate": 8.25699912642064e-05,
|
| 1373 |
-
"loss": 0.5697,
|
| 1374 |
-
"step": 19500
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 0.5723129033199988,
|
| 1378 |
-
"grad_norm": 3.4557132720947266,
|
| 1379 |
-
"learning_rate": 8.163953953631926e-05,
|
| 1380 |
-
"loss": 0.5883,
|
| 1381 |
-
"step": 19600
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 0.5752328671124478,
|
| 1385 |
-
"grad_norm": 5.705073833465576,
|
| 1386 |
-
"learning_rate": 8.071072993628694e-05,
|
| 1387 |
-
"loss": 0.5408,
|
| 1388 |
-
"step": 19700
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 0.5781528309048968,
|
| 1392 |
-
"grad_norm": 4.233023166656494,
|
| 1393 |
-
"learning_rate": 7.978364553523492e-05,
|
| 1394 |
-
"loss": 0.564,
|
| 1395 |
-
"step": 19800
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 0.5810727946973457,
|
| 1399 |
-
"grad_norm": 4.051882266998291,
|
| 1400 |
-
"learning_rate": 7.885836924998998e-05,
|
| 1401 |
-
"loss": 0.5367,
|
| 1402 |
-
"step": 19900
|
| 1403 |
-
},
|
| 1404 |
-
{
|
| 1405 |
-
"epoch": 0.5839927584897947,
|
| 1406 |
-
"grad_norm": 6.308223724365234,
|
| 1407 |
-
"learning_rate": 7.793498383566398e-05,
|
| 1408 |
-
"loss": 0.5494,
|
| 1409 |
-
"step": 20000
|
| 1410 |
-
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 0.5869127222822437,
|
| 1413 |
-
"grad_norm": 3.083712577819824,
|
| 1414 |
-
"learning_rate": 7.701357187825272e-05,
|
| 1415 |
-
"loss": 0.5712,
|
| 1416 |
-
"step": 20100
|
| 1417 |
-
},
|
| 1418 |
-
{
|
| 1419 |
-
"epoch": 0.5898326860746926,
|
| 1420 |
-
"grad_norm": 4.624127388000488,
|
| 1421 |
-
"learning_rate": 7.609421578724932e-05,
|
| 1422 |
-
"loss": 0.6324,
|
| 1423 |
-
"step": 20200
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 0.5927526498671416,
|
| 1427 |
-
"grad_norm": 2.8166234493255615,
|
| 1428 |
-
"learning_rate": 7.517699778827388e-05,
|
| 1429 |
-
"loss": 0.5447,
|
| 1430 |
-
"step": 20300
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 0.5956726136595906,
|
| 1434 |
-
"grad_norm": 5.028948783874512,
|
| 1435 |
-
"learning_rate": 7.42619999157191e-05,
|
| 1436 |
-
"loss": 0.5499,
|
| 1437 |
-
"step": 20400
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 0.5985925774520396,
|
| 1441 |
-
"grad_norm": 2.2335968017578125,
|
| 1442 |
-
"learning_rate": 7.334930400541347e-05,
|
| 1443 |
-
"loss": 0.5178,
|
| 1444 |
-
"step": 20500
|
| 1445 |
-
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 0.6015125412444886,
|
| 1448 |
-
"grad_norm": 2.491394281387329,
|
| 1449 |
-
"learning_rate": 7.243899168730202e-05,
|
| 1450 |
-
"loss": 0.5571,
|
| 1451 |
-
"step": 20600
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 0.6044325050369376,
|
| 1455 |
-
"grad_norm": 5.571224212646484,
|
| 1456 |
-
"learning_rate": 7.153114437814525e-05,
|
| 1457 |
-
"loss": 0.5365,
|
| 1458 |
-
"step": 20700
|
| 1459 |
-
},
|
| 1460 |
-
{
|
| 1461 |
-
"epoch": 0.6073524688293865,
|
| 1462 |
-
"grad_norm": 3.0990610122680664,
|
| 1463 |
-
"learning_rate": 7.062584327423772e-05,
|
| 1464 |
-
"loss": 0.505,
|
| 1465 |
-
"step": 20800
|
| 1466 |
-
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 0.6102724326218355,
|
| 1469 |
-
"grad_norm": 2.507359266281128,
|
| 1470 |
-
"learning_rate": 6.972316934414562e-05,
|
| 1471 |
-
"loss": 0.5801,
|
| 1472 |
-
"step": 20900
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.6131923964142845,
|
| 1476 |
-
"grad_norm": 2.0391249656677246,
|
| 1477 |
-
"learning_rate": 6.882320332146542e-05,
|
| 1478 |
-
"loss": 0.4986,
|
| 1479 |
-
"step": 21000
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.6161123602067334,
|
| 1483 |
-
"grad_norm": 6.61077356338501,
|
| 1484 |
-
"learning_rate": 6.793498340758233e-05,
|
| 1485 |
-
"loss": 0.5985,
|
| 1486 |
-
"step": 21100
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 0.6190323239991824,
|
| 1490 |
-
"grad_norm": 2.313990831375122,
|
| 1491 |
-
"learning_rate": 6.704064534178382e-05,
|
| 1492 |
-
"loss": 0.5378,
|
| 1493 |
-
"step": 21200
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 0.6219522877916314,
|
| 1497 |
-
"grad_norm": 3.221607208251953,
|
| 1498 |
-
"learning_rate": 6.614925510370827e-05,
|
| 1499 |
-
"loss": 0.5767,
|
| 1500 |
-
"step": 21300
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 0.6248722515840803,
|
| 1504 |
-
"grad_norm": 2.1119728088378906,
|
| 1505 |
-
"learning_rate": 6.526089241775799e-05,
|
| 1506 |
-
"loss": 0.5344,
|
| 1507 |
-
"step": 21400
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 0.6277922153765293,
|
| 1511 |
-
"grad_norm": 6.011439323425293,
|
| 1512 |
-
"learning_rate": 6.437563673755653e-05,
|
| 1513 |
-
"loss": 0.6197,
|
| 1514 |
-
"step": 21500
|
| 1515 |
-
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 0.6307121791689783,
|
| 1518 |
-
"grad_norm": 1.3864915370941162,
|
| 1519 |
-
"learning_rate": 6.349356723884206e-05,
|
| 1520 |
-
"loss": 0.5489,
|
| 1521 |
-
"step": 21600
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 0.6336321429614272,
|
| 1525 |
-
"grad_norm": 2.4754810333251953,
|
| 1526 |
-
"learning_rate": 6.261476281238625e-05,
|
| 1527 |
-
"loss": 0.4908,
|
| 1528 |
-
"step": 21700
|
| 1529 |
-
},
|
| 1530 |
-
{
|
| 1531 |
-
"epoch": 0.6365521067538763,
|
| 1532 |
-
"grad_norm": 3.065643787384033,
|
| 1533 |
-
"learning_rate": 6.173930205693829e-05,
|
| 1534 |
-
"loss": 0.5829,
|
| 1535 |
-
"step": 21800
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 0.6394720705463253,
|
| 1539 |
-
"grad_norm": 4.440421104431152,
|
| 1540 |
-
"learning_rate": 6.086726327219525e-05,
|
| 1541 |
-
"loss": 0.5175,
|
| 1542 |
-
"step": 21900
|
| 1543 |
-
},
|
| 1544 |
-
{
|
| 1545 |
-
"epoch": 0.6423920343387742,
|
| 1546 |
-
"grad_norm": 2.5433878898620605,
|
| 1547 |
-
"learning_rate": 5.999872445179904e-05,
|
| 1548 |
-
"loss": 0.604,
|
| 1549 |
-
"step": 22000
|
| 1550 |
-
},
|
| 1551 |
-
{
|
| 1552 |
-
"epoch": 0.6453119981312232,
|
| 1553 |
-
"grad_norm": 5.749415397644043,
|
| 1554 |
-
"learning_rate": 5.9133763276360855e-05,
|
| 1555 |
-
"loss": 0.5274,
|
| 1556 |
-
"step": 22100
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 0.6482319619236722,
|
| 1560 |
-
"grad_norm": 4.130051136016846,
|
| 1561 |
-
"learning_rate": 5.8272457106513435e-05,
|
| 1562 |
-
"loss": 0.524,
|
| 1563 |
-
"step": 22200
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 0.6511519257161211,
|
| 1567 |
-
"grad_norm": 4.023646831512451,
|
| 1568 |
-
"learning_rate": 5.741488297599216e-05,
|
| 1569 |
-
"loss": 0.5641,
|
| 1570 |
-
"step": 22300
|
| 1571 |
-
},
|
| 1572 |
-
{
|
| 1573 |
-
"epoch": 0.6540718895085701,
|
| 1574 |
-
"grad_norm": 3.071547746658325,
|
| 1575 |
-
"learning_rate": 5.656111758474528e-05,
|
| 1576 |
-
"loss": 0.5693,
|
| 1577 |
-
"step": 22400
|
| 1578 |
-
},
|
| 1579 |
-
{
|
| 1580 |
-
"epoch": 0.6569918533010191,
|
| 1581 |
-
"grad_norm": 2.316500663757324,
|
| 1582 |
-
"learning_rate": 5.571123729207408e-05,
|
| 1583 |
-
"loss": 0.4671,
|
| 1584 |
-
"step": 22500
|
| 1585 |
-
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 0.659911817093468,
|
| 1588 |
-
"grad_norm": 2.743230104446411,
|
| 1589 |
-
"learning_rate": 5.4865318109803064e-05,
|
| 1590 |
-
"loss": 0.5422,
|
| 1591 |
-
"step": 22600
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 0.662831780885917,
|
| 1595 |
-
"grad_norm": 4.95118522644043,
|
| 1596 |
-
"learning_rate": 5.402343569548214e-05,
|
| 1597 |
-
"loss": 0.5305,
|
| 1598 |
-
"step": 22700
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 0.665751744678366,
|
| 1602 |
-
"grad_norm": 5.621161460876465,
|
| 1603 |
-
"learning_rate": 5.318566534561956e-05,
|
| 1604 |
-
"loss": 0.5706,
|
| 1605 |
-
"step": 22800
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 0.6686717084708149,
|
| 1609 |
-
"grad_norm": 6.144950866699219,
|
| 1610 |
-
"learning_rate": 5.236039685178543e-05,
|
| 1611 |
-
"loss": 0.5763,
|
| 1612 |
-
"step": 22900
|
| 1613 |
-
},
|
| 1614 |
-
{
|
| 1615 |
-
"epoch": 0.6715916722632639,
|
| 1616 |
-
"grad_norm": 2.8691632747650146,
|
| 1617 |
-
"learning_rate": 5.1531032059287074e-05,
|
| 1618 |
-
"loss": 0.5099,
|
| 1619 |
-
"step": 23000
|
| 1620 |
-
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 0.674511636055713,
|
| 1623 |
-
"grad_norm": 1.8906828165054321,
|
| 1624 |
-
"learning_rate": 5.070600224752108e-05,
|
| 1625 |
-
"loss": 0.5564,
|
| 1626 |
-
"step": 23100
|
| 1627 |
-
},
|
| 1628 |
-
{
|
| 1629 |
-
"epoch": 0.6774315998481619,
|
| 1630 |
-
"grad_norm": 2.1284193992614746,
|
| 1631 |
-
"learning_rate": 4.9885381205728124e-05,
|
| 1632 |
-
"loss": 0.5256,
|
| 1633 |
-
"step": 23200
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 0.6803515636406109,
|
| 1637 |
-
"grad_norm": 8.401644706726074,
|
| 1638 |
-
"learning_rate": 4.9069242328835974e-05,
|
| 1639 |
-
"loss": 0.5609,
|
| 1640 |
-
"step": 23300
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 0.6832715274330599,
|
| 1644 |
-
"grad_norm": 3.2185475826263428,
|
| 1645 |
-
"learning_rate": 4.8257658610895485e-05,
|
| 1646 |
-
"loss": 0.5466,
