Upload folder using huggingface_hub
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +264 -907
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -13,7 +13,7 @@
|
|
| 13 |
"layers_pattern": null,
|
| 14 |
"layers_to_transform": null,
|
| 15 |
"loftq_config": {},
|
| 16 |
-
"lora_alpha":
|
| 17 |
"lora_bias": false,
|
| 18 |
"lora_dropout": 0.1,
|
| 19 |
"megatron_config": null,
|
|
@@ -26,11 +26,11 @@
|
|
| 26 |
"revision": null,
|
| 27 |
"target_modules": [
|
| 28 |
"o_proj",
|
| 29 |
-
"down_proj",
|
| 30 |
-
"v_proj",
|
| 31 |
-
"k_proj",
|
| 32 |
"up_proj",
|
| 33 |
"q_proj",
|
|
|
|
|
|
|
|
|
|
| 34 |
"gate_proj"
|
| 35 |
],
|
| 36 |
"target_parameters": null,
|
|
|
|
| 13 |
"layers_pattern": null,
|
| 14 |
"layers_to_transform": null,
|
| 15 |
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 16,
|
| 17 |
"lora_bias": false,
|
| 18 |
"lora_dropout": 0.1,
|
| 19 |
"megatron_config": null,
|
|
|
|
| 26 |
"revision": null,
|
| 27 |
"target_modules": [
|
| 28 |
"o_proj",
|
|
|
|
|
|
|
|
|
|
| 29 |
"up_proj",
|
| 30 |
"q_proj",
|
| 31 |
+
"down_proj",
|
| 32 |
+
"k_proj",
|
| 33 |
+
"v_proj",
|
| 34 |
"gate_proj"
|
| 35 |
],
|
| 36 |
"target_parameters": null,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 645975704
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7910da3c4957e71800c849379173ba91d8f5e700436be19029cac09f62f6d4a8
|
| 3 |
size 645975704
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1292087499
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8dd8ce59e89a9efc32937df859f80518d82028400a830a9b29264b30c291ebd1
|
| 3 |
size 1292087499
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5844d24e029460a14b821935a1464ceba31cf265535853016e1f6652d0694907
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:954c3b26f534b41d23b4604b2fd25ad07029756854b57c8ddfc09c8a621110f7
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,1126 +2,483 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.060350030175015085,
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
-
"learning_rate": 0.
|
| 16 |
-
"loss": 1.
|
| 17 |
-
"mean_token_accuracy": 0.
|
| 18 |
-
"num_tokens":
|
| 19 |
"step": 25
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 0.12070006035003017,
|
| 23 |
-
"grad_norm": 0.
|
| 24 |
-
"learning_rate": 0.
|
| 25 |
-
"loss": 0.
|
| 26 |
-
"mean_token_accuracy": 0.
|
| 27 |
-
"num_tokens":
|
| 28 |
"step": 50
|
| 29 |
},
|
| 30 |
{
|
| 31 |
"epoch": 0.18105009052504525,
|
| 32 |
-
"grad_norm": 0.
|
| 33 |
-
"learning_rate": 0.
|
| 34 |
-
"loss": 0.
|
| 35 |
-
"mean_token_accuracy": 0.
|
| 36 |
-
"num_tokens":
|
| 37 |
"step": 75
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.24140012070006034,
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
-
"learning_rate": 0.
|
| 43 |
-
"loss": 0.
|
| 44 |
-
"mean_token_accuracy": 0.
|
| 45 |
-
"num_tokens":
|
| 46 |
"step": 100
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.30175015087507545,
|
| 50 |
-
"grad_norm": 0.
|
| 51 |
-
"learning_rate": 0.
|
| 52 |
-
"loss": 0.
|
| 53 |
-
"mean_token_accuracy": 0.
|
| 54 |
-
"num_tokens":
|
| 55 |
"step": 125
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 0.3621001810500905,
|
| 59 |
-
"grad_norm": 0.
|
| 60 |
-
"learning_rate": 0.
|
| 61 |
-
"loss": 0.
|
| 62 |
-
"mean_token_accuracy": 0.
|
| 63 |
-
"num_tokens":
|
| 64 |
"step": 150
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"epoch": 0.4224502112251056,
|
| 68 |
-
"grad_norm": 0.
|
| 69 |
-
"learning_rate": 0.
|
| 70 |
-
"loss": 0.
|
| 71 |
-
"mean_token_accuracy": 0.
|
| 72 |
-
"num_tokens":
|
| 73 |
"step": 175
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.4828002414001207,
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 0.
|
| 80 |
-
"mean_token_accuracy": 0.
|
| 81 |
-
"num_tokens":
|
| 82 |
"step": 200
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 0.5431502715751357,
|
| 86 |
-
"grad_norm": 0.
|
| 87 |
-
"learning_rate": 0.
|
| 88 |
-
"loss": 0.
|
| 89 |
-
"mean_token_accuracy": 0.
|
| 90 |
-
"num_tokens":
|
| 91 |
"step": 225
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 0.6035003017501509,
|
| 95 |
-
"grad_norm": 0.
|
| 96 |
-
"learning_rate": 0.
|
| 97 |
-
"loss": 0.
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 250
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.663850331925166,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
-
"learning_rate": 0.
|
| 106 |
-
"loss": 0.
|
| 107 |
-
"mean_token_accuracy": 0.
|
| 108 |
-
"num_tokens":
|
| 109 |
"step": 275
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 0.724200362100181,
|
| 113 |
-
"grad_norm": 0.
|
| 114 |
-
"learning_rate": 0.
|
| 115 |
-
"loss": 0.
|
| 116 |
-
"mean_token_accuracy": 0.
|
| 117 |
-
"num_tokens":
|
| 118 |
"step": 300
|
| 119 |
},
|
| 120 |
{
|
| 121 |
"epoch": 0.7845503922751962,
|
| 122 |
-
"grad_norm": 0.
|
| 123 |
-
"learning_rate": 0.
|
| 124 |
-
"loss": 0.
|
| 125 |
-
"mean_token_accuracy": 0.
|
| 126 |
-
"num_tokens":
|
| 127 |
"step": 325
|
| 128 |
},
|
| 129 |
{
|
| 130 |
"epoch": 0.8449004224502112,
|
| 131 |
-
"grad_norm": 0.
|
| 132 |
-
"learning_rate": 0.
|
| 133 |
-
"loss": 0.
|
| 134 |
-
"mean_token_accuracy": 0.
|
| 135 |
-
"num_tokens":
|
| 136 |
"step": 350
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.9052504526252263,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
-
"learning_rate": 0.
|
| 142 |
-
"loss": 0.
|
| 143 |
-
"mean_token_accuracy": 0.
|
| 144 |
-
"num_tokens":
|
| 145 |
"step": 375
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 0.9656004828002414,
|
| 149 |
-
"grad_norm": 0.
|
| 150 |
-
"learning_rate": 0.
|
| 151 |
-
"loss": 0.
|
| 152 |
-
"mean_token_accuracy": 0.
|
| 153 |
-
"num_tokens":
|
| 154 |
"step": 400
|
| 155 |
},
|
| 156 |
{
|
| 157 |
"epoch": 1.0,
|
| 158 |
-
"eval_loss": 0.
|
| 159 |
-
"eval_mean_token_accuracy": 0.
|
| 160 |
"eval_num_tokens": 2354180.0,
|
| 161 |
-
"eval_runtime": 15.
|
| 162 |
-
"eval_samples_per_second": 23.
|
| 163 |
-
"eval_steps_per_second": 11.
|
| 164 |
"step": 415
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 1.024140012070006,
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
-
"learning_rate": 0.
|
| 170 |
-
"loss": 0.
|
| 171 |
-
"mean_token_accuracy": 0.
|
| 172 |
-
"num_tokens":
|
| 173 |
"step": 425
|
| 174 |
},
|
| 175 |
{
|
| 176 |
"epoch": 1.0844900422450212,
|
| 177 |
-
"grad_norm": 0.
|
| 178 |
-
"learning_rate": 0.
|
| 179 |
-
"loss": 0.
|
| 180 |
-
"mean_token_accuracy": 0.
|
| 181 |
-
"num_tokens":
|
| 182 |
"step": 450
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 1.1448400724200363,
|
| 186 |
-
"grad_norm": 0.
|
| 187 |
-
"learning_rate": 0.
|
| 188 |
-
"loss": 0.
|
| 189 |
-
"mean_token_accuracy": 0.
|
| 190 |
-
"num_tokens":
|
| 191 |
"step": 475
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 1.2051901025950513,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate": 0.
|
| 197 |
-
"loss": 0.
|
| 198 |
-
"mean_token_accuracy": 0.
|
| 199 |
-
"num_tokens":
|
| 200 |
"step": 500
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.2655401327700664,
|
| 204 |
-
"grad_norm": 0.
|
| 205 |
-
"learning_rate": 0.
|
| 206 |
-
"loss": 0.
|
| 207 |
-
"mean_token_accuracy": 0.
|
| 208 |
-
"num_tokens":
|
| 209 |
"step": 525
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 1.3258901629450814,
|
| 213 |
-
"grad_norm": 0.
|
| 214 |
-
"learning_rate": 0.
|
| 215 |
-
"loss": 0.
|
| 216 |
-
"mean_token_accuracy": 0.
