Commit ·
dbe0879
1
Parent(s): 6d3e2ce
updated model files
Browse files- config.json +1 -1
- generation_config.json +5 -0
- optimizer.pt +2 -2
- pytorch_model.bin +2 -2
- rng_state.pth +2 -2
- scheduler.pt +2 -2
- tokenizer.json +2 -16
- tokenizer_config.json +1 -2
- trainer_state.json +2058 -397
- training_args.bin +2 -2
config.json
CHANGED
|
@@ -19,7 +19,7 @@
|
|
| 19 |
"pad_token_id": 0,
|
| 20 |
"position_embedding_type": "absolute",
|
| 21 |
"torch_dtype": "float32",
|
| 22 |
-
"transformers_version": "4.
|
| 23 |
"type_vocab_size": 2,
|
| 24 |
"use_cache": true,
|
| 25 |
"vocab_size": 30522
|
|
|
|
| 19 |
"pad_token_id": 0,
|
| 20 |
"position_embedding_type": "absolute",
|
| 21 |
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.33.3",
|
| 23 |
"type_vocab_size": 2,
|
| 24 |
"use_cache": true,
|
| 25 |
"vocab_size": 30522
|
generation_config.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"pad_token_id": 0,
|
| 4 |
+
"transformers_version": "4.33.3"
|
| 5 |
+
}
|
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:a3f31febe1dd0cc26397d30f312c151db6acc0fd9b12d444e3ae813896e2da0e
|
| 3 |
+
size 1555717
|
pytorch_model.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 sha256:526e1af0a8d0f41e583eff223781a02d05475392c007c54436302e577c644a7b
|
| 3 |
+
size 438126133
|
rng_state.pth
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:d11b1e775e8d24ee8aa73e2be08c7d3442a71a9b3343c20d8cea19f6c0499c28
|
| 3 |
+
size 14575
|
scheduler.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:ab3169029a398e76d6013cadd402227ef5f345365c777edfd7ad382de6b99270
|
| 3 |
+
size 627
|
tokenizer.json
CHANGED
|
@@ -1,21 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
| 4 |
-
|
| 5 |
-
"max_length": 512,
|
| 6 |
-
"strategy": "LongestFirst",
|
| 7 |
-
"stride": 0
|
| 8 |
-
},
|
| 9 |
-
"padding": {
|
| 10 |
-
"strategy": {
|
| 11 |
-
"Fixed": 512
|
| 12 |
-
},
|
| 13 |
-
"direction": "Right",
|
| 14 |
-
"pad_to_multiple_of": null,
|
| 15 |
-
"pad_id": 0,
|
| 16 |
-
"pad_type_id": 0,
|
| 17 |
-
"pad_token": "[PAD]"
|
| 18 |
-
},
|
| 19 |
"added_tokens": [
|
| 20 |
{
|
| 21 |
"id": 0,
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
| 7 |
"id": 0,
|
tokenizer_config.json
CHANGED
|
@@ -1,14 +1,13 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"cls_token": "[CLS]",
|
| 3 |
"do_lower_case": true,
|
| 4 |
"mask_token": "[MASK]",
|
| 5 |
"model_max_length": 512,
|
| 6 |
-
"name_or_path": "bert-base-uncased",
|
| 7 |
"pad_token": "[PAD]",
|
| 8 |
"padding": "max_length",
|
| 9 |
"return_special_tokens_mask": true,
|
| 10 |
"sep_token": "[SEP]",
|
| 11 |
-
"special_tokens_map_file": null,
|
| 12 |
"strip_accents": null,
|
| 13 |
"tokenize_chinese_chars": true,
|
| 14 |
"tokenizer_class": "BertTokenizer",
|
|
|
|
| 1 |
{
|
| 2 |
+
"clean_up_tokenization_spaces": true,
|
| 3 |
"cls_token": "[CLS]",
|
| 4 |
"do_lower_case": true,
|
| 5 |
"mask_token": "[MASK]",
|
| 6 |
"model_max_length": 512,
|
|
|
|
| 7 |
"pad_token": "[PAD]",
|
| 8 |
"padding": "max_length",
|
| 9 |
"return_special_tokens_mask": true,
|
| 10 |
"sep_token": "[SEP]",
|
|
|
|
| 11 |
"strip_accents": null,
|
| 12 |
"tokenize_chinese_chars": true,
|
| 13 |
"tokenizer_class": "BertTokenizer",
|
trainer_state.json
CHANGED
|
@@ -1,796 +1,2457 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"
|
|
|
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
-
"epoch": 0.
|
| 12 |
-
"learning_rate": 9.
|
| 13 |
-
"loss":
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0.
|
| 18 |
-
"learning_rate": 9.
|
| 19 |
-
"loss":
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0.
|
| 24 |
-
"learning_rate": 9.
|
| 25 |
-
"loss":
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0.
|
| 30 |
-
"learning_rate": 9.
|
| 31 |
-
"loss":
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"learning_rate": 9.
|
| 37 |
-
"loss":
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"
|
| 43 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"learning_rate": 9.
|
| 49 |
-
"loss":
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0.
|
| 54 |
-
"learning_rate": 9.
|
| 55 |
-
"loss":
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0.
|
| 60 |
-
"learning_rate": 9.
|
| 61 |
-
"loss":
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
-
"epoch": 0.
|
| 66 |
-
"learning_rate": 9.
|
| 67 |
-
"loss":
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch": 0.
|
| 72 |
-
"
|
| 73 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 0.
|
| 78 |
-
"learning_rate": 9.
|
| 79 |
-
"loss":
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"learning_rate": 9.
|
| 85 |
-
"loss":
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"learning_rate": 9.
|
| 91 |
-
"loss":
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
-
"learning_rate": 9.
|
| 97 |
-
"loss":
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch": 0.
|
| 102 |
-
"
|
| 103 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch": 0.
|
| 108 |
-
"learning_rate": 9.
|
| 109 |
-
"loss":
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch": 0.
|
| 114 |
-
"learning_rate": 9.
|
| 115 |
-
"loss":
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 0.
|
| 120 |
-
"learning_rate": 9.
|
| 121 |
-
"loss":
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
-
"learning_rate": 9.
|
| 127 |
-
"loss":
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"
|
| 133 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 134 |
"step": 2100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch": 0.
|
| 138 |
-
"learning_rate": 9.
|
| 139 |
-
"loss":
|
| 140 |
"step": 2200
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch": 0.
|
| 144 |
-
"learning_rate": 9.
|
| 145 |
-
"loss":
|
| 146 |
"step": 2300
|
| 147 |
},
|
| 148 |
{
|
| 149 |
-
"epoch": 0.
|
| 150 |
-
"learning_rate": 9.
|
| 151 |
-
"loss":
|
| 152 |
"step": 2400
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch": 0.
|
| 156 |
-
"learning_rate": 9.
|
| 157 |
-
"loss":
|
| 158 |
"step": 2500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch": 0.
|
| 162 |
-
"
|
| 163 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 164 |
"step": 2600
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"learning_rate": 9.
|
| 169 |
-
"loss":
|
| 170 |
"step": 2700
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"learning_rate": 9.
|
| 175 |
-
"loss":
|
| 176 |
"step": 2800
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch": 0.
|
| 180 |
-
"learning_rate": 9.
|
| 181 |
-
"loss":
|
| 182 |
"step": 2900
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch": 0.
|
| 186 |
-
"learning_rate": 9.
|
| 187 |
-
"loss":
|
| 188 |
"step": 3000
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch": 0.
|
| 192 |
-
"
|
| 193 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 194 |
"step": 3100
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch": 0.
|
| 198 |
-
"learning_rate": 9.
|
| 199 |
-
"loss":
|
| 200 |
"step": 3200
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch": 0.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss":
|
| 206 |
"step": 3300
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
-
"learning_rate":
|
| 211 |
-
"loss":
|
| 212 |
"step": 3400
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"learning_rate":
|
| 217 |
-
"loss":
|
| 218 |
"step": 3500
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch": 0.
|
| 222 |
-
"
|
| 223 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 224 |
"step": 3600
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"epoch": 0.
|
| 228 |
-
"learning_rate":
|
| 229 |
-
"loss":
|
| 230 |
"step": 3700
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch": 0.
|
| 234 |
-
"learning_rate":
|
| 235 |
-
"loss":
|
| 236 |
"step": 3800
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch": 0.
|
| 240 |
-
"learning_rate":
|
| 241 |
-
"loss":
|
| 242 |
"step": 3900
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch": 0.
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss":
|
| 248 |
"step": 4000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
-
"
|
| 253 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 254 |
"step": 4100
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"learning_rate":
|
| 259 |
-
"loss":
|
| 260 |
"step": 4200
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"epoch": 0.
|
| 264 |
-
"learning_rate":
|
| 265 |
-
"loss":
|
| 266 |
"step": 4300
|
| 267 |
},
|
| 268 |
{
|
| 269 |
-
"epoch": 0.
|
| 270 |
-
"learning_rate":
|
| 271 |
-
"loss":
|
| 272 |
"step": 4400
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch": 0.
|
| 276 |
-
"learning_rate":
|
| 277 |
-
"loss":
|
| 278 |
"step": 4500
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch": 0.
|
| 282 |
-
"
|
| 283 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 284 |
"step": 4600
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch": 0.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss":
|
| 290 |
"step": 4700
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 0.
|
| 294 |
-
"learning_rate":
|
| 295 |
-
"loss":
|
| 296 |
"step": 4800
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
-
"learning_rate":
|
| 301 |
-
"loss":
|
| 302 |
"step": 4900
|
| 303 |
},
|
| 304 |
{
|
| 305 |
-
"epoch": 0.
|
| 306 |
-
"learning_rate":
|
| 307 |
-
"loss":
|
| 308 |
"step": 5000
|
| 309 |
},
|
| 310 |
{
|
| 311 |
-
"epoch": 0.
|
| 312 |
-
"
|
| 313 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
"step": 5100
|
| 315 |
},
|
| 316 |
{
|
| 317 |
-
"epoch": 0.
|
| 318 |
-
"learning_rate":
|
| 319 |
-
"loss":
|
| 320 |
"step": 5200
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"epoch": 0.
|
| 324 |
-
"learning_rate":
|
| 325 |
-
"loss":
|
| 326 |
"step": 5300
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch": 0.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss":
|
| 332 |
"step": 5400
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
-
"learning_rate":
|
| 337 |
-
"loss":
|
| 338 |
"step": 5500
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"
|
| 343 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 344 |
"step": 5600
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"epoch": 0.
|
| 348 |
-
"learning_rate":
|
| 349 |
-
"loss":
|
| 350 |
"step": 5700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
-
"epoch": 0.
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss":
|
| 356 |
"step": 5800
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch": 0.
|
| 360 |
-
"learning_rate":
|
| 361 |
-
"loss":
|
| 362 |
"step": 5900
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"epoch":
|
| 366 |
-
"learning_rate":
|
| 367 |
-
"loss":
|
| 368 |
"step": 6000
|
| 369 |
},
|
| 370 |
{
|
| 371 |
-
"epoch":
|
| 372 |
-
"
|
| 373 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 374 |
"step": 6100
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch":
|
| 378 |
-
"learning_rate": 8.
|
| 379 |
-
"loss":
|
| 380 |
"step": 6200
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch":
|
| 384 |
-
"learning_rate": 8.
|
| 385 |
-
"loss":
|
| 386 |
"step": 6300
|
| 387 |
},
|
| 388 |
{
|
| 389 |
-
"epoch":
|
| 390 |
-
"learning_rate": 8.