|
| 1647 |
-
"step": 23400
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 0.6861914912255088,
|
| 1651 |
-
"grad_norm": 1.7117905616760254,
|
| 1652 |
-
"learning_rate": 4.7450702638551795e-05,
|
| 1653 |
-
"loss": 0.532,
|
| 1654 |
-
"step": 23500
|
| 1655 |
-
},
|
| 1656 |
-
{
|
| 1657 |
-
"epoch": 0.6891114550179578,
|
| 1658 |
-
"grad_norm": 5.279407501220703,
|
| 1659 |
-
"learning_rate": 4.664844658455262e-05,
|
| 1660 |
-
"loss": 0.4883,
|
| 1661 |
-
"step": 23600
|
| 1662 |
-
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 0.6920314188104068,
|
| 1665 |
-
"grad_norm": 3.187605619430542,
|
| 1666 |
-
"learning_rate": 4.5850962201292956e-05,
|
| 1667 |
-
"loss": 0.5377,
|
| 1668 |
-
"step": 23700
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 0.6949513826028557,
|
| 1672 |
-
"grad_norm": 3.679347515106201,
|
| 1673 |
-
"learning_rate": 4.5058320814397984e-05,
|
| 1674 |
-
"loss": 0.5999,
|
| 1675 |
-
"step": 23800
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 0.6978713463953047,
|
| 1679 |
-
"grad_norm": 2.252751588821411,
|
| 1680 |
-
"learning_rate": 4.427059331634348e-05,
|
| 1681 |
-
"loss": 0.479,
|
| 1682 |
-
"step": 23900
|
| 1683 |
-
},
|
| 1684 |
-
{
|
| 1685 |
-
"epoch": 0.7007913101877536,
|
| 1686 |
-
"grad_norm": 4.126784801483154,
|
| 1687 |
-
"learning_rate": 4.348785016011566e-05,
|
| 1688 |
-
"loss": 0.5708,
|
| 1689 |
-
"step": 24000
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 0.7037112739802026,
|
| 1693 |
-
"grad_norm": 2.982395648956299,
|
| 1694 |
-
"learning_rate": 4.2710161352909674e-05,
|
| 1695 |
-
"loss": 0.4497,
|
| 1696 |
-
"step": 24100
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 0.7066312377726516,
|
| 1700 |
-
"grad_norm": 2.803725242614746,
|
| 1701 |
-
"learning_rate": 4.193759644986852e-05,
|
| 1702 |
-
"loss": 0.526,
|
| 1703 |
-
"step": 24200
|
| 1704 |
-
},
|
| 1705 |
-
{
|
| 1706 |
-
"epoch": 0.7095512015651005,
|
| 1707 |
-
"grad_norm": 2.7953624725341797,
|
| 1708 |
-
"learning_rate": 4.117022454786209e-05,
|
| 1709 |
-
"loss": 0.5743,
|
| 1710 |
-
"step": 24300
|
| 1711 |
-
},
|
| 1712 |
-
{
|
| 1713 |
-
"epoch": 0.7124711653575496,
|
| 1714 |
-
"grad_norm": 3.2696502208709717,
|
| 1715 |
-
"learning_rate": 4.040811427930715e-05,
|
| 1716 |
-
"loss": 0.531,
|
| 1717 |
-
"step": 24400
|
| 1718 |
-
},
|
| 1719 |
-
{
|
| 1720 |
-
"epoch": 0.7153911291499986,
|
| 1721 |
-
"grad_norm": 5.39483118057251,
|
| 1722 |
-
"learning_rate": 3.9651333806029066e-05,
|
| 1723 |
-
"loss": 0.6086,
|
| 1724 |
-
"step": 24500
|
| 1725 |
-
},
|
| 1726 |
-
{
|
| 1727 |
-
"epoch": 0.7183110929424475,
|
| 1728 |
-
"grad_norm": 2.236680030822754,
|
| 1729 |
-
"learning_rate": 3.8899950813165595e-05,
|
| 1730 |
-
"loss": 0.538,
|
| 1731 |
-
"step": 24600
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 0.7212310567348965,
|
| 1735 |
-
"grad_norm": 5.033971786499023,
|
| 1736 |
-
"learning_rate": 3.815403250311317e-05,
|
| 1737 |
-
"loss": 0.5001,
|
| 1738 |
-
"step": 24700
|
| 1739 |
-
},
|
| 1740 |
-
{
|
| 1741 |
-
"epoch": 0.7241510205273455,
|
| 1742 |
-
"grad_norm": 3.3246283531188965,
|
| 1743 |
-
"learning_rate": 3.741364558951638e-05,
|
| 1744 |
-
"loss": 0.5278,
|
| 1745 |
-
"step": 24800
|
| 1746 |
-
},
|
| 1747 |
-
{
|
| 1748 |
-
"epoch": 0.7270709843197944,
|
| 1749 |
-
"grad_norm": 1.9075288772583008,
|
| 1750 |
-
"learning_rate": 3.667885629130124e-05,
|
| 1751 |
-
"loss": 0.5903,
|
| 1752 |
-
"step": 24900
|
| 1753 |
-
},
|
| 1754 |
-
{
|
| 1755 |
-
"epoch": 0.7299909481122434,
|
| 1756 |
-
"grad_norm": 3.299224853515625,
|
| 1757 |
-
"learning_rate": 3.5949730326752854e-05,
|
| 1758 |
-
"loss": 0.4959,
|
| 1759 |
-
"step": 25000
|
| 1760 |
-
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 0.7329109119046924,
|
| 1763 |
-
"grad_norm": 3.502321720123291,
|
| 1764 |
-
"learning_rate": 3.522633290763746e-05,
|
| 1765 |
-
"loss": 0.5437,
|
| 1766 |
-
"step": 25100
|
| 1767 |
-
},
|
| 1768 |
-
{
|
| 1769 |
-
"epoch": 0.7358308756971413,
|
| 1770 |
-
"grad_norm": 6.021385192871094,
|
| 1771 |
-
"learning_rate": 3.45087287333701e-05,
|
| 1772 |
-
"loss": 0.5548,
|
| 1773 |
-
"step": 25200
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 0.7387508394895903,
|
| 1777 |
-
"grad_norm": 2.9654223918914795,
|
| 1778 |
-
"learning_rate": 3.379698198522798e-05,
|
| 1779 |
-
"loss": 0.5051,
|
| 1780 |
-
"step": 25300
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 0.7416708032820393,
|
| 1784 |
-
"grad_norm": 5.164859294891357,
|
| 1785 |
-
"learning_rate": 3.30911563206103e-05,
|
| 1786 |
-
"loss": 0.5165,
|
| 1787 |
-
"step": 25400
|
| 1788 |
-
},
|
| 1789 |
-
{
|
| 1790 |
-
"epoch": 0.7445907670744882,
|
| 1791 |
-
"grad_norm": 3.9957275390625,
|
| 1792 |
-
"learning_rate": 3.239131486734478e-05,
|
| 1793 |
-
"loss": 0.4851,
|
| 1794 |
-
"step": 25500
|
| 1795 |
-
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 0.7475107308669372,
|
| 1798 |
-
"grad_norm": 4.626624584197998,
|
| 1799 |
-
"learning_rate": 3.1697520218041555e-05,
|
| 1800 |
-
"loss": 0.5368,
|
| 1801 |
-
"step": 25600
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 0.7504306946593863,
|
| 1805 |
-
"grad_norm": 2.424008846282959,
|
| 1806 |
-
"learning_rate": 3.100983442449506e-05,
|
| 1807 |
-
"loss": 0.4919,
|
| 1808 |
-
"step": 25700
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 0.7533506584518352,
|
| 1812 |
-
"grad_norm": 2.843548536300659,
|
| 1813 |
-
"learning_rate": 3.032831899213433e-05,
|
| 1814 |
-
"loss": 0.5479,
|
| 1815 |
-
"step": 25800
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 0.7562706222442842,
|
| 1819 |
-
"grad_norm": 2.585984945297241,
|
| 1820 |
-
"learning_rate": 2.965303487452189e-05,
|
| 1821 |
-
"loss": 0.4914,
|
| 1822 |
-
"step": 25900
|
| 1823 |
-
},
|
| 1824 |
-
{
|
| 1825 |
-
"epoch": 0.7591905860367332,
|
| 1826 |
-
"grad_norm": 2.7971699237823486,
|
| 1827 |
-
"learning_rate": 2.8984042467902207e-05,
|
| 1828 |
-
"loss": 0.5433,
|
| 1829 |
-
"step": 26000
|
| 1830 |
-
},
|
| 1831 |
-
{
|
| 1832 |
-
"epoch": 0.7621105498291821,
|
| 1833 |
-
"grad_norm": 2.5568065643310547,
|
| 1834 |
-
"learning_rate": 2.832140160580007e-05,
|
| 1835 |
-
"loss": 0.5143,
|
| 1836 |
-
"step": 26100
|
| 1837 |
-
},
|
| 1838 |
-
{
|
| 1839 |
-
"epoch": 0.7650305136216311,
|
| 1840 |
-
"grad_norm": 2.1987948417663574,
|
| 1841 |
-
"learning_rate": 2.7665171553669023e-05,
|
| 1842 |
-
"loss": 0.47,
|
| 1843 |
-
"step": 26200
|
| 1844 |
-
},
|
| 1845 |
-
{
|
| 1846 |
-
"epoch": 0.7679504774140801,
|
| 1847 |
-
"grad_norm": 2.724154233932495,
|
| 1848 |
-
"learning_rate": 2.701541100359093e-05,
|
| 1849 |
-
"loss": 0.4984,
|
| 1850 |
-
"step": 26300
|
| 1851 |
-
},
|
| 1852 |
-
{
|
| 1853 |
-
"epoch": 0.770870441206529,
|
| 1854 |
-
"grad_norm": 3.9086155891418457,
|
| 1855 |
-
"learning_rate": 2.6372178069026464e-05,
|
| 1856 |
-
"loss": 0.5805,
|
| 1857 |
-
"step": 26400
|
| 1858 |
-
},
|
| 1859 |
-
{
|
| 1860 |
-
"epoch": 0.773790404998978,
|
| 1861 |
-
"grad_norm": 3.4147980213165283,
|
| 1862 |
-
"learning_rate": 2.573553027961778e-05,
|
| 1863 |
-
"loss": 0.4967,
|
| 1864 |
-
"step": 26500
|
| 1865 |
-
},
|
| 1866 |
-
{
|
| 1867 |
-
"epoch": 0.776710368791427,
|
| 1868 |
-
"grad_norm": 6.612050533294678,
|
| 1869 |
-
"learning_rate": 2.5105524576042894e-05,
|
| 1870 |
-
"loss": 0.5695,
|
| 1871 |
-
"step": 26600
|
| 1872 |
-
},
|
| 1873 |
-
{
|
| 1874 |
-
"epoch": 0.7796303325838759,
|
| 1875 |
-
"grad_norm": 3.5038676261901855,
|
| 1876 |
-
"learning_rate": 2.4482217304923273e-05,
|
| 1877 |
-
"loss": 0.5575,
|
| 1878 |
-
"step": 26700
|
| 1879 |
-
},
|
| 1880 |
-
{
|
| 1881 |
-
"epoch": 0.7825502963763249,
|
| 1882 |
-
"grad_norm": 1.891401767730713,
|
| 1883 |
-
"learning_rate": 2.3865664213784024e-05,
|
| 1884 |
-
"loss": 0.5505,
|
| 1885 |
-
"step": 26800
|
| 1886 |
-
},
|
| 1887 |
-
{
|
| 1888 |
-
"epoch": 0.7854702601687739,
|
| 1889 |
-
"grad_norm": 2.7758102416992188,
|
| 1890 |
-
"learning_rate": 2.32559204460682e-05,
|
| 1891 |
-
"loss": 0.548,
|
| 1892 |
-
"step": 26900
|
| 1893 |
-
},
|
| 1894 |
-
{
|
| 1895 |
-
"epoch": 0.7883902239612229,
|
| 1896 |
-
"grad_norm": 8.392036437988281,
|
| 1897 |
-
"learning_rate": 2.2653040536204616e-05,
|
| 1898 |
-
"loss": 0.5254,
|
| 1899 |
-
"step": 27000
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 0.7913101877536719,
|
| 1903 |
-
"grad_norm": 1.787404179573059,
|
| 1904 |
-
"learning_rate": 2.2057078404730637e-05,
|
| 1905 |
-
"loss": 0.5049,
|
| 1906 |
-
"step": 27100
|
| 1907 |
-
},
|
| 1908 |
-
{
|
| 1909 |
-
"epoch": 0.7942301515461209,
|
| 1910 |
-
"grad_norm": 4.141063213348389,
|
| 1911 |
-
"learning_rate": 2.1468087353469336e-05,
|
| 1912 |
-
"loss": 0.5384,
|
| 1913 |
-
"step": 27200
|
| 1914 |
-
},
|
| 1915 |
-
{
|
| 1916 |
-
"epoch": 0.7971501153385698,
|
| 1917 |
-
"grad_norm": 4.328978538513184,
|
| 1918 |
-
"learning_rate": 2.0886120060762516e-05,
|
| 1919 |
-
"loss": 0.5508,
|
| 1920 |
-
"step": 27300
|
| 1921 |
-
},
|
| 1922 |
-
{
|
| 1923 |
-
"epoch": 0.8000700791310188,