|
| 217 |
-
"num_tokens":
|
| 218 |
"step": 550
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 1.3862401931200965,
|
| 222 |
-
"grad_norm": 0.
|
| 223 |
-
"learning_rate": 0.
|
| 224 |
-
"loss": 0.
|
| 225 |
-
"mean_token_accuracy": 0.
|
| 226 |
-
"num_tokens":
|
| 227 |
"step": 575
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 1.4465902232951118,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate": 0.
|
| 233 |
-
"loss": 0.
|
| 234 |
-
"mean_token_accuracy": 0.
|
| 235 |
-
"num_tokens":
|
| 236 |
"step": 600
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 1.5069402534701268,
|
| 240 |
-
"grad_norm": 0.
|
| 241 |
-
"learning_rate": 0.
|
| 242 |
-
"loss": 0.
|
| 243 |
-
"mean_token_accuracy": 0.
|
| 244 |
-
"num_tokens":
|
| 245 |
"step": 625
|
| 246 |
},
|
| 247 |
{
|
| 248 |
"epoch": 1.567290283645142,
|
| 249 |
-
"grad_norm": 0.
|
| 250 |
-
"learning_rate": 0.
|
| 251 |
-
"loss": 0.
|
| 252 |
-
"mean_token_accuracy": 0.
|
| 253 |
-
"num_tokens":
|
| 254 |
"step": 650
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 1.627640313820157,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
-
"learning_rate": 0.
|
| 260 |
-
"loss": 0.
|
| 261 |
-
"mean_token_accuracy": 0.
|
| 262 |
-
"num_tokens":
|
| 263 |
"step": 675
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 1.687990343995172,
|
| 267 |
-
"grad_norm": 0.
|
| 268 |
-
"learning_rate": 0.
|
| 269 |
-
"loss": 0.
|
| 270 |
-
"mean_token_accuracy": 0.
|
| 271 |
-
"num_tokens":
|
| 272 |
"step": 700
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 1.748340374170187,
|
| 276 |
-
"grad_norm": 0.
|
| 277 |
-
"learning_rate": 0.
|
| 278 |
-
"loss": 0.
|
| 279 |
-
"mean_token_accuracy": 0.
|
| 280 |
-
"num_tokens":
|
| 281 |
"step": 725
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 1.8086904043452021,
|
| 285 |
-
"grad_norm": 0.
|
| 286 |
-
"learning_rate": 0.
|
| 287 |
-
"loss": 0.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 750
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 1.8690404345202172,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate": 0.
|
| 296 |
-
"loss": 0.
|
| 297 |
-
"mean_token_accuracy": 0.
|
| 298 |
-
"num_tokens":
|
| 299 |
"step": 775
|
| 300 |
},
|
| 301 |
{
|
| 302 |
"epoch": 1.9293904646952322,
|
| 303 |
-
"grad_norm": 0.
|
| 304 |
-
"learning_rate": 0.
|
| 305 |
-
"loss": 0.
|
| 306 |
-
"mean_token_accuracy": 0.
|
| 307 |
-
"num_tokens":
|
| 308 |
"step": 800
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 1.9897404948702473,
|
| 312 |
-
"grad_norm": 0.
|
| 313 |
-
"learning_rate": 0.
|
| 314 |
-
"loss": 0.
|
| 315 |
-
"mean_token_accuracy": 0.
|
| 316 |
-
"num_tokens":
|
| 317 |
"step": 825
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 2.0,
|
| 321 |
-
"eval_loss": 0.
|
| 322 |
-
"eval_mean_token_accuracy": 0.
|
| 323 |
"eval_num_tokens": 4708360.0,
|
| 324 |
-
"eval_runtime": 15.
|
| 325 |
-
"eval_samples_per_second": 23.
|
| 326 |
-
"eval_steps_per_second": 11.
|
| 327 |
"step": 830
|
| 328 |
},
|
| 329 |
{
|
| 330 |
"epoch": 2.048280024140012,
|
| 331 |
-
"grad_norm": 0.
|
| 332 |
-
"learning_rate": 0.
|
| 333 |
-
"loss": 0.
|
| 334 |
-
"mean_token_accuracy": 0.
|
| 335 |
-
"num_tokens":
|
| 336 |
"step": 850
|
| 337 |
},
|
| 338 |
{
|
| 339 |
"epoch": 2.1086300543150274,
|
| 340 |
-
"grad_norm": 0.
|
| 341 |
-
"learning_rate": 0.
|
| 342 |
-
"loss": 0.
|
| 343 |
-
"mean_token_accuracy": 0.
|
| 344 |
-
"num_tokens":
|
| 345 |
"step": 875
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 2.1689800844900424,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
-
"learning_rate": 0.
|
| 351 |
-
"loss": 0.
|
| 352 |
-
"mean_token_accuracy": 0.
|
| 353 |
-
"num_tokens":
|
| 354 |
"step": 900
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 2.2293301146650575,
|
| 358 |
-
"grad_norm": 0.
|
| 359 |
-
"learning_rate": 0.
|
| 360 |
-
"loss": 0.
|
| 361 |
-
"mean_token_accuracy": 0.
|
| 362 |
-
"num_tokens":
|
| 363 |
"step": 925
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 2.2896801448400725,
|
| 367 |
-
"grad_norm": 0.
|
| 368 |
-
"learning_rate": 0.
|
| 369 |
-
"loss": 0.
|
| 370 |
-
"mean_token_accuracy": 0.
|
| 371 |
-
"num_tokens":
|
| 372 |
"step": 950
|
| 373 |
},
|
| 374 |
{
|
| 375 |
"epoch": 2.3500301750150876,
|
| 376 |
-
"grad_norm": 0.
|
| 377 |
-
"learning_rate": 0.
|
| 378 |
-
"loss": 0.
|
| 379 |
-
"mean_token_accuracy": 0.
|
| 380 |
-
"num_tokens":
|
| 381 |
"step": 975
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 2.4103802051901027,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
-
"learning_rate": 0.
|
| 387 |
-
"loss": 0.
|
| 388 |
-
"mean_token_accuracy": 0.
|
| 389 |
-
"num_tokens":
|
| 390 |
"step": 1000
|
| 391 |
},
|
| 392 |
{
|
| 393 |
"epoch": 2.4707302353651177,
|
| 394 |
-
"grad_norm": 0.
|
| 395 |
-
"learning_rate": 0.
|
| 396 |
-
"loss": 0.
|
| 397 |
-
"mean_token_accuracy": 0.
|
| 398 |
-
"num_tokens":
|
| 399 |
"step": 1025
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 2.5310802655401328,
|
| 403 |
-
"grad_norm": 0.
|
| 404 |
-
"learning_rate": 0.
|
| 405 |
-
"loss": 0.
|
| 406 |
-
"mean_token_accuracy": 0.
|
| 407 |
-
"num_tokens":
|
| 408 |
"step": 1050
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 2.591430295715148,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
-
"learning_rate": 0.
|
| 414 |
-
"loss": 0.
|
| 415 |
-
"mean_token_accuracy": 0.
|
| 416 |
-
"num_tokens":
|
| 417 |
"step": 1075
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 2.651780325890163,
|
| 421 |
-
"grad_norm": 0.
|
| 422 |
-
"learning_rate": 0.
|
| 423 |
-
"loss": 0.
|
| 424 |
-
"mean_token_accuracy": 0.
|
| 425 |
-
"num_tokens":
|
| 426 |
"step": 1100
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 2.712130356065178,
|
| 430 |
-
"grad_norm": 0.
|
| 431 |
-
"learning_rate": 0.
|
| 432 |
-
"loss": 0.
|
| 433 |
-
"mean_token_accuracy": 0.
|
| 434 |
-
"num_tokens":
|
| 435 |
"step": 1125
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 2.772480386240193,
|
| 439 |
-
"grad_norm": 0.
|
| 440 |
-
"learning_rate": 0.
|
| 441 |
-
"loss": 0.
|
| 442 |
-
"mean_token_accuracy": 0.
|
| 443 |
-
"num_tokens":
|
| 444 |
"step": 1150
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 2.832830416415208,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
-
"learning_rate": 0.
|
| 450 |
-
"loss": 0.
|
| 451 |
-
"mean_token_accuracy": 0.
|
| 452 |
-
"num_tokens":
|
| 453 |
"step": 1175
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 2.8931804465902236,
|
| 457 |
-
"grad_norm": 0.
|
| 458 |
-
"learning_rate": 0.
|
| 459 |
-
"loss": 0.
|
| 460 |
-
"mean_token_accuracy": 0.
|
| 461 |
-
"num_tokens":
|
| 462 |
"step": 1200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
"epoch": 2.9535304767652386,
|
| 466 |
-
"grad_norm": 0.
|
| 467 |
-
"learning_rate": 0.
|
| 468 |
-
"loss": 0.
|
| 469 |
-
"mean_token_accuracy": 0.
|
| 470 |
-
"num_tokens":
|
| 471 |
"step": 1225
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 3.0,
|
| 475 |
-
"eval_loss": 0.
|
| 476 |
-
"eval_mean_token_accuracy": 0.
|
| 477 |
"eval_num_tokens": 7062540.0,
|
| 478 |
-
"eval_runtime": 15.
|
| 479 |
-
"eval_samples_per_second": 23.
|
| 480 |
-
"eval_steps_per_second": 11.