|
| 391 |
-
"loss":
|
| 392 |
"step": 6400
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"epoch":
|
| 396 |
-
"learning_rate":
|
| 397 |
-
"loss":
|
| 398 |
"step": 6500
|
| 399 |
},
|
| 400 |
{
|
| 401 |
-
"epoch":
|
| 402 |
-
"
|
| 403 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 404 |
"step": 6600
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch":
|
| 408 |
-
"learning_rate":
|
| 409 |
-
"loss":
|
| 410 |
"step": 6700
|
| 411 |
},
|
| 412 |
{
|
| 413 |
-
"epoch":
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss":
|
| 416 |
"step": 6800
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch":
|
| 420 |
-
"learning_rate":
|
| 421 |
-
"loss":
|
| 422 |
"step": 6900
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch":
|
| 426 |
-
"learning_rate":
|
| 427 |
-
"loss":
|
| 428 |
"step": 7000
|
| 429 |
},
|
| 430 |
{
|
| 431 |
-
"epoch":
|
| 432 |
-
"
|
| 433 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 434 |
"step": 7100
|
| 435 |
},
|
| 436 |
{
|
| 437 |
-
"epoch":
|
| 438 |
-
"learning_rate":
|
| 439 |
-
"loss":
|
| 440 |
"step": 7200
|
| 441 |
},
|
| 442 |
{
|
| 443 |
-
"epoch":
|
| 444 |
-
"learning_rate":
|
| 445 |
-
"loss":
|
| 446 |
"step": 7300
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch":
|
| 450 |
-
"learning_rate":
|
| 451 |
-
"loss":
|
| 452 |
"step": 7400
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch":
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss":
|
| 458 |
"step": 7500
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"
|
| 463 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 464 |
"step": 7600
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch":
|
| 468 |
-
"learning_rate":
|
| 469 |
-
"loss":
|
| 470 |
"step": 7700
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch":
|
| 474 |
-
"learning_rate":
|
| 475 |
-
"loss":
|
| 476 |
"step": 7800
|
| 477 |
},
|
| 478 |
{
|
| 479 |
-
"epoch":
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss":
|
| 482 |
"step": 7900
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"epoch":
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss":
|
| 488 |
"step": 8000
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch":
|
| 492 |
-
"
|
| 493 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 494 |
"step": 8100
|
| 495 |
},
|
| 496 |
{
|
| 497 |
-
"epoch":
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss":
|
| 500 |
"step": 8200
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch":
|
| 504 |
-
"learning_rate":
|
| 505 |
-
"loss":
|
| 506 |
"step": 8300
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch":
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss":
|
| 512 |
"step": 8400
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch":
|
| 516 |
-
"learning_rate":
|
| 517 |
-
"loss":
|
| 518 |
"step": 8500
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch":
|
| 522 |
-
"
|
| 523 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 524 |
"step": 8600
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch":
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss":
|
| 530 |
"step": 8700
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch":
|
| 534 |
-
"learning_rate":
|
| 535 |
-
"loss":
|
| 536 |
"step": 8800
|
| 537 |
},
|
| 538 |
{
|
| 539 |
-
"epoch":
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss":
|
| 542 |
"step": 8900
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch":
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss":
|
| 548 |
"step": 9000
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch":
|
| 552 |
-
"
|
| 553 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 554 |
"step": 9100
|
| 555 |
},
|
| 556 |
{
|
| 557 |
-
"epoch":
|
| 558 |
-
"learning_rate":
|
| 559 |
-
"loss":
|
| 560 |
"step": 9200
|
| 561 |
},
|
| 562 |
{
|
| 563 |
-
"epoch":
|
| 564 |
-
"learning_rate":
|
| 565 |
-
"loss":
|
| 566 |
"step": 9300
|
| 567 |
},
|
| 568 |
{
|
| 569 |
-
"epoch":
|
| 570 |
-
"learning_rate":
|
| 571 |
-
"loss":
|
| 572 |
"step": 9400
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch":
|
| 576 |
-
"learning_rate":
|
| 577 |
-
"loss":
|
| 578 |
"step": 9500
|
| 579 |
},
|
| 580 |
{
|
| 581 |
-
"epoch":
|
| 582 |
-
"
|
| 583 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 584 |
"step": 9600
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch":
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss":
|
| 590 |
"step": 9700
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch":
|
| 594 |
-
"learning_rate":
|
| 595 |
-
"loss":
|
| 596 |
"step": 9800
|
| 597 |
},
|
| 598 |
{
|
| 599 |
-
"epoch":
|
| 600 |
-
"learning_rate":
|
| 601 |
-
"loss":
|
| 602 |
"step": 9900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch":
|
| 606 |
-
"learning_rate":
|
| 607 |
-
"loss":
|
| 608 |
"step": 10000
|
| 609 |
},
|
| 610 |
{
|
| 611 |
-
"epoch":
|
| 612 |
-
"
|
| 613 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 614 |
"step": 10100
|
| 615 |
},
|
| 616 |
{
|
| 617 |
-
"epoch":
|
| 618 |
-
"learning_rate":
|
| 619 |
-
"loss":
|
| 620 |
"step": 10200
|
| 621 |
},
|
| 622 |
{
|
| 623 |
-
"epoch":
|
| 624 |
-
"learning_rate":
|
| 625 |
-
"loss":
|
| 626 |
"step": 10300
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch":
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss":
|
| 632 |
"step": 10400
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch":
|
| 636 |
-
"learning_rate":
|
| 637 |
-
"loss":
|
| 638 |
"step": 10500
|
| 639 |
},
|
| 640 |
{
|
| 641 |
-
"epoch":
|
| 642 |
-
"
|
| 643 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 644 |
"step": 10600
|
| 645 |
},
|
| 646 |
{
|
| 647 |
-
"epoch":
|
| 648 |
-
"learning_rate":
|
| 649 |
-
"loss":
|
| 650 |
"step": 10700
|
| 651 |
},
|
| 652 |
{
|
| 653 |
-
"epoch":
|
| 654 |
-
"learning_rate":
|
| 655 |
-
"loss":
|
| 656 |
"step": 10800
|
| 657 |
},
|
| 658 |
{
|
| 659 |
-
"epoch":
|
| 660 |
-
"learning_rate":
|
| 661 |
-
"loss":
|
| 662 |
"step": 10900
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"epoch":
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss":
|
| 668 |
"step": 11000
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch":
|
| 672 |
-
"
|
| 673 |
-
"
|
| 674 |
-
"
|
| 675 |
-
|
|
|
|
|
|
|
| 676 |
{
|
| 677 |
-
"epoch":
|
| 678 |
-
"learning_rate":
|
| 679 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 680 |
"step": 11200
|
| 681 |
},
|
| 682 |
{
|
| 683 |
-
"epoch":
|
| 684 |
-
"learning_rate":
|
| 685 |
-
"loss":
|
| 686 |
"step": 11300
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch":
|
| 690 |
-
"learning_rate":
|
| 691 |
-
"loss":
|
| 692 |
"step": 11400
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"epoch":
|
| 696 |
-
"learning_rate":
|
| 697 |
-
"loss":
|
| 698 |
"step": 11500
|
| 699 |
},
|
| 700 |
{
|
| 701 |
-
"epoch":
|
| 702 |
-
"
|
| 703 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 704 |
"step": 11600
|
| 705 |
},
|
| 706 |
{
|
| 707 |
-
"epoch":
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss":
|
| 710 |
"step": 11700
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch":
|
| 714 |
-
"learning_rate":
|
| 715 |
-
"loss":
|
| 716 |
"step": 11800
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch":
|
| 720 |
-
"learning_rate":
|
| 721 |
-
"loss":
|
| 722 |
"step": 11900
|
| 723 |
},
|
| 724 |
{
|
| 725 |
-
"epoch":
|
| 726 |
-
"learning_rate":
|
| 727 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 728 |
"step": 12000
|
| 729 |
},
|
| 730 |
{
|
| 731 |
-
"epoch":
|
| 732 |
-
"learning_rate":
|
| 733 |
-
"loss":
|
| 734 |
"step": 12100
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"epoch":
|
| 738 |
-
"learning_rate":
|
| 739 |
-
"loss":
|
| 740 |
"step": 12200
|
| 741 |
},
|
| 742 |
{
|
| 743 |
-
"epoch":
|
| 744 |
-
"learning_rate":
|
| 745 |
-
"loss":
|
| 746 |
"step": 12300
|
| 747 |
},
|
| 748 |
{
|
| 749 |
-
"epoch":
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss":
|
| 752 |
"step": 12400
|
| 753 |
},
|
| 754 |
{
|
| 755 |
-
"epoch":
|
| 756 |
-
"learning_rate":
|
| 757 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 758 |
"step": 12500
|
| 759 |
},
|
| 760 |
{
|
| 761 |
-
"epoch":
|
| 762 |
-
"learning_rate":
|
| 763 |
-
"loss":
|
| 764 |
"step": 12600
|
| 765 |
},
|
| 766 |
{
|
| 767 |
-
"epoch":
|
| 768 |
-
"learning_rate":
|
| 769 |
-
"loss":
|
| 770 |
"step": 12700
|
| 771 |
},
|
| 772 |
{
|
| 773 |
-
"epoch":
|
| 774 |
-
"learning_rate":
|
| 775 |
-
"loss":
|
| 776 |
"step": 12800
|
| 777 |
},
|
| 778 |
{
|
| 779 |
-
"epoch":
|
| 780 |
-
"learning_rate":
|
| 781 |
-
"loss":
|
| 782 |
"step": 12900
|
| 783 |
},
|
| 784 |
{
|
| 785 |
-
"epoch":
|
| 786 |
-
"learning_rate":
|
| 787 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 788 |
"step": 13000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 789 |
}
|
| 790 |
],
|
| 791 |
-
"
|
| 792 |
-
"
|
| 793 |
-
"
|
|
|
|
|
|
|
| 794 |
"trial_name": null,
|
| 795 |
"trial_params": null
|
| 796 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9993226593794158,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 32100,
|
| 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": 9.969479912799752e-06,
|
| 14 |
+
"loss": 5.7053,
|
| 15 |
"step": 100
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 9.938336966677049e-06,
|
| 20 |
+
"loss": 4.2137,
|
| 21 |
"step": 200
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.02,
|
| 25 |
+
"learning_rate": 9.907194020554345e-06,
|
| 26 |
+
"loss": 3.7382,
|
| 27 |
"step": 300
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.02,
|
| 31 |
+
"learning_rate": 9.876051074431642e-06,
|
| 32 |
+
"loss": 3.768,
|
| 33 |
"step": 400
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.03,
|
| 37 |
+
"learning_rate": 9.844908128308939e-06,
|
| 38 |
+
"loss": 4.0831,
|
| 39 |
"step": 500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.03,
|
| 43 |
+
"eval_loss": NaN,
|
| 44 |
+
"eval_runtime": 1241.4479,
|
| 45 |
+
"eval_samples_per_second": 183.933,
|
| 46 |
+
"eval_steps_per_second": 22.992,
|
| 47 |
+
"step": 500
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.04,
|
| 51 |
+
"learning_rate": 9.813765182186236e-06,
|
| 52 |
+
"loss": 3.9938,
|
| 53 |
"step": 600
|
| 54 |
},
|
| 55 |
{
|
| 56 |
+
"epoch": 0.04,
|
| 57 |
+
"learning_rate": 9.782622236063532e-06,
|
| 58 |
+
"loss": 3.6619,
|
| 59 |
"step": 700
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.05,
|
| 63 |
+
"learning_rate": 9.751790719402056e-06,
|
| 64 |
+
"loss": 3.9896,
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.06,
|
| 69 |
+
"learning_rate": 9.720647773279352e-06,
|
| 70 |
+
"loss": 3.9662,
|
| 71 |
"step": 900
|
| 72 |
},
|
| 73 |
{
|
| 74 |
+
"epoch": 0.06,
|
| 75 |
+
"learning_rate": 9.68950482715665e-06,
|
| 76 |
+
"loss": 4.0557,
|
| 77 |
"step": 1000
|
| 78 |
},
|
| 79 |
{
|
| 80 |
+
"epoch": 0.06,
|
| 81 |
+
"eval_loss": NaN,
|
| 82 |
+
"eval_runtime": 1241.0879,
|
| 83 |
+
"eval_samples_per_second": 183.986,
|
| 84 |
+
"eval_steps_per_second": 22.998,
|
| 85 |
+
"step": 1000
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.07,
|
| 89 |
+
"learning_rate": 9.658361881033946e-06,
|
| 90 |
+