|
| 1924 |
-
"grad_norm": 2.4056620597839355,
|
| 1925 |
-
"learning_rate": 2.0311228576759157e-05,
|
| 1926 |
-
"loss": 0.5034,
|
| 1927 |
-
"step": 27400
|
| 1928 |
-
},
|
| 1929 |
-
{
|
| 1930 |
-
"epoch": 0.8029900429234678,
|
| 1931 |
-
"grad_norm": 2.209656000137329,
|
| 1932 |
-
"learning_rate": 1.9743464318760087e-05,
|
| 1933 |
-
"loss": 0.585,
|
| 1934 |
-
"step": 27500
|
| 1935 |
-
},
|
| 1936 |
-
{
|
| 1937 |
-
"epoch": 0.8059100067159167,
|
| 1938 |
-
"grad_norm": 3.6636762619018555,
|
| 1939 |
-
"learning_rate": 1.9182878066619337e-05,
|
| 1940 |
-
"loss": 0.5427,
|
| 1941 |
-
"step": 27600
|
| 1942 |
-
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 0.8088299705083657,
|
| 1945 |
-
"grad_norm": 3.2900936603546143,
|
| 1946 |
-
"learning_rate": 1.862951995820257e-05,
|
| 1947 |
-
"loss": 0.546,
|
| 1948 |
-
"step": 27700
|
| 1949 |
-
},
|
| 1950 |
-
{
|
| 1951 |
-
"epoch": 0.8117499343008147,
|
| 1952 |
-
"grad_norm": 1.6237975358963013,
|
| 1953 |
-
"learning_rate": 1.8083439484902754e-05,
|
| 1954 |
-
"loss": 0.51,
|
| 1955 |
-
"step": 27800
|
| 1956 |
-
},
|
| 1957 |
-
{
|
| 1958 |
-
"epoch": 0.8146698980932636,
|
| 1959 |
-
"grad_norm": 4.96632719039917,
|
| 1960 |
-
"learning_rate": 1.7544685487213718e-05,
|
| 1961 |
-
"loss": 0.5784,
|
| 1962 |
-
"step": 27900
|
| 1963 |
-
},
|
| 1964 |
-
{
|
| 1965 |
-
"epoch": 0.8175898618857126,
|
| 1966 |
-
"grad_norm": 4.221993446350098,
|
| 1967 |
-
"learning_rate": 1.701330615036194e-05,
|
| 1968 |
-
"loss": 0.5209,
|
| 1969 |
-
"step": 28000
|
| 1970 |
-
},
|
| 1971 |
-
{
|
| 1972 |
-
"epoch": 0.8205098256781616,
|
| 1973 |
-
"grad_norm": 3.626847982406616,
|
| 1974 |
-
"learning_rate": 1.6489348999997068e-05,
|
| 1975 |
-
"loss": 0.4913,
|
| 1976 |
-
"step": 28100
|
| 1977 |
-
},
|
| 1978 |
-
{
|
| 1979 |
-
"epoch": 0.8234297894706105,
|
| 1980 |
-
"grad_norm": 4.361175537109375,
|
| 1981 |
-
"learning_rate": 1.5972860897941167e-05,
|
| 1982 |
-
"loss": 0.5247,
|
| 1983 |
-
"step": 28200
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 0.8263497532630595,
|
| 1987 |
-
"grad_norm": 4.415472507476807,
|
| 1988 |
-
"learning_rate": 1.5463888037997522e-05,
|
| 1989 |
-
"loss": 0.5272,
|
| 1990 |
-
"step": 28300
|
| 1991 |
-
},
|
| 1992 |
-
{
|
| 1993 |
-
"epoch": 0.8292697170555086,
|
| 1994 |
-
"grad_norm": 3.3726589679718018,
|
| 1995 |
-
"learning_rate": 1.4962475941819132e-05,
|
| 1996 |
-
"loss": 0.474,
|
| 1997 |
-
"step": 28400
|
| 1998 |
-
},
|
| 1999 |
-
{
|
| 2000 |
-
"epoch": 0.8321896808479575,
|
| 2001 |
-
"grad_norm": 1.9652336835861206,
|
| 2002 |
-
"learning_rate": 1.4468669454837402e-05,
|
| 2003 |
-
"loss": 0.5134,
|
| 2004 |
-
"step": 28500
|
| 2005 |
-
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 0.8351096446404065,
|
| 2008 |
-
"grad_norm": 2.4251391887664795,
|
| 2009 |
-
"learning_rate": 1.3982512742251197e-05,
|
| 2010 |
-
"loss": 0.5433,
|
| 2011 |
-
"step": 28600
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 0.8380296084328555,
|
| 2015 |
-
"grad_norm": 1.2398502826690674,
|
| 2016 |
-
"learning_rate": 1.3504049285076725e-05,
|
| 2017 |
-
"loss": 0.5545,
|
| 2018 |
-
"step": 28700
|
| 2019 |
-
},
|
| 2020 |
-
{
|
| 2021 |
-
"epoch": 0.8409495722253044,
|
| 2022 |
-
"grad_norm": 3.1720731258392334,
|
| 2023 |
-
"learning_rate": 1.3033321876258742e-05,
|
| 2024 |
-
"loss": 0.5603,
|
| 2025 |
-
"step": 28800
|
| 2026 |
-
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 0.8438695360177534,
|
| 2029 |
-
"grad_norm": 2.1568796634674072,
|
| 2030 |
-
"learning_rate": 1.2570372616843229e-05,
|
| 2031 |
-
"loss": 0.472,
|
| 2032 |
-
"step": 28900
|
| 2033 |
-
},
|
| 2034 |
-
{
|
| 2035 |
-
"epoch": 0.8467894998102024,
|
| 2036 |
-
"grad_norm": 2.8056724071502686,
|
| 2037 |
-
"learning_rate": 1.211975536822222e-05,
|
| 2038 |
-
"loss": 0.5052,
|
| 2039 |
-
"step": 29000
|
| 2040 |
-
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 0.8497094636026513,
|
| 2043 |
-
"grad_norm": 1.3424701690673828,
|
| 2044 |
-
"learning_rate": 1.1672407122591966e-05,
|
| 2045 |
-
"loss": 0.4589,
|
| 2046 |
-
"step": 29100
|
| 2047 |
-
},
|
| 2048 |
-
{
|
| 2049 |
-
"epoch": 0.8526294273951003,
|
| 2050 |
-
"grad_norm": 2.5995755195617676,
|
| 2051 |
-
"learning_rate": 1.123295874422794e-05,
|
| 2052 |
-
"loss": 0.5132,
|
| 2053 |
-
"step": 29200
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 0.8555493911875492,
|
| 2057 |
-
"grad_norm": 3.993947744369507,
|
| 2058 |
-
"learning_rate": 1.0801449536633657e-05,
|
| 2059 |
-
"loss": 0.4967,
|
| 2060 |
-
"step": 29300
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 0.8584693549799982,
|
| 2064 |
-
"grad_norm": 1.6350364685058594,
|
| 2065 |
-
"learning_rate": 1.0377918093246986e-05,
|
| 2066 |
-
"loss": 0.4799,
|
| 2067 |
-
"step": 29400
|
| 2068 |
-
},
|
| 2069 |
-
{
|
| 2070 |
-
"epoch": 0.8613893187724472,
|
| 2071 |
-
"grad_norm": 3.297436237335205,
|
| 2072 |
-
"learning_rate": 9.96240229398826e-06,
|
| 2073 |
-
"loss": 0.4838,
|
| 2074 |
-
"step": 29500
|
| 2075 |
-
},
|
| 2076 |
-
{
|
| 2077 |
-
"epoch": 0.8643092825648961,
|
| 2078 |
-
"grad_norm": 2.4876105785369873,
|
| 2079 |
-
"learning_rate": 9.554939301872545e-06,
|
| 2080 |
-
"loss": 0.4792,
|
| 2081 |
-
"step": 29600
|
| 2082 |
-
},
|
| 2083 |
-
{
|
| 2084 |
-
"epoch": 0.8672292463573452,
|
| 2085 |
-
"grad_norm": 3.8208444118499756,
|
| 2086 |
-
"learning_rate": 9.155565559685697e-06,
|
| 2087 |
-
"loss": 0.5394,
|
| 2088 |
-
"step": 29700
|
| 2089 |
-
},
|
| 2090 |
-
{
|
| 2091 |
-
"epoch": 0.8701492101497942,
|
| 2092 |
-
"grad_norm": 1.8198366165161133,
|
| 2093 |
-
"learning_rate": 8.764316786725013e-06,
|
| 2094 |
-
"loss": 0.4818,
|
| 2095 |
-
"step": 29800
|
| 2096 |
-
},
|
| 2097 |
-
{
|
| 2098 |
-
"epoch": 0.8730691739422431,
|
| 2099 |
-
"grad_norm": 4.755416393280029,
|
| 2100 |
-
"learning_rate": 8.381227975604677e-06,
|
| 2101 |
-
"loss": 0.4975,
|
| 2102 |
-
"step": 29900
|
| 2103 |
-
},
|
| 2104 |
-
{
|
| 2105 |
-
"epoch": 0.8759891377346921,
|
| 2106 |
-
"grad_norm": 2.5053181648254395,
|
| 2107 |
-
"learning_rate": 8.006333389125908e-06,
|
| 2108 |
-
"loss": 0.4599,
|
| 2109 |
-
"step": 30000
|
| 2110 |
-
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 0.8789091015271411,
|
| 2113 |
-
"grad_norm": NaN,
|
| 2114 |
-
"learning_rate": 7.643292389855882e-06,
|
| 2115 |
-
"loss": 0.5384,
|
| 2116 |
-
"step": 30100
|
| 2117 |
-
},
|
| 2118 |
-
{
|
| 2119 |
-
"epoch": 0.88182906531959,
|
| 2120 |
-
"grad_norm": 5.616009712219238,
|
| 2121 |
-
"learning_rate": 7.28480334116276e-06,
|
| 2122 |
-
"loss": 0.5446,
|
| 2123 |
-
"step": 30200
|
| 2124 |
-
},
|
| 2125 |
-
{
|
| 2126 |
-
"epoch": 0.884749029112039,
|
| 2127 |
-
"grad_norm": 2.7977347373962402,
|
| 2128 |
-
"learning_rate": 6.934606579435754e-06,
|
| 2129 |
-
"loss": 0.5179,
|
| 2130 |
-
"step": 30300
|
| 2131 |
-
},
|
| 2132 |
-
{
|
| 2133 |
-
"epoch": 0.887668992904488,
|
| 2134 |
-
"grad_norm": 2.461843490600586,
|
| 2135 |
-
"learning_rate": 6.5927334256680735e-06,
|
| 2136 |
-
"loss": 0.5078,
|
| 2137 |
-
"step": 30400
|
| 2138 |
-
},
|
| 2139 |
-
{
|
| 2140 |
-
"epoch": 0.8905889566969369,
|
| 2141 |
-
"grad_norm": 5.445867538452148,
|
| 2142 |
-
"learning_rate": 6.259214456403772e-06,
|
| 2143 |
-
"loss": 0.6466,
|
| 2144 |
-
"step": 30500
|
| 2145 |
-
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 0.8935089204893859,
|
| 2148 |
-
"grad_norm": 2.652655601501465,
|
| 2149 |
-
"learning_rate": 5.934079501003087e-06,
|
| 2150 |
-
"loss": 0.5021,
|
| 2151 |
-
"step": 30600
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 0.8964288842818349,
|
| 2155 |
-
"grad_norm": 2.0053634643554688,
|
| 2156 |
-
"learning_rate": 5.617357638974574e-06,
|
| 2157 |
-
"loss": 0.4928,
|
| 2158 |
-
"step": 30700
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 0.8993488480742838,
|
| 2162 |
-
"grad_norm": 2.91902232170105,
|
| 2163 |
-
"learning_rate": 5.309077197374246e-06,
|
| 2164 |
-
"loss": 0.5492,
|
| 2165 |
-
"step": 30800
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 0.9022688118667328,
|
| 2169 |
-
"grad_norm": 2.4080049991607666,
|
| 2170 |
-
"learning_rate": 5.009265748272041e-06,
|
| 2171 |
-
"loss": 0.4833,
|
| 2172 |
-
"step": 30900
|
| 2173 |
-
},
|
| 2174 |
-
{
|
| 2175 |
-
"epoch": 0.9051887756591819,
|
| 2176 |
-
"grad_norm": 2.9708127975463867,
|
| 2177 |
-
"learning_rate": 4.717950106285862e-06,
|
| 2178 |
-
"loss": 0.5559,
|
| 2179 |
-
"step": 31000
|
| 2180 |
-
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 0.9081087394516308,
|
| 2183 |
-
"grad_norm": 1.8519601821899414,
|
| 2184 |
-
"learning_rate": 4.4351563261832984e-06,
|
| 2185 |
-
"loss": 0.5401,
|
| 2186 |
-
"step": 31100
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 0.9110287032440798,
|
| 2190 |
-
"grad_norm": 4.341070652008057,
|
| 2191 |
-
"learning_rate": 4.160909700551341e-06,
|
| 2192 |
-
"loss": 0.4997,
|
| 2193 |
-
"step": 31200
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 0.9139486670365288,
|
| 2197 |
-
"grad_norm": 2.3663878440856934,
|
