Upload folder using huggingface_hub
Browse files- config.json +1 -0
- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- test.py +17 -0
- trainer_state.json +600 -1200
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"activation_dropout": 0.0,
|
| 3 |
"activation_fn": "swish",
|
| 4 |
"architectures": [
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "retnet-tinystories/checkpoint-2000",
|
| 3 |
"activation_dropout": 0.0,
|
| 4 |
"activation_fn": "swish",
|
| 5 |
"architectures": [
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1615784368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a4476df93c13330ac9bd204b05aa25cadc8aa9356e295e4a1a4c330cb7e6f15
|
| 3 |
size 1615784368
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3231643962
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60e51b3393d17c358c62db87d47057d81911db30098b095117c2224a95990545
|
| 3 |
size 3231643962
|
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 sha256:386fcc8cc1089aade9450d86fb239ea3483f455fd2d78d8378645feecfec9d69
|
| 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 sha256:04cc9f4ab547746c5f2b075d31622dd3c792964f62b6ce351d90227b6dd977bc
|
| 3 |
size 1064
|
test.py
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from transformers import AutoTokenizer
|
| 2 |
+
from retnet.modeling_retnet import RetNetForCausalLM
|
| 3 |
+
|
| 4 |
+
model = RetNetForCausalLM.from_pretrained("retnet-tinystories")
|
| 5 |
+
tokenizer = AutoTokenizer.from_pretrained('gpt2')
|
| 6 |
+
tokenizer.model_max_length = 16384
|
| 7 |
+
tokenizer.pad_token = tokenizer.eos_token
|
| 8 |
+
tokenizer.unk_token = tokenizer.eos_token
|
| 9 |
+
tokenizer.bos_token = tokenizer.eos_token
|
| 10 |
+
|
| 11 |
+
inputs = tokenizer("Hello, my dog is cute and ", return_tensors="pt")
|
| 12 |
+
|
| 13 |
+
# Generate output with max_length parameter
|
| 14 |
+
generation_output = model.generate(**inputs, max_length=50) # Adjust max_length as needed
|
| 15 |
+
output = tokenizer.decode(generation_output[0], skip_special_tokens=True)
|
| 16 |
+
|
| 17 |
+
print(output)
|
trainer_state.json
CHANGED
|
@@ -1,1821 +1,1221 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 1.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.01,
|
| 13 |
-
"learning_rate": 0.
|
| 14 |
-
"loss":
|
| 15 |
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.01,
|
| 19 |
-
"learning_rate": 0.
|
| 20 |
-
"loss":
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 0.
|
| 26 |
-
"loss":
|
| 27 |
"step": 15
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.02,
|
| 31 |
-
"learning_rate": 0.
|
| 32 |
-
"loss":
|
| 33 |
"step": 20
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.03,
|
| 37 |
-
"learning_rate": 0.
|
| 38 |
-
"loss":
|
| 39 |
"step": 25
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss":
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.04,
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss":
|
| 51 |
"step": 35
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate": 0.
|
| 56 |
-
"loss":
|
| 57 |
"step": 40
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 0.
|
| 62 |
-
"loss":
|
| 63 |
"step": 45
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate": 0.
|
| 68 |
-
"loss":
|
| 69 |
"step": 50
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate": 0.
|
| 74 |
-
"loss":
|
| 75 |
"step": 55
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate": 0.
|
| 80 |
-
"loss":
|
| 81 |
"step": 60
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss":
|
| 87 |
"step": 65
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate": 0.
|
| 92 |
-
"loss":
|
| 93 |
"step": 70
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate": 0.
|
| 98 |
-
"loss":
|
| 99 |
"step": 75
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate": 0.
|
| 104 |
-
"loss":
|
| 105 |
"step": 80
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate": 0.
|
| 110 |
-
"loss":
|
| 111 |
"step": 85
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate": 0.
|
| 116 |
-
"loss":
|
| 117 |
"step": 90
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate": 0.
|
| 122 |
-
"loss":
|
| 123 |
"step": 95
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
-
"learning_rate": 0.
|
| 128 |
-
"loss":
|
| 129 |
"step": 100
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"learning_rate": 0.
|
| 134 |
-
"loss":
|
| 135 |
"step": 105
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"learning_rate": 0.
|
| 140 |
-
"loss":
|
| 141 |
"step": 110
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
-
"learning_rate": 0.
|
| 146 |
-
"loss":
|
| 147 |
"step": 115
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
-
"learning_rate": 0.
|
| 152 |
-
"loss":
|
| 153 |
"step": 120
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
-
"learning_rate": 0.
|
| 158 |
-
"loss":
|
| 159 |
"step": 125
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
-
"learning_rate": 0.
|
| 164 |
-
"loss":
|
| 165 |
"step": 130
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 0.
|
| 169 |
-
"learning_rate": 0.
|
| 170 |
-
"loss":
|
| 171 |
"step": 135
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"learning_rate": 0.
|
| 176 |
-
"loss":
|
| 177 |
"step": 140
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"learning_rate": 0.
|
| 182 |
-
"loss":
|
| 183 |
"step": 145
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
-
"learning_rate": 0.
|
| 188 |
-
"loss":
|
| 189 |
"step": 150
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 0.
|
| 193 |
-
"learning_rate": 0.
|
| 194 |
-
"loss":
|
| 195 |
"step": 155
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 0.
|
| 199 |
-
"learning_rate": 0.
|
| 200 |
-
"loss":
|
| 201 |
"step": 160
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 0.
|
| 205 |
-
"learning_rate": 0.
|
| 206 |
-
"loss":
|
| 207 |
"step": 165
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch": 0.
|
| 211 |
-
"learning_rate": 0.
|
| 212 |
-
"loss":
|
| 213 |
"step": 170
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 0.
|
| 217 |
-
"learning_rate": 0.
|
| 218 |
-
"loss":
|
| 219 |
"step": 175
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"learning_rate": 0.
|
| 224 |
-
"loss":
|
| 225 |
"step": 180
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch": 0.
|
| 229 |
-
"learning_rate": 0.
|
| 230 |
-
"loss":
|
| 231 |
"step": 185
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch": 0.
|
| 235 |
-
"learning_rate": 0.
|
| 236 |
-
"loss":
|
| 237 |
"step": 190
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch": 0.
|
| 241 |
-
"learning_rate": 0.
|
| 242 |
-
"loss":
|
| 243 |
"step": 195
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch": 0.
|
| 247 |
-
"learning_rate": 0.
|
| 248 |
-
"loss":
|
| 249 |
"step": 200
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 0.
|
| 253 |
-
"learning_rate": 0.
|
| 254 |
-
"loss":
|
| 255 |
"step": 205
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 0.
|
| 259 |
-
"learning_rate": 0.
|
| 260 |
-
"loss":
|
| 261 |
"step": 210
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"learning_rate": 0.
|
| 266 |
-
"loss":
|
| 267 |
"step": 215
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch": 0.
|
| 271 |
-
"learning_rate": 0.
|
| 272 |
-
"loss":
|
| 273 |
"step": 220
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch": 0.
|
| 277 |
-
"learning_rate": 0.
|
| 278 |
-
"loss":
|
| 279 |
"step": 225
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch": 0.
|
| 283 |
-
"learning_rate": 0.
|
| 284 |
-
"loss":
|
| 285 |
"step": 230
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch": 0.
|
| 289 |
-
"learning_rate": 0.
|
| 290 |
-
"loss":
|
| 291 |
"step": 235
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch": 0.
|
| 295 |
-
"learning_rate": 0.
|
| 296 |
-
"loss":
|
| 297 |
"step": 240
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 0.
|
| 301 |
-
"learning_rate": 0.
|
| 302 |
-
"loss":
|
| 303 |
"step": 245
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 0.
|
| 307 |
-
"learning_rate": 0.
|
| 308 |
-
"loss":
|
| 309 |
"step": 250
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch": 0.
|
| 313 |
-
"learning_rate": 0.
|
| 314 |
-
"loss":
|
| 315 |
"step": 255
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch": 0.
|
| 319 |
-
"learning_rate": 0.
|
| 320 |
-
"loss":
|
| 321 |
"step": 260
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch": 0.
|
| 325 |
-
"learning_rate": 0.
|
| 326 |
-
"loss": 2.
|
| 327 |
"step": 265
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch": 0.
|
| 331 |
-
"learning_rate": 0.
|
| 332 |
-
"loss": 2.
|
| 333 |
"step": 270
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 0.
|
| 337 |
-
"learning_rate": 0.
|
| 338 |
-
"loss":
|
| 339 |
"step": 275
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 0.
|
| 343 |
-
"learning_rate": 0.
|
| 344 |
-
"loss": 2.
|
| 345 |
"step": 280
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 0.
|
| 349 |
-
"learning_rate": 0.
|
| 350 |
-
"loss":
|
| 351 |
"step": 285
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch": 0.
|
| 355 |
-
"learning_rate": 0.
|
| 356 |
-
"loss": 2.
|
| 357 |
"step": 290
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch": 0.
|
| 361 |
-
"learning_rate": 0.
|
| 362 |
-
"loss": 2.
|
| 363 |
"step": 295
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch": 0.
|
| 367 |
-
"learning_rate": 0.
|
| 368 |
-
"loss":
|
| 369 |
"step": 300
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 0.
|
| 373 |
-
"learning_rate": 0.
|
| 374 |
-
"loss":
|
| 375 |
"step": 305
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch": 0.
|
| 379 |
-
"learning_rate": 0.
|
| 380 |
-
"loss":
|
| 381 |
"step": 310
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 0.
|
| 385 |
-
"learning_rate": 0.
|
| 386 |
-
"loss":
|
| 387 |
"step": 315
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 0.
|
| 391 |
-
"learning_rate": 0.
|
| 392 |
-
"loss":
|
| 393 |
"step": 320
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch": 0.
|
| 397 |
-
"learning_rate": 0.
|
| 398 |
-
"loss":
|
| 399 |
"step": 325
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch": 0.
|
| 403 |
-
"learning_rate": 0.
|
| 404 |
-
"loss":
|
| 405 |
"step": 330
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch": 0.
|
| 409 |
-
"learning_rate": 0.
|
| 410 |
-
"loss":
|
| 411 |
"step": 335
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch": 0.
|
| 415 |
-
"learning_rate": 0.
|
| 416 |
-
"loss":
|
| 417 |
"step": 340
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch": 0.
|
| 421 |
-
"learning_rate": 0.
|
| 422 |
-
"loss": 2.
|
| 423 |
"step": 345
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 0.
|
| 427 |
-
"learning_rate": 0.
|
| 428 |
-
"loss":
|
| 429 |
"step": 350
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 0.
|
| 433 |
-
"learning_rate": 0.
|
| 434 |
-
"loss":
|
| 435 |
"step": 355
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch": 0.
|
| 439 |
-
"learning_rate": 0.
|
| 440 |
-
"loss":
|
| 441 |
"step": 360
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch": 0.
|
| 445 |
-
"learning_rate": 0.
|
| 446 |
-
"loss":
|
| 447 |
"step": 365
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch": 0.
|
| 451 |
-
"learning_rate": 0.
|
| 452 |
-
"loss":
|
| 453 |
"step": 370
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch": 0.
|
| 457 |
-
"learning_rate": 0.
|
| 458 |
-
"loss":
|
| 459 |
"step": 375
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch": 0.
|
| 463 |
-
"learning_rate": 0.
|
| 464 |
-
"loss": 2.
|
| 465 |
"step": 380
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch": 0.
|
| 469 |
-
"learning_rate": 0.
|
| 470 |
-
"loss":
|
| 471 |
"step": 385
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch": 0.
|
| 475 |
-
"learning_rate": 0.
|
| 476 |
-
"loss":
|
| 477 |
"step": 390
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch": 0.
|
| 481 |
-
"learning_rate": 0.
|
| 482 |
-
"loss":
|
| 483 |
"step": 395
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch": 0.