|
| 481 |
"step": 1245
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 3.012070006035003,
|
| 485 |
-
"grad_norm": 0.04459620267152786,
|
| 486 |
-
"learning_rate": 0.00038175811469422905,
|
| 487 |
-
"loss": 0.04,
|
| 488 |
-
"mean_token_accuracy": 0.9881644027749288,
|
| 489 |
-
"num_tokens": 7099773.0,
|
| 490 |
-
"step": 1250
|
| 491 |
-
},
|
| 492 |
-
{
|
| 493 |
-
"epoch": 3.0724200362100182,
|
| 494 |
-
"grad_norm": 0.025668803602457047,
|
| 495 |
-
"learning_rate": 0.00037367978565773226,
|
| 496 |
-
"loss": 0.027,
|
| 497 |
-
"mean_token_accuracy": 0.9916459822654724,
|
| 498 |
-
"num_tokens": 7247178.0,
|
| 499 |
-
"step": 1275
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 3.1327700663850333,
|
| 503 |
-
"grad_norm": 0.048161111772060394,
|
| 504 |
-
"learning_rate": 0.0003655441867390346,
|
| 505 |
-
"loss": 0.0296,
|
| 506 |
-
"mean_token_accuracy": 0.990880873799324,
|
| 507 |
-
"num_tokens": 7383662.0,
|
| 508 |
-
"step": 1300
|
| 509 |
-
},
|
| 510 |
-
{
|
| 511 |
-
"epoch": 3.1931200965600484,
|
| 512 |
-
"grad_norm": 0.04576310142874718,
|
| 513 |
-
"learning_rate": 0.00035735764158291254,
|
| 514 |
-
"loss": 0.0296,
|
| 515 |
-
"mean_token_accuracy": 0.9907885307073593,
|
| 516 |
-
"num_tokens": 7533651.0,
|
| 517 |
-
"step": 1325
|
| 518 |
-
},
|
| 519 |
-
{
|
| 520 |
-
"epoch": 3.2534701267350634,
|
| 521 |
-
"grad_norm": 0.04082406684756279,
|
| 522 |
-
"learning_rate": 0.0003491265134336745,
|
| 523 |
-
"loss": 0.0327,
|
| 524 |
-
"mean_token_accuracy": 0.9898776888847352,
|
| 525 |
-
"num_tokens": 7671505.0,
|
| 526 |
-
"step": 1350
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 3.3138201569100785,
|
| 530 |
-
"grad_norm": 0.04111408442258835,
|
| 531 |
-
"learning_rate": 0.00034085720018913276,
|
| 532 |
-
"loss": 0.0264,
|
| 533 |
-
"mean_token_accuracy": 0.9915948182344436,
|
| 534 |
-
"num_tokens": 7819430.0,
|
| 535 |
-
"step": 1375
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 3.3741701870850935,
|
| 539 |
-
"grad_norm": 0.05790106952190399,
|
| 540 |
-
"learning_rate": 0.0003325561294276413,
|
| 541 |
-
"loss": 0.0286,
|
| 542 |
-
"mean_token_accuracy": 0.9912853974103928,
|
| 543 |
-
"num_tokens": 7954433.0,
|
| 544 |
-
"step": 1400
|
| 545 |
-
},
|
| 546 |
-
{
|
| 547 |
-
"epoch": 3.4345202172601086,
|
| 548 |
-
"grad_norm": 0.027405843138694763,
|
| 549 |
-
"learning_rate": 0.00032422975341206157,
|
| 550 |
-
"loss": 0.0281,
|
| 551 |
-
"mean_token_accuracy": 0.991134095788002,
|
| 552 |
-
"num_tokens": 8102229.0,
|
| 553 |
-
"step": 1425
|
| 554 |
-
},
|
| 555 |
-
{
|
| 556 |
-
"epoch": 3.4948702474351236,
|
| 557 |
-
"grad_norm": 0.043464187532663345,
|
| 558 |
-
"learning_rate": 0.000315884544074543,
|
| 559 |
-
"loss": 0.0292,
|
| 560 |
-
"mean_token_accuracy": 0.9907919150590897,
|
| 561 |
-
"num_tokens": 8237057.0,
|
| 562 |
-
"step": 1450
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 3.5552202776101387,
|
| 566 |
-
"grad_norm": 0.04483381658792496,
|
| 567 |
-
"learning_rate": 0.0003075269879860149,
|
| 568 |
-
"loss": 0.0255,
|
| 569 |
-
"mean_token_accuracy": 0.9919383651018143,
|
| 570 |
-
"num_tokens": 8385229.0,
|
| 571 |
-
"step": 1475
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 3.6155703077851538,
|
| 575 |
-
"grad_norm": 0.03719889372587204,
|
| 576 |
-
"learning_rate": 0.0002991635813142984,
|
| 577 |
-
"loss": 0.0305,
|
| 578 |
-
"mean_token_accuracy": 0.9901222789287567,
|
| 579 |
-
"num_tokens": 8521392.0,
|
| 580 |
-
"step": 1500
|
| 581 |
-
},
|
| 582 |
-
{
|
| 583 |
-
"epoch": 3.675920337960169,
|
| 584 |
-
"grad_norm": 0.026905681937932968,
|
| 585 |
-
"learning_rate": 0.0002908008247747611,
|
| 586 |
-
"loss": 0.025,
|
| 587 |
-
"mean_token_accuracy": 0.9921067571640014,
|
| 588 |
-
"num_tokens": 8669874.0,
|
| 589 |
-
"step": 1525
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 3.736270368135184,
|
| 593 |
-
"grad_norm": 0.05321066826581955,
|
| 594 |
-
"learning_rate": 0.00028244521857743467,
|
| 595 |
-
"loss": 0.0309,
|
| 596 |
-
"mean_token_accuracy": 0.9903623193502427,
|
| 597 |
-
"num_tokens": 8804899.0,
|
| 598 |
-
"step": 1550
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 3.796620398310199,
|
| 602 |
-
"grad_norm": 0.02997380495071411,
|
| 603 |
-
"learning_rate": 0.00027410325737452793,
|
| 604 |
-
"loss": 0.0247,
|
| 605 |
-
"mean_token_accuracy": 0.9921429508924484,
|
| 606 |
-
"num_tokens": 8952948.0,
|
| 607 |
-
"step": 1575
|
| 608 |
-
},
|
| 609 |
-
{
|
| 610 |
-
"epoch": 3.856970428485214,
|
| 611 |
-
"grad_norm": 0.042058832943439484,
|
| 612 |
-
"learning_rate": 0.0002657814252122571,
|
| 613 |
-
"loss": 0.0276,
|
| 614 |
-
"mean_token_accuracy": 0.9911820942163467,
|
| 615 |
-
"num_tokens": 9088804.0,
|
| 616 |
-
"step": 1600
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 3.9173204586602295,
|
| 620 |
-
"grad_norm": 0.028029246255755424,
|
| 621 |
-
"learning_rate": 0.00025748619049092167,
|
| 622 |
-
"loss": 0.0246,
|
| 623 |
-
"mean_token_accuracy": 0.9920519244670868,
|
| 624 |
-
"num_tokens": 9237788.0,
|
| 625 |
-
"step": 1625
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 3.9776704888352445,
|
| 629 |
-
"grad_norm": 0.02371133305132389,
|
| 630 |
-
"learning_rate": 0.0002492240009371417,
|
| 631 |
-
"loss": 0.0245,
|
| 632 |
-
"mean_token_accuracy": 0.9923285436630249,
|
| 633 |
-
"num_tokens": 9369559.0,
|
| 634 |
-
"step": 1650
|
| 635 |
-
},
|
| 636 |
-
{
|
| 637 |
-
"epoch": 4.0,
|
| 638 |
-
"eval_loss": 0.041972678154706955,
|
| 639 |
-
"eval_mean_token_accuracy": 0.989047556632274,
|
| 640 |
-
"eval_num_tokens": 9416720.0,
|
| 641 |
-
"eval_runtime": 15.7616,
|
| 642 |
-
"eval_samples_per_second": 23.411,
|
| 643 |
-
"eval_steps_per_second": 11.737,
|
| 644 |
-
"step": 1660
|
| 645 |
-
},
|
| 646 |
-
{
|
| 647 |
-
"epoch": 4.036210018105009,
|
| 648 |
-
"grad_norm": 0.027629472315311432,
|
| 649 |
-
"learning_rate": 0.00024100127859216246,
|
| 650 |
-
"loss": 0.0242,
|
| 651 |
-
"mean_token_accuracy": 0.9923697267611002,
|
| 652 |
-
"num_tokens": 9517456.0,
|
| 653 |
-
"step": 1675
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 4.096560048280024,
|
| 657 |
-
"grad_norm": 0.018458090722560883,
|
| 658 |
-
"learning_rate": 0.0002328244148201266,
|
| 659 |
-
"loss": 0.0198,
|
| 660 |
-
"mean_token_accuracy": 0.9934357500076294,
|
| 661 |
-
"num_tokens": 9655983.0,
|
| 662 |
-
"step": 1700
|
| 663 |
-
},
|
| 664 |
-
{
|
| 665 |
-
"epoch": 4.15691007845504,
|
| 666 |
-
"grad_norm": 0.025151990354061127,
|
| 667 |
-
"learning_rate": 0.0002246997653401883,
|
| 668 |
-
"loss": 0.0252,
|
| 669 |
-
"mean_token_accuracy": 0.9922802877426148,
|
| 670 |
-
"num_tokens": 9802927.0,
|
| 671 |
-