| 2198 |
-
"learning_rate": 3.895234757534249e-06,
|
| 2199 |
-
"loss": 0.4632,
|
| 2200 |
-
"step": 31300
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 0.9168686308289777,
|
| 2204 |
-
"grad_norm": 7.448829174041748,
|
| 2205 |
-
"learning_rate": 3.6381552586398193e-06,
|
| 2206 |
-
"loss": 0.4853,
|
| 2207 |
-
"step": 31400
|
| 2208 |
-
},
|
| 2209 |
-
{
|
| 2210 |
-
"epoch": 0.9197885946214267,
|
| 2211 |
-
"grad_norm": 5.03635311126709,
|
| 2212 |
-
"learning_rate": 3.392136072366148e-06,
|
| 2213 |
-
"loss": 0.57,
|
| 2214 |
-
"step": 31500
|
| 2215 |
-
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 0.9227085584138757,
|
| 2218 |
-
"grad_norm": 1.9699945449829102,
|
| 2219 |
-
"learning_rate": 3.1522291550866503e-06,
|
| 2220 |
-
"loss": 0.5089,
|
| 2221 |
-
"step": 31600
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 0.9256285222063246,
|
| 2225 |
-
"grad_norm": 2.2074334621429443,
|
| 2226 |
-
"learning_rate": 2.9209841350668955e-06,
|
| 2227 |
-
"loss": 0.5101,
|
| 2228 |
-
"step": 31700
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 0.9285484859987736,
|
| 2232 |
-
"grad_norm": 4.090612888336182,
|
| 2233 |
-
"learning_rate": 2.6984216944618502e-06,
|
| 2234 |
-
"loss": 0.5134,
|
| 2235 |
-
"step": 31800
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 0.9314684497912226,
|
| 2239 |
-
"grad_norm": 4.454416751861572,
|
| 2240 |
-
"learning_rate": 2.4845617388715823e-06,
|
| 2241 |
-
"loss": 0.4935,
|
| 2242 |
-
"step": 31900
|
| 2243 |
-
},
|
| 2244 |
-
{
|
| 2245 |
-
"epoch": 0.9343884135836715,
|
| 2246 |
-
"grad_norm": 2.1300718784332275,
|
| 2247 |
-
"learning_rate": 2.2794233955609425e-06,
|
| 2248 |
-
"loss": 0.5529,
|
| 2249 |
-
"step": 32000
|
| 2250 |
-
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 0.9373083773761205,
|
| 2253 |
-
"grad_norm": 2.274291753768921,
|
| 2254 |
-
"learning_rate": 2.0830250117488627e-06,
|
| 2255 |
-
"loss": 0.5482,
|
| 2256 |
-
"step": 32100
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 0.9402283411685695,
|
| 2260 |
-
"grad_norm": 8.231354713439941,
|
| 2261 |
-
"learning_rate": 1.8953841529673722e-06,
|
| 2262 |
-
"loss": 0.5735,
|
| 2263 |
-
"step": 32200
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 0.9431483049610185,
|
| 2267 |
-
"grad_norm": 5.990418910980225,
|
| 2268 |
-
"learning_rate": 1.7165176014906415e-06,
|
| 2269 |
-
"loss": 0.5678,
|
| 2270 |
-
"step": 32300
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 0.9460682687534675,
|
| 2274 |
-
"grad_norm": 2.1669516563415527,
|
| 2275 |
-
"learning_rate": 1.5464413548339162e-06,
|
| 2276 |
-
"loss": 0.4608,
|
| 2277 |
-
"step": 32400
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 0.9489882325459165,
|
| 2281 |
-
"grad_norm": 4.036328315734863,
|
| 2282 |
-
"learning_rate": 1.3851706243227736e-06,
|
| 2283 |
-
"loss": 0.486,
|
| 2284 |
-
"step": 32500
|
| 2285 |
-
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 0.9519081963383654,
|
| 2288 |
-
"grad_norm": 3.1790454387664795,
|
| 2289 |
-
"learning_rate": 1.2327198337326762e-06,
|
| 2290 |
-
"loss": 0.5589,
|
| 2291 |
-
"step": 32600
|
| 2292 |
-
},
|
| 2293 |
-
{
|
| 2294 |
-
"epoch": 0.9548281601308144,
|
| 2295 |
-
"grad_norm": 4.643735408782959,
|
| 2296 |
-
"learning_rate": 1.0891026179988718e-06,
|
| 2297 |
-
"loss": 0.5459,
|
| 2298 |
-
"step": 32700
|
| 2299 |
-
},
|
| 2300 |
-
{
|
| 2301 |
-
"epoch": 0.9577481239232634,
|
| 2302 |
-
"grad_norm": 2.757411479949951,
|
| 2303 |
-
"learning_rate": 9.543318219969565e-07,
|
| 2304 |
-
"loss": 0.5064,
|
| 2305 |
-
"step": 32800
|
| 2306 |
-
},
|
| 2307 |
-
{
|
| 2308 |
-
"epoch": 0.9606680877157123,
|
| 2309 |
-
"grad_norm": 9.022377014160156,
|
| 2310 |
-
"learning_rate": 8.284194993940065e-07,
|
| 2311 |
-
"loss": 0.6064,
|
| 2312 |
-
"step": 32900
|
| 2313 |
-
},
|
| 2314 |
-
{
|
| 2315 |
-
"epoch": 0.9635880515081613,
|
| 2316 |
-
"grad_norm": 4.3898210525512695,
|
| 2317 |
-
"learning_rate": 7.113769115705715e-07,
|
| 2318 |
-
"loss": 0.5154,
|
| 2319 |
-
"step": 33000
|
| 2320 |
-
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 0.9665080153006103,
|
| 2323 |
-
"grad_norm": 1.2006958723068237,
|
| 2324 |
-
"learning_rate": 6.032145266134493e-07,
|
| 2325 |
-
"loss": 0.5003,
|
| 2326 |
-
"step": 33100
|
| 2327 |
-
},
|
| 2328 |
-
{
|
| 2329 |
-
"epoch": 0.9694279790930592,
|
| 2330 |
-
"grad_norm": 3.930570602416992,
|
| 2331 |
-
"learning_rate": 5.03942018379433e-07,
|
| 2332 |
-
"loss": 0.5025,
|
| 2333 |
-
"step": 33200
|
| 2334 |
-
},
|
| 2335 |
-
{
|
| 2336 |
-
"epoch": 0.9723479428855082,
|
| 2337 |
-
"grad_norm": 1.9354479312896729,
|
| 2338 |
-
"learning_rate": 4.135682656300821e-07,
|
| 2339 |
-
"loss": 0.5043,
|
| 2340 |
-
"step": 33300
|
| 2341 |
-
},
|
| 2342 |
-
{
|
| 2343 |
-
"epoch": 0.9752679066779572,
|
| 2344 |
-
"grad_norm": 3.8160500526428223,
|
| 2345 |
-
"learning_rate": 3.3210135123770137e-07,
|
| 2346 |
-
"loss": 0.5009,
|
| 2347 |
-
"step": 33400
|
| 2348 |
-
},
|
| 2349 |
-
{
|
| 2350 |
-
"epoch": 0.9781878704704061,
|
| 2351 |
-
"grad_norm": 6.7679123878479,
|
| 2352 |
-
"learning_rate": 2.5954856146229723e-07,
|
| 2353 |
-
"loss": 0.5898,
|
| 2354 |
-
"step": 33500
|
| 2355 |
-
},
|
| 2356 |
-
{
|
| 2357 |
-
"epoch": 0.9811078342628552,
|
| 2358 |
-
"grad_norm": 2.734801769256592,
|
| 2359 |
-
"learning_rate": 1.9591638530003232e-07,
|
| 2360 |
-
"loss": 0.54,
|
| 2361 |
-
"step": 33600
|
| 2362 |
-
},
|
| 2363 |
-
{
|
| 2364 |
-
"epoch": 0.9840277980553042,
|
| 2365 |
-
"grad_norm": 3.867943048477173,
|
| 2366 |
-
"learning_rate": 1.4121051390276752e-07,
|
| 2367 |
-
"loss": 0.5046,
|
| 2368 |
-
"step": 33700
|
| 2369 |
-
},
|
| 2370 |
-
{
|
| 2371 |
-
"epoch": 0.9869477618477531,
|
| 2372 |
-
"grad_norm": 2.5671937465667725,
|
| 2373 |
-
"learning_rate": 9.543584006909134e-08,
|
| 2374 |
-
"loss": 0.609,
|
| 2375 |
-
"step": 33800
|
| 2376 |
-
},
|
| 2377 |
-
{
|
| 2378 |
-
"epoch": 0.9898677256402021,
|
| 2379 |
-
"grad_norm": 3.0211405754089355,
|
| 2380 |
-
"learning_rate": 5.859645780674772e-08,
|
| 2381 |
-
"loss": 0.5742,
|
| 2382 |
-
"step": 33900
|
| 2383 |
-
},
|
| 2384 |
-
{
|
| 2385 |
-
"epoch": 0.992787689432651,
|
| 2386 |
-
"grad_norm": 2.964229106903076,
|
| 2387 |
-
"learning_rate": 3.069566196640672e-08,
|
| 2388 |
-
"loss": 0.5081,
|
| 2389 |
-
"step": 34000
|
| 2390 |
-
},
|
| 2391 |
-
{
|
| 2392 |
-
"epoch": 0.9957076532251,
|
| 2393 |
-
"grad_norm": 2.8453125953674316,
|
| 2394 |
-
"learning_rate": 1.173594794704469e-08,
|
| 2395 |
-
"loss": 0.5484,
|
| 2396 |
-
"step": 34100
|
| 2397 |
-
},
|
| 2398 |
-
{
|
| 2399 |
-
"epoch": 0.998627617017549,
|
| 2400 |
-
"grad_norm": 2.5372684001922607,
|
| 2401 |
-
"learning_rate": 1.7190114727116957e-09,
|
| 2402 |
-
"loss": 0.4831,
|
| 2403 |
-
"step": 34200
|
| 2404 |
}
|
| 2405 |
],
|
| 2406 |
"logging_steps": 100,
|
| 2407 |
-
"max_steps":
|
| 2408 |
"num_input_tokens_seen": 0,
|
| 2409 |
-
"num_train_epochs":
|
| 2410 |
"save_steps": 0,
|
| 2411 |
"stateful_callbacks": {
|
| 2412 |
"TrainerControl": {
|
|
@@ -2420,7 +1223,7 @@
|
|
| 2420 |
"attributes": {}
|
| 2421 |
}
|
| 2422 |
},
|
| 2423 |
-
"total_flos":
|
| 2424 |
"train_batch_size": 8,
|
| 2425 |
"trial_name": null,
|
| 2426 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 17125,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.029197080291970802,
|
| 13 |
+
"grad_norm": 0.8470640182495117,
|
| 14 |
+
"learning_rate": 3.891050583657588e-05,
|
| 15 |
+
"loss": 4.8508,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.058394160583941604,
|
| 20 |
+
"grad_norm": 1.3564280271530151,
|
| 21 |
+
"learning_rate": 7.782101167315176e-05,
|
| 22 |
+
"loss": 4.4557,
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.08759124087591241,
|
| 27 |
+
"grad_norm": 1.3530114889144897,
|
| 28 |
+
"learning_rate": 0.00011673151750972763,
|
| 29 |
+
"loss": 4.1905,
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.11678832116788321,
|
| 34 |
+
"grad_norm": 1.328213095664978,
|
| 35 |
+
"learning_rate": 0.0001556420233463035,
|
| 36 |
+
"loss": 4.1219,
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.145985401459854,
|
| 41 |
+
"grad_norm": 1.3318136930465698,
|
| 42 |
+
"learning_rate": 0.0001945525291828794,
|
| 43 |
+
"loss": 3.9169,
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.17518248175182483,
|
| 48 |
+
"grad_norm": 1.7535721063613892,
|
| 49 |
+
"learning_rate": 0.00019998677287306197,
|
| 50 |
+
"loss": 3.8635,
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.20437956204379562,
|
| 55 |
+
"grad_norm": 1.6469675302505493,
|
| 56 |
+
"learning_rate": 0.00019993813296600207,
|
| 57 |
+
"loss": 3.775,
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.23357664233576642,
|
| 62 |
+
"grad_norm": 1.637732744216919,
|
| 63 |
+
"learning_rate": 0.00019985374713924855,
|
| 64 |
+
"loss": 3.6853,
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.26277372262773724,
|
| 69 |
+
"grad_norm": 1.2182079553604126,
|
| 70 |
+
"learning_rate": 0.00019973364557596465,
|
| 71 |
+
"loss": 3.6172,