|
| 487 |
-
"learning_rate": 0.
|
| 488 |
-
"loss": 2.
|
| 489 |
"step": 400
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch": 0.
|
| 493 |
-
"learning_rate": 0.
|
| 494 |
-
"loss":
|
| 495 |
"step": 405
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch": 0.
|
| 499 |
-
"learning_rate": 0.
|
| 500 |
-
"loss":
|
| 501 |
"step": 410
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch": 0.
|
| 505 |
-
"learning_rate": 0.
|
| 506 |
-
"loss":
|
| 507 |
"step": 415
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch": 0.
|
| 511 |
-
"learning_rate": 0.
|
| 512 |
-
"loss":
|
| 513 |
"step": 420
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch": 0.
|
| 517 |
-
"learning_rate": 0.
|
| 518 |
-
"loss":
|
| 519 |
"step": 425
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch": 0.
|
| 523 |
-
"learning_rate": 0.
|
| 524 |
-
"loss":
|
| 525 |
"step": 430
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch": 0.
|
| 529 |
-
"learning_rate": 0.
|
| 530 |
-
"loss":
|
| 531 |
"step": 435
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch": 0.
|
| 535 |
-
"learning_rate": 0.
|
| 536 |
-
"loss":
|
| 537 |
"step": 440
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch": 0.
|
| 541 |
-
"learning_rate": 0.
|
| 542 |
-
"loss":
|
| 543 |
"step": 445
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch": 0.
|
| 547 |
-
"learning_rate": 0.
|
| 548 |
-
"loss":
|
| 549 |
"step": 450
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch": 0.
|
| 553 |
-
"learning_rate": 0.
|
| 554 |
-
"loss":
|
| 555 |
"step": 455
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch": 0.
|
| 559 |
-
"learning_rate": 0.
|
| 560 |
-
"loss":
|
| 561 |
"step": 460
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch": 0.
|
| 565 |
-
"learning_rate": 0.
|
| 566 |
-
"loss":
|
| 567 |
"step": 465
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch": 0.
|
| 571 |
-
"learning_rate": 0.
|
| 572 |
-
"loss":
|
| 573 |
"step": 470
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch": 0.
|
| 577 |
-
"learning_rate": 0.
|
| 578 |
-
"loss":
|
| 579 |
"step": 475
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch": 0.
|
| 583 |
-
"learning_rate": 0.
|
| 584 |
-
"loss":
|
| 585 |
"step": 480
|
| 586 |
},
|
| 587 |
{
|
| 588 |
-
"epoch": 0.
|
| 589 |
-
"learning_rate": 0.
|
| 590 |
-
"loss":
|
| 591 |
"step": 485
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch": 0.
|
| 595 |
-
"learning_rate": 0.
|
| 596 |
-
"loss":
|
| 597 |
"step": 490
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch": 0.
|
| 601 |
-
"learning_rate": 0.
|
| 602 |
-
"loss":
|
| 603 |
"step": 495
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch": 0.
|
| 607 |
-
"learning_rate": 0.
|
| 608 |
-
"loss":
|
| 609 |
"step": 500
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch": 0.
|
| 613 |
-
"learning_rate": 0.
|
| 614 |
-
"loss":
|
| 615 |
"step": 505
|
| 616 |
},
|
| 617 |
{
|
| 618 |
-
"epoch": 0.
|
| 619 |
-
"learning_rate": 0.
|
| 620 |
-
"loss":
|
| 621 |
"step": 510
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch": 0.
|
| 625 |
-
"learning_rate": 0.
|
| 626 |
-
"loss":
|
| 627 |
"step": 515
|
| 628 |
},
|
| 629 |
{
|
| 630 |
-
"epoch": 0.
|
| 631 |
-
"learning_rate": 0.
|
| 632 |
-
"loss":
|
| 633 |
"step": 520
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch": 0.
|
| 637 |
-
"learning_rate": 0.
|
| 638 |
-
"loss":
|
| 639 |
"step": 525
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch": 0.
|
| 643 |
-
"learning_rate": 0.
|
| 644 |
-
"loss":
|
| 645 |
"step": 530
|
| 646 |
},
|
| 647 |
{
|
| 648 |
-
"epoch": 0.
|
| 649 |
-
"learning_rate": 0.
|
| 650 |
-
"loss":
|
| 651 |
"step": 535
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch": 0.
|
| 655 |
-
"learning_rate": 0.
|
| 656 |
-
"loss":
|
| 657 |
"step": 540
|
| 658 |
},
|
| 659 |
{
|
| 660 |
-
"epoch": 0.
|
| 661 |
-
"learning_rate": 0.
|
| 662 |
-
"loss":
|
| 663 |
"step": 545
|
| 664 |
},
|
| 665 |
{
|
| 666 |
-
"epoch": 0.
|
| 667 |
-
"learning_rate": 0.
|
| 668 |
-
"loss": 2.
|
| 669 |
"step": 550
|
| 670 |
},
|
| 671 |
{
|
| 672 |
-
"epoch": 0.
|
| 673 |
-
"learning_rate": 0.
|
| 674 |
-
"loss":
|
| 675 |
"step": 555
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch": 0.
|
| 679 |
-
"learning_rate": 0.
|
| 680 |
-
"loss":
|
| 681 |
"step": 560
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch": 0.
|
| 685 |
-
"learning_rate": 0.
|
| 686 |
-
"loss":
|
| 687 |
"step": 565
|
| 688 |
},
|
| 689 |
{
|
| 690 |
-
"epoch": 0.
|
| 691 |
-
"learning_rate": 0.
|
| 692 |
-
"loss":
|
| 693 |
"step": 570
|
| 694 |
},
|
| 695 |
{
|
| 696 |
-
"epoch": 0.
|
| 697 |
-
"learning_rate": 0.
|
| 698 |
-
"loss":
|
| 699 |
"step": 575
|
| 700 |
},
|
| 701 |
{
|
| 702 |
-
"epoch": 0.
|
| 703 |
-
"learning_rate": 0.
|
| 704 |
-
"loss":
|
| 705 |
"step": 580
|
| 706 |
},
|
| 707 |
{
|
| 708 |
-
"epoch": 0.
|
| 709 |
-
"learning_rate": 0.
|
| 710 |
-
"loss":
|
| 711 |
"step": 585
|
| 712 |
},
|
| 713 |
{
|
| 714 |
-
"epoch": 0.
|
| 715 |
-
"learning_rate": 0.
|
| 716 |
-
"loss":
|
| 717 |
"step": 590
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch": 0.
|
| 721 |
-
"learning_rate": 0.
|
| 722 |
-
"loss":
|
| 723 |
"step": 595
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch": 0.
|
| 727 |
-
"learning_rate": 0.
|
| 728 |
-
"loss":
|
| 729 |
"step": 600
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch": 0.
|
| 733 |
-
"learning_rate": 0.
|
| 734 |
-
"loss":
|
| 735 |
"step": 605
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch": 0.
|
| 739 |
-
"learning_rate": 0.
|
| 740 |
-
"loss":
|
| 741 |
"step": 610
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch": 0.
|
| 745 |
-
"learning_rate": 0.
|
| 746 |
-
"loss":
|
| 747 |
"step": 615
|
| 748 |
},
|
| 749 |
{
|
| 750 |
-
"epoch": 0.
|
| 751 |
-
"learning_rate": 0.
|
| 752 |
-
"loss":
|
| 753 |
"step": 620
|
| 754 |
},
|
| 755 |
{
|
| 756 |
-
"epoch": 0.
|
| 757 |
-
"learning_rate": 0.
|
| 758 |
-
"loss":
|
| 759 |
"step": 625
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch": 0.
|
| 763 |
-
"learning_rate": 0.
|
| 764 |
-
"loss":
|
| 765 |
"step": 630
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch": 0.
|
| 769 |
-
"learning_rate": 0.
|
| 770 |
-
"loss":
|
| 771 |
"step": 635
|
| 772 |
},
|
| 773 |
{
|
| 774 |
-
"epoch": 0.
|
| 775 |
-
"learning_rate": 0.
|
| 776 |
-
"loss":
|
| 777 |
"step": 640
|
| 778 |
},
|
| 779 |
{
|
| 780 |
-
"epoch": 0.
|
| 781 |
-
"learning_rate": 0.
|
| 782 |
-
"loss":
|
| 783 |
"step": 645
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch": 0.
|
| 787 |
-
"learning_rate": 0.
|
| 788 |
-
"loss":
|
| 789 |
"step": 650
|
| 790 |
},
|
| 791 |
{
|
| 792 |
-
"epoch": 0.
|
| 793 |
-
"learning_rate": 0.
|
| 794 |
-
"loss":
|
| 795 |
"step": 655
|
| 796 |
},
|
| 797 |
{
|
| 798 |
-
"epoch": 0.
|
| 799 |
-
"learning_rate": 0.
|
| 800 |
-
"loss":
|
| 801 |
"step": 660
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch": 0.
|
| 805 |
-
"learning_rate": 0.
|
| 806 |
-
"loss":
|
| 807 |
"step": 665
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch": 0.
|
| 811 |
-
"learning_rate": 0.
|
| 812 |
-
"loss":
|
| 813 |
"step": 670
|
| 814 |
},
|
| 815 |
{
|
| 816 |
-
"epoch": 0.
|
| 817 |
-
"learning_rate": 0.
|
| 818 |
-
"loss":
|
| 819 |
"step": 675
|
| 820 |
},
|
| 821 |
{
|
| 822 |
-
"epoch": 0.
|
| 823 |
-
"learning_rate": 0.
|
| 824 |
-
"loss":
|
| 825 |
"step": 680
|
| 826 |
},
|
| 827 |
{
|
| 828 |
-
"epoch": 0.
|
| 829 |
-
"learning_rate": 0.
|
| 830 |
-
"loss":
|
| 831 |
"step": 685
|
| 832 |
},
|
| 833 |
{
|
| 834 |
-
"epoch": 0.
|
| 835 |
-
"learning_rate": 0.
|
| 836 |
-
"loss":
|
| 837 |
"step": 690
|
| 838 |
},
|
| 839 |
{
|
| 840 |
-
"epoch": 0.
|
| 841 |
-
"learning_rate": 0.
|
| 842 |
-
"loss":
|
| 843 |
"step": 695
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch": 0.
|
| 847 |
-
"learning_rate": 0.
|
| 848 |
-
"loss":
|
| 849 |
"step": 700
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch": 0.
|
| 853 |
-
"learning_rate": 0.
|
| 854 |
-
"loss":
|
| 855 |
"step": 705
|
| 856 |
},
|
| 857 |
{
|
| 858 |
-
"epoch": 0.
|
| 859 |
-
"learning_rate": 0.
|
| 860 |
-
"loss":
|
| 861 |
"step": 710
|
| 862 |
},
|
| 863 |
{
|
| 864 |
-
"epoch": 0.
|
| 865 |
-
"learning_rate": 0.
|
| 866 |
-
"loss":
|
| 867 |
"step": 715
|
| 868 |
},
|
| 869 |
{
|
| 870 |
-
"epoch": 0.
|
| 871 |
-
"learning_rate": 0.
|
| 872 |
-
"loss":
|
| 873 |
"step": 720
|
| 874 |
},
|
| 875 |
{
|
| 876 |
-
"epoch": 0.
|
| 877 |
-
"learning_rate": 0.
|
| 878 |
-
"loss":
|
| 879 |
"step": 725
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch": 0.
|
| 883 |
-
"learning_rate": 0.
|
| 884 |
-
"loss":
|
| 885 |
"step": 730
|
| 886 |
},
|
| 887 |
{
|
| 888 |
-
"epoch": 0.
|
| 889 |
-
"learning_rate": 0.
|
| 890 |
-
"loss":
|
| 891 |
"step": 735
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch": 0.
|
| 895 |
-
"learning_rate": 0.