"loss": 3.4231,
|
| 91 |
"step": 1100
|
| 92 |
},
|
| 93 |
{
|
| 94 |
+
"epoch": 0.07,
|
| 95 |
+
"learning_rate": 9.627218934911242e-06,
|
| 96 |
+
"loss": 3.4517,
|
| 97 |
"step": 1200
|
| 98 |
},
|
| 99 |
{
|
| 100 |
+
"epoch": 0.08,
|
| 101 |
+
"learning_rate": 9.59607598878854e-06,
|
| 102 |
+
"loss": 4.235,
|
| 103 |
"step": 1300
|
| 104 |
},
|
| 105 |
{
|
| 106 |
+
"epoch": 0.09,
|
| 107 |
+
"learning_rate": 9.564933042665836e-06,
|
| 108 |
+
"loss": 3.3104,
|
| 109 |
"step": 1400
|
| 110 |
},
|
| 111 |
{
|
| 112 |
+
"epoch": 0.09,
|
| 113 |
+
"learning_rate": 9.533790096543134e-06,
|
| 114 |
+
"loss": 3.6169,
|
| 115 |
"step": 1500
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 0.09,
|
| 119 |
+
"eval_loss": NaN,
|
| 120 |
+
"eval_runtime": 1240.8951,
|
| 121 |
+
"eval_samples_per_second": 184.015,
|
| 122 |
+
"eval_steps_per_second": 23.002,
|
| 123 |
+
"step": 1500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1,
|
| 127 |
+
"learning_rate": 9.50264715042043e-06,
|
| 128 |
+
"loss": 3.3399,
|
| 129 |
"step": 1600
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.11,
|
| 133 |
+
"learning_rate": 9.471504204297728e-06,
|
| 134 |
+
"loss": 3.9392,
|
| 135 |
"step": 1700
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.11,
|
| 139 |
+
"learning_rate": 9.440361258175024e-06,
|
| 140 |
+
"loss": 4.2941,
|
| 141 |
"step": 1800
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.12,
|
| 145 |
+
"learning_rate": 9.40921831205232e-06,
|
| 146 |
+
"loss": 4.2293,
|
| 147 |
"step": 1900
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.12,
|
| 151 |
+
"learning_rate": 9.378075365929618e-06,
|
| 152 |
+
"loss": 3.411,
|
| 153 |
"step": 2000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.12,
|
| 157 |
+
"eval_loss": NaN,
|
| 158 |
+
"eval_runtime": 1240.9958,
|
| 159 |
+
"eval_samples_per_second": 184.0,
|
| 160 |
+
"eval_steps_per_second": 23.0,
|
| 161 |
+
"step": 2000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.13,
|
| 165 |
+
"learning_rate": 9.346932419806914e-06,
|
| 166 |
+
"loss": 3.1979,
|
| 167 |
"step": 2100
|
| 168 |
},
|
| 169 |
{
|
| 170 |
+
"epoch": 0.14,
|
| 171 |
+
"learning_rate": 9.315789473684212e-06,
|
| 172 |
+
"loss": 3.303,
|
| 173 |
"step": 2200
|
| 174 |
},
|
| 175 |
{
|
| 176 |
+
"epoch": 0.14,
|
| 177 |
+
"learning_rate": 9.284646527561508e-06,
|
| 178 |
+
"loss": 3.2433,
|
| 179 |
"step": 2300
|
| 180 |
},
|
| 181 |
{
|
| 182 |
+
"epoch": 0.15,
|
| 183 |
+
"learning_rate": 9.253503581438806e-06,
|
| 184 |
+
"loss": 3.0607,
|
| 185 |
"step": 2400
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 0.16,
|
| 189 |
+
"learning_rate": 9.222360635316102e-06,
|
| 190 |
+
"loss": 2.8986,
|
| 191 |
"step": 2500
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.16,
|
| 195 |
+
"eval_loss": NaN,
|
| 196 |
+
"eval_runtime": 1242.4997,
|
| 197 |
+
"eval_samples_per_second": 183.777,
|
| 198 |
+
"eval_steps_per_second": 22.972,
|
| 199 |
+
"step": 2500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.16,
|
| 203 |
+
"learning_rate": 9.191217689193398e-06,
|
| 204 |
+
"loss": 3.5778,
|
| 205 |
"step": 2600
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.17,
|
| 209 |
+
"learning_rate": 9.160074743070694e-06,
|
| 210 |
+
"loss": 3.2627,
|
| 211 |
"step": 2700
|
| 212 |
},
|
| 213 |
{
|
| 214 |
+
"epoch": 0.17,
|
| 215 |
+
"learning_rate": 9.128931796947992e-06,
|
| 216 |
+
"loss": 2.8115,
|
| 217 |
"step": 2800
|
| 218 |
},
|
| 219 |
{
|
| 220 |
+
"epoch": 0.18,
|
| 221 |
+
"learning_rate": 9.097788850825288e-06,
|
| 222 |
+
"loss": 3.1633,
|
| 223 |
"step": 2900
|
| 224 |
},
|
| 225 |
{
|
| 226 |
+
"epoch": 0.19,
|
| 227 |
+
"learning_rate": 9.066645904702586e-06,
|
| 228 |
+
"loss": 3.3414,
|
| 229 |
"step": 3000
|
| 230 |
},
|
| 231 |
{
|
| 232 |
+
"epoch": 0.19,
|
| 233 |
+
"eval_loss": NaN,
|
| 234 |
+
"eval_runtime": 1241.065,
|
| 235 |
+
"eval_samples_per_second": 183.99,
|
| 236 |
+
"eval_steps_per_second": 22.999,
|
| 237 |
+
"step": 3000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.19,
|
| 241 |
+
"learning_rate": 9.035502958579882e-06,
|
| 242 |
+
"loss": 3.1895,
|
| 243 |
"step": 3100
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 0.2,
|
| 247 |
+
"learning_rate": 9.00436001245718e-06,
|
| 248 |
+
"loss": 3.2972,
|
| 249 |
"step": 3200
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 0.21,
|
| 253 |
+
"learning_rate": 8.973217066334476e-06,
|
| 254 |
+
"loss": 3.3117,
|
| 255 |
"step": 3300
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.21,
|
| 259 |
+
"learning_rate": 8.942074120211772e-06,
|
| 260 |
+
"loss": 3.8183,
|
| 261 |
"step": 3400
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.22,
|
| 265 |
+
"learning_rate": 8.911242603550296e-06,
|
| 266 |
+
"loss": 3.0445,
|
| 267 |
"step": 3500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 0.22,
|
| 271 |
+
"eval_loss": NaN,
|
| 272 |
+
"eval_runtime": 1241.4692,
|
| 273 |
+
"eval_samples_per_second": 183.93,
|
| 274 |
+
"eval_steps_per_second": 22.991,
|
| 275 |
+
"step": 3500
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.22,
|
| 279 |
+
"learning_rate": 8.880099657427594e-06,
|
| 280 |
+
"loss": 2.716,
|
| 281 |
"step": 3600
|
| 282 |
},
|
| 283 |
{
|
| 284 |
+
"epoch": 0.23,
|
| 285 |
+
"learning_rate": 8.84895671130489e-06,
|
| 286 |
+
"loss": 2.7579,
|
| 287 |
"step": 3700
|
| 288 |
},
|
| 289 |
{
|
| 290 |
+
"epoch": 0.24,
|
| 291 |
+
"learning_rate": 8.817813765182188e-06,
|
| 292 |
+
"loss": 3.6821,
|
| 293 |
"step": 3800
|
| 294 |
},
|
| 295 |
{
|
| 296 |
+
"epoch": 0.24,
|
| 297 |
+
"learning_rate": 8.786670819059484e-06,
|
| 298 |
+
"loss": 2.9784,
|
| 299 |
"step": 3900
|
| 300 |
},
|
| 301 |
{
|
| 302 |
+
"epoch": 0.25,
|
| 303 |
+
"learning_rate": 8.755527872936782e-06,
|
| 304 |
+
"loss": 3.1284,
|
| 305 |
"step": 4000
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 0.25,
|
| 309 |
+
"eval_loss": NaN,
|
| 310 |
+
"eval_runtime": 1241.3243,
|
| 311 |
+
"eval_samples_per_second": 183.951,
|
| 312 |
+
"eval_steps_per_second": 22.994,
|
| 313 |
+
"step": 4000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.26,
|
| 317 |
+
"learning_rate": 8.724384926814076e-06,
|
| 318 |
+
"loss": 3.3944,
|
| 319 |
"step": 4100
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 0.26,
|
| 323 |
+
"learning_rate": 8.693241980691374e-06,
|
| 324 |
+
"loss": 4.3378,
|
| 325 |
"step": 4200
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 0.27,
|
| 329 |
+
"learning_rate": 8.66209903456867e-06,
|
| 330 |
+
"loss": 3.4281,
|
| 331 |
"step": 4300
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 0.27,
|
| 335 |
+
"learning_rate": 8.630956088445968e-06,
|
| 336 |
+
"loss": 3.4173,
|
| 337 |
"step": 4400
|
| 338 |
},
|
| 339 |
{
|
| 340 |
+
"epoch": 0.28,
|
| 341 |
+
"learning_rate": 8.599813142323264e-06,
|
| 342 |
+
"loss": 2.987,
|
| 343 |
"step": 4500
|
| 344 |
},
|
| 345 |
{
|
| 346 |
+
"epoch": 0.28,
|
| 347 |
+
"eval_loss": NaN,
|
| 348 |
+
"eval_runtime": 1242.9007,
|
| 349 |
+
"eval_samples_per_second": 183.718,
|
| 350 |
+
"eval_steps_per_second": 22.965,
|
| 351 |
+
"step": 4500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.29,
|
| 355 |
+
"learning_rate": 8.568670196200562e-06,
|
| 356 |
+
"loss": 2.9806,
|
| 357 |
"step": 4600
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 0.29,
|
| 361 |
+
"learning_rate": 8.537527250077858e-06,
|
| 362 |
+
"loss": 3.2971,
|
| 363 |
"step": 4700
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 0.3,
|
| 367 |
+
"learning_rate": 8.506384303955156e-06,
|
| 368 |
+
"loss": 2.9308,
|
| 369 |
"step": 4800
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 0.31,
|
| 373 |
+
"learning_rate": 8.475241357832452e-06,
|
| 374 |
+
"loss": 3.95,
|
| 375 |
"step": 4900
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 0.31,
|
| 379 |
+
"learning_rate": 8.444098411709748e-06,
|
| 380 |
+
"loss": 3.6195,
|
| 381 |
"step": 5000
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 0.31,
|
| 385 |
+
"eval_loss": NaN,
|
| 386 |
+
"eval_runtime": 12249.9104,
|
| 387 |
+
"eval_samples_per_second": 18.64,
|
| 388 |
+
"eval_steps_per_second": 2.33,
|
| 389 |
+
"step": 5000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.32,
|
| 393 |
+
"learning_rate": 8.412955465587044e-06,
|
| 394 |
+
"loss": 3.3623,
|
| 395 |
"step": 5100
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 0.32,
|
| 399 |
+
"learning_rate": 8.381812519464342e-06,
|
| 400 |
+
"loss": 2.7048,
|
| 401 |
"step": 5200
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 0.33,
|
| 405 |
+
"learning_rate": 8.350669573341638e-06,
|
| 406 |
+
"loss": 2.9451,
|
| 407 |
"step": 5300
|
| 408 |
},
|
| 409 |
{
|
| 410 |
+
"epoch": 0.34,
|
| 411 |
+
"learning_rate": 8.319526627218936e-06,
|
| 412 |
+
"loss": 3.9476,
|
| 413 |
"step": 5400
|
| 414 |
},
|
| 415 |
{
|
| 416 |
+
"epoch": 0.34,
|
| 417 |
+
"learning_rate": 8.288383681096232e-06,
|
| 418 |
+
"loss": 3.0161,
|
| 419 |
"step": 5500
|
| 420 |
},
|
| 421 |
{
|
| 422 |
+
"epoch": 0.34,
|
| 423 |
+
"eval_loss": NaN,
|
| 424 |
+
"eval_runtime": 1242.1829,
|
| 425 |
+
"eval_samples_per_second": 183.824,
|
| 426 |
+
"eval_steps_per_second": 22.978,
|
| 427 |
+
"step": 5500
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.35,
|
| 431 |
+
"learning_rate": 8.25724073497353e-06,
|
| 432 |
+
"loss": 2.7617,
|
| 433 |
"step": 5600
|
| 434 |
},
|
| 435 |
{
|
| 436 |
+
"epoch": 0.36,
|
| 437 |
+
"learning_rate": 8.226097788850826e-06,
|
| 438 |
+
"loss": 2.6001,
|
| 439 |
"step": 5700
|
| 440 |
},
|
| 441 |
{
|
| 442 |
+
"epoch": 0.36,
|
| 443 |
+
"learning_rate": 8.194954842728122e-06,
|
| 444 |
+
"loss": 2.8024,
|
| 445 |
"step": 5800
|
| 446 |
},
|
| 447 |
{
|
| 448 |
+
"epoch": 0.37,
|
| 449 |
+
"learning_rate": 8.16381189660542e-06,
|
| 450 |
+
"loss": 2.6335,
|
| 451 |
"step": 5900
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 0.37,
|
| 455 |
+
"learning_rate": 8.132668950482716e-06,
|
| 456 |
+
"loss": 3.0251,
|
| 457 |
"step": 6000
|
| 458 |
},
|
| 459 |
{
|
| 460 |
+
"epoch": 0.37,
|
| 461 |
+
"eval_loss": NaN,
|
| 462 |
+
"eval_runtime": 1241.5477,
|
| 463 |
+
"eval_samples_per_second": 183.918,
|
| 464 |
+
"eval_steps_per_second": 22.99,
|
| 465 |
+
"step": 6000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.38,
|
| 469 |
+
"learning_rate": 8.101526004360014e-06,
|
| 470 |
+
"loss": 4.0077,
|
| 471 |
"step": 6100
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 0.39,
|
| 475 |
+
"learning_rate": 8.07038305823731e-06,
|
| 476 |
+
"loss": 4.0494,
|
| 477 |
"step": 6200
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 0.39,
|
| 481 |
+
"learning_rate": 8.039240112114608e-06,
|
| 482 |
+
"loss": 2.8992,
|
| 483 |
"step": 6300
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 0.4,
|
| 487 |
+
"learning_rate": 8.008097165991904e-06,
|
| 488 |
+
"loss": 3.0616,
|
| 489 |
"step": 6400
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 0.4,
|
| 493 |
+
"learning_rate": 7.9769542198692e-06,
|
| 494 |
+
"loss": 3.4257,
|
| 495 |
"step": 6500
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 0.4,
|
| 499 |
+
"eval_loss": NaN,
|
| 500 |
+