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.291970802919708,
|
| 76 |
+
"grad_norm": 1.8510034084320068,
|
| 77 |
+
"learning_rate": 0.00019957787123413558,
|
| 78 |
+
"loss": 3.5044,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.32116788321167883,
|
| 83 |
+
"grad_norm": 1.4988285303115845,
|
| 84 |
+
"learning_rate": 0.00019938647983120316,
|
| 85 |
+
"loss": 3.445,
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.35036496350364965,
|
| 90 |
+
"grad_norm": 2.2497406005859375,
|
| 91 |
+
"learning_rate": 0.0001991595398241369,
|
| 92 |
+
"loss": 3.4068,
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.3795620437956204,
|
| 97 |
+
"grad_norm": 2.1076061725616455,
|
| 98 |
+
"learning_rate": 0.00019889713238494824,
|
| 99 |
+
"loss": 3.275,
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.40875912408759124,
|
| 104 |
+
"grad_norm": 2.4178290367126465,
|
| 105 |
+
"learning_rate": 0.0001985993513716568,
|
| 106 |
+
"loss": 3.1974,
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.43795620437956206,
|
| 111 |
+
"grad_norm": 1.8352422714233398,
|
| 112 |
+
"learning_rate": 0.00019826630329471928,
|
| 113 |
+
"loss": 3.1553,
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.46715328467153283,
|
| 118 |
+
"grad_norm": 1.7513710260391235,
|
| 119 |
+
"learning_rate": 0.00019789810727893284,
|
| 120 |
+
"loss": 3.0717,
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.49635036496350365,
|
| 125 |
+
"grad_norm": 1.8882313966751099,
|
| 126 |
+
"learning_rate": 0.00019749489502082632,
|
| 127 |
+
"loss": 3.0037,
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.5255474452554745,
|
| 132 |
+
"grad_norm": 2.062272310256958,
|
| 133 |
+
"learning_rate": 0.000197056810741555,
|
| 134 |
+
"loss": 2.9222,
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.5547445255474452,
|
| 139 |
+
"grad_norm": 1.6502212285995483,
|
| 140 |
+
"learning_rate": 0.00019658401113531565,
|
| 141 |
+
"loss": 2.8871,
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.583941605839416,
|
| 146 |
+
"grad_norm": 1.6748732328414917,
|
| 147 |
+
"learning_rate": 0.0001960766653132999,
|
| 148 |
+
"loss": 2.8554,
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.6131386861313869,
|
| 153 |
+
"grad_norm": 1.9615834951400757,
|
| 154 |
+
"learning_rate": 0.0001955349547432065,
|
| 155 |
+
"loss": 2.7841,
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.6423357664233577,
|
| 160 |
+
"grad_norm": 3.604374885559082,
|
| 161 |
+
"learning_rate": 0.000194959073184334,
|
| 162 |
+
"loss": 2.6938,
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.6715328467153284,
|
| 167 |
+
"grad_norm": 2.6139862537384033,
|
| 168 |
+
"learning_rate": 0.00019434922661827663,
|
| 169 |
+
"loss": 2.634,
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.7007299270072993,
|
| 174 |
+
"grad_norm": 2.5986483097076416,
|
| 175 |
+
"learning_rate": 0.00019370563317524882,
|
| 176 |
+
"loss": 2.5901,
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.7299270072992701,
|
| 181 |
+
"grad_norm": 2.7669074535369873,
|
| 182 |
+
"learning_rate": 0.00019302852305606432,
|
| 183 |
+
"loss": 2.5001,
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.7591240875912408,
|
| 188 |
+
"grad_norm": 2.4265894889831543,
|
| 189 |
+
"learning_rate": 0.00019231813844979777,
|
| 190 |
+
"loss": 2.4595,
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.7883211678832117,
|
| 195 |
+
"grad_norm": 2.589481830596924,
|
| 196 |
+
"learning_rate": 0.0001915747334471584,
|
| 197 |
+
"loss": 2.3496,
|
| 198 |
"step": 2700
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 0.8175182481751825,
|
| 202 |
+
"grad_norm": 2.969148874282837,
|
| 203 |
+
"learning_rate": 0.0001907985739496068,
|
| 204 |
+
"loss": 2.3099,
|
| 205 |
"step": 2800
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.8467153284671532,
|
| 209 |
+
"grad_norm": 2.7729787826538086,
|
| 210 |
+
"learning_rate": 0.00018998993757424713,
|
| 211 |
+
"loss": 2.2896,
|
| 212 |
"step": 2900
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.8759124087591241,
|
| 216 |
+
"grad_norm": 3.3500335216522217,
|
| 217 |
+
"learning_rate": 0.00018914911355452895,
|
| 218 |
+
"loss": 2.2433,
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.9051094890510949,
|
| 223 |
+
"grad_norm": 2.8922252655029297,
|
| 224 |
+
"learning_rate": 0.00018827640263679394,
|
| 225 |
+
"loss": 2.1564,
|
| 226 |
"step": 3100
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 0.9343065693430657,
|
| 230 |
+
"grad_norm": 2.678309679031372,
|
| 231 |
+
"learning_rate": 0.0001873721169727048,
|
| 232 |
+
"loss": 2.1654,
|
| 233 |
"step": 3200
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 0.9635036496350365,
|
| 237 |
+
"grad_norm": 3.6411967277526855,
|
| 238 |
+
"learning_rate": 0.00018643658000759493,
|
| 239 |
+
"loss": 2.1212,
|
| 240 |
"step": 3300
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 0.9927007299270073,
|
| 244 |
+
"grad_norm": 2.9299466609954834,
|
| 245 |
+
"learning_rate": 0.0001854701263647781,
|
| 246 |
+
"loss": 1.9858,
|
| 247 |
"step": 3400
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 1.0218978102189782,
|
| 251 |
+
"grad_norm": 1.6852861642837524,
|
| 252 |
+
"learning_rate": 0.0001844731017258603,
|
| 253 |
+
"loss": 1.953,
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 1.051094890510949,
|
| 258 |
+
"grad_norm": 2.1536929607391357,
|
| 259 |
+
"learning_rate": 0.00018344586270709613,
|
| 260 |
+
"loss": 1.8913,
|
| 261 |
"step": 3600
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 1.0802919708029197,
|
| 265 |
+
"grad_norm": 1.5574983358383179,
|
| 266 |
+
"learning_rate": 0.00018238877673183428,
|
| 267 |
+
"loss": 1.8142,
|
| 268 |
"step": 3700
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 1.1094890510948905,
|
| 272 |
+
"grad_norm": 1.319333791732788,
|
| 273 |
+
"learning_rate": 0.0001813022218990972,
|
| 274 |
+
"loss": 1.7464,
|
| 275 |
"step": 3800
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 1.1386861313868613,
|
| 279 |
+
"grad_norm": 1.4836809635162354,
|
| 280 |
+
"learning_rate": 0.00018018658684834256,
|
| 281 |
+
"loss": 1.7574,
|
| 282 |
"step": 3900
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 1.167883211678832,
|
| 286 |
+
"grad_norm": 1.7886255979537964,
|
| 287 |
+
"learning_rate": 0.00017904227062045437,
|
| 288 |
+
"loss": 1.7255,
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 1.197080291970803,
|
| 293 |
+
"grad_norm": 1.7846744060516357,
|
| 294 |
+
"learning_rate": 0.00017786968251501406,
|
| 295 |
+
"loss": 1.7593,
|
| 296 |
"step": 4100
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 1.2262773722627738,
|
| 300 |
+
"grad_norm": 1.5439305305480957,
|
| 301 |
+
"learning_rate": 0.00017666924194390183,
|
| 302 |
+
"loss": 1.7175,
|
| 303 |
"step": 4200
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 1.2554744525547445,
|
| 307 |
+
"grad_norm": 1.6620721817016602,
|
| 308 |
+
"learning_rate": 0.0001754413782812812,
|
| 309 |
+
"loss": 1.6345,
|
| 310 |
"step": 4300
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 1.2846715328467153,
|
| 314 |
+
"grad_norm": 1.7942287921905518,
|
| 315 |
+
"learning_rate": 0.00017418653071002047,
|
| 316 |
+
"loss": 1.6425,
|
| 317 |
"step": 4400
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 1.313868613138686,
|
| 321 |
+
"grad_norm": 1.7287062406539917,
|
| 322 |
+
"learning_rate": 0.0001729051480646052,
|
| 323 |
+
"loss": 1.5037,
|
| 324 |
"step": 4500
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 1.343065693430657,
|
| 328 |
+
"grad_norm": 1.340302586555481,
|
| 329 |
+
"learning_rate": 0.00017159768867059936,
|
| 330 |
+
"loss": 1.6512,
|
| 331 |
"step": 4600
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 1.3722627737226278,
|
| 335 |
+
"grad_norm": 1.3054938316345215,
|
| 336 |
+
"learning_rate": 0.0001702646201807107,
|
| 337 |
+
"loss": 1.5379,
|
| 338 |
"step": 4700
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 1.4014598540145986,
|
| 342 |
+
"grad_norm": 1.4920105934143066,
|
| 343 |
+
"learning_rate": 0.00016890641940752095,
|
| 344 |
+
"loss": 1.4965,
|
| 345 |
"step": 4800
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 1.4306569343065694,
|
| 349 |
+
"grad_norm": 1.6876699924468994,
|
| 350 |
+
"learning_rate": 0.00016752357215293897,
|
| 351 |
+
"loss": 1.5866,
|
| 352 |
"step": 4900
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 1.4598540145985401,
|
| 356 |
+
"grad_norm": 1.3976876735687256,
|
| 357 |
+
"learning_rate": 0.00016611657303443903,
|
| 358 |
+
"loss": 1.4894,
|
| 359 |
"step": 5000
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 1.489051094890511,
|
| 363 |
+
"grad_norm": 1.5993022918701172,
|
| 364 |
+
"learning_rate": 0.0001646859253081458,
|
| 365 |
+
"loss": 1.5393,
|
| 366 |
"step": 5100
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 1.5182481751824817,
|
| 370 |
+
"grad_norm": 1.5950144529342651,
|
| 371 |
+
"learning_rate": 0.00016323214068882935,
|
| 372 |
+
"loss": 1.4881,
|