|
| 896 |
-
"loss":
|
| 897 |
"step": 740
|
| 898 |
},
|
| 899 |
{
|
| 900 |
-
"epoch": 0.
|
| 901 |
-
"learning_rate": 0.
|
| 902 |
-
"loss":
|
| 903 |
"step": 745
|
| 904 |
},
|
| 905 |
{
|
| 906 |
-
"epoch": 0.
|
| 907 |
-
"learning_rate": 0.
|
| 908 |
-
"loss":
|
| 909 |
"step": 750
|
| 910 |
},
|
| 911 |
{
|
| 912 |
-
"epoch": 0.
|
| 913 |
-
"learning_rate": 0.
|
| 914 |
-
"loss":
|
| 915 |
"step": 755
|
| 916 |
},
|
| 917 |
{
|
| 918 |
-
"epoch": 0.
|
| 919 |
-
"learning_rate": 0.
|
| 920 |
-
"loss":
|
| 921 |
"step": 760
|
| 922 |
},
|
| 923 |
{
|
| 924 |
-
"epoch": 0.
|
| 925 |
-
"learning_rate": 0.
|
| 926 |
-
"loss":
|
| 927 |
"step": 765
|
| 928 |
},
|
| 929 |
{
|
| 930 |
-
"epoch": 0.
|
| 931 |
-
"learning_rate": 0.
|
| 932 |
-
"loss":
|
| 933 |
"step": 770
|
| 934 |
},
|
| 935 |
{
|
| 936 |
-
"epoch": 0.
|
| 937 |
-
"learning_rate": 0.
|
| 938 |
-
"loss":
|
| 939 |
"step": 775
|
| 940 |
},
|
| 941 |
{
|
| 942 |
-
"epoch": 0.
|
| 943 |
-
"learning_rate": 0.
|
| 944 |
-
"loss":
|
| 945 |
"step": 780
|
| 946 |
},
|
| 947 |
{
|
| 948 |
-
"epoch": 0.
|
| 949 |
-
"learning_rate": 0.
|
| 950 |
-
"loss":
|
| 951 |
"step": 785
|
| 952 |
},
|
| 953 |
{
|
| 954 |
-
"epoch": 0.
|
| 955 |
-
"learning_rate": 0.
|
| 956 |
-
"loss":
|
| 957 |
"step": 790
|
| 958 |
},
|
| 959 |
{
|
| 960 |
-
"epoch": 0.
|
| 961 |
-
"learning_rate": 0.
|
| 962 |
-
"loss":
|
| 963 |
"step": 795
|
| 964 |
},
|
| 965 |
{
|
| 966 |
-
"epoch": 0.
|
| 967 |
-
"learning_rate": 0.
|
| 968 |
-
"loss":
|
| 969 |
"step": 800
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch": 0.
|
| 973 |
-
"learning_rate": 0.
|
| 974 |
-
"loss":
|
| 975 |
"step": 805
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch": 0.
|
| 979 |
-
"learning_rate": 0.
|
| 980 |
-
"loss":
|
| 981 |
"step": 810
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch": 0.
|
| 985 |
-
"learning_rate": 0.
|
| 986 |
-
"loss":
|
| 987 |
"step": 815
|
| 988 |
},
|
| 989 |
{
|
| 990 |
-
"epoch": 0.
|
| 991 |
-
"learning_rate": 0.
|
| 992 |
-
"loss":
|
| 993 |
"step": 820
|
| 994 |
},
|
| 995 |
{
|
| 996 |
-
"epoch": 0.
|
| 997 |
-
"learning_rate": 0.
|
| 998 |
-
"loss":
|
| 999 |
"step": 825
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
-
"epoch":
|
| 1003 |
-
"learning_rate": 0.
|
| 1004 |
-
"loss":
|
| 1005 |
"step": 830
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
-
"epoch":
|
| 1009 |
-
"learning_rate": 0.
|
| 1010 |
-
"loss":
|
| 1011 |
"step": 835
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
-
"epoch":
|
| 1015 |
-
"learning_rate": 0.
|
| 1016 |
-
"loss":
|
| 1017 |
"step": 840
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
-
"epoch":
|
| 1021 |
-
"learning_rate": 0.
|
| 1022 |
-
"loss":
|
| 1023 |
"step": 845
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
-
"epoch":
|
| 1027 |
-
"learning_rate": 0.
|
| 1028 |
-
"loss":
|
| 1029 |
"step": 850
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
-
"epoch":
|
| 1033 |
-
"learning_rate": 0.
|
| 1034 |
-
"loss":
|
| 1035 |
"step": 855
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
-
"epoch":
|
| 1039 |
-
"learning_rate": 0.
|
| 1040 |
-
"loss":
|
| 1041 |
"step": 860
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
-
"epoch":
|
| 1045 |
-
"learning_rate": 0.
|
| 1046 |
-
"loss":
|
| 1047 |
"step": 865
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
-
"epoch":
|
| 1051 |
-
"learning_rate": 0.
|
| 1052 |
-
"loss":
|
| 1053 |
"step": 870
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
-
"epoch":
|
| 1057 |
-
"learning_rate": 0.
|
| 1058 |
-
"loss":
|
| 1059 |
"step": 875
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch":
|
| 1063 |
-
"learning_rate": 0.
|
| 1064 |
-
"loss":
|
| 1065 |
"step": 880
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
-
"epoch":
|
| 1069 |
-
"learning_rate": 0.
|
| 1070 |
-
"loss":
|
| 1071 |
"step": 885
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
-
"epoch":
|
| 1075 |
-
"learning_rate": 0.
|
| 1076 |
-
"loss":
|
| 1077 |
"step": 890
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
-
"epoch":
|
| 1081 |
-
"learning_rate": 0.
|
| 1082 |
-
"loss":
|
| 1083 |
"step": 895
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
-
"epoch":
|
| 1087 |
-
"learning_rate": 0.
|
| 1088 |
-
"loss":
|
| 1089 |
"step": 900
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
-
"epoch":
|
| 1093 |
-
"learning_rate": 0.
|
| 1094 |
-
"loss":
|
| 1095 |
"step": 905
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
-
"epoch":
|
| 1099 |
-
"learning_rate": 0.
|
| 1100 |
-
"loss":
|
| 1101 |
"step": 910
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch":
|
| 1105 |
-
"learning_rate": 0.
|
| 1106 |
-
"loss":
|
| 1107 |
"step": 915
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
-
"epoch":
|
| 1111 |
-
"learning_rate": 0.
|
| 1112 |
-
"loss":
|
| 1113 |
"step": 920
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
-
"epoch":
|
| 1117 |
-
"learning_rate": 0.
|
| 1118 |
-
"loss":
|
| 1119 |
"step": 925
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
-
"epoch":
|
| 1123 |
-
"learning_rate": 0.
|
| 1124 |
-
"loss":
|
| 1125 |
"step": 930
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
-
"epoch":
|
| 1129 |
-
"learning_rate": 0.
|
| 1130 |
-
"loss":
|
| 1131 |
"step": 935
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
-
"epoch":
|
| 1135 |
-
"learning_rate": 0.
|
| 1136 |
-
"loss":
|
| 1137 |
"step": 940
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
-
"epoch":
|
| 1141 |
-
"learning_rate": 0.
|
| 1142 |
-
"loss":
|
| 1143 |
"step": 945
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch":
|
| 1147 |
-
"learning_rate": 0.
|
| 1148 |
-
"loss":
|
| 1149 |
"step": 950
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
-
"epoch":
|
| 1153 |
-
"learning_rate": 0.
|
| 1154 |
-
"loss":
|
| 1155 |
"step": 955
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
-
"epoch":
|
| 1159 |
-
"learning_rate": 0.
|
| 1160 |
-
"loss":
|
| 1161 |
"step": 960
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
-
"epoch":
|
| 1165 |
-
"learning_rate": 0.
|
| 1166 |
-
"loss":
|
| 1167 |
"step": 965
|
| 1168 |
},
|
| 1169 |
{
|
| 1170 |
-
"epoch":
|
| 1171 |
-
"learning_rate": 0.
|
| 1172 |
-
"loss":
|
| 1173 |
"step": 970
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
-
"epoch":
|
| 1177 |
-
"learning_rate": 0.
|
| 1178 |
-
"loss":
|
| 1179 |
"step": 975
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
-
"epoch":
|
| 1183 |
-
"learning_rate": 0.
|
| 1184 |
-
"loss":
|
| 1185 |
"step": 980
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch":
|
| 1189 |
-
"learning_rate": 0.
|
| 1190 |
-
"loss":
|
| 1191 |
"step": 985
|
| 1192 |
},
|
| 1193 |
{
|
| 1194 |
-
"epoch":
|
| 1195 |
-
"learning_rate": 0.
|
| 1196 |
-
"loss":
|
| 1197 |
"step": 990
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
-
"epoch":
|
| 1201 |
-
"learning_rate": 0.
|
| 1202 |
-
"loss":
|
| 1203 |
"step": 995
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
-
"epoch": 1.
|
| 1207 |
-
"learning_rate": 0.
|
| 1208 |
-
"loss": 1.