"eval_runtime": 1242.4356,
|
| 501 |
+
"eval_samples_per_second": 183.787,
|
| 502 |
+
"eval_steps_per_second": 22.973,
|
| 503 |
+
"step": 6500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.41,
|
| 507 |
+
"learning_rate": 7.945811273746496e-06,
|
| 508 |
+
"loss": 2.9658,
|
| 509 |
"step": 6600
|
| 510 |
},
|
| 511 |
{
|
| 512 |
+
"epoch": 0.42,
|
| 513 |
+
"learning_rate": 7.914668327623794e-06,
|
| 514 |
+
"loss": 3.288,
|
| 515 |
"step": 6700
|
| 516 |
},
|
| 517 |
{
|
| 518 |
+
"epoch": 0.42,
|
| 519 |
+
"learning_rate": 7.88352538150109e-06,
|
| 520 |
+
"loss": 3.1517,
|
| 521 |
"step": 6800
|
| 522 |
},
|
| 523 |
{
|
| 524 |
+
"epoch": 0.43,
|
| 525 |
+
"learning_rate": 7.852382435378388e-06,
|
| 526 |
+
"loss": 3.3913,
|
| 527 |
"step": 6900
|
| 528 |
},
|
| 529 |
{
|
| 530 |
+
"epoch": 0.44,
|
| 531 |
+
"learning_rate": 7.821239489255684e-06,
|
| 532 |
+
"loss": 2.7966,
|
| 533 |
"step": 7000
|
| 534 |
},
|
| 535 |
{
|
| 536 |
+
"epoch": 0.44,
|
| 537 |
+
"eval_loss": NaN,
|
| 538 |
+
"eval_runtime": 1239.9554,
|
| 539 |
+
"eval_samples_per_second": 184.154,
|
| 540 |
+
"eval_steps_per_second": 23.019,
|
| 541 |
+
"step": 7000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.44,
|
| 545 |
+
"learning_rate": 7.790096543132982e-06,
|
| 546 |
+
"loss": 3.372,
|
| 547 |
"step": 7100
|
| 548 |
},
|
| 549 |
{
|
| 550 |
+
"epoch": 0.45,
|
| 551 |
+
"learning_rate": 7.758953597010278e-06,
|
| 552 |
+
"loss": 2.8074,
|
| 553 |
"step": 7200
|
| 554 |
},
|
| 555 |
{
|
| 556 |
+
"epoch": 0.45,
|
| 557 |
+
"learning_rate": 7.727810650887576e-06,
|
| 558 |
+
"loss": 2.6364,
|
| 559 |
"step": 7300
|
| 560 |
},
|
| 561 |
{
|
| 562 |
+
"epoch": 0.46,
|
| 563 |
+
"learning_rate": 7.696667704764872e-06,
|
| 564 |
+
"loss": 3.0556,
|
| 565 |
"step": 7400
|
| 566 |
},
|
| 567 |
{
|
| 568 |
+
"epoch": 0.47,
|
| 569 |
+
"learning_rate": 7.665524758642168e-06,
|
| 570 |
+
"loss": 2.4966,
|
| 571 |
"step": 7500
|
| 572 |
},
|
| 573 |
{
|
| 574 |
+
"epoch": 0.47,
|
| 575 |
+
"eval_loss": NaN,
|
| 576 |
+
"eval_runtime": 1238.3825,
|
| 577 |
+
"eval_samples_per_second": 184.388,
|
| 578 |
+
"eval_steps_per_second": 23.049,
|
| 579 |
+
"step": 7500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.47,
|
| 583 |
+
"learning_rate": 7.634381812519464e-06,
|
| 584 |
+
"loss": 3.562,
|
| 585 |
"step": 7600
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 0.48,
|
| 589 |
+
"learning_rate": 7.603238866396762e-06,
|
| 590 |
+
"loss": 4.0498,
|
| 591 |
"step": 7700
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 0.49,
|
| 595 |
+
"learning_rate": 7.572095920274059e-06,
|
| 596 |
+
"loss": 2.8504,
|
| 597 |
"step": 7800
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 0.49,
|
| 601 |
+
"learning_rate": 7.540952974151356e-06,
|
| 602 |
+
"loss": 2.7274,
|
| 603 |
"step": 7900
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 0.5,
|
| 607 |
+
"learning_rate": 7.509810028028653e-06,
|
| 608 |
+
"loss": 2.9404,
|
| 609 |
"step": 8000
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 0.5,
|
| 613 |
+
"eval_loss": NaN,
|
| 614 |
+
"eval_runtime": 1242.6775,
|
| 615 |
+
"eval_samples_per_second": 183.751,
|
| 616 |
+
"eval_steps_per_second": 22.969,
|
| 617 |
+
"step": 8000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.5,
|
| 621 |
+
"learning_rate": 7.47866708190595e-06,
|
| 622 |
+
"loss": 2.6808,
|
| 623 |
"step": 8100
|
| 624 |
},
|
| 625 |
{
|
| 626 |
+
"epoch": 0.51,
|
| 627 |
+
"learning_rate": 7.447835565244473e-06,
|
| 628 |
+
"loss": 2.7214,
|
| 629 |
"step": 8200
|
| 630 |
},
|
| 631 |
{
|
| 632 |
+
"epoch": 0.52,
|
| 633 |
+
"learning_rate": 7.4166926191217695e-06,
|
| 634 |
+
"loss": 2.7306,
|
| 635 |
"step": 8300
|
| 636 |
},
|
| 637 |
{
|
| 638 |
+
"epoch": 0.52,
|
| 639 |
+
"learning_rate": 7.3855496729990665e-06,
|
| 640 |
+
"loss": 2.7578,
|
| 641 |
"step": 8400
|
| 642 |
},
|
| 643 |
{
|
| 644 |
+
"epoch": 0.53,
|
| 645 |
+
"learning_rate": 7.354406726876363e-06,
|
| 646 |
+
"loss": 2.6296,
|
| 647 |
"step": 8500
|
| 648 |
},
|
| 649 |
{
|
| 650 |
+
"epoch": 0.53,
|
| 651 |
+
"eval_loss": NaN,
|
| 652 |
+
"eval_runtime": 1242.0617,
|
| 653 |
+
"eval_samples_per_second": 183.842,
|
| 654 |
+
"eval_steps_per_second": 22.98,
|
| 655 |
+
"step": 8500
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.54,
|
| 659 |
+
"learning_rate": 7.32326378075366e-06,
|
| 660 |
+
"loss": 2.7712,
|
| 661 |
"step": 8600
|
| 662 |
},
|
| 663 |
{
|
| 664 |
+
"epoch": 0.54,
|
| 665 |
+
"learning_rate": 7.2921208346309565e-06,
|
| 666 |
+
"loss": 2.9979,
|
| 667 |
"step": 8700
|
| 668 |
},
|
| 669 |
{
|
| 670 |
+
"epoch": 0.55,
|
| 671 |
+
"learning_rate": 7.2609778885082535e-06,
|
| 672 |
+
"loss": 2.8045,
|
| 673 |
"step": 8800
|
| 674 |
},
|
| 675 |
{
|
| 676 |
+
"epoch": 0.55,
|
| 677 |
+
"learning_rate": 7.2298349423855505e-06,
|
| 678 |
+
"loss": 3.1612,
|
| 679 |
"step": 8900
|
| 680 |
},
|
| 681 |
{
|
| 682 |
+
"epoch": 0.56,
|
| 683 |
+
"learning_rate": 7.198691996262847e-06,
|
| 684 |
+
"loss": 2.8809,
|
| 685 |
"step": 9000
|
| 686 |
},
|
| 687 |
{
|
| 688 |
+
"epoch": 0.56,
|
| 689 |
+
"eval_loss": NaN,
|
| 690 |
+
"eval_runtime": 1242.2163,
|
| 691 |
+
"eval_samples_per_second": 183.819,
|
| 692 |
+
"eval_steps_per_second": 22.977,
|
| 693 |
+
"step": 9000
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.57,
|
| 697 |
+
"learning_rate": 7.1675490501401435e-06,
|
| 698 |
+
"loss": 2.8204,
|
| 699 |
"step": 9100
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 0.57,
|
| 703 |
+
"learning_rate": 7.1364061040174405e-06,
|
| 704 |
+
"loss": 2.508,
|
| 705 |
"step": 9200
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 0.58,
|
| 709 |
+
"learning_rate": 7.1052631578947375e-06,
|
| 710 |
+
"loss": 2.3787,
|
| 711 |
"step": 9300
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 0.59,
|
| 715 |
+
"learning_rate": 7.0741202117720344e-06,
|
| 716 |
+
"loss": 2.5762,
|
| 717 |
"step": 9400
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 0.59,
|
| 721 |
+
"learning_rate": 7.042977265649331e-06,
|
| 722 |
+
"loss": 2.9636,
|
| 723 |
"step": 9500
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 0.59,
|
| 727 |
+
"eval_loss": NaN,
|
| 728 |
+
"eval_runtime": 1242.6873,
|
| 729 |
+
"eval_samples_per_second": 183.749,
|
| 730 |
+
"eval_steps_per_second": 22.969,
|
| 731 |
+
"step": 9500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.6,
|
| 735 |
+
"learning_rate": 7.011834319526628e-06,
|
| 736 |
+
"loss": 2.6669,
|
| 737 |
"step": 9600
|
| 738 |
},
|
| 739 |
{
|
| 740 |
+
"epoch": 0.6,
|
| 741 |
+
"learning_rate": 6.980691373403925e-06,
|
| 742 |
+
"loss": 2.9417,
|
| 743 |
"step": 9700
|
| 744 |
},
|
| 745 |
{
|
| 746 |
+
"epoch": 0.61,
|
| 747 |
+
"learning_rate": 6.949548427281222e-06,
|
| 748 |
+
"loss": 3.1507,
|
| 749 |
"step": 9800
|
| 750 |
},
|
| 751 |
{
|
| 752 |
+
"epoch": 0.62,
|
| 753 |
+
"learning_rate": 6.9184054811585175e-06,
|
| 754 |
+
"loss": 3.0434,
|
| 755 |
"step": 9900
|
| 756 |
},
|
| 757 |
{
|
| 758 |
+
"epoch": 0.62,
|
| 759 |
+
"learning_rate": 6.8872625350358145e-06,
|
| 760 |
+
"loss": 3.0385,
|
| 761 |
"step": 10000
|
| 762 |
},
|
| 763 |
{
|
| 764 |
+
"epoch": 0.62,
|
| 765 |
+
"eval_loss": NaN,
|
| 766 |
+
"eval_runtime": 1242.5564,
|
| 767 |
+
"eval_samples_per_second": 183.769,
|
| 768 |
+
"eval_steps_per_second": 22.971,
|
| 769 |
+
"step": 10000
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.63,
|
| 773 |
+
"learning_rate": 6.8561195889131115e-06,
|
| 774 |
+
"loss": 2.6385,
|
| 775 |
"step": 10100
|
| 776 |
},
|
| 777 |
{
|
| 778 |
+
"epoch": 0.64,
|
| 779 |
+
"learning_rate": 6.8249766427904084e-06,
|
| 780 |
+
"loss": 2.5347,
|
| 781 |
"step": 10200
|
| 782 |
},
|
| 783 |
{
|
| 784 |
+
"epoch": 0.64,
|
| 785 |
+
"learning_rate": 6.793833696667705e-06,
|
| 786 |
+
"loss": 3.0458,
|
| 787 |
"step": 10300
|
| 788 |
},
|
| 789 |
{
|
| 790 |
+
"epoch": 0.65,
|
| 791 |
+
"learning_rate": 6.763002180006229e-06,
|
| 792 |
+
"loss": 3.0999,
|
| 793 |
"step": 10400
|
| 794 |
},
|
| 795 |
{
|
| 796 |
+
"epoch": 0.65,
|
| 797 |
+
"learning_rate": 6.731859233883526e-06,
|
| 798 |
+
"loss": 2.5865,
|
| 799 |
"step": 10500
|
| 800 |
},
|
| 801 |
{
|
| 802 |
+
"epoch": 0.65,
|
| 803 |
+
"eval_loss": NaN,
|
| 804 |
+
"eval_runtime": 1242.285,
|
| 805 |
+
"eval_samples_per_second": 183.809,
|
| 806 |
+
"eval_steps_per_second": 22.976,
|
| 807 |
+
"step": 10500
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.66,
|
| 811 |
+
"learning_rate": 6.700716287760822e-06,
|
| 812 |
+
"loss": 2.7499,
|
| 813 |
"step": 10600
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 0.67,
|
| 817 |
+
"learning_rate": 6.669573341638119e-06,
|
| 818 |
+
"loss": 2.9551,
|
| 819 |
"step": 10700
|
| 820 |
},
|
| 821 |
{
|
| 822 |
+
"epoch": 0.67,
|
| 823 |
+
"learning_rate": 6.638430395515416e-06,
|
| 824 |
+
"loss": 2.8254,
|
| 825 |
"step": 10800
|
| 826 |
},
|
| 827 |
{
|
| 828 |
+
"epoch": 0.68,
|
| 829 |
+
"learning_rate": 6.607287449392713e-06,
|
| 830 |
+
"loss": 2.8981,
|
| 831 |
"step": 10900
|
| 832 |
},
|
| 833 |
{
|
| 834 |
+
"epoch": 0.69,
|
| 835 |
+
"learning_rate": 6.57614450327001e-06,
|
| 836 |
+
"loss": 2.7491,
|
| 837 |
"step": 11000
|
| 838 |
},
|
| 839 |
{
|
| 840 |
+
"epoch": 0.69,
|
| 841 |
+
"eval_loss": NaN,
|
| 842 |
+
"eval_runtime": 1242.9348,
|
| 843 |
+
"eval_samples_per_second": 183.713,
|
| 844 |
+
"eval_steps_per_second": 22.964,
|
| 845 |
+
"step": 11000
|
| 846 |
+
},
|
| 847 |
{
|
| 848 |
+
"epoch": 0.69,
|
| 849 |
+
"learning_rate": 6.545001557147307e-06,
|
| 850 |
+
"loss": 2.8125,
|
| 851 |
+
"step": 11100
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.7,
|
| 855 |
+
"learning_rate": 6.513858611024604e-06,
|
| 856 |
+
"loss": 2.6231,
|
| 857 |
"step": 11200
|
| 858 |
},
|
| 859 |
{
|
| 860 |
+
"epoch": 0.7,
|
| 861 |
+
"learning_rate": 6.482715664901901e-06,
|
| 862 |
+
"loss": 2.8729,
|
| 863 |
"step": 11300
|
| 864 |
},
|
| 865 |
{
|
| 866 |
+
"epoch": 0.71,
|
| 867 |
+
"learning_rate": 6.451572718779196e-06,
|
| 868 |
+
"loss": 2.9655,
|
| 869 |
"step": 11400
|
| 870 |
},
|
| 871 |
{
|
| 872 |
+
"epoch": 0.72,
|
| 873 |
+
"learning_rate": 6.420429772656493e-06,
|
| 874 |
+
"loss": 3.1339,
|
| 875 |
"step": 11500
|
| 876 |
},
|
| 877 |
{
|
| 878 |
+
"epoch": 0.72,
|
| 879 |
+
"eval_loss": NaN,
|
| 880 |
+
"eval_runtime": 1242.7737,
|
| 881 |
+
"eval_samples_per_second": 183.737,
|
| 882 |
+
"eval_steps_per_second": 22.967,
|
| 883 |
+
"step": 11500
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 0.72,
|
| 887 |
+
"learning_rate": 6.38928682653379e-06,
|
| 888 |
+
"loss": 3.2655,
|
| 889 |
"step": 11600
|
| 890 |
},
|
| 891 |
{
|
| 892 |
+
"epoch": 0.73,
|
| 893 |
+
"learning_rate": 6.358143880411087e-06,
|
| 894 |
+
"loss": 2.7888,
|
| 895 |
"step": 11700
|
| 896 |
},
|
| 897 |
{
|
| 898 |
+
"epoch": 0.73,
|
| 899 |
+
"learning_rate": 6.327000934288384e-06,
|
| 900 |
+
"loss": 3.7292,
|
| 901 |
"step": 11800
|
| 902 |
},
|
| 903 |
{
|
| 904 |
+
"epoch": 0.74,
|
| 905 |
+
"learning_rate": 6.295857988165681e-06,
|
| 906 |
+
"loss": 2.6393,
|
| 907 |
"step": 11900
|
| 908 |
},
|
| 909 |
{
|
| 910 |
+
"epoch": 0.75,
|
| 911 |
+