| 373 |
"step": 5200
|
| 374 |
},
|
| 375 |
{
|
| 376 |
+
"epoch": 1.5474452554744524,
|
| 377 |
+
"grad_norm": 1.4699268341064453,
|
| 378 |
+
"learning_rate": 0.00016175573916687484,
|
| 379 |
+
"loss": 1.5266,
|
| 380 |
"step": 5300
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 1.5766423357664232,
|
| 384 |
+
"grad_norm": 1.494767427444458,
|
| 385 |
+
"learning_rate": 0.00016025724882229208,
|
| 386 |
+
"loss": 1.4496,
|
| 387 |
"step": 5400
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 1.6058394160583942,
|
| 391 |
+
"grad_norm": 1.589040756225586,
|
| 392 |
+
"learning_rate": 0.00015873720563583165,
|
| 393 |
+
"loss": 1.4994,
|
| 394 |
"step": 5500
|
| 395 |
},
|
| 396 |
{
|
| 397 |
+
"epoch": 1.635036496350365,
|
| 398 |
+
"grad_norm": 1.1471683979034424,
|
| 399 |
+
"learning_rate": 0.00015719615329727512,
|
| 400 |
+
"loss": 1.4464,
|
| 401 |
"step": 5600
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 1.6642335766423357,
|
| 405 |
+
"grad_norm": 1.5427780151367188,
|
| 406 |
+
"learning_rate": 0.00015563464301096756,
|
| 407 |
+
"loss": 1.3936,
|
| 408 |
"step": 5700
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 1.6934306569343067,
|
| 412 |
+
"grad_norm": 1.4828369617462158,
|
| 413 |
+
"learning_rate": 0.0001540532332986628,
|
| 414 |
+
"loss": 1.3653,
|
| 415 |
"step": 5800
|
| 416 |
},
|
| 417 |
{
|
| 418 |
+
"epoch": 1.7226277372262775,
|
| 419 |
+
"grad_norm": 1.3401451110839844,
|
| 420 |
+
"learning_rate": 0.0001524524897997509,
|
| 421 |
+
"loss": 1.4141,
|
| 422 |
"step": 5900
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 1.7518248175182483,
|
| 426 |
+
"grad_norm": 1.289553165435791,
|
| 427 |
+
"learning_rate": 0.00015083298506894015,
|
| 428 |
+
"loss": 1.3119,
|
| 429 |
"step": 6000
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 1.781021897810219,
|
| 433 |
+
"grad_norm": 1.2549211978912354,
|
| 434 |
+
"learning_rate": 0.00014919529837146528,
|
| 435 |
+
"loss": 1.3384,
|
| 436 |
"step": 6100
|
| 437 |
},
|
| 438 |
{
|
| 439 |
+
"epoch": 1.8102189781021898,
|
| 440 |
+
"grad_norm": 1.71195387840271,
|
| 441 |
+
"learning_rate": 0.00014754001547589564,
|
| 442 |
+
"loss": 1.3231,
|
| 443 |
"step": 6200
|
| 444 |
},
|
| 445 |
{
|
| 446 |
+
"epoch": 1.8394160583941606,
|
| 447 |
+
"grad_norm": 1.4366215467453003,
|
| 448 |
+
"learning_rate": 0.0001458677284446172,
|
| 449 |
+
"loss": 1.3196,
|
| 450 |
"step": 6300
|
| 451 |
},
|
| 452 |
{
|
| 453 |
+
"epoch": 1.8686131386861313,
|
| 454 |
+
"grad_norm": 1.5663456916809082,
|
| 455 |
+
"learning_rate": 0.00014417903542206342,
|
| 456 |
+
"loss": 1.4379,
|
| 457 |
"step": 6400
|
| 458 |
},
|
| 459 |
{
|
| 460 |
+
"epoch": 1.897810218978102,
|
| 461 |
+
"grad_norm": 2.2416155338287354,
|
| 462 |
+
"learning_rate": 0.00014247454042077068,
|
| 463 |
+
"loss": 1.3796,
|
| 464 |
"step": 6500
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 1.9270072992700729,
|
| 468 |
+
"grad_norm": 1.9517488479614258,
|
| 469 |
+
"learning_rate": 0.00014075485310533473,
|
| 470 |
+
"loss": 1.3097,
|
| 471 |
"step": 6600
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 1.9562043795620438,
|
| 475 |
+
"grad_norm": 1.7964544296264648,
|
| 476 |
+
"learning_rate": 0.00013902058857434557,
|
| 477 |
+
"loss": 1.3401,
|
| 478 |
"step": 6700
|
| 479 |
},
|
| 480 |
{
|
| 481 |
+
"epoch": 1.9854014598540146,
|
| 482 |
+
"grad_norm": 1.0145196914672852,
|
| 483 |
+
"learning_rate": 0.00013727236714037872,
|
| 484 |
+
"loss": 1.161,
|
| 485 |
"step": 6800
|
| 486 |
},
|
| 487 |
{
|
| 488 |
+
"epoch": 2.0145985401459856,
|
| 489 |
+
"grad_norm": 3.1186952590942383,
|
| 490 |
+
"learning_rate": 0.00013551081410812147,
|
| 491 |
+
"loss": 1.1588,
|
| 492 |
"step": 6900
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"epoch": 2.0437956204379564,
|
| 496 |
+
"grad_norm": 2.3698039054870605,
|
| 497 |
+
"learning_rate": 0.0001337365595507137,
|
| 498 |
+
"loss": 1.0564,
|
| 499 |
"step": 7000
|
| 500 |
},
|
| 501 |
{
|
| 502 |
+
"epoch": 2.072992700729927,
|
| 503 |
+
"grad_norm": 3.22995924949646,
|
| 504 |
+
"learning_rate": 0.0001319502380843829,
|
| 505 |
+
"loss": 1.1415,
|
| 506 |
"step": 7100
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 2.102189781021898,
|
| 510 |
+
"grad_norm": 3.0756795406341553,
|
| 511 |
+
"learning_rate": 0.00013015248864145434,
|
| 512 |
+
"loss": 1.3046,
|
| 513 |
"step": 7200
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 2.1313868613138687,
|
| 517 |
+
"grad_norm": 3.943307399749756,
|
| 518 |
+
"learning_rate": 0.00012834395424181748,
|
| 519 |
+
"loss": 1.1222,
|
| 520 |
"step": 7300
|
| 521 |
},
|
| 522 |
{
|
| 523 |
+
"epoch": 2.1605839416058394,
|
| 524 |
+
"grad_norm": 2.424041271209717,
|
| 525 |
+
"learning_rate": 0.00012652528176293042,
|
| 526 |
+
"loss": 1.1787,
|
| 527 |
"step": 7400
|
| 528 |
},
|
| 529 |
{
|
| 530 |
+
"epoch": 2.18978102189781,
|
| 531 |
+
"grad_norm": 2.5731918811798096,
|
| 532 |
+
"learning_rate": 0.0001246971217084443,
|
| 533 |
+
"loss": 1.1214,
|
| 534 |
"step": 7500
|
| 535 |
},
|
| 536 |
{
|
| 537 |
+
"epoch": 2.218978102189781,
|
| 538 |
+
"grad_norm": 1.8488351106643677,
|
| 539 |
+
"learning_rate": 0.00012286012797553075,
|
| 540 |
+
"loss": 1.0242,
|
| 541 |
"step": 7600
|
| 542 |
},
|
| 543 |
{
|
| 544 |
+
"epoch": 2.2481751824817517,
|
| 545 |
+
"grad_norm": 3.49588680267334,
|
| 546 |
+
"learning_rate": 0.0001210149576209959,
|
| 547 |
+
"loss": 1.1995,
|
| 548 |
"step": 7700
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 2.2773722627737225,
|
| 552 |
+
"grad_norm": 1.9914380311965942,
|
| 553 |
+
"learning_rate": 0.00011916227062626388,
|
| 554 |
+
"loss": 1.2218,
|
| 555 |
"step": 7800
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 2.3065693430656933,
|
| 559 |
+
"grad_norm": 2.816720485687256,
|
| 560 |
+
"learning_rate": 0.00011730272966131422,
|
| 561 |
+
"loss": 1.12,
|
| 562 |
"step": 7900
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"epoch": 2.335766423357664,
|
| 566 |
+
"grad_norm": 1.5543715953826904,
|
| 567 |
+
"learning_rate": 0.00011543699984765788,
|
| 568 |
+
"loss": 1.0535,
|
| 569 |
"step": 8000
|
| 570 |
},
|
| 571 |
{
|
| 572 |
+
"epoch": 2.3649635036496353,
|
| 573 |
+
"grad_norm": 2.867494821548462,
|
| 574 |
+
"learning_rate": 0.00011356574852043617,
|
| 575 |
+
"loss": 1.1174,
|
| 576 |
"step": 8100
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 2.394160583941606,
|
| 580 |
+
"grad_norm": 3.366215944290161,
|
| 581 |
+
"learning_rate": 0.00011168964498972818,
|
| 582 |
+
"loss": 1.3649,
|
| 583 |
"step": 8200
|
| 584 |
},
|
| 585 |
{
|
| 586 |
+
"epoch": 2.423357664233577,
|
| 587 |
+
"grad_norm": 5.301421165466309,
|
| 588 |
+
"learning_rate": 0.00010980936030115132,
|
| 589 |
+
"loss": 1.0221,
|
| 590 |
"step": 8300
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 2.4525547445255476,
|
| 594 |
+
"grad_norm": 2.44589900970459,
|
| 595 |
+
"learning_rate": 0.0001079255669958416,
|
| 596 |
+
"loss": 1.1887,
|
| 597 |
"step": 8400
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 2.4817518248175183,
|
| 601 |
+
"grad_norm": 1.8434127569198608,
|
| 602 |
+
"learning_rate": 0.00010603893886989883,
|
| 603 |
+
"loss": 1.0379,
|
| 604 |
"step": 8500
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 2.510948905109489,
|
| 608 |
+
"grad_norm": 5.0544257164001465,
|
| 609 |
+
"learning_rate": 0.00010415015073338286,
|
| 610 |
+
"loss": 1.0467,
|
| 611 |
"step": 8600
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 2.54014598540146,
|
| 615 |
+
"grad_norm": 2.266094446182251,
|
| 616 |
+
"learning_rate": 0.00010225987816894698,
|
| 617 |
+
"loss": 1.0284,
|
| 618 |
"step": 8700
|
| 619 |
},
|
| 620 |
{
|
| 621 |
+
"epoch": 2.5693430656934306,
|
| 622 |
+
"grad_norm": 4.524600028991699,
|
| 623 |
+
"learning_rate": 0.00010036879729019559,
|
| 624 |
+
"loss": 1.0921,
|
| 625 |
"step": 8800
|
| 626 |
},
|
| 627 |
{
|
| 628 |
+
"epoch": 2.5985401459854014,
|
| 629 |
+
"grad_norm": 3.9681153297424316,
|
| 630 |
+
"learning_rate": 9.847758449985124e-05,
|
| 631 |
+
"loss": 1.0259,
|
| 632 |
"step": 8900
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 2.627737226277372,
|
| 636 |
+
"grad_norm": 4.628920078277588,
|
| 637 |
+
"learning_rate": 9.658691624781866e-05,
|
| 638 |
+
"loss": 1.1767,
|
| 639 |
"step": 9000
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 2.656934306569343,
|
| 643 |
+
"grad_norm": 3.457036256790161,
|
| 644 |
+
"learning_rate": 9.469746878923188e-05,
|
| 645 |
+
"loss": 1.0662,
|
| 646 |
"step": 9100
|
| 647 |
},
|
| 648 |
{
|
| 649 |
+
"epoch": 2.686131386861314,
|
| 650 |
+
"grad_norm": 3.112386703491211,
|
| 651 |
+
"learning_rate": 9.280991794257103e-05,
|
| 652 |
+
"loss": 1.0129,
|
| 653 |
"step": 9200
|
| 654 |
},
|
| 655 |
{
|
| 656 |
+
"epoch": 2.7153284671532845,
|
| 657 |
+
"grad_norm": 2.2668728828430176,
|
| 658 |
+
"learning_rate": 9.092493884793501e-05,
|
| 659 |
+
"loss": 1.0139,
|
| 660 |
"step": 9300
|
| 661 |
},
|
| 662 |
{
|
| 663 |
+
"epoch": 2.7445255474452557,