|
| 1209 |
"step": 1000
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 1.21,
|
| 1213 |
-
"learning_rate": 0.00037935174069627856,
|
| 1214 |
-
"loss": 2.0543,
|
| 1215 |
-
"step": 1005
|
| 1216 |
-
},
|
| 1217 |
-
{
|
| 1218 |
-
"epoch": 1.21,
|
| 1219 |
-
"learning_rate": 0.0003787515006002401,
|
| 1220 |
-
"loss": 2.1018,
|
| 1221 |
-
"step": 1010
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 1.22,
|
| 1225 |
-
"learning_rate": 0.00037815126050420167,
|
| 1226 |
-
"loss": 2.1016,
|
| 1227 |
-
"step": 1015
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 1.22,
|
| 1231 |
-
"learning_rate": 0.00037755102040816323,
|
| 1232 |
-
"loss": 2.0779,
|
| 1233 |
-
"step": 1020
|
| 1234 |
-
},
|
| 1235 |
-
{
|
| 1236 |
-
"epoch": 1.23,
|
| 1237 |
-
"learning_rate": 0.0003769507803121249,
|
| 1238 |
-
"loss": 2.1172,
|
| 1239 |
-
"step": 1025
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 1.24,
|
| 1243 |
-
"learning_rate": 0.00037635054021608645,
|
| 1244 |
-
"loss": 2.0815,
|
| 1245 |
-
"step": 1030
|
| 1246 |
-
},
|
| 1247 |
-
{
|
| 1248 |
-
"epoch": 1.24,
|
| 1249 |
-
"learning_rate": 0.000375750300120048,
|
| 1250 |
-
"loss": 2.0598,
|
| 1251 |
-
"step": 1035
|
| 1252 |
-
},
|
| 1253 |
-
{
|
| 1254 |
-
"epoch": 1.25,
|
| 1255 |
-
"learning_rate": 0.00037515006002400956,
|
| 1256 |
-
"loss": 2.0456,
|
| 1257 |
-
"step": 1040
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 1.25,
|
| 1261 |
-
"learning_rate": 0.00037454981992797123,
|
| 1262 |
-
"loss": 2.1081,
|
| 1263 |
-
"step": 1045
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 1.26,
|
| 1267 |
-
"learning_rate": 0.0003739495798319328,
|
| 1268 |
-
"loss": 2.0235,
|
| 1269 |
-
"step": 1050
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 1.27,
|
| 1273 |
-
"learning_rate": 0.00037334933973589434,
|
| 1274 |
-
"loss": 2.0866,
|
| 1275 |
-
"step": 1055
|
| 1276 |
-
},
|
| 1277 |
-
{
|
| 1278 |
-
"epoch": 1.27,
|
| 1279 |
-
"learning_rate": 0.0003727490996398559,
|
| 1280 |
-
"loss": 2.0599,
|
| 1281 |
-
"step": 1060
|
| 1282 |
-
},
|
| 1283 |
-
{
|
| 1284 |
-
"epoch": 1.28,
|
| 1285 |
-
"learning_rate": 0.00037214885954381757,
|
| 1286 |
-
"loss": 2.0281,
|
| 1287 |
-
"step": 1065
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 1.28,
|
| 1291 |
-
"learning_rate": 0.0003715486194477791,
|
| 1292 |
-
"loss": 2.0481,
|
| 1293 |
-
"step": 1070
|
| 1294 |
-
},
|
| 1295 |
-
{
|
| 1296 |
-
"epoch": 1.29,
|
| 1297 |
-
"learning_rate": 0.0003709483793517407,
|
| 1298 |
-
"loss": 2.0223,
|
| 1299 |
-
"step": 1075
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 1.3,
|
| 1303 |
-
"learning_rate": 0.0003703481392557023,
|
| 1304 |
-
"loss": 2.0303,
|
| 1305 |
-
"step": 1080
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 1.3,
|
| 1309 |
-
"learning_rate": 0.0003697478991596639,
|
| 1310 |
-
"loss": 2.13,
|
| 1311 |
-
"step": 1085
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 1.31,
|
| 1315 |
-
"learning_rate": 0.00036914765906362546,
|
| 1316 |
-
"loss": 2.1372,
|
| 1317 |
-
"step": 1090
|
| 1318 |
-
},
|
| 1319 |
-
{
|
| 1320 |
-
"epoch": 1.31,
|
| 1321 |
-
"learning_rate": 0.000368547418967587,
|
| 1322 |
-
"loss": 2.0192,
|
| 1323 |
-
"step": 1095
|
| 1324 |
-
},
|
| 1325 |
-
{
|
| 1326 |
-
"epoch": 1.32,
|
| 1327 |
-
"learning_rate": 0.00036794717887154863,
|
| 1328 |
-
"loss": 2.0754,
|
| 1329 |
-
"step": 1100
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 1.33,
|
| 1333 |
-
"learning_rate": 0.00036734693877551024,
|
| 1334 |
-
"loss": 2.123,
|
| 1335 |
-
"step": 1105
|
| 1336 |
-
},
|
| 1337 |
-
{
|
| 1338 |
-
"epoch": 1.33,
|
| 1339 |
-
"learning_rate": 0.0003667466986794718,
|
| 1340 |
-
"loss": 2.0884,
|
| 1341 |
-
"step": 1110
|
| 1342 |
-
},
|
| 1343 |
-
{
|
| 1344 |
-
"epoch": 1.34,
|
| 1345 |
-
"learning_rate": 0.00036614645858343335,
|
| 1346 |
-
"loss": 2.0835,
|
| 1347 |
-
"step": 1115
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 1.34,
|
| 1351 |
-
"learning_rate": 0.00036554621848739496,
|
| 1352 |
-
"loss": 2.0276,
|
| 1353 |
-
"step": 1120
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 1.35,
|
| 1357 |
-
"learning_rate": 0.0003649459783913566,
|
| 1358 |
-
"loss": 2.0106,
|
| 1359 |
-
"step": 1125
|
| 1360 |
-
},
|
| 1361 |
-
{
|
| 1362 |
-
"epoch": 1.36,
|
| 1363 |
-
"learning_rate": 0.00036434573829531813,
|
| 1364 |
-
"loss": 1.9971,
|
| 1365 |
-
"step": 1130
|
| 1366 |
-
},
|
| 1367 |
-
{
|
| 1368 |
-
"epoch": 1.36,
|
| 1369 |
-
"learning_rate": 0.0003637454981992797,
|
| 1370 |
-
"loss": 1.9734,
|
| 1371 |
-
"step": 1135
|
| 1372 |
-
},
|
| 1373 |
-
{
|
| 1374 |
-
"epoch": 1.37,
|
| 1375 |
-
"learning_rate": 0.0003631452581032413,
|
| 1376 |
-
"loss": 2.0989,
|
| 1377 |
-
"step": 1140
|
| 1378 |
-
},
|
| 1379 |
-
{
|
| 1380 |
-
"epoch": 1.37,
|
| 1381 |
-
"learning_rate": 0.0003625450180072029,
|
| 1382 |
-
"loss": 2.0539,
|
| 1383 |
-
"step": 1145
|
| 1384 |
-
},
|
| 1385 |
-
{
|
| 1386 |
-
"epoch": 1.38,
|
| 1387 |
-
"learning_rate": 0.00036194477791116447,
|
| 1388 |
-
"loss": 1.9613,
|
| 1389 |
-
"step": 1150
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 1.39,
|
| 1393 |
-
"learning_rate": 0.000361344537815126,
|
| 1394 |
-
"loss": 2.0324,
|
| 1395 |
-
"step": 1155
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 1.39,
|
| 1399 |
-
"learning_rate": 0.00036074429771908764,
|
| 1400 |
-
"loss": 2.0144,
|
| 1401 |
-
"step": 1160
|
| 1402 |
-
},
|
| 1403 |
-
{
|
| 1404 |
-
"epoch": 1.4,
|
| 1405 |
-
"learning_rate": 0.00036014405762304925,
|
| 1406 |
-
"loss": 2.0124,
|
| 1407 |
-
"step": 1165
|
| 1408 |
-
},
|
| 1409 |
-
{
|
| 1410 |
-
"epoch": 1.4,
|
| 1411 |
-
"learning_rate": 0.0003595438175270108,
|
| 1412 |
-
"loss": 2.0638,
|
| 1413 |
-
"step": 1170
|
| 1414 |
-
},
|
| 1415 |
-
{
|
| 1416 |
-
"epoch": 1.41,
|
| 1417 |
-
"learning_rate": 0.00035894357743097236,
|
| 1418 |
-
"loss": 2.0049,
|
| 1419 |
-
"step": 1175
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 1.42,
|
| 1423 |
-
"learning_rate": 0.000358343337334934,
|
| 1424 |
-
"loss": 2.0731,
|
| 1425 |
-
"step": 1180
|
| 1426 |
-
},
|
| 1427 |
-
{
|
| 1428 |
-
"epoch": 1.42,
|
| 1429 |
-
"learning_rate": 0.0003577430972388956,
|
| 1430 |
-
"loss": 2.0804,
|
| 1431 |
-
"step": 1185
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 1.43,
|
| 1435 |
-
"learning_rate": 0.00035714285714285714,
|
| 1436 |
-
"loss": 2.0652,
|
| 1437 |
-
"step": 1190
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 1.43,
|
| 1441 |
-
"learning_rate": 0.00035654261704681875,
|
| 1442 |
-
"loss": 1.9836,
|
| 1443 |
-
"step": 1195
|
| 1444 |
-
},
|
| 1445 |
-
{
|
| 1446 |
-
"epoch": 1.44,
|
| 1447 |
-
"learning_rate": 0.0003559423769507803,
|
| 1448 |
-
"loss": 2.0288,
|
| 1449 |
-
"step": 1200
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 1.45,
|
| 1453 |
-
"learning_rate": 0.0003553421368547419,
|
| 1454 |
-
"loss": 1.9811,
|
| 1455 |
-
"step": 1205
|
| 1456 |
-
},
|
| 1457 |
-
{
|
| 1458 |
-
"epoch": 1.45,
|
| 1459 |
-
"learning_rate": 0.0003547418967587035,
|
| 1460 |
-
"loss": 1.9923,
|
| 1461 |
-
"step": 1210
|
| 1462 |
-
},
|
| 1463 |
-
{
|
| 1464 |
-
"epoch": 1.46,
|
| 1465 |
-
"learning_rate": 0.0003541416566626651,
|
| 1466 |
-
"loss": 2.0066,
|
| 1467 |
-
"step": 1215
|
| 1468 |
-
},
|
| 1469 |
-
{
|
| 1470 |
-
"epoch": 1.46,
|
| 1471 |
-
"learning_rate": 0.00035354141656662665,
|
| 1472 |
-
"loss": 1.992,
|
| 1473 |
-
"step": 1220
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 1.47,
|
| 1477 |
-
"learning_rate": 0.00035294117647058826,
|
| 1478 |
-
"loss": 2.0675,
|
| 1479 |
-
"step": 1225
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 1.48,
|
| 1483 |
-
"learning_rate": 0.0003523409363745498,
|
| 1484 |
-
"loss": 2.0066,
|
| 1485 |
-
"step": 1230
|
| 1486 |
-
},
|
| 1487 |
-
{
|
| 1488 |
-
"epoch": 1.48,
|
| 1489 |
-
"learning_rate": 0.0003517406962785114,
|
| 1490 |
-
"loss": 2.0111,
|
| 1491 |
-
"step": 1235
|
| 1492 |
-
},
|
| 1493 |
-
{
|
| 1494 |
-
"epoch": 1.49,
|
| 1495 |
-
"learning_rate": 0.000351140456182473,
|
| 1496 |
-
"loss": 1.9843,
|
| 1497 |
-
"step": 1240
|
| 1498 |
-
},
|
| 1499 |
-
{
|
| 1500 |
-
"epoch": 1.49,
|
| 1501 |
-
"learning_rate": 0.0003505402160864346,
|
| 1502 |
-
"loss": 2.0322,
|
| 1503 |
-
"step": 1245
|
| 1504 |
-
},
|
| 1505 |
-
{
|
| 1506 |
-
"epoch": 1.5,
|
| 1507 |
-
"learning_rate": 0.00034993997599039615,
|
| 1508 |
-
"loss": 2.0266,
|
| 1509 |
-
"step": 1250
|
| 1510 |
-
},
|
| 1511 |
-
{
|
| 1512 |
-
"epoch": 1.51,
|
| 1513 |
-
