Model save
Browse files- README.md +1 -2
- all_results.json +7 -7
- config.json +1 -1
- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- train_results.json +7 -7
- trainer_state.json +146 -356
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -4,7 +4,6 @@ library_name: transformers
|
|
| 4 |
model_name: llama_tuned
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
-
- alignment-handbook
|
| 8 |
- trl
|
| 9 |
- sft
|
| 10 |
licence: license
|
|
@@ -28,7 +27,7 @@ print(output["generated_text"])
|
|
| 28 |
|
| 29 |
## Training procedure
|
| 30 |
|
| 31 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/vedaantj/ft-llm/runs/
|
| 32 |
|
| 33 |
This model was trained with SFT.
|
| 34 |
|
|
|
|
| 4 |
model_name: llama_tuned
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
| 7 |
- trl
|
| 8 |
- sft
|
| 9 |
licence: license
|
|
|
|
| 27 |
|
| 28 |
## Training procedure
|
| 29 |
|
| 30 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/vedaantj/ft-llm/runs/oqwx6fvg)
|
| 31 |
|
| 32 |
This model was trained with SFT.
|
| 33 |
|
all_results.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
"eval_loss": 2.1213064193725586,
|
| 4 |
"eval_runtime": 12.4292,
|
| 5 |
"eval_samples": 1124,
|
| 6 |
"eval_samples_per_second": 90.432,
|
| 7 |
"eval_steps_per_second": 15.126,
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples":
|
| 12 |
-
"train_samples_per_second":
|
| 13 |
-
"train_steps_per_second": 0.
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
"eval_loss": 2.1213064193725586,
|
| 4 |
"eval_runtime": 12.4292,
|
| 5 |
"eval_samples": 1124,
|
| 6 |
"eval_samples_per_second": 90.432,
|
| 7 |
"eval_steps_per_second": 15.126,
|
| 8 |
+
"total_flos": 4.93214716919808e+16,
|
| 9 |
+
"train_loss": 0.09651691963275273,
|
| 10 |
+
"train_runtime": 472.7445,
|
| 11 |
+
"train_samples": 280,
|
| 12 |
+
"train_samples_per_second": 11.846,
|
| 13 |
+
"train_steps_per_second": 0.508
|
| 14 |
}
|
config.json
CHANGED
|
@@ -36,6 +36,6 @@
|
|
| 36 |
"tie_word_embeddings": true,
|
| 37 |
"torch_dtype": "bfloat16",
|
| 38 |
"transformers_version": "4.46.3",
|
| 39 |
-
"use_cache":
|
| 40 |
"vocab_size": 128256
|
| 41 |
}
|
|
|
|
| 36 |
"tie_word_embeddings": true,
|
| 37 |
"torch_dtype": "bfloat16",
|
| 38 |
"transformers_version": "4.46.3",
|
| 39 |
+
"use_cache": false,
|
| 40 |
"vocab_size": 128256
|
| 41 |
}
|
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4965799096
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d4922ba0d6c33910c0b33eab909a52ade69b4e118d8867a1baa2079be805da7e
|
| 3 |
size 4965799096
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2247734992
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6cbd84ce4fefdb9dcdcc33bde9a5722809b3a5d2a3ca4f0a36128f1de844cc77
|
| 3 |
size 2247734992
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 4.93214716919808e+16,
|
| 4 |
+
"train_loss": 0.09651691963275273,
|
| 5 |
+
"train_runtime": 472.7445,
|
| 6 |
+
"train_samples": 280,
|
| 7 |
+
"train_samples_per_second": 11.846,
|
| 8 |
+
"train_steps_per_second": 0.508
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,424 +1,214 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0
|
| 20 |
-
"
|
| 21 |
-
"
|
| 22 |
-
"
|
| 23 |
-
"
|
| 24 |
-
"
|
|
|
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 2.0,
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"
|
|
|
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch":
|
| 36 |
-
"
|
| 37 |
-
"
|
| 38 |
-
"
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
-
|
| 42 |
-
{
|
| 43 |
-
"epoch": 3.7333333333333334,
|
| 44 |
-
"grad_norm": 0.7766222357749939,
|
| 45 |
-
"learning_rate": 9.358288770053476e-05,
|
| 46 |
-
"loss": 0.582,
|
| 47 |
-
"step": 700
|
| 48 |
},
|
| 49 |
{
|
| 50 |
"epoch": 4.0,
|
| 51 |
-
"
|
| 52 |
-
"
|
| 53 |
-
"
|
| 54 |
-
"
|
| 55 |
-
"
|
|
|
|
| 56 |
},
|
| 57 |
{
|
| 58 |
-
"epoch":
|
| 59 |
-
"
|
| 60 |
-
"
|
| 61 |
-
"
|
| 62 |
-
"
|
| 63 |
-
"
|
|
|
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 6.0,
|
| 67 |
-
"
|
| 68 |
-
"
|
| 69 |
-
"
|
| 70 |
-
"
|
| 71 |
-
"
|
| 72 |
-
|
| 73 |
-
{
|
| 74 |
-
"epoch": 6.997333333333334,
|
| 75 |
-
"eval_loss": 1.200727939605713,
|
| 76 |
-
"eval_runtime": 12.5034,
|
| 77 |
-
"eval_samples_per_second": 89.896,
|
| 78 |
-
"eval_steps_per_second": 15.036,
|
| 79 |
-
"step": 1312
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 7.