|
| 664 |
+
"grad_norm": 3.580484390258789,
|
| 665 |
+
"learning_rate": 8.904320572555734e-05,
|
| 666 |
+
"loss": 1.0144,
|
| 667 |
"step": 9400
|
| 668 |
},
|
| 669 |
{
|
| 670 |
+
"epoch": 2.7737226277372264,
|
| 671 |
+
"grad_norm": 1.5489946603775024,
|
| 672 |
+
"learning_rate": 8.71653916346505e-05,
|
| 673 |
+
"loss": 0.9979,
|
| 674 |
"step": 9500
|
| 675 |
},
|
| 676 |
{
|
| 677 |
+
"epoch": 2.802919708029197,
|
| 678 |
+
"grad_norm": 1.7365363836288452,
|
| 679 |
+
"learning_rate": 8.529216823266606e-05,
|
| 680 |
+
"loss": 1.0874,
|
| 681 |
"step": 9600
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 2.832116788321168,
|
| 685 |
+
"grad_norm": 3.3894410133361816,
|
| 686 |
+
"learning_rate": 8.342420553505559e-05,
|
| 687 |
+
"loss": 1.1251,
|
| 688 |
"step": 9700
|
| 689 |
},
|
| 690 |
{
|
| 691 |
+
"epoch": 2.8613138686131387,
|
| 692 |
+
"grad_norm": 4.5942912101745605,
|
| 693 |
+
"learning_rate": 8.15621716756195e-05,
|
| 694 |
+
"loss": 1.1814,
|
| 695 |
"step": 9800
|
| 696 |
},
|
| 697 |
{
|
| 698 |
+
"epoch": 2.8905109489051095,
|
| 699 |
+
"grad_norm": 1.3230512142181396,
|
| 700 |
+
"learning_rate": 7.970673266752838e-05,
|
| 701 |
+
"loss": 1.1201,
|
| 702 |
"step": 9900
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"epoch": 2.9197080291970803,
|
| 706 |
+
"grad_norm": 1.1530183553695679,
|
| 707 |
+
"learning_rate": 7.785855216510337e-05,
|
| 708 |
+
"loss": 1.0758,
|
| 709 |
"step": 10000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
+
"epoch": 2.948905109489051,
|
| 713 |
+
"grad_norm": 2.9893460273742676,
|
| 714 |
+
"learning_rate": 7.601829122643957e-05,
|
| 715 |
+
"loss": 1.02,
|
| 716 |
"step": 10100
|
| 717 |
},
|
| 718 |
{
|
| 719 |
+
"epoch": 2.978102189781022,
|
| 720 |
+
"grad_norm": 2.8600292205810547,
|
| 721 |
+
"learning_rate": 7.418660807695897e-05,
|
| 722 |
+
"loss": 1.1625,
|
| 723 |
"step": 10200
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 3.0072992700729926,
|
| 727 |
+
"grad_norm": 0.6397629976272583,
|
| 728 |
+
"learning_rate": 7.236415787397548e-05,
|
| 729 |
+
"loss": 1.2091,
|
| 730 |
"step": 10300
|
| 731 |
},
|
| 732 |
{
|
| 733 |
+
"epoch": 3.0364963503649633,
|
| 734 |
+
"grad_norm": 0.5129945278167725,
|
| 735 |
+
"learning_rate": 7.055159247235844e-05,
|
| 736 |
+
"loss": 1.1668,
|
| 737 |
"step": 10400
|
| 738 |
},
|
| 739 |
{
|
| 740 |
+
"epoch": 3.065693430656934,
|
| 741 |
+
"grad_norm": 0.49799928069114685,
|
| 742 |
+
"learning_rate": 6.874956019137669e-05,
|
| 743 |
+
"loss": 0.8265,
|
| 744 |
"step": 10500
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"epoch": 3.094890510948905,
|
| 748 |
+
"grad_norm": 0.8108125925064087,
|
| 749 |
+
"learning_rate": 6.695870558280718e-05,
|
| 750 |
+
"loss": 1.0216,
|
| 751 |
"step": 10600
|
| 752 |
},
|
| 753 |
{
|
| 754 |
+
"epoch": 3.124087591240876,
|
| 755 |
+
"grad_norm": 0.45047125220298767,
|
| 756 |
+
"learning_rate": 6.51796692003918e-05,
|
| 757 |
+
"loss": 1.1076,
|
| 758 |
"step": 10700
|
| 759 |
},
|
| 760 |
{
|
| 761 |
+
"epoch": 3.153284671532847,
|
| 762 |
+
"grad_norm": 0.700933039188385,
|
| 763 |
+
"learning_rate": 6.341308737072349e-05,
|
| 764 |
+
"loss": 0.9756,
|
| 765 |
"step": 10800
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 3.1824817518248176,
|
| 769 |
+
"grad_norm": 0.5597018003463745,
|
| 770 |
+
"learning_rate": 6.165959196564481e-05,
|
| 771 |
+
"loss": 0.854,
|
| 772 |
"step": 10900
|
| 773 |
},
|
| 774 |
{
|
| 775 |
+
"epoch": 3.2116788321167884,
|
| 776 |
+
"grad_norm": 0.5288468599319458,
|
| 777 |
+
"learning_rate": 5.991981017623955e-05,
|
| 778 |
+
"loss": 0.9882,
|
| 779 |
"step": 11000
|
| 780 |
},
|
| 781 |
{
|
| 782 |
+
"epoch": 3.240875912408759,
|
| 783 |
+
"grad_norm": 0.7121827602386475,
|
| 784 |
+
"learning_rate": 5.819436428849896e-05,
|
| 785 |
+
"loss": 1.0701,
|
| 786 |
"step": 11100
|
| 787 |
},
|
| 788 |
{
|
| 789 |
+
"epoch": 3.27007299270073,
|
| 790 |
+
"grad_norm": 0.7537636756896973,
|
| 791 |
+
"learning_rate": 5.648387146074192e-05,
|
| 792 |
+
"loss": 0.9367,
|
| 793 |
"step": 11200
|
| 794 |
},
|
| 795 |
{
|
| 796 |
+
"epoch": 3.2992700729927007,
|
| 797 |
+
"grad_norm": 0.33666905760765076,
|
| 798 |
+
"learning_rate": 5.478894350286965e-05,
|
| 799 |
+
"loss": 1.0724,
|
| 800 |
"step": 11300
|
| 801 |
},
|
| 802 |
{
|
| 803 |
+
"epoch": 3.3284671532846715,
|
| 804 |
+
"grad_norm": 0.4632696509361267,
|
| 805 |
+
"learning_rate": 5.311018665753318e-05,
|
| 806 |
+
"loss": 1.0971,
|
| 807 |
"step": 11400
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 3.3576642335766422,
|
| 811 |
+
"grad_norm": 0.5160499811172485,
|
| 812 |
+
"learning_rate": 5.144820138329223e-05,
|
| 813 |
+
"loss": 1.0371,
|
| 814 |
"step": 11500
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"epoch": 3.386861313868613,
|
| 818 |
+
"grad_norm": 0.5635733008384705,
|
| 819 |
+
"learning_rate": 4.980358213984282e-05,
|
| 820 |
+
"loss": 0.978,
|
| 821 |
"step": 11600
|
| 822 |
},
|
| 823 |
{
|
| 824 |
+
"epoch": 3.4160583941605838,
|
| 825 |
+
"grad_norm": 0.6345902681350708,
|
| 826 |
+
"learning_rate": 4.8176917175390656e-05,
|
| 827 |
+
"loss": 0.9646,
|
| 828 |
"step": 11700
|
| 829 |
},
|
| 830 |
{
|
| 831 |
+
"epoch": 3.445255474452555,
|
| 832 |
+
"grad_norm": 0.6311865448951721,
|
| 833 |
+
"learning_rate": 4.656878831624636e-05,
|
| 834 |
+
"loss": 0.9127,
|
| 835 |
"step": 11800
|
| 836 |
},
|
| 837 |
{
|
| 838 |
+
"epoch": 3.4744525547445253,
|
| 839 |
+
"grad_norm": 0.4736361503601074,
|
| 840 |
+
"learning_rate": 4.497977075871738e-05,
|
| 841 |
+
"loss": 1.2787,
|
| 842 |
"step": 11900
|
| 843 |
},
|
| 844 |
{
|
| 845 |
+
"epoch": 3.5036496350364965,
|
| 846 |
+
"grad_norm": 0.3800007402896881,
|
| 847 |
+
"learning_rate": 4.341043286337153e-05,
|
| 848 |
+
"loss": 0.9448,
|
| 849 |
"step": 12000
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 3.5328467153284673,
|
| 853 |
+
"grad_norm": 0.4484248161315918,
|
| 854 |
+
"learning_rate": 4.1861335951745594e-05,
|
| 855 |
+
"loss": 0.9813,
|
| 856 |
"step": 12100
|
| 857 |
},
|
| 858 |
{
|
| 859 |
+
"epoch": 3.562043795620438,
|
| 860 |
+
"grad_norm": 0.4944378733634949,
|
| 861 |
+
"learning_rate": 4.0333034105571565e-05,
|
| 862 |
+
"loss": 1.1448,
|
| 863 |
"step": 12200
|
| 864 |
},
|
| 865 |
{
|
| 866 |
+
"epoch": 3.591240875912409,
|
| 867 |
+
"grad_norm": 0.42330047488212585,
|
| 868 |
+
"learning_rate": 3.882607396859229e-05,
|
| 869 |
+
"loss": 1.0001,
|
| 870 |
"step": 12300
|
| 871 |
},
|
| 872 |
{
|
| 873 |
+
"epoch": 3.6204379562043796,
|
| 874 |
+
"grad_norm": 0.492709219455719,
|
| 875 |
+
"learning_rate": 3.734099455103779e-05,
|
| 876 |
+
"loss": 1.1352,
|
| 877 |
"step": 12400
|
| 878 |
},
|
| 879 |
{
|
| 880 |
+
"epoch": 3.6496350364963503,
|
| 881 |
+
"grad_norm": 0.42140600085258484,
|
| 882 |
+
"learning_rate": 3.587832703683175e-05,
|
| 883 |
+
"loss": 1.0103,
|
| 884 |
"step": 12500
|
| 885 |
},
|
| 886 |
{
|
| 887 |
+
"epoch": 3.678832116788321,
|
| 888 |
+
"grad_norm": 0.393767774105072,
|
| 889 |
+
"learning_rate": 3.4438594593597596e-05,
|
| 890 |
+
"loss": 0.9709,
|
| 891 |
"step": 12600
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 3.708029197080292,
|
| 895 |
+
"grad_norm": 0.5599704384803772,
|
| 896 |
+
"learning_rate": 3.3022312185531214e-05,
|
| 897 |
+
"loss": 0.9962,
|
| 898 |
"step": 12700
|
| 899 |
},
|
| 900 |
{
|
| 901 |
+
"epoch": 3.7372262773722627,
|
| 902 |
+
"grad_norm": 0.3712032735347748,
|
| 903 |
+
"learning_rate": 3.16299863892088e-05,
|
| 904 |
+
"loss": 1.1135,
|
| 905 |
"step": 12800
|
| 906 |
},
|
| 907 |
{
|
| 908 |
+
"epoch": 3.7664233576642334,
|
| 909 |
+
"grad_norm": 0.3675720989704132,
|
| 910 |
+
"learning_rate": 3.026211521239408e-05,
|
| 911 |
+
"loss": 0.9385,
|
| 912 |
"step": 12900
|
| 913 |
},
|
| 914 |
{
|
| 915 |
+
"epoch": 3.795620437956204,
|
| 916 |
+
"grad_norm": 0.3175857961177826,
|
| 917 |
+
"learning_rate": 2.891918791591046e-05,
|
| 918 |
+
"loss": 1.0536,
|
| 919 |
"step": 13000
|
| 920 |
},
|
| 921 |
{
|
| 922 |
+
"epoch": 3.8248175182481754,
|
| 923 |
+
"grad_norm": 0.42325085401535034,
|
| 924 |
+
"learning_rate": 2.7601684838642405e-05,
|
| 925 |
+
"loss": 1.0483,
|
| 926 |
"step": 13100
|
| 927 |
},
|
| 928 |
{
|
| 929 |
+
"epoch": 3.8540145985401457,
|
| 930 |
+
"grad_norm": 0.41798830032348633,
|
| 931 |
+
"learning_rate": 2.6310077225727224e-05,
|
| 932 |
+
"loss": 1.068,
|
| 933 |
"step": 13200
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 3.883211678832117,
|
| 937 |
+
"grad_norm": 0.4633995294570923,
|
| 938 |
+
"learning_rate": 2.5044827060000085e-05,
|
| 939 |
+
"loss": 1.0305,
|
| 940 |
"step": 13300
|
| 941 |
},
|
| 942 |
{
|
| 943 |
+
"epoch": 3.9124087591240877,
|
| 944 |
+
"grad_norm": 0.38265103101730347,
|
| 945 |
+
"learning_rate": 2.380638689675164e-05,
|
| 946 |
+
"loss": 0.9594,
|
| 947 |
"step": 13400
|
| 948 |
},
|
| 949 |
{
|
| 950 |
+
"epoch": 3.9416058394160585,
|
| 951 |
+
"grad_norm": 0.3224177062511444,
|
| 952 |
+
"learning_rate": 2.2595199701858026e-05,