"learning_rate": 0.00034933973589435776,
|
| 1514 |
-
"loss": 2.0346,
|
| 1515 |
-
"step": 1255
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 1.51,
|
| 1519 |
-
"learning_rate": 0.0003487394957983193,
|
| 1520 |
-
"loss": 1.9676,
|
| 1521 |
-
"step": 1260
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 1.52,
|
| 1525 |
-
"learning_rate": 0.00034813925570228093,
|
| 1526 |
-
"loss": 2.0872,
|
| 1527 |
-
"step": 1265
|
| 1528 |
-
},
|
| 1529 |
-
{
|
| 1530 |
-
"epoch": 1.52,
|
| 1531 |
-
"learning_rate": 0.0003475390156062425,
|
| 1532 |
-
"loss": 1.9265,
|
| 1533 |
-
"step": 1270
|
| 1534 |
-
},
|
| 1535 |
-
{
|
| 1536 |
-
"epoch": 1.53,
|
| 1537 |
-
"learning_rate": 0.0003469387755102041,
|
| 1538 |
-
"loss": 2.0074,
|
| 1539 |
-
"step": 1275
|
| 1540 |
-
},
|
| 1541 |
-
{
|
| 1542 |
-
"epoch": 1.54,
|
| 1543 |
-
"learning_rate": 0.00034633853541416566,
|
| 1544 |
-
"loss": 2.0108,
|
| 1545 |
-
"step": 1280
|
| 1546 |
-
},
|
| 1547 |
-
{
|
| 1548 |
-
"epoch": 1.54,
|
| 1549 |
-
"learning_rate": 0.00034573829531812727,
|
| 1550 |
-
"loss": 1.9766,
|
| 1551 |
-
"step": 1285
|
| 1552 |
-
},
|
| 1553 |
-
{
|
| 1554 |
-
"epoch": 1.55,
|
| 1555 |
-
"learning_rate": 0.0003451380552220889,
|
| 1556 |
-
"loss": 1.9319,
|
| 1557 |
-
"step": 1290
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 1.55,
|
| 1561 |
-
"learning_rate": 0.00034453781512605044,
|
| 1562 |
-
"loss": 2.0803,
|
| 1563 |
-
"step": 1295
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 1.56,
|
| 1567 |
-
"learning_rate": 0.000343937575030012,
|
| 1568 |
-
"loss": 2.0196,
|
| 1569 |
-
"step": 1300
|
| 1570 |
-
},
|
| 1571 |
-
{
|
| 1572 |
-
"epoch": 1.57,
|
| 1573 |
-
"learning_rate": 0.0003433373349339736,
|
| 1574 |
-
"loss": 2.0166,
|
| 1575 |
-
"step": 1305
|
| 1576 |
-
},
|
| 1577 |
-
{
|
| 1578 |
-
"epoch": 1.57,
|
| 1579 |
-
"learning_rate": 0.0003427370948379352,
|
| 1580 |
-
"loss": 2.0094,
|
| 1581 |
-
"step": 1310
|
| 1582 |
-
},
|
| 1583 |
-
{
|
| 1584 |
-
"epoch": 1.58,
|
| 1585 |
-
"learning_rate": 0.0003421368547418968,
|
| 1586 |
-
"loss": 1.9954,
|
| 1587 |
-
"step": 1315
|
| 1588 |
-
},
|
| 1589 |
-
{
|
| 1590 |
-
"epoch": 1.58,
|
| 1591 |
-
"learning_rate": 0.00034153661464585833,
|
| 1592 |
-
"loss": 1.9782,
|
| 1593 |
-
"step": 1320
|
| 1594 |
-
},
|
| 1595 |
-
{
|
| 1596 |
-
"epoch": 1.59,
|
| 1597 |
-
"learning_rate": 0.00034093637454981994,
|
| 1598 |
-
"loss": 2.0126,
|
| 1599 |
-
"step": 1325
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 1.6,
|
| 1603 |
-
"learning_rate": 0.00034033613445378155,
|
| 1604 |
-
"loss": 1.9949,
|
| 1605 |
-
"step": 1330
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 1.6,
|
| 1609 |
-
"learning_rate": 0.0003397358943577431,
|
| 1610 |
-
"loss": 1.9744,
|
| 1611 |
-
"step": 1335
|
| 1612 |
-
},
|
| 1613 |
-
{
|
| 1614 |
-
"epoch": 1.61,
|
| 1615 |
-
"learning_rate": 0.00033913565426170467,
|
| 1616 |
-
"loss": 2.0246,
|
| 1617 |
-
"step": 1340
|
| 1618 |
-
},
|
| 1619 |
-
{
|
| 1620 |
-
"epoch": 1.61,
|
| 1621 |
-
"learning_rate": 0.0003385354141656663,
|
| 1622 |
-
"loss": 1.9461,
|
| 1623 |
-
"step": 1345
|
| 1624 |
-
},
|
| 1625 |
-
{
|
| 1626 |
-
"epoch": 1.62,
|
| 1627 |
-
"learning_rate": 0.0003379351740696279,
|
| 1628 |
-
"loss": 1.9792,
|
| 1629 |
-
"step": 1350
|
| 1630 |
-
},
|
| 1631 |
-
{
|
| 1632 |
-
"epoch": 1.63,
|
| 1633 |
-
"learning_rate": 0.00033733493397358945,
|
| 1634 |
-
"loss": 1.9813,
|
| 1635 |
-
"step": 1355
|
| 1636 |
-
},
|
| 1637 |
-
{
|
| 1638 |
-
"epoch": 1.63,
|
| 1639 |
-
"learning_rate": 0.000336734693877551,
|
| 1640 |
-
"loss": 2.0243,
|
| 1641 |
-
"step": 1360
|
| 1642 |
-
},
|
| 1643 |
-
{
|
| 1644 |
-
"epoch": 1.64,
|
| 1645 |
-
"learning_rate": 0.0003361344537815126,
|
| 1646 |
-
"loss": 1.9371,
|
| 1647 |
-
"step": 1365
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 1.64,
|
| 1651 |
-
"learning_rate": 0.0003355342136854742,
|
| 1652 |
-
"loss": 2.0438,
|
| 1653 |
-
"step": 1370
|
| 1654 |
-
},
|
| 1655 |
-
{
|
| 1656 |
-
"epoch": 1.65,
|
| 1657 |
-
"learning_rate": 0.0003349339735894358,
|
| 1658 |
-
"loss": 1.9632,
|
| 1659 |
-
"step": 1375
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 1.66,
|
| 1663 |
-
"learning_rate": 0.00033433373349339734,
|
| 1664 |
-
"loss": 2.0137,
|
| 1665 |
-
"step": 1380
|
| 1666 |
-
},
|
| 1667 |
-
{
|
| 1668 |
-
"epoch": 1.66,
|
| 1669 |
-
"learning_rate": 0.000333733493397359,
|
| 1670 |
-
"loss": 1.9849,
|
| 1671 |
-
"step": 1385
|
| 1672 |
-
},
|
| 1673 |
-
{
|
| 1674 |
-
"epoch": 1.67,
|
| 1675 |
-
"learning_rate": 0.00033313325330132056,
|
| 1676 |
-
"loss": 1.9578,
|
| 1677 |
-
"step": 1390
|
| 1678 |
-
},
|
| 1679 |
-
{
|
| 1680 |
-
"epoch": 1.67,
|
| 1681 |
-
"learning_rate": 0.0003325330132052821,
|
| 1682 |
-
"loss": 1.942,
|
| 1683 |
-
"step": 1395
|
| 1684 |
-
},
|
| 1685 |
-
{
|
| 1686 |
-
"epoch": 1.68,
|
| 1687 |
-
"learning_rate": 0.0003319327731092437,
|
| 1688 |
-
"loss": 1.9424,
|
| 1689 |
-
"step": 1400
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 1.69,
|
| 1693 |
-
"learning_rate": 0.0003313325330132053,
|
| 1694 |
-
"loss": 2.0015,
|
| 1695 |
-
"step": 1405
|
| 1696 |
-
},
|
| 1697 |
-
{
|
| 1698 |
-
"epoch": 1.69,
|
| 1699 |
-
"learning_rate": 0.0003307322929171669,
|
| 1700 |
-
"loss": 2.0385,
|
| 1701 |
-
"step": 1410
|
| 1702 |
-
},
|
| 1703 |
-
{
|
| 1704 |
-
"epoch": 1.7,
|
| 1705 |
-
"learning_rate": 0.00033013205282112846,
|
| 1706 |
-
"loss": 2.019,
|
| 1707 |
-
"step": 1415
|
| 1708 |
-
},
|
| 1709 |
-
{
|
| 1710 |
-
"epoch": 1.7,
|
| 1711 |
-
"learning_rate": 0.00032953181272509,
|
| 1712 |
-
"loss": 1.99,
|
| 1713 |
-
"step": 1420
|
| 1714 |
-
},
|
| 1715 |
-
{
|
| 1716 |
-
"epoch": 1.71,
|
| 1717 |
-
"learning_rate": 0.0003289315726290516,
|
| 1718 |
-
"loss": 1.9583,
|
| 1719 |
-
"step": 1425
|
| 1720 |
-
},
|
| 1721 |
-
{
|
| 1722 |
-
"epoch": 1.72,
|
| 1723 |
-
"learning_rate": 0.00032833133253301324,
|
| 1724 |
-
"loss": 1.9645,
|
| 1725 |
-
"step": 1430
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 1.72,
|
| 1729 |
-
"learning_rate": 0.0003277310924369748,
|
| 1730 |
-
"loss": 2.0095,
|
| 1731 |
-
"step": 1435
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 1.73,
|
| 1735 |
-
"learning_rate": 0.00032713085234093635,
|
| 1736 |
-
"loss": 1.9755,
|
| 1737 |
-
"step": 1440
|
| 1738 |
-
},
|
| 1739 |
-
{
|
| 1740 |
-
"epoch": 1.73,
|
| 1741 |
-
"learning_rate": 0.00032653061224489796,
|
| 1742 |
-
"loss": 1.8859,
|
| 1743 |
-
"step": 1445
|
| 1744 |
-
},
|
| 1745 |
-
{
|
| 1746 |
-
"epoch": 1.74,
|
| 1747 |
-
"learning_rate": 0.00032593037214885957,
|
| 1748 |
-
"loss": 1.9807,
|
| 1749 |
-
"step": 1450
|
| 1750 |
-
},
|
| 1751 |
-
{
|
| 1752 |
-
"epoch": 1.75,
|
| 1753 |
-
"learning_rate": 0.00032533013205282113,
|
| 1754 |
-
"loss": 1.9688,
|
| 1755 |
-
"step": 1455
|
| 1756 |
-
},
|
| 1757 |
-
{
|
| 1758 |
-
"epoch": 1.75,
|
| 1759 |
-
"learning_rate": 0.0003247298919567827,
|
| 1760 |
-
"loss": 1.9898,
|
| 1761 |
-
"step": 1460
|
| 1762 |
-
},
|
| 1763 |
-
{
|
| 1764 |
-
"epoch": 1.76,
|
| 1765 |
-
"learning_rate": 0.0003241296518607443,
|
| 1766 |
-
"loss": 1.9136,
|
| 1767 |
-
"step": 1465
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 1.76,
|
| 1771 |
-
"learning_rate": 0.0003235294117647059,
|
| 1772 |
-
"loss": 1.9058,
|
| 1773 |
-
"step": 1470
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 1.77,
|
| 1777 |
-
"learning_rate": 0.00032292917166866747,
|
| 1778 |
-
"loss": 2.001,
|
| 1779 |
-
"step": 1475
|
| 1780 |
-
},
|
| 1781 |
-
{
|
| 1782 |
-
"epoch": 1.78,
|
| 1783 |
-
"learning_rate": 0.000322328931572629,
|
| 1784 |
-
"loss": 1.9529,
|
| 1785 |
-
"step": 1480
|
| 1786 |
-
},
|
| 1787 |
-
{
|
| 1788 |
-
"epoch": 1.78,
|
| 1789 |
-
"learning_rate": 0.00032172869147659063,
|
| 1790 |
-
"loss": 1.9859,
|
| 1791 |
-
"step": 1485
|
| 1792 |
-
},
|
| 1793 |
-
{
|
| 1794 |
-
"epoch": 1.79,
|
| 1795 |
-
"learning_rate": 0.00032112845138055224,
|
| 1796 |
-
"loss": 1.9973,
|
| 1797 |
-
"step": 1490
|
| 1798 |
-
},
|
| 1799 |
-
{
|
| 1800 |
-
"epoch": 1.79,
|
| 1801 |
-
"learning_rate": 0.0003205282112845138,
|
| 1802 |
-
"loss": 1.9371,
|
| 1803 |
-
"step": 1495
|
| 1804 |
-
},
|
| 1805 |
-
{
|
| 1806 |
-
"epoch": 1.8,
|
| 1807 |
-
"learning_rate": 0.0003199279711884754,
|
| 1808 |
-
"loss": 1.949,
|
| 1809 |
-
"step": 1500
|
| 1810 |
}
|
| 1811 |
],
|
| 1812 |
"logging_steps": 5,
|
| 1813 |
-
"max_steps":
|
| 1814 |
"num_input_tokens_seen": 0,
|
| 1815 |
"num_train_epochs": 5,
|
| 1816 |
"save_steps": 500,
|
| 1817 |
-
"total_flos": 2.