"step": 1725
|
| 672 |
-
},
|
| 673 |
-
{
|
| 674 |
-
"epoch": 4.217260108630055,
|
| 675 |
-
"grad_norm": 0.02359866164624691,
|
| 676 |
-
"learning_rate": 0.00021663364528633574,
|
| 677 |
-
"loss": 0.0192,
|
| 678 |
-
"mean_token_accuracy": 0.9935739403963089,
|
| 679 |
-
"num_tokens": 9939784.0,
|
| 680 |
-
"step": 1750
|
| 681 |
-
},
|
| 682 |
-
{
|
| 683 |
-
"epoch": 4.27761013880507,
|
| 684 |
-
"grad_norm": 0.029893064871430397,
|
| 685 |
-
"learning_rate": 0.00020863232429875822,
|
| 686 |
-
"loss": 0.0219,
|
| 687 |
-
"mean_token_accuracy": 0.9929185563325882,
|
| 688 |
-
"num_tokens": 10085314.0,
|
| 689 |
-
"step": 1775
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 4.337960168980085,
|
| 693 |
-
"grad_norm": 0.01731249690055847,
|
| 694 |
-
"learning_rate": 0.00020070202165057554,
|
| 695 |
-
"loss": 0.0191,
|
| 696 |
-
"mean_token_accuracy": 0.9936865222454071,
|
| 697 |
-
"num_tokens": 10223073.0,
|
| 698 |
-
"step": 1800
|
| 699 |
-
},
|
| 700 |
-
{
|
| 701 |
-
"epoch": 4.3983101991551,
|
| 702 |
-
"grad_norm": 0.01831655018031597,
|
| 703 |
-
"learning_rate": 0.00019284890141371618,
|
| 704 |
-
"loss": 0.0228,
|
| 705 |
-
"mean_token_accuracy": 0.992409136891365,
|
| 706 |
-
"num_tokens": 10368536.0,
|
| 707 |
-
"step": 1825
|
| 708 |
-
},
|
| 709 |
-
{
|
| 710 |
-
"epoch": 4.458660229330115,
|
| 711 |
-
"grad_norm": 0.018383309245109558,
|
| 712 |
-
"learning_rate": 0.00018507906766770314,
|
| 713 |
-
"loss": 0.0186,
|
| 714 |
-
"mean_token_accuracy": 0.99377023935318,
|
| 715 |
-
"num_tokens": 10507164.0,
|
| 716 |
-
"step": 1850
|
| 717 |
-
},
|
| 718 |
-
{
|
| 719 |
-
"epoch": 4.51901025950513,
|
| 720 |
-
"grad_norm": 0.022526893764734268,
|
| 721 |
-
"learning_rate": 0.00017739855975506917,
|
| 722 |
-
"loss": 0.024,
|
| 723 |
-
"mean_token_accuracy": 0.9923234033584595,
|
| 724 |
-
"num_tokens": 10653645.0,
|
| 725 |
-
"step": 1875
|
| 726 |
-
},
|
| 727 |
-
{
|
| 728 |
-
"epoch": 4.579360289680145,
|
| 729 |
-
"grad_norm": 0.01562182791531086,
|
| 730 |
-
"learning_rate": 0.00016981334758709322,
|
| 731 |
-
"loss": 0.0187,
|
| 732 |
-
"mean_token_accuracy": 0.9941441065073013,
|
| 733 |
-
"num_tokens": 10791134.0,
|
| 734 |
-
"step": 1900
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 4.63971031985516,
|
| 738 |
-
"grad_norm": 0.02683549001812935,
|
| 739 |
-
"learning_rate": 0.00016232932700350157,
|
| 740 |
-
"loss": 0.0214,
|
| 741 |
-
"mean_token_accuracy": 0.9926981467008591,
|
| 742 |
-
"num_tokens": 10938557.0,
|
| 743 |
-
"step": 1925
|
| 744 |
-
},
|
| 745 |
-
{
|
| 746 |
-
"epoch": 4.700060350030175,
|
| 747 |
-
"grad_norm": 0.03423510119318962,
|
| 748 |
-
"learning_rate": 0.00015495231518974608,
|
| 749 |
-
"loss": 0.0193,
|
| 750 |
-
"mean_token_accuracy": 0.9936627286672592,
|
| 751 |
-
"num_tokens": 11076634.0,
|
| 752 |
-
"step": 1950
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 4.76041038020519,
|
| 756 |
-
"grad_norm": 0.029518209397792816,
|
| 757 |
-
"learning_rate": 0.000147688046155417,
|
| 758 |
-
"loss": 0.0226,
|
| 759 |
-
"mean_token_accuracy": 0.9926369667053223,
|
| 760 |
-
"num_tokens": 11224638.0,
|
| 761 |
-
"step": 1975
|
| 762 |
-
},
|
| 763 |
-
{
|
| 764 |
-
"epoch": 4.820760410380205,
|
| 765 |
-
"grad_norm": 0.021354857832193375,
|
| 766 |
-
"learning_rate": 0.00014054216627730755,
|
| 767 |
-
"loss": 0.0188,
|
| 768 |
-
"mean_token_accuracy": 0.9936299502849579,
|
| 769 |
-
"num_tokens": 11362368.0,
|
| 770 |
-
"step": 2000
|
| 771 |
-
},
|
| 772 |
-
{
|
| 773 |
-
"epoch": 4.88111044055522,
|
| 774 |
-
"grad_norm": 0.033567801117897034,
|
| 775 |
-
"learning_rate": 0.00013352022991059375,
|
| 776 |
-
"loss": 0.0236,
|
| 777 |
-
"mean_token_accuracy": 0.9924637532234192,
|
| 778 |
-
"num_tokens": 11509364.0,
|
| 779 |
-
"step": 2025
|
| 780 |
-
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 4.941460470730235,
|
| 783 |
-
"grad_norm": 0.02483433112502098,
|
| 784 |
-
"learning_rate": 0.00012662769507154113,
|
| 785 |
-
"loss": 0.0183,
|
| 786 |
-
"mean_token_accuracy": 0.9939434814453125,
|
| 787 |
-
"num_tokens": 11645477.0,
|
| 788 |
-
"step": 2050
|
| 789 |
-
},
|
| 790 |
-
{
|
| 791 |
-
"epoch": 5.0,
|
| 792 |
-
"grad_norm": 0.04451137036085129,
|
| 793 |
-
"learning_rate": 0.00011986991919509261,
|
| 794 |
-
"loss": 0.0203,
|
| 795 |
-
"mean_token_accuracy": 0.9933847133646306,
|
| 796 |
-
"num_tokens": 11770900.0,
|
| 797 |
-
"step": 2075
|
| 798 |
-
},
|
| 799 |
-
{
|
| 800 |
-
"epoch": 5.0,
|
| 801 |
-
"eval_loss": 0.04097144305706024,
|
| 802 |
-
"eval_mean_token_accuracy": 0.9897108348640236,
|
| 803 |
-
"eval_num_tokens": 11770900.0,
|
| 804 |
-
"eval_runtime": 15.7693,
|
| 805 |
-
"eval_samples_per_second": 23.4,
|
| 806 |
-
"eval_steps_per_second": 11.732,
|
| 807 |
-
"step": 2075
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 5.060350030175015,
|
| 811 |
-
"grad_norm": 0.01309128850698471,
|
| 812 |
-
"learning_rate": 0.00011325215497063858,
|
| 813 |
-
"loss": 0.0169,
|
| 814 |
-
"mean_token_accuracy": 0.9943383944034576,
|
| 815 |
-
"num_tokens": 11929319.0,
|
| 816 |
-
"step": 2100
|
| 817 |
-
},
|
| 818 |
-
{
|
| 819 |
-
"epoch": 5.12070006035003,
|
| 820 |
-
"grad_norm": 0.018124833703041077,
|
| 821 |
-
"learning_rate": 0.00010677954625920108,
|
| 822 |
-
"loss": 0.0165,
|
| 823 |
-
"mean_token_accuracy": 0.994610543847084,
|
| 824 |
-
"num_tokens": 12054482.0,
|
| 825 |
-
"step": 2125
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 5.181050090525045,
|
| 829 |
-
"grad_norm": 0.014541085809469223,
|
| 830 |
-
"learning_rate": 0.00010045712409521008,
|
| 831 |
-
"loss": 0.017,
|
| 832 |
-
"mean_token_accuracy": 0.9943281805515289,
|
| 833 |
-
"num_tokens": 12211531.0,
|
| 834 |
-
"step": 2150
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 5.24140012070006,
|
| 838 |
-
"grad_norm": 0.024412041530013084,
|
| 839 |
-
"learning_rate": 9.428980277597675e-05,
|
| 840 |
-
"loss": 0.0164,
|
| 841 |
-
"mean_token_accuracy": 0.9944160294532776,
|
| 842 |
-
"num_tokens": 12338886.0,
|
| 843 |
-
"step": 2175
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 5.301750150875075,
|
| 847 |
-
"grad_norm": 0.01901293359696865,
|
| 848 |
-
"learning_rate": 8.828237604190513e-05,
|
| 849 |
-
"loss": 0.0175,
|
| 850 |
-
"mean_token_accuracy": 0.9941371762752533,
|
| 851 |
-
"num_tokens": 12495549.0,
|
| 852 |
-
"step": 2200
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 5.36210018105009,
|
| 856 |
-
"grad_norm": 0.017199428752064705,
|
| 857 |
-
"learning_rate": 8.243951335040891e-05,
|
| 858 |
-
"loss": 0.0166,
|
| 859 |
-
"mean_token_accuracy": 0.9944494879245758,
|
| 860 |
-
"num_tokens": 12622459.0,
|
| 861 |
-
"step": 2225
|
| 862 |
-