|
| 953 |
+
"loss": 0.9919,
|
| 954 |
"step": 13500
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"epoch": 3.9708029197080292,
|
| 958 |
+
"grad_norm": 0.44377565383911133,
|
| 959 |
+
"learning_rate": 2.1411698693340355e-05,
|
| 960 |
+
"loss": 1.0119,
|
| 961 |
"step": 13600
|
| 962 |
},
|
| 963 |
{
|
| 964 |
+
"epoch": 4.0,
|
| 965 |
+
"grad_norm": 2.024462938308716,
|
| 966 |
+
"learning_rate": 2.0256307186411295e-05,
|
| 967 |
+
"loss": 0.8573,
|
| 968 |
"step": 13700
|
| 969 |
},
|
| 970 |
{
|
| 971 |
+
"epoch": 4.029197080291971,
|
| 972 |
+
"grad_norm": 2.932596206665039,
|
| 973 |
+
"learning_rate": 1.912943844206333e-05,
|
| 974 |
+
"loss": 0.9969,
|
| 975 |
"step": 13800
|
| 976 |
},
|
| 977 |
{
|
| 978 |
+
"epoch": 4.0583941605839415,
|
| 979 |
+
"grad_norm": 2.8395912647247314,
|
| 980 |
+
"learning_rate": 1.803149551925356e-05,
|
| 981 |
+
"loss": 0.8543,
|
| 982 |
"step": 13900
|
| 983 |
},
|
| 984 |
{
|
| 985 |
+
"epoch": 4.087591240875913,
|
| 986 |
+
"grad_norm": 3.7360405921936035,
|
| 987 |
+
"learning_rate": 1.6962871130737168e-05,
|
| 988 |
+
"loss": 0.9961,
|
| 989 |
"step": 14000
|
| 990 |
},
|
| 991 |
{
|
| 992 |
+
"epoch": 4.116788321167883,
|
| 993 |
+
"grad_norm": 1.6590383052825928,
|
| 994 |
+
"learning_rate": 1.59239475026018e-05,
|
| 995 |
+
"loss": 0.839,
|
| 996 |
"step": 14100
|
| 997 |
},
|
| 998 |
{
|
| 999 |
+
"epoch": 4.145985401459854,
|
| 1000 |
+
"grad_norm": 2.4709715843200684,
|
| 1001 |
+
"learning_rate": 1.4915096237552873e-05,
|
| 1002 |
+
"loss": 1.0494,
|
| 1003 |
"step": 14200
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
+
"epoch": 4.175182481751825,
|
| 1007 |
+
"grad_norm": 3.2636585235595703,
|
| 1008 |
+
"learning_rate": 1.3936678181998374e-05,
|
| 1009 |
+
"loss": 1.0201,
|
| 1010 |
"step": 14300
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
+
"epoch": 4.204379562043796,
|
| 1014 |
+
"grad_norm": 1.440502405166626,
|
| 1015 |
+
"learning_rate": 1.298904329698123e-05,
|
| 1016 |
+
"loss": 1.0966,
|
| 1017 |
"step": 14400
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
+
"epoch": 4.233576642335766,
|
| 1021 |
+
"grad_norm": 2.1606736183166504,
|
| 1022 |
+
"learning_rate": 1.2072530533005012e-05,
|
| 1023 |
+
"loss": 1.0813,
|
| 1024 |
"step": 14500
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"epoch": 4.262773722627737,
|
| 1028 |
+
"grad_norm": 1.3628579378128052,
|
| 1029 |
+
"learning_rate": 1.1187467708798116e-05,
|
| 1030 |
+
"loss": 1.0362,
|
| 1031 |
"step": 14600
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
+
"epoch": 4.291970802919708,
|
| 1035 |
+
"grad_norm": 2.560506820678711,
|
| 1036 |
+
"learning_rate": 1.0334171394059122e-05,
|
| 1037 |
+
"loss": 1.0291,
|
| 1038 |
"step": 14700
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
+
"epoch": 4.321167883211679,
|
| 1042 |
+
"grad_norm": 1.9132288694381714,
|
| 1043 |
+
"learning_rate": 9.512946796226296e-06,
|
| 1044 |
+
"loss": 1.0084,
|
| 1045 |
"step": 14800
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
+
"epoch": 4.350364963503649,
|
| 1049 |
+
"grad_norm": 2.191732406616211,
|
| 1050 |
+
"learning_rate": 8.724087651310609e-06,
|
| 1051 |
+
"loss": 0.9277,
|
| 1052 |
"step": 14900
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
+
"epoch": 4.37956204379562,
|
| 1056 |
+
"grad_norm": 1.6962709426879883,
|
| 1057 |
+
"learning_rate": 7.967876118832229e-06,
|
| 1058 |
+
"loss": 0.9741,
|
| 1059 |
"step": 15000
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 4.408759124087592,
|
| 1063 |
+
"grad_norm": 1.9875869750976562,
|
| 1064 |
+
"learning_rate": 7.244582680897527e-06,
|
| 1065 |
+
"loss": 0.9878,
|
| 1066 |
"step": 15100
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
+
"epoch": 4.437956204379562,
|
| 1070 |
+
"grad_norm": 3.515268087387085,
|
| 1071 |
+
"learning_rate": 6.554466045452923e-06,
|
| 1072 |
+
"loss": 1.0505,
|
| 1073 |
"step": 15200
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
+
"epoch": 4.467153284671533,
|
| 1077 |
+
"grad_norm": 3.007796287536621,
|
| 1078 |
+
"learning_rate": 5.897773053750066e-06,
|
| 1079 |
+
"loss": 1.0459,
|
| 1080 |
"step": 15300
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
+
"epoch": 4.4963503649635035,
|
| 1084 |
+
"grad_norm": 1.8410799503326416,
|
| 1085 |
+
"learning_rate": 5.274738592055573e-06,
|
| 1086 |
+
"loss": 0.9201,
|
| 1087 |
"step": 15400
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
+
"epoch": 4.525547445255475,
|
| 1091 |
+
"grad_norm": 2.346604824066162,
|
| 1092 |
+
"learning_rate": 4.6855855076367804e-06,
|
| 1093 |
+
"loss": 1.0322,
|
| 1094 |
"step": 15500
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
+
"epoch": 4.554744525547445,
|
| 1098 |
+
"grad_norm": 2.057304859161377,
|
| 1099 |
+
"learning_rate": 4.130524529053626e-06,
|
| 1100 |
+
"loss": 1.001,
|
| 1101 |
"step": 15600
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 4.583941605839416,
|
| 1105 |
+
"grad_norm": 4.114916801452637,
|
| 1106 |
+
"learning_rate": 3.609754190785164e-06,
|
| 1107 |
+
"loss": 0.9103,
|
| 1108 |
"step": 15700
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
+
"epoch": 4.613138686131387,
|
| 1112 |
+
"grad_norm": 3.7978780269622803,
|
| 1113 |
+
"learning_rate": 3.1234607622176227e-06,
|
| 1114 |
+
"loss": 1.052,
|
| 1115 |
"step": 15800
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
+
"epoch": 4.642335766423358,
|
| 1119 |
+
"grad_norm": 1.945447564125061,
|
| 1120 |
+
"learning_rate": 2.6718181810195696e-06,
|
| 1121 |
+
"loss": 0.9763,
|
| 1122 |
"step": 15900
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
+
"epoch": 4.671532846715328,
|
| 1126 |
+
"grad_norm": 2.225212812423706,
|
| 1127 |
+
"learning_rate": 2.2549879909276593e-06,
|
| 1128 |
+
"loss": 1.0564,
|
| 1129 |
"step": 16000
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
+
"epoch": 4.700729927007299,
|
| 1133 |
+
"grad_norm": 1.6726568937301636,
|
| 1134 |
+
"learning_rate": 1.8731192839657407e-06,
|
| 1135 |
+
"loss": 1.042,
|
| 1136 |
"step": 16100
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"epoch": 4.7299270072992705,
|
| 1140 |
+
"grad_norm": 1.6171780824661255,
|
| 1141 |
+
"learning_rate": 1.5263486471174482e-06,
|
| 1142 |
+
"loss": 0.8253,
|
| 1143 |
"step": 16200
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
+
"epoch": 4.759124087591241,
|
| 1147 |
+
"grad_norm": 3.288196086883545,
|
| 1148 |
+
"learning_rate": 1.2148001134717369e-06,
|
| 1149 |
+
"loss": 1.0111,
|
| 1150 |
"step": 16300
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
+
"epoch": 4.788321167883212,
|
| 1154 |
+
"grad_norm": 3.768450975418091,
|
| 1155 |
+
"learning_rate": 9.385851178586924e-07,
|
| 1156 |
+
"loss": 0.8975,
|
| 1157 |
"step": 16400
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
+
"epoch": 4.817518248175182,
|
| 1161 |
+
"grad_norm": 3.030186176300049,
|
| 1162 |
+
"learning_rate": 6.978024569914032e-07,
|
| 1163 |
+
"loss": 1.0036,
|
| 1164 |
"step": 16500
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
+
"epoch": 4.846715328467154,
|
| 1168 |
+
"grad_norm": 0.843503475189209,
|
| 1169 |
+
"learning_rate": 4.92538254128383e-07,
|
| 1170 |
+
"loss": 1.011,
|
| 1171 |
"step": 16600
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
+
"epoch": 4.875912408759124,
|
| 1175 |
+
"grad_norm": 2.515411376953125,
|
| 1176 |
+
"learning_rate": 3.2286592826888953e-07,
|
| 1177 |
+
"loss": 1.2158,
|
| 1178 |
"step": 16700
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
+
"epoch": 4.905109489051095,
|
| 1182 |
+
"grad_norm": 1.0402921438217163,
|
| 1183 |
+
"learning_rate": 1.8884616789244248e-07,
|
| 1184 |
+
"loss": 0.9555,
|
| 1185 |
"step": 16800
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 4.934306569343065,
|
| 1189 |
+
"grad_norm": 0.927361249923706,
|
| 1190 |
+
"learning_rate": 9.052690925168695e-08,
|
| 1191 |
+
"loss": 1.0264,
|
| 1192 |
"step": 16900
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
+
"epoch": 4.963503649635037,
|
| 1196 |
+
"grad_norm": 1.8087859153747559,
|
| 1197 |
+
"learning_rate": 2.7943319226564346e-08,
|
| 1198 |
+
"loss": 1.0197,
|
| 1199 |
"step": 17000
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
+
"epoch": 4.992700729927007,
|
| 1203 |
+
"grad_norm": 2.9732730388641357,
|
| 1204 |
+
"learning_rate": 1.1177827458075386e-09,
|
| 1205 |
+
"loss": 1.0122,
|
| 1206 |
"step": 17100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1207 |
}
|
| 1208 |
],
|
| 1209 |
"logging_steps": 100,
|
| 1210 |
+
"max_steps": 17125,
|
| 1211 |
"num_input_tokens_seen": 0,
|
| 1212 |
+
"num_train_epochs": 5,
|
| 1213 |
"save_steps": 0,
|
| 1214 |
"stateful_callbacks": {
|
| 1215 |
"TrainerControl": {
|
|
|
|
| 1223 |
"attributes": {}
|
| 1224 |
}
|
| 1225 |
},
|
| 1226 |
+
"total_flos": 1.2724762270100429e+17,
|
| 1227 |
"train_batch_size": 8,
|
| 1228 |
"trial_name": null,
|
| 1229 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5abbcc02e7630f5ef822b6e6a8955573defc6c815565874ce5885a2c0ef54915
|
| 3 |
+
size 5560
|