"learning_rate": 6.264715042042978e-06,
|
| 912 |
+
"loss": 2.9632,
|
| 913 |
+
"step": 12000
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.75,
|
| 917 |
+
"eval_loss": NaN,
|
| 918 |
+
"eval_runtime": 1242.1518,
|
| 919 |
+
"eval_samples_per_second": 183.829,
|
| 920 |
+
"eval_steps_per_second": 22.979,
|
| 921 |
"step": 12000
|
| 922 |
},
|
| 923 |
{
|
| 924 |
+
"epoch": 0.75,
|
| 925 |
+
"learning_rate": 6.233572095920275e-06,
|
| 926 |
+
"loss": 2.4857,
|
| 927 |
"step": 12100
|
| 928 |
},
|
| 929 |
{
|
| 930 |
+
"epoch": 0.76,
|
| 931 |
+
"learning_rate": 6.202429149797571e-06,
|
| 932 |
+
"loss": 2.9144,
|
| 933 |
"step": 12200
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 0.77,
|
| 937 |
+
"learning_rate": 6.171286203674868e-06,
|
| 938 |
+
"loss": 2.7133,
|
| 939 |
"step": 12300
|
| 940 |
},
|
| 941 |
{
|
| 942 |
+
"epoch": 0.77,
|
| 943 |
+
"learning_rate": 6.140143257552165e-06,
|
| 944 |
+
"loss": 3.8753,
|
| 945 |
"step": 12400
|
| 946 |
},
|
| 947 |
{
|
| 948 |
+
"epoch": 0.78,
|
| 949 |
+
"learning_rate": 6.109000311429461e-06,
|
| 950 |
+
"loss": 3.0089,
|
| 951 |
+
"step": 12500
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.78,
|
| 955 |
+
"eval_loss": NaN,
|
| 956 |
+
"eval_runtime": 1242.2159,
|
| 957 |
+
"eval_samples_per_second": 183.819,
|
| 958 |
+
"eval_steps_per_second": 22.977,
|
| 959 |
"step": 12500
|
| 960 |
},
|
| 961 |
{
|
| 962 |
+
"epoch": 0.78,
|
| 963 |
+
"learning_rate": 6.077857365306758e-06,
|
| 964 |
+
"loss": 2.9934,
|
| 965 |
"step": 12600
|
| 966 |
},
|
| 967 |
{
|
| 968 |
+
"epoch": 0.79,
|
| 969 |
+
"learning_rate": 6.047025848645283e-06,
|
| 970 |
+
"loss": 2.5078,
|
| 971 |
"step": 12700
|
| 972 |
},
|
| 973 |
{
|
| 974 |
+
"epoch": 0.8,
|
| 975 |
+
"learning_rate": 6.01588290252258e-06,
|
| 976 |
+
"loss": 2.7188,
|
| 977 |
"step": 12800
|
| 978 |
},
|
| 979 |
{
|
| 980 |
+
"epoch": 0.8,
|
| 981 |
+
"learning_rate": 5.984739956399875e-06,
|
| 982 |
+
"loss": 2.6552,
|
| 983 |
"step": 12900
|
| 984 |
},
|
| 985 |
{
|
| 986 |
+
"epoch": 0.81,
|
| 987 |
+
"learning_rate": 5.953597010277172e-06,
|
| 988 |
+
"loss": 2.3908,
|
| 989 |
+
"step": 13000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.81,
|
| 993 |
+
"eval_loss": NaN,
|
| 994 |
+
"eval_runtime": 1242.3282,
|
| 995 |
+
"eval_samples_per_second": 183.802,
|
| 996 |
+
"eval_steps_per_second": 22.975,
|
| 997 |
"step": 13000
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.82,
|
| 1001 |
+
"learning_rate": 5.922454064154469e-06,
|
| 1002 |
+
"loss": 2.7332,
|
| 1003 |
+
"step": 13100
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.82,
|
| 1007 |
+
"learning_rate": 5.891311118031766e-06,
|
| 1008 |
+
"loss": 2.3849,
|
| 1009 |
+
"step": 13200
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 0.83,
|
| 1013 |
+
"learning_rate": 5.860168171909063e-06,
|
| 1014 |
+
"loss": 3.4628,
|
| 1015 |
+
"step": 13300
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.83,
|
| 1019 |
+
"learning_rate": 5.82902522578636e-06,
|
| 1020 |
+
"loss": 2.6728,
|
| 1021 |
+
"step": 13400
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.84,
|
| 1025 |
+
"learning_rate": 5.797882279663657e-06,
|
| 1026 |
+
"loss": 2.8807,
|
| 1027 |
+
"step": 13500
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.84,
|
| 1031 |
+
"eval_loss": NaN,
|
| 1032 |
+
"eval_runtime": 1241.8007,
|
| 1033 |
+
"eval_samples_per_second": 183.881,
|
| 1034 |
+
"eval_steps_per_second": 22.985,
|
| 1035 |
+
"step": 13500
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.85,
|
| 1039 |
+
"learning_rate": 5.766739333540954e-06,
|
| 1040 |
+
"loss": 2.9648,
|
| 1041 |
+
"step": 13600
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.85,
|
| 1045 |
+
"learning_rate": 5.735596387418251e-06,
|
| 1046 |
+
"loss": 2.8685,
|
| 1047 |
+
"step": 13700
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.86,
|
| 1051 |
+
"learning_rate": 5.704453441295547e-06,
|
| 1052 |
+
"loss": 2.7226,
|
| 1053 |
+
"step": 13800
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.87,
|
| 1057 |
+
"learning_rate": 5.673310495172844e-06,
|
| 1058 |
+
"loss": 2.7493,
|
| 1059 |
+
"step": 13900
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.87,
|
| 1063 |
+
"learning_rate": 5.642167549050141e-06,
|
| 1064 |
+
"loss": 2.4787,
|
| 1065 |
+
"step": 14000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.87,
|
| 1069 |
+
"eval_loss": NaN,
|
| 1070 |
+
"eval_runtime": 1242.1112,
|
| 1071 |
+
"eval_samples_per_second": 183.835,
|
| 1072 |
+
"eval_steps_per_second": 22.979,
|
| 1073 |
+
"step": 14000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.88,
|
| 1077 |
+
"learning_rate": 5.611024602927437e-06,
|
| 1078 |
+
"loss": 2.6621,
|
| 1079 |
+
"step": 14100
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.88,
|
| 1083 |
+
"learning_rate": 5.580193086265962e-06,
|
| 1084 |
+
"loss": 2.5317,
|
| 1085 |
+
"step": 14200
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 0.89,
|
| 1089 |
+
"learning_rate": 5.549050140143259e-06,
|
| 1090 |
+
"loss": 2.8094,
|
| 1091 |
+
"step": 14300
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9,
|
| 1095 |
+
"learning_rate": 5.5179071940205556e-06,
|
| 1096 |
+
"loss": 2.9184,
|
| 1097 |
+
"step": 14400
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.9,
|
| 1101 |
+
"learning_rate": 5.486764247897851e-06,
|
| 1102 |
+
"loss": 3.0641,
|
| 1103 |
+
"step": 14500
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.9,
|
| 1107 |
+
"eval_loss": NaN,
|
| 1108 |
+
"eval_runtime": 1242.2939,
|
| 1109 |
+
"eval_samples_per_second": 183.808,
|
| 1110 |
+
"eval_steps_per_second": 22.976,
|
| 1111 |
+
"step": 14500
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.91,
|
| 1115 |
+
"learning_rate": 5.455621301775148e-06,
|
| 1116 |
+
"loss": 2.707,
|
| 1117 |
+
"step": 14600
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.92,
|
| 1121 |
+
"learning_rate": 5.424478355652445e-06,
|
| 1122 |
+
"loss": 2.8484,
|
| 1123 |
+
"step": 14700
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.92,
|
| 1127 |
+
"learning_rate": 5.393335409529742e-06,
|
| 1128 |
+
"loss": 2.7637,
|
| 1129 |
+
"step": 14800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.93,
|
| 1133 |
+
"learning_rate": 5.362192463407039e-06,
|
| 1134 |
+
"loss": 3.2172,
|
| 1135 |
+
"step": 14900
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 0.93,
|
| 1139 |
+
"learning_rate": 5.331049517284336e-06,
|
| 1140 |
+
"loss": 3.8572,
|
| 1141 |
+
"step": 15000
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 0.93,
|
| 1145 |
+
"eval_loss": NaN,
|
| 1146 |
+
"eval_runtime": 1242.3426,
|
| 1147 |
+
"eval_samples_per_second": 183.8,
|
| 1148 |
+
"eval_steps_per_second": 22.975,
|
| 1149 |
+
"step": 15000
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.94,
|
| 1153 |
+
"learning_rate": 5.299906571161633e-06,
|
| 1154 |
+
"loss": 2.6719,
|
| 1155 |
+
"step": 15100
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.95,
|
| 1159 |
+
"learning_rate": 5.2687636250389296e-06,
|
| 1160 |
+
"loss": 2.6299,
|
| 1161 |
+
"step": 15200
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.95,
|
| 1165 |
+
"learning_rate": 5.237620678916226e-06,
|
| 1166 |
+
"loss": 2.7315,
|
| 1167 |
+
"step": 15300
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.96,
|
| 1171 |
+
"learning_rate": 5.206477732793523e-06,
|
| 1172 |
+
"loss": 2.5421,
|
| 1173 |
+
"step": 15400
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.97,
|
| 1177 |
+
"learning_rate": 5.17533478667082e-06,
|
| 1178 |
+
"loss": 2.6612,
|
| 1179 |
+
"step": 15500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.97,
|
| 1183 |
+
"eval_loss": NaN,
|
| 1184 |
+
"eval_runtime": 1242.7472,
|
| 1185 |
+
"eval_samples_per_second": 183.741,
|
| 1186 |
+
"eval_steps_per_second": 22.968,
|
| 1187 |
+
"step": 15500
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.97,
|
| 1191 |
+
"learning_rate": 5.1441918405481166e-06,
|
| 1192 |
+
"loss": 2.4232,
|
| 1193 |
+
"step": 15600
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.98,
|
| 1197 |
+
"learning_rate": 5.113048894425413e-06,
|
| 1198 |
+
"loss": 2.4607,
|
| 1199 |
+
"step": 15700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.98,
|
| 1203 |
+
"learning_rate": 5.08190594830271e-06,
|
| 1204 |
+
"loss": 2.5239,
|
| 1205 |
+
"step": 15800
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 0.99,
|
| 1209 |
+
"learning_rate": 5.050763002180007e-06,
|
| 1210 |
+
"loss": 2.7479,
|
| 1211 |
+
"step": 15900
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 1.0,
|
| 1215 |
+
"learning_rate": 5.0196200560573036e-06,
|
| 1216 |
+
"loss": 2.3672,
|
| 1217 |
+
"step": 16000
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.0,
|
| 1221 |
+
"eval_loss": NaN,
|
| 1222 |
+
"eval_runtime": 1242.0429,
|
| 1223 |
+
"eval_samples_per_second": 183.845,
|
| 1224 |
+
"eval_steps_per_second": 22.981,
|
| 1225 |
+
"step": 16000
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 1.0,
|
| 1229 |
+
"learning_rate": 4.9884771099346005e-06,
|
| 1230 |
+
"loss": 2.8632,
|
| 1231 |
+
"step": 16100
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 1.01,
|
| 1235 |
+
"learning_rate": 4.9573341638118975e-06,
|
| 1236 |
+
"loss": 2.6653,
|
| 1237 |
+
"step": 16200
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 1.02,
|
| 1241 |
+
"learning_rate": 4.9261912176891945e-06,
|
| 1242 |
+
"loss": 2.539,
|
| 1243 |
+
"step": 16300
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.02,
|
| 1247 |
+
"learning_rate": 4.8950482715664906e-06,
|
| 1248 |
+
"loss": 2.8172,
|
| 1249 |
+
"step": 16400
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.03,
|
| 1253 |
+
"learning_rate": 4.8639053254437875e-06,
|
| 1254 |
+
"loss": 2.5157,
|
| 1255 |
+
"step": 16500
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.03,
|
| 1259 |
+
"eval_loss": NaN,
|
| 1260 |
+
"eval_runtime": 11659.1974,
|
| 1261 |
+
"eval_samples_per_second": 19.585,
|
| 1262 |
+
"eval_steps_per_second": 2.448,
|
| 1263 |
+
"step": 16500
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.03,
|
| 1267 |
+
"learning_rate": 4.8327623793210845e-06,
|
| 1268 |
+
"loss": 2.5209,
|
| 1269 |
+
"step": 16600
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.04,
|
| 1273 |
+
"learning_rate": 4.8016194331983815e-06,
|
| 1274 |
+
"loss": 2.9114,
|
| 1275 |
+
"step": 16700
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 1.05,
|
| 1279 |
+
"learning_rate": 4.7704764870756776e-06,
|
| 1280 |
+
"loss": 2.6053,
|
| 1281 |
+
"step": 16800
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 1.05,
|
| 1285 |
+
"learning_rate": 4.7393335409529745e-06,
|
| 1286 |
+
"loss": 2.8304,
|
| 1287 |
+
"step": 16900
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.06,
|
| 1291 |
+
"learning_rate": 4.7081905948302715e-06,
|
| 1292 |
+
"loss": 2.7648,
|
| 1293 |
+
"step": 17000
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 1.06,
|
| 1297 |
+
"eval_loss": NaN,
|
| 1298 |
+
"eval_runtime": 1241.8221,
|
| 1299 |
+
"eval_samples_per_second": 183.877,
|
| 1300 |
+
"eval_steps_per_second": 22.985,
|
| 1301 |
+
"step": 17000
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.07,