},
|
| 863 |
-
{
|
| 864 |
-
"epoch": 5.422450211225105,
|
| 865 |
-
"grad_norm": 0.03323078528046608,
|
| 866 |
-
"learning_rate": 7.67657562464325e-05,
|
| 867 |
-
"loss": 0.0174,
|
| 868 |
-
"mean_token_accuracy": 0.9939983171224595,
|
| 869 |
-
"num_tokens": 12780223.0,
|
| 870 |
-
"step": 2250
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 5.4828002414001205,
|
| 874 |
-
"grad_norm": 0.02575470507144928,
|
| 875 |
-
"learning_rate": 7.126551483239509e-05,
|
| 876 |
-
"loss": 0.0165,
|
| 877 |
-
"mean_token_accuracy": 0.9944772917032242,
|
| 878 |
-
"num_tokens": 12907099.0,
|
| 879 |
-
"step": 2275
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 5.5431502715751355,
|
| 883 |
-
"grad_norm": 0.01993207074701786,
|
| 884 |
-
"learning_rate": 6.59430643403031e-05,
|
| 885 |
-
"loss": 0.0172,
|
| 886 |
-
"mean_token_accuracy": 0.9943294197320938,
|
| 887 |
-
"num_tokens": 13065287.0,
|
| 888 |
-
"step": 2300
|
| 889 |
-
},
|
| 890 |
-
{
|
| 891 |
-
"epoch": 5.603500301750151,
|
| 892 |
-
"grad_norm": 0.021883873268961906,
|
| 893 |
-
"learning_rate": 6.0802541808693984e-05,
|
| 894 |
-
"loss": 0.0167,
|
| 895 |
-
"mean_token_accuracy": 0.9944075202941894,
|
| 896 |
-
"num_tokens": 13191570.0,
|
| 897 |
-
"step": 2325
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 5.663850331925166,
|
| 901 |
-
"grad_norm": 0.02263251692056656,
|
| 902 |
-
"learning_rate": 5.584794286699583e-05,
|
| 903 |
-
"loss": 0.017,
|
| 904 |
-
"mean_token_accuracy": 0.9944959133863449,
|
| 905 |
-
"num_tokens": 13348858.0,
|
| 906 |
-
"step": 2350
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 5.724200362100181,
|
| 910 |
-
"grad_norm": 0.01838994212448597,
|
| 911 |
-
"learning_rate": 5.10831186298017e-05,
|
| 912 |
-
"loss": 0.0165,
|
| 913 |
-
"mean_token_accuracy": 0.994613738656044,
|
| 914 |
-
"num_tokens": 13475838.0,
|
| 915 |
-
"step": 2375
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 5.784550392275197,
|
| 919 |
-
"grad_norm": 0.015676449984312057,
|
| 920 |
-
"learning_rate": 4.6511772703471414e-05,
|
| 921 |
-
"loss": 0.0169,
|
| 922 |
-
"mean_token_accuracy": 0.9943792551755906,
|
| 923 |
-
"num_tokens": 13634866.0,
|
| 924 |
-
"step": 2400
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 5.844900422450211,
|
| 928 |
-
"grad_norm": 0.024185990914702415,
|
| 929 |
-
"learning_rate": 4.2137458307390404e-05,
|
| 930 |
-
"loss": 0.0166,
|
| 931 |
-
"mean_token_accuracy": 0.9944125992059708,
|
| 932 |
-
"num_tokens": 13761458.0,
|
| 933 |
-
"step": 2425
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 5.905250452625227,
|
| 937 |
-
"grad_norm": 0.017580190673470497,
|
| 938 |
-
"learning_rate": 3.796357551211986e-05,
|
| 939 |
-
"loss": 0.0163,
|
| 940 |
-
"mean_token_accuracy": 0.9945475596189499,
|
| 941 |
-
"num_tokens": 13918561.0,
|
| 942 |
-
"step": 2450
|
| 943 |
-
},
|
| 944 |
-
{
|
| 945 |
-
"epoch": 5.965600482800241,
|
| 946 |
-
"grad_norm": 0.018194038420915604,
|
| 947 |
-
"learning_rate": 3.399336859658742e-05,
|
| 948 |
-
"loss": 0.016,
|
| 949 |
-
"mean_token_accuracy": 0.9947626197338104,
|
| 950 |
-
"num_tokens": 14046227.0,
|
| 951 |
-
"step": 2475
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 6.0,
|
| 955 |
-
"eval_loss": 0.04244406521320343,
|
| 956 |
-
"eval_mean_token_accuracy": 0.9900913918340528,
|
| 957 |
-
"eval_num_tokens": 14125080.0,
|
| 958 |
-
"eval_runtime": 15.774,
|
| 959 |
-
"eval_samples_per_second": 23.393,
|
| 960 |
-
"eval_steps_per_second": 11.728,
|
| 961 |
-
"step": 2490
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 6.024140012070006,
|
| 965 |
-
"grad_norm": 0.019757593050599098,
|
| 966 |
-
"learning_rate": 3.0229923526371535e-05,
|
| 967 |
-
"loss": 0.0165,
|
| 968 |
-
"mean_token_accuracy": 0.9946722658639101,
|
| 969 |
-
"num_tokens": 14194026.0,
|
| 970 |
-
"step": 2500
|
| 971 |
-
},
|
| 972 |
-
{
|
| 973 |
-
"epoch": 6.084490042245021,
|
| 974 |
-
"grad_norm": 0.015921050682663918,
|
| 975 |
-
"learning_rate": 2.667616555503964e-05,
|
| 976 |
-
"loss": 0.0139,
|
| 977 |
-
"mean_token_accuracy": 0.9953104478120803,
|
| 978 |
-
"num_tokens": 14337164.0,
|
| 979 |
-
"step": 2525
|
| 980 |
-
},
|
| 981 |
-
{
|
| 982 |
-
"epoch": 6.1448400724200365,
|
| 983 |
-
"grad_norm": 0.01795029267668724,
|
| 984 |
-
"learning_rate": 2.333485695040469e-05,
|
| 985 |
-
"loss": 0.0153,
|
| 986 |
-
"mean_token_accuracy": 0.9950164467096329,
|
| 987 |
-
"num_tokens": 14478393.0,
|
| 988 |
-
"step": 2550
|
| 989 |
-
},
|
| 990 |
-
{
|
| 991 |
-
"epoch": 6.2051901025950515,
|
| 992 |
-
"grad_norm": 0.01615080237388611,
|
| 993 |
-
"learning_rate": 2.0208594847467508e-05,
|
| 994 |
-
"loss": 0.0142,
|
| 995 |
-
"mean_token_accuracy": 0.9951647013425827,
|
| 996 |
-
"num_tokens": 14620740.0,
|
| 997 |
-
"step": 2575
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 6.265540132770067,
|
| 1001 |
-
"grad_norm": 0.02364126406610012,
|
| 1002 |
-
"learning_rate": 1.729980922971349e-05,
|
| 1003 |
-
"loss": 0.0154,
|
| 1004 |
-
"mean_token_accuracy": 0.9948040336370468,
|
| 1005 |
-
"num_tokens": 14762878.0,
|
| 1006 |
-
"step": 2600
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 6.325890162945082,
|
| 1010 |
-
"grad_norm": 0.018266433849930763,
|
| 1011 |
-
"learning_rate": 1.4610761040333573e-05,
|
| 1012 |
-
"loss": 0.0142,
|
| 1013 |
-
"mean_token_accuracy": 0.9954759681224823,
|
| 1014 |
-
"num_tokens": 14905328.0,
|
| 1015 |
-
"step": 2625
|
| 1016 |
-
},
|
| 1017 |
-
{
|
| 1018 |
-
"epoch": 6.386240193120097,
|
| 1019 |
-
"grad_norm": 0.022892849519848824,
|
| 1020 |
-
"learning_rate": 1.214354042483573e-05,
|
| 1021 |
-
"loss": 0.0148,
|
| 1022 |
-
"mean_token_accuracy": 0.9951095223426819,
|
| 1023 |
-
"num_tokens": 15047420.0,
|
| 1024 |
-
"step": 2650
|
| 1025 |
-
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 6.446590223295112,
|
| 1028 |
-
"grad_norm": 0.013281609863042831,
|
| 1029 |
-
"learning_rate": 9.900065106415866e-06,
|
| 1030 |
-
"loss": 0.0139,
|
| 1031 |
-
"mean_token_accuracy": 0.9951714134216308,
|
| 1032 |
-
"num_tokens": 15191784.0,
|
| 1033 |
-
"step": 2675
|
| 1034 |
-
},
|
| 1035 |
-
{
|
| 1036 |
-
"epoch": 6.506940253470127,
|
| 1037 |
-
"grad_norm": 0.015181739814579487,
|
| 1038 |
-
"learning_rate": 7.882078895347798e-06,
|
| 1039 |
-
"loss": 0.016,
|
| 1040 |
-
"mean_token_accuracy": 0.9947543793916702,
|
| 1041 |
-
"num_tokens": 15336195.0,
|
| 1042 |
-
"step": 2700
|
| 1043 |
-
},
|
| 1044 |
-
{
|
| 1045 |
-
"epoch": 6.567290283645142,
|
| 1046 |
-
"grad_norm": 0.013374337926506996,
|
| 1047 |
-
"learning_rate": 6.091150333553274e-06,
|
| 1048 |
-
"loss": 0.0136,
|
| 1049 |
-
"mean_token_accuracy": 0.995315499305725,
|
| 1050 |
-