|
| 1818 |
-
"train_batch_size":
|
| 1819 |
"trial_name": null,
|
| 1820 |
"trial_params": null
|
| 1821 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 0.0004995,
|
| 14 |
+
"loss": 2.0736,
|
| 15 |
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 0.000499,
|
| 20 |
+
"loss": 2.165,
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 0.0004985,
|
| 26 |
+
"loss": 2.084,
|
| 27 |
"step": 15
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.02,
|
| 31 |
+
"learning_rate": 0.000498,
|
| 32 |
+
"loss": 2.1784,
|
| 33 |
"step": 20
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.03,
|
| 37 |
+
"learning_rate": 0.0004975,
|
| 38 |
+
"loss": 2.0932,
|
| 39 |
"step": 25
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.03,
|
| 43 |
+
"learning_rate": 0.000497,
|
| 44 |
+
"loss": 2.056,
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.04,
|
| 49 |
+
"learning_rate": 0.0004965,
|
| 50 |
+
"loss": 2.0944,
|
| 51 |
"step": 35
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.04,
|
| 55 |
+
"learning_rate": 0.000496,
|
| 56 |
+
"loss": 2.1092,
|
| 57 |
"step": 40
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.04,
|
| 61 |
+
"learning_rate": 0.0004955,
|
| 62 |
+
"loss": 2.1146,
|
| 63 |
"step": 45
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.05,
|
| 67 |
+
"learning_rate": 0.000495,
|
| 68 |
+
"loss": 2.1368,
|
| 69 |
"step": 50
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.06,
|
| 73 |
+
"learning_rate": 0.0004945,
|
| 74 |
+
"loss": 2.0964,
|
| 75 |
"step": 55
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.06,
|
| 79 |
+
"learning_rate": 0.000494,
|
| 80 |
+
"loss": 2.0192,
|
| 81 |
"step": 60
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.07,
|
| 85 |
+
"learning_rate": 0.0004935,
|
| 86 |
+
"loss": 2.0938,
|
| 87 |
"step": 65
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.07,
|
| 91 |
+
"learning_rate": 0.0004930000000000001,
|
| 92 |
+
"loss": 2.0871,
|
| 93 |
"step": 70
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.07,
|
| 97 |
+
"learning_rate": 0.0004925,
|
| 98 |
+
"loss": 2.0773,
|
| 99 |
"step": 75
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.08,
|
| 103 |
+
"learning_rate": 0.000492,
|
| 104 |
+
"loss": 2.0618,
|
| 105 |
"step": 80
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.09,
|
| 109 |
+
"learning_rate": 0.0004915,
|
| 110 |
+
"loss": 2.0788,
|
| 111 |
"step": 85
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.09,
|
| 115 |
+
"learning_rate": 0.000491,
|
| 116 |
+
"loss": 2.1009,
|
| 117 |
"step": 90
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.1,
|
| 121 |
+
"learning_rate": 0.0004905,
|
| 122 |
+
"loss": 2.0138,
|
| 123 |
"step": 95
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.1,
|
| 127 |
+
"learning_rate": 0.00049,
|
| 128 |
+
"loss": 2.0689,
|
| 129 |
"step": 100
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.1,
|
| 133 |
+
"learning_rate": 0.0004895,
|
| 134 |
+
"loss": 2.0381,
|
| 135 |
"step": 105
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.11,
|
| 139 |
+
"learning_rate": 0.000489,
|
| 140 |
+
"loss": 2.1174,
|
| 141 |
"step": 110
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.12,
|
| 145 |
+
"learning_rate": 0.0004885,
|
| 146 |
+
"loss": 2.0315,
|
| 147 |
"step": 115
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.12,
|
| 151 |
+
"learning_rate": 0.000488,
|
| 152 |
+
"loss": 2.0351,
|
| 153 |
"step": 120
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.12,
|
| 157 |
+
"learning_rate": 0.0004875,
|
| 158 |
+
"loss": 2.1047,
|
| 159 |
"step": 125
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.13,
|
| 163 |
+
"learning_rate": 0.000487,
|
| 164 |
+
"loss": 2.0413,
|
| 165 |
"step": 130
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.14,
|
| 169 |
+
"learning_rate": 0.0004865,
|
| 170 |
+
"loss": 2.0978,
|
| 171 |
"step": 135
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.14,
|
| 175 |
+
"learning_rate": 0.000486,
|
| 176 |
+
"loss": 2.1018,
|
| 177 |
"step": 140
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.14,
|
| 181 |
+
"learning_rate": 0.0004855,
|
| 182 |
+
"loss": 2.0679,
|
| 183 |
"step": 145
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.15,
|
| 187 |
+
"learning_rate": 0.00048499999999999997,
|
| 188 |
+
"loss": 1.974,
|
| 189 |
"step": 150
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.15,
|
| 193 |
+
"learning_rate": 0.0004845,
|
| 194 |
+
"loss": 2.0155,
|
| 195 |
"step": 155
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 0.16,
|
| 199 |
+
"learning_rate": 0.000484,
|
| 200 |
+
"loss": 2.0396,
|
| 201 |
"step": 160
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 0.17,
|
| 205 |
+
"learning_rate": 0.0004835,
|
| 206 |
+
"loss": 2.0207,
|
| 207 |
"step": 165
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 0.17,
|
| 211 |
+
"learning_rate": 0.000483,
|
| 212 |
+
"loss": 2.017,
|
| 213 |
"step": 170
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 0.17,
|
| 217 |
+
"learning_rate": 0.0004825,
|
| 218 |
+
"loss": 2.0844,
|
| 219 |
"step": 175
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.18,
|
| 223 |
+
"learning_rate": 0.000482,
|
| 224 |
+
"loss": 2.0593,
|
| 225 |
"step": 180
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 0.18,
|
| 229 |
+
"learning_rate": 0.0004815,
|
| 230 |
+
"loss": 2.0018,
|
| 231 |
"step": 185
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 0.19,
|
| 235 |
+
"learning_rate": 0.000481,
|
| 236 |
+
"loss": 1.9544,
|
| 237 |
"step": 190
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 0.2,
|
| 241 |
+
"learning_rate": 0.00048049999999999997,
|
| 242 |
+
"loss": 2.063,
|
| 243 |
"step": 195
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 0.2,
|
| 247 |
+
"learning_rate": 0.00048,
|
| 248 |
+
"loss": 2.0345,
|
| 249 |
"step": 200
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 0.2,
|
| 253 |
+
"learning_rate": 0.0004795,
|
| 254 |
+
"loss": 2.0164,
|
| 255 |
"step": 205
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.21,
|
| 259 |
+
"learning_rate": 0.000479,
|
| 260 |
+
"loss": 1.9801,
|
| 261 |
"step": 210
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.21,
|
| 265 |
+
"learning_rate": 0.0004785,
|
| 266 |
+
"loss": 2.038,
|
| 267 |
"step": 215
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 0.22,
|
| 271 |
+
"learning_rate": 0.00047799999999999996,
|
| 272 |
+
"loss": 2.0919,
|
| 273 |
"step": 220
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 0.23,
|
| 277 |
+
"learning_rate": 0.0004775,
|
| 278 |
+
"loss": 2.0151,
|
| 279 |
"step": 225
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 0.23,
|
| 283 |
+
"learning_rate": 0.000477,
|
| 284 |
+
"loss": 1.9425,
|
| 285 |
"step": 230
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 0.23,
|
| 289 |
+
"learning_rate": 0.0004765,
|
| 290 |
+
"loss": 2.0252,
|
| 291 |
"step": 235
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 0.24,
|
| 295 |
+
"learning_rate": 0.00047599999999999997,
|
| 296 |
+
"loss": 1.965,
|
| 297 |
"step": 240
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 0.24,
|
| 301 |
+
"learning_rate": 0.0004755,
|
| 302 |
+
"loss": 1.9773,
|
| 303 |
"step": 245
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.25,
|
| 307 |
+
"learning_rate": 0.000475,
|
| 308 |
+
"loss": 1.9828,
|
| 309 |
"step": 250
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 0.26,
|
| 313 |
+
"learning_rate": 0.0004745,
|
| 314 |
+
"loss": 1.9247,
|
| 315 |
"step": 255
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 0.26,
|
| 319 |
+
"learning_rate": 0.000474,
|
| 320 |
+
"loss": 1.9878,
|
| 321 |
"step": 260
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 0.27,
|
| 325 |
+
"learning_rate": 0.00047349999999999996,
|
| 326 |
+
"loss": 2.0097,
|
| 327 |
"step": 265
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 0.27,
|
| 331 |
+
"learning_rate": 0.000473,
|
| 332 |
+
"loss": 2.0276,
|
| 333 |
"step": 270
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 0.28,
|
| 337 |
+
"learning_rate": 0.0004725,
|
| 338 |
+
"loss": 1.949,
|
| 339 |
"step": 275
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 0.28,
|
| 343 |
+
"learning_rate": 0.000472,
|
| 344 |
+
"loss": 2.0144,
|
| 345 |
"step": 280
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.28,
|
| 349 |
+
"learning_rate": 0.00047149999999999997,
|
| 350 |
+
"loss": 1.9787,
|
| 351 |
"step": 285
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 0.29,
|
| 355 |
+
"learning_rate": 0.000471,
|
| 356 |
+
"loss": 2.0019,
|
| 357 |
"step": 290
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 0.29,
|
| 361 |
+
"learning_rate": 0.0004705,
|
| 362 |
+
"loss": 2.0421,
|
| 363 |
"step": 295
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 0.3,
|
| 367 |
+
"learning_rate": 0.00047,
|
| 368 |
+
"loss": 1.9843,
|
| 369 |
"step": 300
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 0.3,
|
| 373 |
+
"learning_rate": 0.0004695,
|
| 374 |
+
"loss": 1.9604,
|
| 375 |
"step": 305
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 0.31,
|
| 379 |
+
"learning_rate": 0.00046899999999999996,
|
| 380 |
+
"loss": 1.9696,
|
| 381 |
"step": 310
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 0.32,
|
| 385 |
+
"learning_rate": 0.00046850000000000006,
|
| 386 |
+
"loss": 1.988,
|
| 387 |
"step": 315
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 0.32,
|
| 391 |
+
"learning_rate": 0.00046800000000000005,
|
| 392 |
+
"loss": 1.9925,
|
| 393 |
"step": 320
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 0.33,
|
| 397 |
+
"learning_rate": 0.00046750000000000003,
|
| 398 |
+
"loss": 1.9881,
|
| 399 |
"step": 325
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 0.33,
|
| 403 |
+
"learning_rate": 0.000467,
|
| 404 |
+
"loss": 1.9817,
|
| 405 |
"step": 330
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 0.34,
|
| 409 |
+
"learning_rate": 0.0004665,
|
| 410 |
+
"loss": 1.9384,
|
| 411 |
"step": 335
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 0.34,
|
| 415 |
+
"learning_rate": 0.00046600000000000005,
|
| 416 |
+
"loss": 1.9971,
|
| 417 |
"step": 340
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 0.34,
|
| 421 |
+
"learning_rate": 0.00046550000000000004,
|
| 422 |
+
"loss": 2.0022,
|
| 423 |
"step": 345
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 0.35,
|
| 427 |