|
| 1305 |
+
"learning_rate": 4.6770476487075685e-06,
|
| 1306 |
+
"loss": 2.9948,
|
| 1307 |
+
"step": 17100
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 1.07,
|
| 1311 |
+
"learning_rate": 4.6459047025848646e-06,
|
| 1312 |
+
"loss": 2.7007,
|
| 1313 |
+
"step": 17200
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.08,
|
| 1317 |
+
"learning_rate": 4.6147617564621615e-06,
|
| 1318 |
+
"loss": 3.0049,
|
| 1319 |
+
"step": 17300
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.08,
|
| 1323 |
+
"learning_rate": 4.5836188103394585e-06,
|
| 1324 |
+
"loss": 2.4344,
|
| 1325 |
+
"step": 17400
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.09,
|
| 1329 |
+
"learning_rate": 4.5524758642167555e-06,
|
| 1330 |
+
"loss": 2.474,
|
| 1331 |
+
"step": 17500
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.09,
|
| 1335 |
+
"eval_loss": NaN,
|
| 1336 |
+
"eval_runtime": 1242.354,
|
| 1337 |
+
"eval_samples_per_second": 183.799,
|
| 1338 |
+
"eval_steps_per_second": 22.975,
|
| 1339 |
+
"step": 17500
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.1,
|
| 1343 |
+
"learning_rate": 4.5213329180940516e-06,
|
| 1344 |
+
"loss": 2.653,
|
| 1345 |
+
"step": 17600
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 1.1,
|
| 1349 |
+
"learning_rate": 4.4901899719713485e-06,
|
| 1350 |
+
"loss": 2.7398,
|
| 1351 |
+
"step": 17700
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 1.11,
|
| 1355 |
+
"learning_rate": 4.4590470258486455e-06,
|
| 1356 |
+
"loss": 2.3064,
|
| 1357 |
+
"step": 17800
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 1.11,
|
| 1361 |
+
"learning_rate": 4.4279040797259425e-06,
|
| 1362 |
+
"loss": 2.8268,
|
| 1363 |
+
"step": 17900
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 1.12,
|
| 1367 |
+
"learning_rate": 4.3967611336032386e-06,
|
| 1368 |
+
"loss": 2.3197,
|
| 1369 |
+
"step": 18000
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.12,
|
| 1373 |
+
"eval_loss": NaN,
|
| 1374 |
+
"eval_runtime": 1244.3942,
|
| 1375 |
+
"eval_samples_per_second": 183.497,
|
| 1376 |
+
"eval_steps_per_second": 22.937,
|
| 1377 |
+
"step": 18000
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.13,
|
| 1381 |
+
"learning_rate": 4.3656181874805355e-06,
|
| 1382 |
+
"loss": 2.6969,
|
| 1383 |
+
"step": 18100
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.13,
|
| 1387 |
+
"learning_rate": 4.33478667081906e-06,
|
| 1388 |
+
"loss": 2.6233,
|
| 1389 |
+
"step": 18200
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.14,
|
| 1393 |
+
"learning_rate": 4.303643724696356e-06,
|
| 1394 |
+
"loss": 2.9522,
|
| 1395 |
+
"step": 18300
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.15,
|
| 1399 |
+
"learning_rate": 4.272500778573653e-06,
|
| 1400 |
+
"loss": 2.5308,
|
| 1401 |
+
"step": 18400
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 1.15,
|
| 1405 |
+
"learning_rate": 4.24135783245095e-06,
|
| 1406 |
+
"loss": 2.9766,
|
| 1407 |
+
"step": 18500
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 1.15,
|
| 1411 |
+
"eval_loss": NaN,
|
| 1412 |
+
"eval_runtime": 1242.8298,
|
| 1413 |
+
"eval_samples_per_second": 183.728,
|
| 1414 |
+
"eval_steps_per_second": 22.966,
|
| 1415 |
+
"step": 18500
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 1.16,
|
| 1419 |
+
"learning_rate": 4.210214886328247e-06,
|
| 1420 |
+
"loss": 2.4952,
|
| 1421 |
+
"step": 18600
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 1.16,
|
| 1425 |
+
"learning_rate": 4.179071940205543e-06,
|
| 1426 |
+
"loss": 3.0581,
|
| 1427 |
+
"step": 18700
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.17,
|
| 1431 |
+
"learning_rate": 4.14792899408284e-06,
|
| 1432 |
+
"loss": 2.3722,
|
| 1433 |
+
"step": 18800
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 1.18,
|
| 1437 |
+
"learning_rate": 4.116786047960137e-06,
|
| 1438 |
+
"loss": 2.6995,
|
| 1439 |
+
"step": 18900
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.18,
|
| 1443 |
+
"learning_rate": 4.085643101837434e-06,
|
| 1444 |
+
"loss": 2.827,
|
| 1445 |
+
"step": 19000
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.18,
|
| 1449 |
+
"eval_loss": NaN,
|
| 1450 |
+
"eval_runtime": 1242.4071,
|
| 1451 |
+
"eval_samples_per_second": 183.791,
|
| 1452 |
+
"eval_steps_per_second": 22.974,
|
| 1453 |
+
"step": 19000
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.19,
|
| 1457 |
+
"learning_rate": 4.054500155714731e-06,
|
| 1458 |
+
"loss": 2.7205,
|
| 1459 |
+
"step": 19100
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.2,
|
| 1463 |
+
"learning_rate": 4.023357209592027e-06,
|
| 1464 |
+
"loss": 2.9195,
|
| 1465 |
+
"step": 19200
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.2,
|
| 1469 |
+
"learning_rate": 3.992214263469324e-06,
|
| 1470 |
+
"loss": 2.8558,
|
| 1471 |
+
"step": 19300
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 1.21,
|
| 1475 |
+
"learning_rate": 3.961071317346621e-06,
|
| 1476 |
+
"loss": 3.2538,
|
| 1477 |
+
"step": 19400
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 1.21,
|
| 1481 |
+
"learning_rate": 3.929928371223918e-06,
|
| 1482 |
+
"loss": 2.7213,
|
| 1483 |
+
"step": 19500
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 1.21,
|
| 1487 |
+
"eval_loss": NaN,
|
| 1488 |
+
"eval_runtime": 1242.5138,
|
| 1489 |
+
"eval_samples_per_second": 183.775,
|
| 1490 |
+
"eval_steps_per_second": 22.972,
|
| 1491 |
+
"step": 19500
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1.22,
|
| 1495 |
+
"learning_rate": 3.898785425101214e-06,
|
| 1496 |
+
"loss": 2.7368,
|
| 1497 |
+
"step": 19600
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.23,
|
| 1501 |
+
"learning_rate": 3.867642478978511e-06,
|
| 1502 |
+
"loss": 2.6442,
|
| 1503 |
+
"step": 19700
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.23,
|
| 1507 |
+
"learning_rate": 3.836499532855808e-06,
|
| 1508 |
+
"loss": 2.6891,
|
| 1509 |
+
"step": 19800
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.24,
|
| 1513 |
+
"learning_rate": 3.8053565867331056e-06,
|
| 1514 |
+
"loss": 2.6657,
|
| 1515 |
+
"step": 19900
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.25,
|
| 1519 |
+
"learning_rate": 3.7742136406104017e-06,
|
| 1520 |
+
"loss": 2.6653,
|
| 1521 |
+
"step": 20000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.25,
|
| 1525 |
+
"eval_loss": NaN,
|
| 1526 |
+
"eval_runtime": 1243.214,
|
| 1527 |
+
"eval_samples_per_second": 183.672,
|
| 1528 |
+
"eval_steps_per_second": 22.959,
|
| 1529 |
+
"step": 20000
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.25,
|
| 1533 |
+
"learning_rate": 3.7430706944876987e-06,
|
| 1534 |
+
"loss": 3.0333,
|
| 1535 |
+
"step": 20100
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.26,
|
| 1539 |
+
"learning_rate": 3.7119277483649957e-06,
|
| 1540 |
+
"loss": 2.2654,
|
| 1541 |
+
"step": 20200
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 1.26,
|
| 1545 |
+
"learning_rate": 3.6807848022422926e-06,
|
| 1546 |
+
"loss": 2.8074,
|
| 1547 |
+
"step": 20300
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1.27,
|
| 1551 |
+
"learning_rate": 3.649641856119589e-06,
|
| 1552 |
+
"loss": 2.5416,
|
| 1553 |
+
"step": 20400
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 1.28,
|
| 1557 |
+
"learning_rate": 3.618498909996886e-06,
|
| 1558 |
+
"loss": 2.671,
|
| 1559 |
+
"step": 20500
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 1.28,
|
| 1563 |
+
"eval_loss": NaN,
|
| 1564 |
+
"eval_runtime": 1242.2133,
|
| 1565 |
+
"eval_samples_per_second": 183.819,
|
| 1566 |
+
"eval_steps_per_second": 22.978,
|
| 1567 |
+
"step": 20500
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 1.28,
|
| 1571 |
+
"learning_rate": 3.5873559638741827e-06,
|
| 1572 |
+
"loss": 2.3197,
|
| 1573 |
+
"step": 20600
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 1.29,
|
| 1577 |
+
"learning_rate": 3.5562130177514796e-06,
|
| 1578 |
+
"loss": 2.6825,
|
| 1579 |
+
"step": 20700
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 1.3,
|
| 1583 |
+
"learning_rate": 3.525070071628776e-06,
|
| 1584 |
+
"loss": 2.9642,
|
| 1585 |
+
"step": 20800
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.3,
|
| 1589 |
+
"learning_rate": 3.493927125506073e-06,
|
| 1590 |
+
"loss": 2.327,
|
| 1591 |
+
"step": 20900
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.31,
|
| 1595 |
+
"learning_rate": 3.463095608844597e-06,
|
| 1596 |
+
"loss": 2.5401,
|
| 1597 |
+
"step": 21000
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 1.31,
|
| 1601 |
+
"eval_loss": NaN,
|
| 1602 |
+
"eval_runtime": 1242.9072,
|
| 1603 |
+
"eval_samples_per_second": 183.717,
|
| 1604 |
+
"eval_steps_per_second": 22.965,
|
| 1605 |
+
"step": 21000
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.31,
|
| 1609 |
+
"learning_rate": 3.4319526627218935e-06,
|
| 1610 |
+
"loss": 2.6376,
|
| 1611 |
+
"step": 21100
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.32,
|
| 1615 |
+
"learning_rate": 3.4008097165991905e-06,
|
| 1616 |
+
"loss": 2.8178,
|
| 1617 |
+
"step": 21200
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.33,
|
| 1621 |
+
"learning_rate": 3.3696667704764874e-06,
|
| 1622 |
+
"loss": 2.2995,
|
| 1623 |
+
"step": 21300
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.33,
|
| 1627 |
+
"learning_rate": 3.3385238243537844e-06,
|
| 1628 |
+
"loss": 3.476,
|
| 1629 |
+
"step": 21400
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.34,
|
| 1633 |
+
"learning_rate": 3.3073808782310805e-06,
|
| 1634 |
+
"loss": 2.7193,
|
| 1635 |
+
"step": 21500
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.34,
|
| 1639 |
+
"eval_loss": NaN,
|
| 1640 |
+
"eval_runtime": 1242.302,
|
| 1641 |
+
"eval_samples_per_second": 183.806,
|
| 1642 |
+
"eval_steps_per_second": 22.976,
|
| 1643 |
+
"step": 21500
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 1.35,
|
| 1647 |
+
"learning_rate": 3.2762379321083775e-06,
|
| 1648 |
+
"loss": 2.2606,
|
| 1649 |
+
"step": 21600
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 1.35,
|
| 1653 |
+
"learning_rate": 3.2450949859856744e-06,
|
| 1654 |
+
"loss": 2.4979,
|
| 1655 |
+
"step": 21700
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.36,
|
| 1659 |
+
"learning_rate": 3.2139520398629714e-06,
|
| 1660 |
+
"loss": 2.8394,
|
| 1661 |
+
"step": 21800
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.36,
|
| 1665 |
+
"learning_rate": 3.1828090937402684e-06,
|
| 1666 |
+
"loss": 2.5935,
|
| 1667 |
+
"step": 21900
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 1.37,
|
| 1671 |
+
"learning_rate": 3.151666147617565e-06,
|
| 1672 |
+
"loss": 2.5924,
|
| 1673 |
+
"step": 22000
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 1.37,
|
| 1677 |
+
"eval_loss": NaN,
|
| 1678 |
+
"eval_runtime": 1239.6995,
|
| 1679 |
+
"eval_samples_per_second": 184.192,
|
| 1680 |
+
"eval_steps_per_second": 23.024,
|
| 1681 |
+
"step": 22000
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 1.38,
|
| 1685 |
+
"learning_rate": 3.120523201494862e-06,
|
| 1686 |
+
"loss": 2.3212,
|
| 1687 |
+
"step": 22100
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 1.38,
|
| 1691 |
+