"num_tokens": 15480763.0,
|
| 1051 |
-
"step": 2725
|
| 1052 |
-
},
|
| 1053 |
-
{
|
| 1054 |
-
"epoch": 6.627640313820157,
|
| 1055 |
-
"grad_norm": 0.014570921659469604,
|
| 1056 |
-
"learning_rate": 4.528671475404433e-06,
|
| 1057 |
-
"loss": 0.0152,
|
| 1058 |
-
"mean_token_accuracy": 0.9951956886053085,
|
| 1059 |
-
"num_tokens": 15623624.0,
|
| 1060 |
-
"step": 2750
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 6.687990343995172,
|
| 1064 |
-
"grad_norm": 0.0181302297860384,
|
| 1065 |
-
"learning_rate": 3.1958568057067313e-06,
|
| 1066 |
-
"loss": 0.0142,
|
| 1067 |
-
"mean_token_accuracy": 0.9953425723314285,
|
| 1068 |
-
"num_tokens": 15765653.0,
|
| 1069 |
-
"step": 2775
|
| 1070 |
-
},
|
| 1071 |
-
{
|
| 1072 |
-
"epoch": 6.748340374170187,
|
| 1073 |
-
"grad_norm": 0.017040640115737915,
|
| 1074 |
-
"learning_rate": 2.093742295703127e-06,
|
| 1075 |
-
"loss": 0.0148,
|
| 1076 |
-
"mean_token_accuracy": 0.9950855672359467,
|
| 1077 |
-
"num_tokens": 15905979.0,
|
| 1078 |
-
"step": 2800
|
| 1079 |
-
},
|
| 1080 |
-
{
|
| 1081 |
-
"epoch": 6.808690404345202,
|
| 1082 |
-
"grad_norm": 0.016313739120960236,
|
| 1083 |
-
"learning_rate": 1.2231845978335708e-06,
|
| 1084 |
-
"loss": 0.0141,
|
| 1085 |
-
"mean_token_accuracy": 0.9952906262874603,
|
| 1086 |
-
"num_tokens": 16047220.0,
|
| 1087 |
-
"step": 2825
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 6.869040434520217,
|
| 1091 |
-
"grad_norm": 0.0156137989833951,
|
| 1092 |
-
"learning_rate": 5.848603798755402e-07,
|
| 1093 |
-
"loss": 0.0152,
|
| 1094 |
-
"mean_token_accuracy": 0.9951014250516892,
|
| 1095 |
-
"num_tokens": 16187543.0,
|
| 1096 |
-
"step": 2850
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 6.929390464695232,
|
| 1100 |
-
"grad_norm": 0.01440385077148676,
|
| 1101 |
-
"learning_rate": 1.7926579898319693e-07,
|
| 1102 |
-
"loss": 0.014,
|
| 1103 |
-
"mean_token_accuracy": 0.9952051192522049,
|
| 1104 |
-
"num_tokens": 16328399.0,
|
| 1105 |
-
"step": 2875
|
| 1106 |
-
},
|
| 1107 |
-
{
|
| 1108 |
-
"epoch": 6.989740494870247,
|
| 1109 |
-
"grad_norm": 0.014613240025937557,
|
| 1110 |
-
"learning_rate": 6.716116033844699e-09,
|
| 1111 |
-
"loss": 0.0148,
|
| 1112 |
-
"mean_token_accuracy": 0.9950145679712296,
|
| 1113 |
-
"num_tokens": 16459625.0,
|
| 1114 |
-
"step": 2900
|
| 1115 |
-
},
|
| 1116 |
-
{
|
| 1117 |
-
"epoch": 7.0,
|
| 1118 |
-
"eval_loss": 0.04441880062222481,
|
| 1119 |
-
"eval_mean_token_accuracy": 0.990020250307547,
|
| 1120 |
-
"eval_num_tokens": 16479260.0,
|
| 1121 |
-
"eval_runtime": 15.7702,
|
| 1122 |
-
"eval_samples_per_second": 23.398,
|
| 1123 |
-
"eval_steps_per_second": 11.731,
|
| 1124 |
-
"step": 2905
|
| 1125 |
}
|
| 1126 |
],
|
| 1127 |
"logging_steps": 25,
|
|
@@ -1136,12 +493,12 @@
|
|
| 1136 |
"should_evaluate": false,
|
| 1137 |
"should_log": false,
|
| 1138 |
"should_save": true,
|
| 1139 |
-
"should_training_stop":
|
| 1140 |
},
|
| 1141 |
"attributes": {}
|
| 1142 |
}
|
| 1143 |
},
|
| 1144 |
-
"total_flos":
|
| 1145 |
"train_batch_size": 2,
|
| 1146 |
"trial_name": null,
|
| 1147 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1245,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.060350030175015085,
|
| 14 |
+
"grad_norm": 0.3043883144855499,
|
| 15 |
+
"learning_rate": 0.00010909090909090909,
|
| 16 |
+
"loss": 1.8586,
|
| 17 |
+
"mean_token_accuracy": 0.6245462906360626,
|
| 18 |
+
"num_tokens": 157786.0,
|
| 19 |
"step": 25
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 0.12070006035003017,
|
| 23 |
+
"grad_norm": 0.24695894122123718,
|
| 24 |
+
"learning_rate": 0.00022272727272727272,
|
| 25 |
+
"loss": 0.9383,
|
| 26 |
+
"mean_token_accuracy": 0.7651280963420868,
|
| 27 |
+
"num_tokens": 283437.0,
|
| 28 |
"step": 50
|
| 29 |
},
|
| 30 |
{
|
| 31 |
"epoch": 0.18105009052504525,
|
| 32 |
+
"grad_norm": 0.19023272395133972,
|
| 33 |
+
"learning_rate": 0.0003363636363636364,
|
| 34 |
+
"loss": 0.649,
|
| 35 |
+
"mean_token_accuracy": 0.8234076803922653,
|
| 36 |
+
"num_tokens": 442299.0,
|
| 37 |
"step": 75
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.24140012070006034,
|
| 41 |
+
"grad_norm": 0.22401368618011475,
|
| 42 |
+
"learning_rate": 0.0003999849510577617,
|
| 43 |
+
"loss": 0.5125,
|
| 44 |
+
"mean_token_accuracy": 0.8562434083223343,
|
| 45 |
+
"num_tokens": 569621.0,
|
| 46 |
"step": 100
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.30175015087507545,
|
| 50 |
+
"grad_norm": 0.23774276673793793,
|
| 51 |
+
"learning_rate": 0.0003998388342637495,
|
| 52 |
+
"loss": 0.3972,
|
| 53 |
+
"mean_token_accuracy": 0.8866757136583329,
|
| 54 |
+
"num_tokens": 725959.0,
|
| 55 |
"step": 125
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 0.3621001810500905,
|
| 59 |
+
"grad_norm": 0.3460250496864319,
|
| 60 |
+
"learning_rate": 0.00039953738658223166,
|
| 61 |
+
"loss": 0.3554,
|
| 62 |
+
"mean_token_accuracy": 0.9000710541009903,
|
| 63 |
+
"num_tokens": 851583.0,
|
| 64 |
"step": 150
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"epoch": 0.4224502112251056,
|
| 68 |
+
"grad_norm": 0.2395441234111786,
|
| 69 |
+
"learning_rate": 0.00039908084232270096,
|
| 70 |
+
"loss": 0.2503,
|
| 71 |
+
"mean_token_accuracy": 0.9276190227270127,
|
| 72 |
+
"num_tokens": 1010006.0,
|
| 73 |
"step": 175
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.4828002414001207,
|
| 77 |
+
"grad_norm": 0.28712666034698486,
|
| 78 |
+
"learning_rate": 0.00039846955634824144,
|
| 79 |
+
"loss": 0.2602,
|
| 80 |
+
"mean_token_accuracy": 0.9267517280578613,
|
| 81 |
+
"num_tokens": 1137732.0,
|
| 82 |
"step": 200
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 0.5431502715751357,
|
| 86 |
+
"grad_norm": 0.17056156694889069,
|
| 87 |
+
"learning_rate": 0.00039770400379969973,
|
| 88 |
+
"loss": 0.1851,
|
| 89 |
+
"mean_token_accuracy": 0.9475770330429077,
|
| 90 |
+
"num_tokens": 1294237.0,
|
| 91 |
"step": 225
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 0.6035003017501509,
|
| 95 |
+
"grad_norm": 0.21380072832107544,
|
| 96 |
+
"learning_rate": 0.00039678477972636774,
|
| 97 |
+
"loss": 0.1622,
|
| 98 |
+
"mean_token_accuracy": 0.9546336072683335,
|
| 99 |
+
"num_tokens": 1419338.0,
|
| 100 |
"step": 250
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.663850331925166,
|
| 104 |
+
"grad_norm": 0.20465914905071259,
|
| 105 |
+
"learning_rate": 0.00039571259862346183,
|
| 106 |
+
"loss": 0.1329,
|
| 107 |
+
"mean_token_accuracy": 0.961662837266922,
|
| 108 |
+
"num_tokens": 1576946.0,
|
| 109 |
"step": 275
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 0.724200362100181,
|
| 113 |
+
"grad_norm": 0.2113582342863083,
|
| 114 |
+
"learning_rate": 0.00039448829387675954,
|