+
"learning_rate": 0.000465,
|
| 428 |
+
"loss": 1.9593,
|
| 429 |
"step": 350
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 0.35,
|
| 433 |
+
"learning_rate": 0.0004645,
|
| 434 |
+
"loss": 1.9979,
|
| 435 |
"step": 355
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 0.36,
|
| 439 |
+
"learning_rate": 0.00046400000000000006,
|
| 440 |
+
"loss": 1.9264,
|
| 441 |
"step": 360
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 0.36,
|
| 445 |
+
"learning_rate": 0.00046350000000000004,
|
| 446 |
+
"loss": 1.9665,
|
| 447 |
"step": 365
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 0.37,
|
| 451 |
+
"learning_rate": 0.00046300000000000003,
|
| 452 |
+
"loss": 1.9307,
|
| 453 |
"step": 370
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 0.38,
|
| 457 |
+
"learning_rate": 0.0004625,
|
| 458 |
+
"loss": 1.9584,
|
| 459 |
"step": 375
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 0.38,
|
| 463 |
+
"learning_rate": 0.000462,
|
| 464 |
+
"loss": 2.0025,
|
| 465 |
"step": 380
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 0.39,
|
| 469 |
+
"learning_rate": 0.00046150000000000005,
|
| 470 |
+
"loss": 1.98,
|
| 471 |
"step": 385
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 0.39,
|
| 475 |
+
"learning_rate": 0.00046100000000000004,
|
| 476 |
+
"loss": 1.9673,
|
| 477 |
"step": 390
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 0.4,
|
| 481 |
+
"learning_rate": 0.0004605,
|
| 482 |
+
"loss": 1.9598,
|
| 483 |
"step": 395
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 0.4,
|
| 487 |
+
"learning_rate": 0.00046,
|
| 488 |
+
"loss": 2.0057,
|
| 489 |
"step": 400
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 0.41,
|
| 493 |
+
"learning_rate": 0.00045950000000000006,
|
| 494 |
+
"loss": 1.9537,
|
| 495 |
"step": 405
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 0.41,
|
| 499 |
+
"learning_rate": 0.00045900000000000004,
|
| 500 |
+
"loss": 1.9726,
|
| 501 |
"step": 410
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 0.41,
|
| 505 |
+
"learning_rate": 0.00045850000000000003,
|
| 506 |
+
"loss": 1.9011,
|
| 507 |
"step": 415
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 0.42,
|
| 511 |
+
"learning_rate": 0.000458,
|
| 512 |
+
"loss": 1.9988,
|
| 513 |
"step": 420
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 0.42,
|
| 517 |
+
"learning_rate": 0.0004575,
|
| 518 |
+
"loss": 1.9626,
|
| 519 |
"step": 425
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 0.43,
|
| 523 |
+
"learning_rate": 0.00045700000000000005,
|
| 524 |
+
"loss": 1.9083,
|
| 525 |
"step": 430
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 0.43,
|
| 529 |
+
"learning_rate": 0.00045650000000000004,
|
| 530 |
+
"loss": 1.9429,
|
| 531 |
"step": 435
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 0.44,
|
| 535 |
+
"learning_rate": 0.000456,
|
| 536 |
+
"loss": 1.9819,
|
| 537 |
"step": 440
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 0.45,
|
| 541 |
+
"learning_rate": 0.0004555,
|
| 542 |
+
"loss": 1.9259,
|
| 543 |
"step": 445
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 0.45,
|
| 547 |
+
"learning_rate": 0.000455,
|
| 548 |
+
"loss": 1.9024,
|
| 549 |
"step": 450
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 0.46,
|
| 553 |
+
"learning_rate": 0.00045450000000000004,
|
| 554 |
+
"loss": 1.983,
|
| 555 |
"step": 455
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 0.46,
|
| 559 |
+
"learning_rate": 0.00045400000000000003,
|
| 560 |
+
"loss": 1.9776,
|
| 561 |
"step": 460
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 0.47,
|
| 565 |
+
"learning_rate": 0.0004535,
|
| 566 |
+
"loss": 1.9816,
|
| 567 |
"step": 465
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 0.47,
|
| 571 |
+
"learning_rate": 0.000453,
|
| 572 |
+
"loss": 1.9402,
|
| 573 |
"step": 470
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 0.47,
|
| 577 |
+
"learning_rate": 0.00045250000000000005,
|
| 578 |
+
"loss": 1.9931,
|
| 579 |
"step": 475
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 0.48,
|
| 583 |
+
"learning_rate": 0.00045200000000000004,
|
| 584 |
+
"loss": 1.9528,
|
| 585 |
"step": 480
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 0.48,
|
| 589 |
+
"learning_rate": 0.0004515,
|
| 590 |
+
"loss": 1.9392,
|
| 591 |
"step": 485
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 0.49,
|
| 595 |
+
"learning_rate": 0.000451,
|
| 596 |
+
"loss": 1.9249,
|
| 597 |
"step": 490
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 0.49,
|
| 601 |
+
"learning_rate": 0.0004505,
|
| 602 |
+
"loss": 1.9711,
|
| 603 |
"step": 495
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 0.5,
|
| 607 |
+
"learning_rate": 0.00045000000000000004,
|
| 608 |
+
"loss": 1.8939,
|
| 609 |
"step": 500
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 0.51,
|
| 613 |
+
"learning_rate": 0.00044950000000000003,
|
| 614 |
+
"loss": 1.96,
|
| 615 |
"step": 505
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 0.51,
|
| 619 |
+
"learning_rate": 0.000449,
|
| 620 |
+
"loss": 1.9014,
|
| 621 |
"step": 510
|
| 622 |
},
|
| 623 |
{
|
| 624 |
+
"epoch": 0.52,
|
| 625 |
+
"learning_rate": 0.0004485,
|
| 626 |
+
"loss": 1.9354,
|
| 627 |
"step": 515
|
| 628 |
},
|
| 629 |
{
|
| 630 |
+
"epoch": 0.52,
|
| 631 |
+
"learning_rate": 0.000448,
|
| 632 |
+
"loss": 1.9597,
|
| 633 |
"step": 520
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 0.53,
|
| 637 |
+
"learning_rate": 0.00044750000000000004,
|
| 638 |
+
"loss": 1.9367,
|
| 639 |
"step": 525
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 0.53,
|
| 643 |
+
"learning_rate": 0.000447,
|
| 644 |
+
"loss": 1.9698,
|
| 645 |
"step": 530
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 0.54,
|
| 649 |
+
"learning_rate": 0.0004465,
|
| 650 |
+
"loss": 1.8743,
|
| 651 |
"step": 535
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 0.54,
|
| 655 |
+
"learning_rate": 0.000446,
|
| 656 |
+
"loss": 1.9083,
|
| 657 |
"step": 540
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 0.55,
|
| 661 |
+
"learning_rate": 0.00044550000000000004,
|
| 662 |
+
"loss": 1.9096,
|
| 663 |
"step": 545
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 0.55,
|
| 667 |
+
"learning_rate": 0.00044500000000000003,
|
| 668 |
+
"loss": 2.0429,
|
| 669 |
"step": 550
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 0.56,
|
| 673 |
+
"learning_rate": 0.0004445,
|
| 674 |
+
"loss": 1.9972,
|
| 675 |
"step": 555
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 0.56,
|
| 679 |
+
"learning_rate": 0.000444,
|
| 680 |
+
"loss": 1.8974,
|
| 681 |
"step": 560
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 0.56,
|
| 685 |
+
"learning_rate": 0.0004435,
|
| 686 |
+
"loss": 1.9467,
|
| 687 |
"step": 565
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 0.57,
|
| 691 |
+
"learning_rate": 0.00044300000000000003,
|
| 692 |
+
"loss": 1.9382,
|
| 693 |
"step": 570
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 0.57,
|
| 697 |
+
"learning_rate": 0.0004425,
|
| 698 |
+
"loss": 1.8893,
|
| 699 |
"step": 575
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 0.58,
|
| 703 |
+
"learning_rate": 0.000442,
|
| 704 |
+
"loss": 1.9392,
|
| 705 |
"step": 580
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 0.58,
|
| 709 |
+
"learning_rate": 0.0004415,
|
| 710 |
+
"loss": 1.9229,
|
| 711 |
"step": 585
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 0.59,
|
| 715 |
+
"learning_rate": 0.000441,
|
| 716 |
+
"loss": 1.9042,
|
| 717 |
"step": 590
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 0.59,
|
| 721 |
+
"learning_rate": 0.00044050000000000003,
|
| 722 |
+
"loss": 1.9104,
|
| 723 |
"step": 595
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 0.6,
|
| 727 |
+
"learning_rate": 0.00044,
|
| 728 |
+
"loss": 1.9541,
|
| 729 |
"step": 600
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 0.6,
|
| 733 |
+
"learning_rate": 0.0004395,
|
| 734 |
+
"loss": 1.8948,
|
| 735 |
"step": 605
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 0.61,
|
| 739 |
+
"learning_rate": 0.000439,
|
| 740 |
+
"loss": 1.8849,
|
| 741 |
"step": 610
|
| 742 |
},
|
| 743 |
{
|
| 744 |
+
"epoch": 0.61,
|
| 745 |
+
"learning_rate": 0.00043850000000000003,
|
| 746 |
+
"loss": 1.9097,
|
| 747 |
"step": 615
|
| 748 |
},
|
| 749 |
{
|
| 750 |
+
"epoch": 0.62,
|
| 751 |
+
"learning_rate": 0.000438,
|
| 752 |
+
"loss": 1.9457,
|
| 753 |
"step": 620
|
| 754 |
},
|
| 755 |
{
|
| 756 |
+
"epoch": 0.62,
|
| 757 |
+
"learning_rate": 0.0004375,
|
| 758 |
+
"loss": 1.9231,
|
| 759 |
"step": 625
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 0.63,
|
| 763 |
+
"learning_rate": 0.000437,
|
| 764 |
+
"loss": 1.9228,
|
| 765 |
"step": 630
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 0.64,
|
| 769 |
+
"learning_rate": 0.0004365,
|
| 770 |
+
"loss": 1.9258,
|
| 771 |
"step": 635
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 0.64,
|
| 775 |
+
"learning_rate": 0.000436,
|
| 776 |
+
"loss": 1.8884,
|
| 777 |
"step": 640
|
| 778 |
},
|
| 779 |
{
|
| 780 |
+
"epoch": 0.65,
|
| 781 |
+
"learning_rate": 0.0004355,
|
| 782 |
+
"loss": 1.8693,
|
| 783 |
"step": 645
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 0.65,
|
| 787 |
+
"learning_rate": 0.000435,
|
| 788 |
+
"loss": 1.8635,
|
| 789 |
"step": 650
|
| 790 |
},
|
| 791 |
{
|
| 792 |
+
"epoch": 0.66,
|
| 793 |
+
"learning_rate": 0.0004345,
|
| 794 |
+
"loss": 1.9319,
|
| 795 |
"step": 655
|
| 796 |
},
|
| 797 |
{
|
| 798 |
+
"epoch": 0.66,
|
| 799 |
+
"learning_rate": 0.00043400000000000003,
|
| 800 |
+
"loss": 1.9116,
|
| 801 |
"step": 660
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 0.67,
|
| 805 |
+
"learning_rate": 0.0004335,
|
| 806 |
+
"loss": 1.9299,
|
| 807 |
"step": 665
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 0.67,
|
| 811 |
+
"learning_rate": 0.000433,
|
| 812 |
+
"loss": 1.8971,
|
| 813 |
"step": 670
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 0.68,
|
| 817 |
+
"learning_rate": 0.0004325,
|
| 818 |
+
"loss": 1.9317,
|
| 819 |
"step": 675
|
| 820 |
},
|
| 821 |
{
|
| 822 |
+
"epoch": 0.68,
|
| 823 |
+
"learning_rate": 0.000432,
|
| 824 |
+
"loss": 1.9523,
|
| 825 |
"step": 680
|
| 826 |
},
|
| 827 |
{
|
| 828 |
+
"epoch": 0.69,
|
| 829 |
+
"learning_rate": 0.0004315,
|
| 830 |
+
"loss": 1.8683,
|
| 831 |
"step": 685
|
| 832 |
},
|
| 833 |
{
|
| 834 |
+