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
"
|
| 86 |
-
"
|
|
|
|
|
|
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 8.0,
|
| 90 |
-
"
|
| 91 |
-
"
|
| 92 |
-
"
|
| 93 |
-
"
|
| 94 |
-
"
|
|
|
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch":
|
| 98 |
-
"
|
| 99 |
-
"
|
| 100 |
-
"
|
| 101 |
-
"
|
| 102 |
-
"
|
|
|
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"epoch": 10.0,
|
| 106 |
-
"
|
| 107 |
-
"
|
| 108 |
-
"
|
| 109 |
-
"
|
| 110 |
-
"
|
|
|
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch":
|
| 114 |
-
"
|
| 115 |
-
"
|
| 116 |
-
"
|
| 117 |
-
"
|
| 118 |
-
"
|
| 119 |
-
|
| 120 |
-
{
|
| 121 |
-
"epoch": 11.2,
|
| 122 |
-
"grad_norm": 0.3173273503780365,
|
| 123 |
-
"learning_rate": 0.00019501438995110992,
|
| 124 |
-
"loss": 0.1313,
|
| 125 |
-
"step": 2100
|
| 126 |
},
|
| 127 |
{
|
| 128 |
"epoch": 12.0,
|
| 129 |
-
"
|
| 130 |
-
"
|
| 131 |
-
"
|
| 132 |
-
"
|
| 133 |
-
"
|
|
|
|
| 134 |
},
|
| 135 |
{
|
| 136 |
-
"epoch":
|
| 137 |
-
"
|
| 138 |
-
"
|
| 139 |
-
"
|
| 140 |
-
"
|
| 141 |
-
"
|
|
|
|
| 142 |
},
|
| 143 |
{
|
| 144 |
"epoch": 14.0,
|
| 145 |
-
"
|
| 146 |
-
"
|
| 147 |
-
"
|
| 148 |
-
"
|
| 149 |
-
"
|
| 150 |
-
|
| 151 |
-
{
|
| 152 |
-
"epoch": 14.933333333333334,
|
| 153 |
-
"grad_norm": 0.269999235868454,
|
| 154 |
-
"learning_rate": 0.000177467306667847,
|
| 155 |
-
"loss": 0.0448,
|
| 156 |
-
"step": 2800
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch":
|
| 160 |
-
"
|
| 161 |
-
"
|
| 162 |
-
"
|
| 163 |
-
"
|
| 164 |
-
"
|
|
|
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 16.0,
|
| 168 |
-
"
|
| 169 |
-
"
|
| 170 |
-
"
|
| 171 |
-
"
|
| 172 |
-
"
|
|
|
|
| 173 |
},
|
| 174 |
{
|
| 175 |
-
"epoch":
|
| 176 |
-
"
|
| 177 |
-
"
|
| 178 |
-
"
|
| 179 |
-
"
|
| 180 |
-
"
|
|
|
|
| 181 |
},
|
| 182 |
{
|
| 183 |
"epoch": 18.0,
|
| 184 |
-
"
|
| 185 |
-
"
|
| 186 |
-
"
|
| 187 |
-
"
|
| 188 |
-
"
|
|
|
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch":
|
| 192 |
-
"
|
| 193 |
-
"
|
| 194 |
-
"
|
| 195 |
-
"
|
| 196 |
-
|
| 197 |
-
|
| 198 |
-
"epoch": 18.997333333333334,
|
| 199 |
-
"eval_loss": 1.7563754320144653,
|
| 200 |
-
"eval_runtime": 12.9626,
|
| 201 |
-
"eval_samples_per_second": 86.711,
|
| 202 |
-
"eval_steps_per_second": 14.503,
|
| 203 |
-
"step": 3562
|
| 204 |
},
|
| 205 |
{
|
| 206 |
"epoch": 20.0,
|
| 207 |
-
"
|
| 208 |
-
"
|
| 209 |
-
"
|
| 210 |
-
"
|
| 211 |
-
"
|
| 212 |
-
|
| 213 |
-
{
|
| 214 |
-
"epoch": 20.997333333333334,
|
| 215 |
-
"eval_loss": 1.7924139499664307,
|
| 216 |
-
"eval_runtime": 12.441,
|
| 217 |
-
"eval_samples_per_second": 90.346,
|
| 218 |
-
"eval_steps_per_second": 15.111,
|
| 219 |
-
"step": 3937
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 22.0,
|
| 223 |
-
"eval_loss": 1.832397222518921,
|
| 224 |
-
"eval_runtime": 12.4418,
|
| 225 |
-