| 115 |
+
"loss": 0.1287,
|
| 116 |
+
"mean_token_accuracy": 0.9641147536039353,
|
| 117 |
+
"num_tokens": 1703683.0,
|
| 118 |
"step": 300
|
| 119 |
},
|
| 120 |
{
|
| 121 |
"epoch": 0.7845503922751962,
|
| 122 |
+
"grad_norm": 0.15074850618839264,
|
| 123 |
+
"learning_rate": 0.0003931128171148249,
|
| 124 |
+
"loss": 0.1232,
|
| 125 |
+
"mean_token_accuracy": 0.9665374368429184,
|
| 126 |
+
"num_tokens": 1860622.0,
|
| 127 |
"step": 325
|
| 128 |
},
|
| 129 |
{
|
| 130 |
"epoch": 0.8449004224502112,
|
| 131 |
+
"grad_norm": 0.4272288978099823,
|
| 132 |
+
"learning_rate": 0.00039158723746932566,
|
| 133 |
+
"loss": 0.0974,
|
| 134 |
+
"mean_token_accuracy": 0.9736644911766053,
|
| 135 |
+
"num_tokens": 1987884.0,
|
| 136 |
"step": 350
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.9052504526252263,
|
| 140 |
+
"grad_norm": 0.13680922985076904,
|
| 141 |
+
"learning_rate": 0.00038991274074401806,
|
| 142 |
+
"loss": 0.1007,
|
| 143 |
+
"mean_token_accuracy": 0.9723848593235016,
|
| 144 |
+
"num_tokens": 2145729.0,
|
| 145 |
"step": 375
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 0.9656004828002414,
|
| 149 |
+
"grad_norm": 0.15923215448856354,
|
| 150 |
+
"learning_rate": 0.00038809062849304407,
|
| 151 |
+
"loss": 0.0747,
|
| 152 |
+
"mean_token_accuracy": 0.9784966939687729,
|
| 153 |
+
"num_tokens": 2273076.0,
|
| 154 |
"step": 400
|
| 155 |
},
|
| 156 |
{
|
| 157 |
"epoch": 1.0,
|
| 158 |
+
"eval_loss": 0.09010029584169388,
|
| 159 |
+
"eval_mean_token_accuracy": 0.9750096942927386,
|
| 160 |
"eval_num_tokens": 2354180.0,
|
| 161 |
+
"eval_runtime": 15.8229,
|
| 162 |
+
"eval_samples_per_second": 23.321,
|
| 163 |
+
"eval_steps_per_second": 11.692,
|
| 164 |
"step": 415
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 1.024140012070006,
|
| 168 |
+
"grad_norm": 0.12324528396129608,
|
| 169 |
+
"learning_rate": 0.0003861223170092585,
|
| 170 |
+
"loss": 0.0923,
|
| 171 |
+
"mean_token_accuracy": 0.9740137457847595,
|
| 172 |
+
"num_tokens": 2422803.0,
|
| 173 |
"step": 425
|
| 174 |
},
|
| 175 |
{
|
| 176 |
"epoch": 1.0844900422450212,
|
| 177 |
+
"grad_norm": 0.21983321011066437,
|
| 178 |
+
"learning_rate": 0.00038400933622337167,
|
| 179 |
+
"loss": 0.0619,
|
| 180 |
+
"mean_token_accuracy": 0.9822656351327896,
|
| 181 |
+
"num_tokens": 2563793.0,
|
| 182 |
"step": 450
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 1.1448400724200363,
|
| 186 |
+
"grad_norm": 0.11064327508211136,
|
| 187 |
+
"learning_rate": 0.00038175332851476387,
|
| 188 |
+
"loss": 0.075,
|
| 189 |
+
"mean_token_accuracy": 0.9794844657182693,
|
| 190 |
+
"num_tokens": 2706197.0,
|
| 191 |
"step": 475
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 1.2051901025950513,
|
| 195 |
+
"grad_norm": 0.1068505346775055,
|
| 196 |
+
"learning_rate": 0.00037935604743489506,
|
| 197 |
+
"loss": 0.0544,
|
| 198 |
+
"mean_token_accuracy": 0.9840904027223587,
|
| 199 |
+
"num_tokens": 2850268.0,
|
| 200 |
"step": 500
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.2655401327700664,
|
| 204 |
+
"grad_norm": 0.15822124481201172,
|
| 205 |
+
"learning_rate": 0.00037681935634430327,
|
| 206 |
+
"loss": 0.078,
|
| 207 |
+
"mean_token_accuracy": 0.9782475352287292,
|
| 208 |
+
"num_tokens": 2992275.0,
|
| 209 |
"step": 525
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 1.3258901629450814,
|
| 213 |
+
"grad_norm": 0.10555300116539001,
|
| 214 |
+
"learning_rate": 0.0003741452269642502,
|
| 215 |
+
"loss": 0.0542,
|
| 216 |
+
"mean_token_accuracy": 0.9846927672624588,
|
| 217 |
+
"num_tokens": 3135599.0,
|
| 218 |
"step": 550
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 1.3862401931200965,
|
| 222 |
+
"grad_norm": 0.07533632218837738,
|
| 223 |
+
"learning_rate": 0.0003713357378441402,
|
| 224 |
+
"loss": 0.0641,
|
| 225 |
+
"mean_token_accuracy": 0.9822721928358078,
|
| 226 |
+
"num_tokens": 3278124.0,
|
| 227 |
"step": 575
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 1.4465902232951118,
|
| 231 |
+
"grad_norm": 0.1403859406709671,
|
| 232 |
+
"learning_rate": 0.00036839307274590355,
|
| 233 |
+
"loss": 0.0491,
|
| 234 |
+
"mean_token_accuracy": 0.985781243443489,
|
| 235 |
+
"num_tokens": 3421679.0,
|
| 236 |
"step": 600
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 1.5069402534701268,
|
| 240 |
+
"grad_norm": 0.10379917174577713,
|
| 241 |
+
"learning_rate": 0.00036531951894660034,
|
| 242 |
+
"loss": 0.069,
|
| 243 |
+
"mean_token_accuracy": 0.9809466338157654,
|
| 244 |
+
"num_tokens": 3564486.0,
|
| 245 |
"step": 625
|
| 246 |
},
|
| 247 |
{
|
| 248 |
"epoch": 1.567290283645142,
|
| 249 |
+
"grad_norm": 0.09850325435400009,
|
| 250 |
+
"learning_rate": 0.00036211746546056415,
|
| 251 |
+
"loss": 0.0516,
|
| 252 |
+
"mean_token_accuracy": 0.9853906160593033,
|
| 253 |
+
"num_tokens": 3707529.0,
|
| 254 |
"step": 650
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 1.627640313820157,
|
| 258 |
+
"grad_norm": 0.11964337527751923,
|
| 259 |
+
"learning_rate": 0.00035878940118246673,
|
| 260 |
+
"loss": 0.0596,
|
| 261 |
+
"mean_token_accuracy": 0.9836755973100663,
|
| 262 |
+
"num_tokens": 3848570.0,
|
| 263 |
"step": 675
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 1.687990343995172,
|
| 267 |
+
"grad_norm": 0.05955597385764122,
|
| 268 |
+
"learning_rate": 0.00035533791295274834,
|
| 269 |
+
"loss": 0.0428,
|
| 270 |
+
"mean_token_accuracy": 0.9875983273983002,
|
| 271 |
+
"num_tokens": 3990912.0,
|
| 272 |
"step": 700
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 1.748340374170187,
|
| 276 |
+
"grad_norm": 0.07114022970199585,
|
| 277 |
+
"learning_rate": 0.0003517656835469161,
|
| 278 |
+
"loss": 0.0627,
|
| 279 |
+
"mean_token_accuracy": 0.9830698877573013,
|
| 280 |
+
"num_tokens": 4131622.0,
|
| 281 |
"step": 725
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 1.8086904043452021,
|
| 285 |
+
"grad_norm": 0.09389431029558182,
|
| 286 |
+
"learning_rate": 0.0003480754895902742,
|
| 287 |
+
"loss": 0.0459,
|
| 288 |
+
"mean_token_accuracy": 0.9864954763650894,
|
| 289 |
+
"num_tokens": 4273640.0,
|
| 290 |
"step": 750
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 1.8690404345202172,
|
| 294 |
+
"grad_norm": 0.06497396528720856,
|
| 295 |
+
"learning_rate": 0.0003442701993997064,
|
| 296 |
+
"loss": 0.0577,
|
| 297 |
+
"mean_token_accuracy": 0.9838357955217362,
|
| 298 |
+
"num_tokens": 4414471.0,
|
| 299 |
"step": 775
|
| 300 |
},
|
| 301 |
{
|
| 302 |
"epoch": 1.9293904646952322,
|
| 303 |
+
"grad_norm": 0.11333642154932022,
|
| 304 |
+
"learning_rate": 0.00034035277075418854,
|
| 305 |
+
"loss": 0.0431,
|
| 306 |
+
"mean_token_accuracy": 0.9871519947052002,
|
| 307 |
+
"num_tokens": 4555376.0,