"learning_rate": 3.0893802553721584e-06,
|
| 1692 |
+
"loss": 2.5372,
|
| 1693 |
+
"step": 22200
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 1.39,
|
| 1697 |
+
"learning_rate": 3.0582373092494554e-06,
|
| 1698 |
+
"loss": 3.17,
|
| 1699 |
+
"step": 22300
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 1.4,
|
| 1703 |
+
"learning_rate": 3.027094363126752e-06,
|
| 1704 |
+
"loss": 2.3103,
|
| 1705 |
+
"step": 22400
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.4,
|
| 1709 |
+
"learning_rate": 2.995951417004049e-06,
|
| 1710 |
+
"loss": 2.5506,
|
| 1711 |
+
"step": 22500
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.4,
|
| 1715 |
+
"eval_loss": NaN,
|
| 1716 |
+
"eval_runtime": 1238.6358,
|
| 1717 |
+
"eval_samples_per_second": 184.35,
|
| 1718 |
+
"eval_steps_per_second": 23.044,
|
| 1719 |
+
"step": 22500
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 1.41,
|
| 1723 |
+
"learning_rate": 2.964808470881346e-06,
|
| 1724 |
+
"loss": 2.3131,
|
| 1725 |
+
"step": 22600
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.41,
|
| 1729 |
+
"learning_rate": 2.9336655247586428e-06,
|
| 1730 |
+
"loss": 2.9797,
|
| 1731 |
+
"step": 22700
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.42,
|
| 1735 |
+
"learning_rate": 2.902522578635939e-06,
|
| 1736 |
+
"loss": 3.3517,
|
| 1737 |
+
"step": 22800
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 1.43,
|
| 1741 |
+
"learning_rate": 2.871379632513236e-06,
|
| 1742 |
+
"loss": 2.3309,
|
| 1743 |
+
"step": 22900
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 1.43,
|
| 1747 |
+
"learning_rate": 2.840236686390533e-06,
|
| 1748 |
+
"loss": 2.9167,
|
| 1749 |
+
"step": 23000
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.43,
|
| 1753 |
+
"eval_loss": NaN,
|
| 1754 |
+
"eval_runtime": 1239.5086,
|
| 1755 |
+
"eval_samples_per_second": 184.221,
|
| 1756 |
+
"eval_steps_per_second": 23.028,
|
| 1757 |
+
"step": 23000
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 1.44,
|
| 1761 |
+
"learning_rate": 2.8090937402678298e-06,
|
| 1762 |
+
"loss": 2.612,
|
| 1763 |
+
"step": 23100
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.44,
|
| 1767 |
+
"learning_rate": 2.7779507941451263e-06,
|
| 1768 |
+
"loss": 2.5327,
|
| 1769 |
+
"step": 23200
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 1.45,
|
| 1773 |
+
"learning_rate": 2.746807848022423e-06,
|
| 1774 |
+
"loss": 2.4244,
|
| 1775 |
+
"step": 23300
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.46,
|
| 1779 |
+
"learning_rate": 2.71566490189972e-06,
|
| 1780 |
+
"loss": 2.6675,
|
| 1781 |
+
"step": 23400
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.46,
|
| 1785 |
+
"learning_rate": 2.6845219557770168e-06,
|
| 1786 |
+
"loss": 2.8272,
|
| 1787 |
+
"step": 23500
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.46,
|
| 1791 |
+
"eval_loss": NaN,
|
| 1792 |
+
"eval_runtime": 1242.2478,
|
| 1793 |
+
"eval_samples_per_second": 183.814,
|
| 1794 |
+
"eval_steps_per_second": 22.977,
|
| 1795 |
+
"step": 23500
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.47,
|
| 1799 |
+
"learning_rate": 2.6533790096543133e-06,
|
| 1800 |
+
"loss": 2.419,
|
| 1801 |
+
"step": 23600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.48,
|
| 1805 |
+
"learning_rate": 2.6222360635316103e-06,
|
| 1806 |
+
"loss": 2.4781,
|
| 1807 |
+
"step": 23700
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 1.48,
|
| 1811 |
+
"learning_rate": 2.5910931174089072e-06,
|
| 1812 |
+
"loss": 3.1415,
|
| 1813 |
+
"step": 23800
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 1.49,
|
| 1817 |
+
"learning_rate": 2.559950171286204e-06,
|
| 1818 |
+
"loss": 2.5226,
|
| 1819 |
+
"step": 23900
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 1.49,
|
| 1823 |
+
"learning_rate": 2.5288072251635003e-06,
|
| 1824 |
+
"loss": 2.4586,
|
| 1825 |
+
"step": 24000
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 1.49,
|
| 1829 |
+
"eval_loss": NaN,
|
| 1830 |
+
"eval_runtime": 1242.6605,
|
| 1831 |
+
"eval_samples_per_second": 183.753,
|
| 1832 |
+
"eval_steps_per_second": 22.969,
|
| 1833 |
+
"step": 24000
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.5,
|
| 1837 |
+
"learning_rate": 2.4976642790407973e-06,
|
| 1838 |
+
"loss": 2.6241,
|
| 1839 |
+
"step": 24100
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.51,
|
| 1843 |
+
"learning_rate": 2.4665213329180942e-06,
|
| 1844 |
+
"loss": 2.9088,
|
| 1845 |
+
"step": 24200
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.51,
|
| 1849 |
+
"learning_rate": 2.435378386795391e-06,
|
| 1850 |
+
"loss": 2.3136,
|
| 1851 |
+
"step": 24300
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.52,
|
| 1855 |
+
"learning_rate": 2.4042354406726877e-06,
|
| 1856 |
+
"loss": 2.5916,
|
| 1857 |
+
"step": 24400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.53,
|
| 1861 |
+
"learning_rate": 2.3730924945499847e-06,
|
| 1862 |
+
"loss": 2.5698,
|
| 1863 |
+
"step": 24500
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 1.53,
|
| 1867 |
+
"eval_loss": NaN,
|
| 1868 |
+
"eval_runtime": 1243.0815,
|
| 1869 |
+
"eval_samples_per_second": 183.691,
|
| 1870 |
+
"eval_steps_per_second": 22.961,
|
| 1871 |
+
"step": 24500
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.53,
|
| 1875 |
+
"learning_rate": 2.3419495484272812e-06,
|
| 1876 |
+
"loss": 2.9015,
|
| 1877 |
+
"step": 24600
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 1.54,
|
| 1881 |
+
"learning_rate": 2.310806602304578e-06,
|
| 1882 |
+
"loss": 3.1771,
|
| 1883 |
+
"step": 24700
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 1.54,
|
| 1887 |
+
"learning_rate": 2.279663656181875e-06,
|
| 1888 |
+
"loss": 3.388,
|
| 1889 |
+
"step": 24800
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 1.55,
|
| 1893 |
+
"learning_rate": 2.2485207100591717e-06,
|
| 1894 |
+
"loss": 2.5991,
|
| 1895 |
+
"step": 24900
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 1.56,
|
| 1899 |
+
"learning_rate": 2.2173777639364687e-06,
|
| 1900 |
+
"loss": 2.6171,
|
| 1901 |
+
"step": 25000
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 1.56,
|
| 1905 |
+
"eval_loss": NaN,
|
| 1906 |
+
"eval_runtime": 1242.7545,
|
| 1907 |
+
"eval_samples_per_second": 183.739,
|
| 1908 |
+
"eval_steps_per_second": 22.968,
|
| 1909 |
+
"step": 25000
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 1.56,
|
| 1913 |
+
"learning_rate": 2.1865462472749925e-06,
|
| 1914 |
+
"loss": 2.506,
|
| 1915 |
+
"step": 25100
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 1.57,
|
| 1919 |
+
"learning_rate": 2.1554033011522895e-06,
|
| 1920 |
+
"loss": 2.6567,
|
| 1921 |
+
"step": 25200
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.58,
|
| 1925 |
+
"learning_rate": 2.124260355029586e-06,
|
| 1926 |
+
"loss": 3.2163,
|
| 1927 |
+
"step": 25300
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.58,
|
| 1931 |
+
"learning_rate": 2.09342883836811e-06,
|
| 1932 |
+
"loss": 2.8622,
|
| 1933 |
+
"step": 25400
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 1.59,
|
| 1937 |
+
"learning_rate": 2.062285892245407e-06,
|
| 1938 |
+
"loss": 3.0785,
|
| 1939 |
+
"step": 25500
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 1.59,
|
| 1943 |
+
"eval_loss": NaN,
|
| 1944 |
+
"eval_runtime": 1241.7925,
|
| 1945 |
+
"eval_samples_per_second": 183.882,
|
| 1946 |
+
"eval_steps_per_second": 22.985,
|
| 1947 |
+
"step": 25500
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.59,
|
| 1951 |
+
"learning_rate": 2.0311429461227034e-06,
|
| 1952 |
+
"loss": 2.735,
|
| 1953 |
+
"step": 25600
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 1.6,
|
| 1957 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 1958 |
+
"loss": 2.61,
|
| 1959 |
+
"step": 25700
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 1.61,
|
| 1963 |
+
"learning_rate": 1.968857053877297e-06,
|
| 1964 |
+
"loss": 2.3942,
|
| 1965 |
+
"step": 25800
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 1.61,
|
| 1969 |
+
"learning_rate": 1.937714107754594e-06,
|
| 1970 |
+
"loss": 2.1723,
|
| 1971 |
+
"step": 25900
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 1.62,
|
| 1975 |
+
"learning_rate": 1.9065711616318906e-06,
|
| 1976 |
+
"loss": 2.856,
|
| 1977 |
+
"step": 26000
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.62,
|
| 1981 |
+
"eval_loss": NaN,
|
| 1982 |
+
"eval_runtime": 1242.4465,
|
| 1983 |
+
"eval_samples_per_second": 183.785,
|
| 1984 |
+
"eval_steps_per_second": 22.973,
|
| 1985 |
+
"step": 26000
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 1.63,
|
| 1989 |
+
"learning_rate": 1.8757396449704144e-06,
|
| 1990 |
+
"loss": 2.9466,
|
| 1991 |
+
"step": 26100
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.63,
|
| 1995 |
+
"learning_rate": 1.8445966988477112e-06,
|
| 1996 |
+
"loss": 3.1076,
|
| 1997 |
+
"step": 26200
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.64,
|
| 2001 |
+
"learning_rate": 1.813453752725008e-06,
|
| 2002 |
+
"loss": 2.2859,
|
| 2003 |
+
"step": 26300
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 1.64,
|
| 2007 |
+
"learning_rate": 1.782310806602305e-06,
|
| 2008 |
+
"loss": 3.0796,
|
| 2009 |
+
"step": 26400
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 1.65,
|
| 2013 |
+
"learning_rate": 1.7511678604796014e-06,
|
| 2014 |
+
"loss": 2.3788,
|
| 2015 |
+
"step": 26500
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 1.65,
|
| 2019 |
+
"eval_loss": NaN,
|
| 2020 |
+
"eval_runtime": 1242.137,
|
| 2021 |
+
"eval_samples_per_second": 183.831,
|
| 2022 |
+
"eval_steps_per_second": 22.979,
|
| 2023 |
+
"step": 26500
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 1.66,
|
| 2027 |
+
"learning_rate": 1.7200249143568984e-06,
|
| 2028 |
+
"loss": 2.7557,
|
| 2029 |
+
"step": 26600
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 1.66,
|
| 2033 |
+
"learning_rate": 1.688881968234195e-06,
|
| 2034 |
+
"loss": 2.328,
|
| 2035 |
+
"step": 26700
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 1.67,
|
| 2039 |
+
"learning_rate": 1.657739022111492e-06,
|
| 2040 |
+
"loss": 2.7651,
|
| 2041 |
+
"step": 26800
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 1.68,
|
| 2045 |
+
"learning_rate": 1.6265960759887886e-06,
|
| 2046 |
+
"loss": 3.1467,
|
| 2047 |
+
"step": 26900
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.68,
|
| 2051 |
+
"learning_rate": 1.5954531298660856e-06,
|
| 2052 |
+
"loss": 2.5629,
|
| 2053 |
+
"step": 27000
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.68,
|
| 2057 |
+
"eval_loss": NaN,
|
| 2058 |
+
"eval_runtime": 1243.6496,
|
| 2059 |
+
"eval_samples_per_second": 183.607,
|
| 2060 |
+
"eval_steps_per_second": 22.951,
|
| 2061 |
+
"step": 27000
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.69,
|
| 2065 |
+
"learning_rate": 1.5643101837433821e-06,
|
| 2066 |
+
"loss": 2.6638,
|
| 2067 |
+
"step": 27100
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.69,
|
| 2071 |
+
"learning_rate": 1.5331672376206791e-06,
|
| 2072 |
+
"loss": 2.7094,
|
| 2073 |
+
"step": 27200
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 1.7,
|
| 2077 |
+
"learning_rate": 1.5020242914979756e-06,