"epoch": 0.69,
|
| 835 |
+
"learning_rate": 0.000431,
|
| 836 |
+
"loss": 1.9321,
|
| 837 |
"step": 690
|
| 838 |
},
|
| 839 |
{
|
| 840 |
+
"epoch": 0.69,
|
| 841 |
+
"learning_rate": 0.0004305,
|
| 842 |
+
"loss": 1.8971,
|
| 843 |
"step": 695
|
| 844 |
},
|
| 845 |
{
|
| 846 |
+
"epoch": 0.7,
|
| 847 |
+
"learning_rate": 0.00043,
|
| 848 |
+
"loss": 1.8666,
|
| 849 |
"step": 700
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 0.7,
|
| 853 |
+
"learning_rate": 0.0004295,
|
| 854 |
+
"loss": 1.888,
|
| 855 |
"step": 705
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 0.71,
|
| 859 |
+
"learning_rate": 0.000429,
|
| 860 |
+
"loss": 1.8946,
|
| 861 |
"step": 710
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 0.71,
|
| 865 |
+
"learning_rate": 0.0004285,
|
| 866 |
+
"loss": 1.9037,
|
| 867 |
"step": 715
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 0.72,
|
| 871 |
+
"learning_rate": 0.000428,
|
| 872 |
+
"loss": 1.8533,
|
| 873 |
"step": 720
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 0.72,
|
| 877 |
+
"learning_rate": 0.0004275,
|
| 878 |
+
"loss": 1.9052,
|
| 879 |
"step": 725
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 0.73,
|
| 883 |
+
"learning_rate": 0.000427,
|
| 884 |
+
"loss": 1.8677,
|
| 885 |
"step": 730
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 0.73,
|
| 889 |
+
"learning_rate": 0.0004265,
|
| 890 |
+
"loss": 1.9164,
|
| 891 |
"step": 735
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 0.74,
|
| 895 |
+
"learning_rate": 0.000426,
|
| 896 |
+
"loss": 1.8823,
|
| 897 |
"step": 740
|
| 898 |
},
|
| 899 |
{
|
| 900 |
+
"epoch": 0.74,
|
| 901 |
+
"learning_rate": 0.0004255,
|
| 902 |
+
"loss": 1.9215,
|
| 903 |
"step": 745
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 0.75,
|
| 907 |
+
"learning_rate": 0.000425,
|
| 908 |
+
"loss": 1.8903,
|
| 909 |
"step": 750
|
| 910 |
},
|
| 911 |
{
|
| 912 |
+
"epoch": 0.76,
|
| 913 |
+
"learning_rate": 0.0004245,
|
| 914 |
+
"loss": 1.8148,
|
| 915 |
"step": 755
|
| 916 |
},
|
| 917 |
{
|
| 918 |
+
"epoch": 0.76,
|
| 919 |
+
"learning_rate": 0.000424,
|
| 920 |
+
"loss": 1.8324,
|
| 921 |
"step": 760
|
| 922 |
},
|
| 923 |
{
|
| 924 |
+
"epoch": 0.77,
|
| 925 |
+
"learning_rate": 0.0004235,
|
| 926 |
+
"loss": 1.9011,
|
| 927 |
"step": 765
|
| 928 |
},
|
| 929 |
{
|
| 930 |
+
"epoch": 0.77,
|
| 931 |
+
"learning_rate": 0.000423,
|
| 932 |
+
"loss": 1.9424,
|
| 933 |
"step": 770
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 0.78,
|
| 937 |
+
"learning_rate": 0.00042249999999999997,
|
| 938 |
+
"loss": 1.9028,
|
| 939 |
"step": 775
|
| 940 |
},
|
| 941 |
{
|
| 942 |
+
"epoch": 0.78,
|
| 943 |
+
"learning_rate": 0.000422,
|
| 944 |
+
"loss": 1.8789,
|
| 945 |
"step": 780
|
| 946 |
},
|
| 947 |
{
|
| 948 |
+
"epoch": 0.79,
|
| 949 |
+
"learning_rate": 0.0004215,
|
| 950 |
+
"loss": 1.9016,
|
| 951 |
"step": 785
|
| 952 |
},
|
| 953 |
{
|
| 954 |
+
"epoch": 0.79,
|
| 955 |
+
"learning_rate": 0.000421,
|
| 956 |
+
"loss": 1.8837,
|
| 957 |
"step": 790
|
| 958 |
},
|
| 959 |
{
|
| 960 |
+
"epoch": 0.8,
|
| 961 |
+
"learning_rate": 0.0004205,
|
| 962 |
+
"loss": 1.8651,
|
| 963 |
"step": 795
|
| 964 |
},
|
| 965 |
{
|
| 966 |
+
"epoch": 0.8,
|
| 967 |
+
"learning_rate": 0.00042,
|
| 968 |
+
"loss": 1.8212,
|
| 969 |
"step": 800
|
| 970 |
},
|
| 971 |
{
|
| 972 |
+
"epoch": 0.81,
|
| 973 |
+
"learning_rate": 0.0004195,
|
| 974 |
+
"loss": 1.9073,
|
| 975 |
"step": 805
|
| 976 |
},
|
| 977 |
{
|
| 978 |
+
"epoch": 0.81,
|
| 979 |
+
"learning_rate": 0.000419,
|
| 980 |
+
"loss": 1.9267,
|
| 981 |
"step": 810
|
| 982 |
},
|
| 983 |
{
|
| 984 |
+
"epoch": 0.81,
|
| 985 |
+
"learning_rate": 0.0004185,
|
| 986 |
+
"loss": 1.8311,
|
| 987 |
"step": 815
|
| 988 |
},
|
| 989 |
{
|
| 990 |
+
"epoch": 0.82,
|
| 991 |
+
"learning_rate": 0.00041799999999999997,
|
| 992 |
+
"loss": 1.8537,
|
| 993 |
"step": 820
|
| 994 |
},
|
| 995 |
{
|
| 996 |
+
"epoch": 0.82,
|
| 997 |
+
"learning_rate": 0.0004175,
|
| 998 |
+
"loss": 1.8717,
|
| 999 |
"step": 825
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
+
"epoch": 0.83,
|
| 1003 |
+
"learning_rate": 0.000417,
|
| 1004 |
+
"loss": 1.8241,
|
| 1005 |
"step": 830
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
+
"epoch": 0.83,
|
| 1009 |
+
"learning_rate": 0.0004165,
|
| 1010 |
+
"loss": 1.8863,
|
| 1011 |
"step": 835
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
+
"epoch": 0.84,
|
| 1015 |
+
"learning_rate": 0.000416,
|
| 1016 |
+
"loss": 1.9053,
|
| 1017 |
"step": 840
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
+
"epoch": 0.84,
|
| 1021 |
+
"learning_rate": 0.00041549999999999996,
|
| 1022 |
+
"loss": 1.87,
|
| 1023 |
"step": 845
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
+
"epoch": 0.85,
|
| 1027 |
+
"learning_rate": 0.000415,
|
| 1028 |
+
"loss": 1.837,
|
| 1029 |
"step": 850
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
+
"epoch": 0.85,
|
| 1033 |
+
"learning_rate": 0.0004145,
|
| 1034 |
+
"loss": 1.8941,
|
| 1035 |
"step": 855
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
+
"epoch": 0.86,
|
| 1039 |
+
"learning_rate": 0.000414,
|
| 1040 |
+
"loss": 1.9237,
|
| 1041 |
"step": 860
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
+
"epoch": 0.86,
|
| 1045 |
+
"learning_rate": 0.00041349999999999997,
|
| 1046 |
+
"loss": 1.8854,
|
| 1047 |
"step": 865
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
+
"epoch": 0.87,
|
| 1051 |
+
"learning_rate": 0.000413,
|
| 1052 |
+
"loss": 1.8401,
|
| 1053 |
"step": 870
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
+
"epoch": 0.88,
|
| 1057 |
+
"learning_rate": 0.0004125,
|
| 1058 |
+
"loss": 1.8314,
|
| 1059 |
"step": 875
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 0.88,
|
| 1063 |
+
"learning_rate": 0.000412,
|
| 1064 |
+
"loss": 1.858,
|
| 1065 |
"step": 880
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 0.89,
|
| 1069 |
+
"learning_rate": 0.0004115,
|
| 1070 |
+
"loss": 1.8806,
|
| 1071 |
"step": 885
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 0.89,
|
| 1075 |
+
"learning_rate": 0.00041099999999999996,
|
| 1076 |
+
"loss": 1.8244,
|
| 1077 |
"step": 890
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
+
"epoch": 0.9,
|
| 1081 |
+
"learning_rate": 0.0004105,
|
| 1082 |
+
"loss": 1.8432,
|
| 1083 |
"step": 895
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
+
"epoch": 0.9,
|
| 1087 |
+
"learning_rate": 0.00041,
|
| 1088 |
+
"loss": 1.8596,
|
| 1089 |
"step": 900
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
+
"epoch": 0.91,
|
| 1093 |
+
"learning_rate": 0.0004095,
|
| 1094 |
+
"loss": 1.8276,
|
| 1095 |
"step": 905
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
+
"epoch": 0.91,
|
| 1099 |
+
"learning_rate": 0.00040899999999999997,
|
| 1100 |
+
"loss": 1.9486,
|
| 1101 |
"step": 910
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 0.92,
|
| 1105 |
+
"learning_rate": 0.0004085,
|
| 1106 |
+
"loss": 1.8382,
|
| 1107 |
"step": 915
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
+
"epoch": 0.92,
|
| 1111 |
+
"learning_rate": 0.000408,
|
| 1112 |
+
"loss": 1.8735,
|
| 1113 |
"step": 920
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
+
"epoch": 0.93,
|
| 1117 |
+
"learning_rate": 0.0004075,
|
| 1118 |
+
"loss": 1.8203,
|
| 1119 |
"step": 925
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
+
"epoch": 0.93,
|
| 1123 |
+
"learning_rate": 0.00040699999999999997,
|
| 1124 |
+
"loss": 1.8045,
|
| 1125 |
"step": 930
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
+
"epoch": 0.94,
|
| 1129 |
+
"learning_rate": 0.00040649999999999996,
|
| 1130 |
+
"loss": 1.8732,
|
| 1131 |
"step": 935
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
+
"epoch": 0.94,
|
| 1135 |
+
"learning_rate": 0.00040600000000000006,
|
| 1136 |
+
"loss": 1.8032,
|
| 1137 |
"step": 940
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
+
"epoch": 0.94,
|
| 1141 |
+
"learning_rate": 0.00040550000000000004,
|
| 1142 |
+
"loss": 1.8079,
|
| 1143 |
"step": 945
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
+
"epoch": 0.95,
|
| 1147 |
+
"learning_rate": 0.00040500000000000003,
|
| 1148 |
+
"loss": 1.8855,
|
| 1149 |
"step": 950
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
+
"epoch": 0.95,
|
| 1153 |
+
"learning_rate": 0.0004045,
|
| 1154 |
+
"loss": 1.8511,
|
| 1155 |
"step": 955
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
+
"epoch": 0.96,
|
| 1159 |
+
"learning_rate": 0.000404,
|
| 1160 |
+
"loss": 1.8529,
|
| 1161 |
"step": 960
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
+
"epoch": 0.96,
|
| 1165 |
+
"learning_rate": 0.00040350000000000005,
|
| 1166 |
+
"loss": 1.8482,
|
| 1167 |
"step": 965
|
| 1168 |
},
|
| 1169 |
{
|
| 1170 |
+
"epoch": 0.97,
|
| 1171 |
+
"learning_rate": 0.00040300000000000004,
|
| 1172 |
+
"loss": 1.8288,
|
| 1173 |
"step": 970
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
+
"epoch": 0.97,
|
| 1177 |
+
"learning_rate": 0.0004025,
|
| 1178 |
+
"loss": 1.8781,
|
| 1179 |
"step": 975
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
+
"epoch": 0.98,
|
| 1183 |
+
"learning_rate": 0.000402,
|
| 1184 |
+
"loss": 1.8458,
|
| 1185 |
"step": 980
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 0.98,
|
| 1189 |
+
"learning_rate": 0.00040150000000000006,
|
| 1190 |
+
"loss": 1.8426,
|
| 1191 |
"step": 985
|
| 1192 |
},
|
| 1193 |
{
|
| 1194 |
+
"epoch": 0.99,
|
| 1195 |
+
"learning_rate": 0.00040100000000000004,
|
| 1196 |
+
"loss": 1.8517,
|
| 1197 |
"step": 990
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
+
"epoch": 0.99,
|
| 1201 |
+
"learning_rate": 0.00040050000000000003,
|
| 1202 |
+
"loss": 1.8426,
|
| 1203 |
"step": 995
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
+
"epoch": 1.0,
|
| 1207 |
+
"learning_rate": 0.0004,
|
| 1208 |
+
"loss": 1.7787,
|
| 1209 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1210 |
}
|
| 1211 |
],
|
| 1212 |
"logging_steps": 5,
|
| 1213 |
+
"max_steps": 5000,
|
| 1214 |
"num_input_tokens_seen": 0,
|
| 1215 |
"num_train_epochs": 5,
|
| 1216 |
"save_steps": 500,
|
| 1217 |
+
"total_flos": 2.65273620205824e+16,
|
| 1218 |
+
"train_batch_size": 10,
|
| 1219 |
"trial_name": null,
|
| 1220 |
"trial_params": null
|
| 1221 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69db0ed0280f7b72e766ea62e28c3f9ca85173e00962b7b7d10e2f20229f6d83
|
| 3 |
size 4728
|