|
| 308 |
"step": 800
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 1.9897404948702473,
|
| 312 |
+
"grad_norm": 0.05740824714303017,
|
| 313 |
+
"learning_rate": 0.0003363262485957633,
|
| 314 |
+
"loss": 0.0418,
|
| 315 |
+
"mean_token_accuracy": 0.9883395010232925,
|
| 316 |
+
"num_tokens": 4688406.0,
|
| 317 |
"step": 825
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 2.0,
|
| 321 |
+
"eval_loss": 0.053012676537036896,
|
| 322 |
+
"eval_mean_token_accuracy": 0.9854460068651147,
|
| 323 |
"eval_num_tokens": 4708360.0,
|
| 324 |
+
"eval_runtime": 15.7874,
|
| 325 |
+
"eval_samples_per_second": 23.373,
|
| 326 |
+
"eval_steps_per_second": 11.718,
|
| 327 |
"step": 830
|
| 328 |
},
|
| 329 |
{
|
| 330 |
"epoch": 2.048280024140012,
|
| 331 |
+
"grad_norm": 0.09828540682792664,
|
| 332 |
+
"learning_rate": 0.00033219376266276594,
|
| 333 |
+
"loss": 0.0495,
|
| 334 |
+
"mean_token_accuracy": 0.9852321922164602,
|
| 335 |
+
"num_tokens": 4836946.0,
|
| 336 |
"step": 850
|
| 337 |
},
|
| 338 |
{
|
| 339 |
"epoch": 2.1086300543150274,
|
| 340 |
+
"grad_norm": 0.10078238695859909,
|
| 341 |
+
"learning_rate": 0.00032795852505713806,
|
| 342 |
+
"loss": 0.0313,
|
| 343 |
+
"mean_token_accuracy": 0.9908820760250091,
|
| 344 |
+
"num_tokens": 4968203.0,
|
| 345 |
"step": 875
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 2.1689800844900424,
|
| 349 |
+
"grad_norm": 0.07217393070459366,
|
| 350 |
+
"learning_rate": 0.0003236238277477231,
|
| 351 |
+
"loss": 0.0471,
|
| 352 |
+
"mean_token_accuracy": 0.985995357632637,
|
| 353 |
+
"num_tokens": 5118541.0,
|
| 354 |
"step": 900
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 2.2293301146650575,
|
| 358 |
+
"grad_norm": 0.09504982829093933,
|
| 359 |
+
"learning_rate": 0.0003191930400114816,
|
| 360 |
+
"loss": 0.0322,
|
| 361 |
+
"mean_token_accuracy": 0.9904332131147384,
|
| 362 |
+
"num_tokens": 5251007.0,
|
| 363 |
"step": 925
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 2.2896801448400725,
|
| 367 |
+
"grad_norm": 0.04393278807401657,
|
| 368 |
+
"learning_rate": 0.0003146696058146176,
|
| 369 |
+
"loss": 0.0481,
|
| 370 |
+
"mean_token_accuracy": 0.9860882490873337,
|
| 371 |
+
"num_tokens": 5403517.0,
|
| 372 |
"step": 950
|
| 373 |
},
|
| 374 |
{
|
| 375 |
"epoch": 2.3500301750150876,
|
| 376 |
+
"grad_norm": 0.12678726017475128,
|
| 377 |
+
"learning_rate": 0.00031005704113564917,
|
| 378 |
+
"loss": 0.0349,
|
| 379 |
+
"mean_token_accuracy": 0.9900296354293823,
|
| 380 |
+
"num_tokens": 5536492.0,
|
| 381 |
"step": 975
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 2.4103802051901027,
|
| 385 |
+
"grad_norm": 0.06465219706296921,
|
| 386 |
+
"learning_rate": 0.00030535893123250635,
|
| 387 |
+
"loss": 0.0484,
|
| 388 |
+
"mean_token_accuracy": 0.9857117992639541,
|
| 389 |
+
"num_tokens": 5688657.0,
|
| 390 |
"step": 1000
|
| 391 |
},
|
| 392 |
{
|
| 393 |
"epoch": 2.4707302353651177,
|
| 394 |
+
"grad_norm": 0.07911183685064316,
|
| 395 |
+
"learning_rate": 0.00030057892785577867,
|
| 396 |
+
"loss": 0.0313,
|
| 397 |
+
"mean_token_accuracy": 0.9907743036746979,
|
| 398 |
+
"num_tokens": 5821424.0,
|
| 399 |
"step": 1025
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 2.5310802655401328,
|
| 403 |
+
"grad_norm": 0.04166734963655472,
|
| 404 |
+
"learning_rate": 0.00029572074641027996,
|
| 405 |
+
"loss": 0.0448,
|
| 406 |
+
"mean_token_accuracy": 0.9870600712299347,
|
| 407 |
+
"num_tokens": 5970144.0,
|
| 408 |
"step": 1050
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 2.591430295715148,
|
| 412 |
+
"grad_norm": 0.09666065871715546,
|
| 413 |
+
"learning_rate": 0.0002907881630671351,
|
| 414 |
+
"loss": 0.0301,
|
| 415 |
+
"mean_token_accuracy": 0.9908553779125213,
|
| 416 |
+
"num_tokens": 6103689.0,
|
| 417 |
"step": 1075
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 2.651780325890163,
|
| 421 |
+
"grad_norm": 0.07375594228506088,
|
| 422 |
+
"learning_rate": 0.00028578501182863507,
|
| 423 |
+
"loss": 0.0425,
|
| 424 |
+
"mean_token_accuracy": 0.9875227802991867,
|
| 425 |
+
"num_tokens": 6256525.0,
|
| 426 |
"step": 1100
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 2.712130356065178,
|
| 430 |
+
"grad_norm": 0.086298368871212,
|
| 431 |
+
"learning_rate": 0.00028071518154814036,
|
| 432 |
+
"loss": 0.0323,
|
| 433 |
+
"mean_token_accuracy": 0.9902477955818176,
|
| 434 |
+
"num_tokens": 6390348.0,
|
| 435 |
"step": 1125
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 2.772480386240193,
|
| 439 |
+
"grad_norm": 0.03587740287184715,
|
| 440 |
+
"learning_rate": 0.0002755826129073503,
|
| 441 |
+
"loss": 0.0427,
|
| 442 |
+
"mean_token_accuracy": 0.9872915095090866,
|
| 443 |
+
"num_tokens": 6540900.0,
|
| 444 |
"step": 1150
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 2.832830416415208,
|
| 448 |
+
"grad_norm": 0.08814697712659836,
|
| 449 |
+
"learning_rate": 0.00027039129535328646,
|
| 450 |
+
"loss": 0.0328,
|
| 451 |
+
"mean_token_accuracy": 0.990228921175003,
|
| 452 |
+
"num_tokens": 6673531.0,
|
| 453 |
"step": 1175
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 2.8931804465902236,
|
| 457 |
+
"grad_norm": 0.054661527276039124,
|
| 458 |
+
"learning_rate": 0.00026514526399737235,
|
| 459 |
+
"loss": 0.0406,
|
| 460 |
+
"mean_token_accuracy": 0.9875594407320023,
|
| 461 |
+
"num_tokens": 6826045.0,
|
| 462 |
"step": 1200
|
| 463 |
},
|
| 464 |
{
|
| 465 |
"epoch": 2.9535304767652386,
|
| 466 |
+
"grad_norm": 0.04500816389918327,
|
| 467 |
+
"learning_rate": 0.00025984859647901865,
|
| 468 |
+
"loss": 0.0282,
|
| 469 |
+
"mean_token_accuracy": 0.9915571695566178,
|
| 470 |
+
"num_tokens": 6959176.0,
|
| 471 |
"step": 1225
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 3.0,
|
| 475 |
+
"eval_loss": 0.04558952525258064,
|
| 476 |
+
"eval_mean_token_accuracy": 0.9875019208804982,
|
| 477 |
"eval_num_tokens": 7062540.0,
|
| 478 |
+
"eval_runtime": 15.8013,
|
| 479 |
+
"eval_samples_per_second": 23.352,
|
| 480 |
+
"eval_steps_per_second": 11.708,
|
| 481 |
"step": 1245
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 482 |
}
|
| 483 |
],
|
| 484 |
"logging_steps": 25,
|
|
|
|
| 493 |
"should_evaluate": false,
|
| 494 |
"should_log": false,
|
| 495 |
"should_save": true,
|
| 496 |
+
"should_training_stop": false
|
| 497 |
},
|
| 498 |
"attributes": {}
|
| 499 |
}
|
| 500 |
},
|
| 501 |
+
"total_flos": 3.0683713883526144e+17,
|
| 502 |
"train_batch_size": 2,
|
| 503 |
"trial_name": null,
|
| 504 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6097
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb3c001872d0d895453223b44ff6bc35437b36d517310a2af89b31048c8561be
|
| 3 |
size 6097
|