|
| 2078 |
+
"loss": 2.6052,
|
| 2079 |
+
"step": 27300
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.71,
|
| 2083 |
+
"learning_rate": 1.4708813453752726e-06,
|
| 2084 |
+
"loss": 2.6647,
|
| 2085 |
+
"step": 27400
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 1.71,
|
| 2089 |
+
"learning_rate": 1.4397383992525694e-06,
|
| 2090 |
+
"loss": 2.5914,
|
| 2091 |
+
"step": 27500
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 1.71,
|
| 2095 |
+
"eval_loss": NaN,
|
| 2096 |
+
"eval_runtime": 1242.5399,
|
| 2097 |
+
"eval_samples_per_second": 183.771,
|
| 2098 |
+
"eval_steps_per_second": 22.971,
|
| 2099 |
+
"step": 27500
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 1.72,
|
| 2103 |
+
"learning_rate": 1.4085954531298663e-06,
|
| 2104 |
+
"loss": 2.6187,
|
| 2105 |
+
"step": 27600
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 1.73,
|
| 2109 |
+
"learning_rate": 1.3774525070071629e-06,
|
| 2110 |
+
"loss": 2.2137,
|
| 2111 |
+
"step": 27700
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 1.73,
|
| 2115 |
+
"learning_rate": 1.3463095608844598e-06,
|
| 2116 |
+
"loss": 2.5718,
|
| 2117 |
+
"step": 27800
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.74,
|
| 2121 |
+
"learning_rate": 1.3151666147617564e-06,
|
| 2122 |
+
"loss": 2.3102,
|
| 2123 |
+
"step": 27900
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.74,
|
| 2127 |
+
"learning_rate": 1.2840236686390533e-06,
|
| 2128 |
+
"loss": 2.5802,
|
| 2129 |
+
"step": 28000
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 1.74,
|
| 2133 |
+
"eval_loss": NaN,
|
| 2134 |
+
"eval_runtime": 1242.0317,
|
| 2135 |
+
"eval_samples_per_second": 183.846,
|
| 2136 |
+
"eval_steps_per_second": 22.981,
|
| 2137 |
+
"step": 28000
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.75,
|
| 2141 |
+
"learning_rate": 1.25288072251635e-06,
|
| 2142 |
+
"loss": 3.2864,
|
| 2143 |
+
"step": 28100
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 1.76,
|
| 2147 |
+
"learning_rate": 1.221737776393647e-06,
|
| 2148 |
+
"loss": 2.4009,
|
| 2149 |
+
"step": 28200
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 1.76,
|
| 2153 |
+
"learning_rate": 1.1905948302709438e-06,
|
| 2154 |
+
"loss": 2.2293,
|
| 2155 |
+
"step": 28300
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 1.77,
|
| 2159 |
+
"learning_rate": 1.1594518841482405e-06,
|
| 2160 |
+
"loss": 2.5808,
|
| 2161 |
+
"step": 28400
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 1.78,
|
| 2165 |
+
"learning_rate": 1.1283089380255373e-06,
|
| 2166 |
+
"loss": 2.2956,
|
| 2167 |
+
"step": 28500
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 1.78,
|
| 2171 |
+
"eval_loss": NaN,
|
| 2172 |
+
"eval_runtime": 1242.0354,
|
| 2173 |
+
"eval_samples_per_second": 183.846,
|
| 2174 |
+
"eval_steps_per_second": 22.981,
|
| 2175 |
+
"step": 28500
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 1.78,
|
| 2179 |
+
"learning_rate": 1.097165991902834e-06,
|
| 2180 |
+
"loss": 2.6273,
|
| 2181 |
+
"step": 28600
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 1.79,
|
| 2185 |
+
"learning_rate": 1.0660230457801308e-06,
|
| 2186 |
+
"loss": 2.5466,
|
| 2187 |
+
"step": 28700
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.79,
|
| 2191 |
+
"learning_rate": 1.0348800996574275e-06,
|
| 2192 |
+
"loss": 2.7805,
|
| 2193 |
+
"step": 28800
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.8,
|
| 2197 |
+
"learning_rate": 1.0037371535347245e-06,
|
| 2198 |
+
"loss": 2.4265,
|
| 2199 |
+
"step": 28900
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 1.81,
|
| 2203 |
+
"learning_rate": 9.725942074120212e-07,
|
| 2204 |
+
"loss": 2.6872,
|
| 2205 |
+
"step": 29000
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 1.81,
|
| 2209 |
+
"eval_loss": NaN,
|
| 2210 |
+
"eval_runtime": 1241.1816,
|
| 2211 |
+
"eval_samples_per_second": 183.972,
|
| 2212 |
+
"eval_steps_per_second": 22.997,
|
| 2213 |
+
"step": 29000
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 1.81,
|
| 2217 |
+
"learning_rate": 9.41451261289318e-07,
|
| 2218 |
+
"loss": 2.8077,
|
| 2219 |
+
"step": 29100
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 1.82,
|
| 2223 |
+
"learning_rate": 9.103083151666147e-07,
|
| 2224 |
+
"loss": 2.7051,
|
| 2225 |
+
"step": 29200
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 1.82,
|
| 2229 |
+
"learning_rate": 8.791653690439116e-07,
|
| 2230 |
+
"loss": 2.6449,
|
| 2231 |
+
"step": 29300
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 1.83,
|
| 2235 |
+
"learning_rate": 8.480224229212085e-07,
|
| 2236 |
+
"loss": 2.2203,
|
| 2237 |
+
"step": 29400
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 1.84,
|
| 2241 |
+
"learning_rate": 8.168794767985053e-07,
|
| 2242 |
+
"loss": 2.7376,
|
| 2243 |
+
"step": 29500
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 1.84,
|
| 2247 |
+
"eval_loss": NaN,
|
| 2248 |
+
"eval_runtime": 1242.2627,
|
| 2249 |
+
"eval_samples_per_second": 183.812,
|
| 2250 |
+
"eval_steps_per_second": 22.977,
|
| 2251 |
+
"step": 29500
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 1.84,
|
| 2255 |
+
"learning_rate": 7.857365306758021e-07,
|
| 2256 |
+
"loss": 2.3105,
|
| 2257 |
+
"step": 29600
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 1.85,
|
| 2261 |
+
"learning_rate": 7.545935845530988e-07,
|
| 2262 |
+
"loss": 3.303,
|
| 2263 |
+
"step": 29700
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.86,
|
| 2267 |
+
"learning_rate": 7.234506384303956e-07,
|
| 2268 |
+
"loss": 2.6327,
|
| 2269 |
+
"step": 29800
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 1.86,
|
| 2273 |
+
"learning_rate": 6.923076923076924e-07,
|
| 2274 |
+
"loss": 2.4727,
|
| 2275 |
+
"step": 29900
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 1.87,
|
| 2279 |
+
"learning_rate": 6.611647461849892e-07,
|
| 2280 |
+
"loss": 2.5736,
|
| 2281 |
+
"step": 30000
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 1.87,
|
| 2285 |
+
"eval_loss": NaN,
|
| 2286 |
+
"eval_runtime": 1242.5556,
|
| 2287 |
+
"eval_samples_per_second": 183.769,
|
| 2288 |
+
"eval_steps_per_second": 22.971,
|
| 2289 |
+
"step": 30000
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 1.87,
|
| 2293 |
+
"learning_rate": 6.300218000622859e-07,
|
| 2294 |
+
"loss": 2.687,
|
| 2295 |
+
"step": 30100
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 1.88,
|
| 2299 |
+
"learning_rate": 5.988788539395828e-07,
|
| 2300 |
+
"loss": 2.664,
|
| 2301 |
+
"step": 30200
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 1.89,
|
| 2305 |
+
"learning_rate": 5.677359078168795e-07,
|
| 2306 |
+
"loss": 2.4285,
|
| 2307 |
+
"step": 30300
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 1.89,
|
| 2311 |
+
"learning_rate": 5.369043911554033e-07,
|
| 2312 |
+
"loss": 2.4898,
|
| 2313 |
+
"step": 30400
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 1.9,
|
| 2317 |
+
"learning_rate": 5.057614450327001e-07,
|
| 2318 |
+
"loss": 2.3551,
|
| 2319 |
+
"step": 30500
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.9,
|
| 2323 |
+
"eval_loss": NaN,
|
| 2324 |
+
"eval_runtime": 1242.2664,
|
| 2325 |
+
"eval_samples_per_second": 183.812,
|
| 2326 |
+
"eval_steps_per_second": 22.977,
|
| 2327 |
+
"step": 30500
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 1.91,
|
| 2331 |
+
"learning_rate": 4.7461849890999693e-07,
|
| 2332 |
+
"loss": 2.8719,
|
| 2333 |
+
"step": 30600
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 1.91,
|
| 2337 |
+
"learning_rate": 4.4347555278729373e-07,
|
| 2338 |
+
"loss": 2.6584,
|
| 2339 |
+
"step": 30700
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 1.92,
|
| 2343 |
+
"learning_rate": 4.1233260666459054e-07,
|
| 2344 |
+
"loss": 2.3366,
|
| 2345 |
+
"step": 30800
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 1.92,
|
| 2349 |
+
"learning_rate": 3.811896605418873e-07,
|
| 2350 |
+
"loss": 2.8612,
|
| 2351 |
+
"step": 30900
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 1.93,
|
| 2355 |
+
"learning_rate": 3.500467144191841e-07,
|
| 2356 |
+
"loss": 2.7175,
|
| 2357 |
+
"step": 31000
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 1.93,
|
| 2361 |
+
"eval_loss": NaN,
|
| 2362 |
+
"eval_runtime": 1242.8604,
|
| 2363 |
+
"eval_samples_per_second": 183.724,
|
| 2364 |
+
"eval_steps_per_second": 22.966,
|
| 2365 |
+
"step": 31000
|
| 2366 |
+
},
|
| 2367 |
+
{
|
| 2368 |
+
"epoch": 1.94,
|
| 2369 |
+
"learning_rate": 3.189037682964809e-07,
|
| 2370 |
+
"loss": 2.9036,
|
| 2371 |
+
"step": 31100
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 1.94,
|
| 2375 |
+
"learning_rate": 2.8776082217377764e-07,
|
| 2376 |
+
"loss": 2.8631,
|
| 2377 |
+
"step": 31200
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 1.95,
|
| 2381 |
+
"learning_rate": 2.5661787605107445e-07,
|
| 2382 |
+
"loss": 2.4594,
|
| 2383 |
+
"step": 31300
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 1.96,
|
| 2387 |
+
"learning_rate": 2.2547492992837125e-07,
|
| 2388 |
+
"loss": 2.9569,
|
| 2389 |
+
"step": 31400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 1.96,
|
| 2393 |
+
"learning_rate": 1.9433198380566805e-07,
|
| 2394 |
+
"loss": 3.0516,
|
| 2395 |
+
"step": 31500
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 1.96,
|
| 2399 |
+
"eval_loss": NaN,
|
| 2400 |
+
"eval_runtime": 1243.9849,
|
| 2401 |
+
"eval_samples_per_second": 183.558,
|
| 2402 |
+
"eval_steps_per_second": 22.945,
|
| 2403 |
+
"step": 31500
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.97,
|
| 2407 |
+
"learning_rate": 1.6318903768296483e-07,
|
| 2408 |
+
"loss": 2.529,
|
| 2409 |
+
"step": 31600
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 1.97,
|
| 2413 |
+
"learning_rate": 1.320460915602616e-07,
|
| 2414 |
+
"loss": 2.4768,
|
| 2415 |
+
"step": 31700
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 1.98,
|
| 2419 |
+
"learning_rate": 1.009031454375584e-07,
|
| 2420 |
+
"loss": 2.5839,
|
| 2421 |
+
"step": 31800
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 1.99,
|
| 2425 |
+
"learning_rate": 6.97601993148552e-08,
|
| 2426 |
+
"loss": 2.6676,
|
| 2427 |
+
"step": 31900
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 1.99,
|
| 2431 |
+
"learning_rate": 3.861725319215198e-08,
|
| 2432 |
+
"loss": 3.2724,
|
| 2433 |
+
"step": 32000
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 1.99,
|
| 2437 |
+
"eval_loss": NaN,
|
| 2438 |
+
"eval_runtime": 1242.7853,
|
| 2439 |
+
"eval_samples_per_second": 183.735,
|
| 2440 |
+
"eval_steps_per_second": 22.967,
|
| 2441 |
+
"step": 32000
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 2.0,
|
| 2445 |
+
"learning_rate": 7.47430706944877e-09,
|
| 2446 |
+
"loss": 2.7629,
|
| 2447 |
+
"step": 32100
|
| 2448 |
}
|
| 2449 |
],
|
| 2450 |
+
"logging_steps": 100,
|
| 2451 |
+
"max_steps": 32110,
|
| 2452 |
+
"num_train_epochs": 2,
|
| 2453 |
+
"save_steps": 100,
|
| 2454 |
+
"total_flos": 1.0814548820133888e+18,
|
| 2455 |
"trial_name": null,
|
| 2456 |
"trial_params": null
|
| 2457 |
}
|
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 sha256:63248d3212eed9dc79dcf28db34d358c07675fbcbcd9e5b5d24121d0ccc9f00a
|
| 3 |
+
size 4027
|