"eval_samples_per_second": 90.34,
|
| 226 |
-
"eval_steps_per_second": 15.11,
|
| 227 |
-
"step": 4125
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 22.4,
|
| 231 |
-
"grad_norm": 0.10293902456760406,
|
| 232 |
-
"learning_rate": 0.0001150624200350232,
|
| 233 |
-
"loss": 0.0157,
|
| 234 |
-
"step": 4200
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch":
|
| 238 |
-
"
|
| 239 |
-
"
|
| 240 |
-
"
|
| 241 |
-
"
|
| 242 |
-
"
|
| 243 |
-
|
| 244 |
-
{
|
| 245 |
-
"epoch": 24.0,
|
| 246 |
-
"eval_loss": 1.8424901962280273,
|
| 247 |
-
"eval_runtime": 12.4549,
|
| 248 |
-
"eval_samples_per_second": 90.246,
|
| 249 |
-
"eval_steps_per_second": 15.094,
|
| 250 |
-
"step": 4500
|
| 251 |
-
},
|
| 252 |
-
{
|
| 253 |
-
"epoch": 24.997333333333334,
|
| 254 |
-
"eval_loss": 1.8898138999938965,
|
| 255 |
-
"eval_runtime": 12.4368,
|
| 256 |
-
"eval_samples_per_second": 90.377,
|
| 257 |
-
"eval_steps_per_second": 15.116,
|
| 258 |
-
"step": 4687
|
| 259 |
-
},
|
| 260 |
-
{
|
| 261 |
-
"epoch": 26.0,
|
| 262 |
-
"eval_loss": 1.9125876426696777,
|
| 263 |
-
"eval_runtime": 12.4412,
|
| 264 |
-
"eval_samples_per_second": 90.345,
|
| 265 |
-
"eval_steps_per_second": 15.111,
|
| 266 |
-
"step": 4875
|
| 267 |
-
},
|
| 268 |
-
{
|
| 269 |
-
"epoch": 26.133333333333333,
|
| 270 |
-
"grad_norm": 0.07291631400585175,
|
| 271 |
-
"learning_rate": 7.853831770392755e-05,
|
| 272 |
-
"loss": 0.0061,
|
| 273 |
-
"step": 4900
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 26.997333333333334,
|
| 277 |
-
"eval_loss": 1.9254367351531982,
|
| 278 |
-
"eval_runtime": 12.4413,
|
| 279 |
-
"eval_samples_per_second": 90.345,
|
| 280 |
-
"eval_steps_per_second": 15.111,
|
| 281 |
-
"step": 5062
|
| 282 |
-
},
|
| 283 |
-
{
|
| 284 |
-
"epoch": 28.0,
|
| 285 |
-
"eval_loss": 1.975632667541504,
|
| 286 |
-
"eval_runtime": 12.4478,
|
| 287 |
-
"eval_samples_per_second": 90.297,
|
| 288 |
-
"eval_steps_per_second": 15.103,
|
| 289 |
-
"step": 5250
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 28.997333333333334,
|
| 293 |
-
"eval_loss": 1.9888043403625488,
|
| 294 |
-
"eval_runtime": 12.4518,
|
| 295 |
-
"eval_samples_per_second": 90.268,
|
| 296 |
-
"eval_steps_per_second": 15.098,
|
| 297 |
-
"step": 5437
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 29.866666666666667,
|
| 301 |
-
"grad_norm": 0.000998605857603252,
|
| 302 |
-
"learning_rate": 4.4880260858357746e-05,
|
| 303 |
-
"loss": 0.0011,
|
| 304 |
-
"step": 5600
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 30.0,
|
| 308 |
-
"eval_loss": 2.0155458450317383,
|
| 309 |
-
"eval_runtime": 12.4346,
|
| 310 |
-
"eval_samples_per_second": 90.393,
|
| 311 |
-
"eval_steps_per_second": 15.119,
|
| 312 |
-
"step": 5625
|
| 313 |
-
},
|
| 314 |
-
{
|
| 315 |
-
"epoch": 30.997333333333334,
|
| 316 |
-
"eval_loss": 2.048884153366089,
|
| 317 |
-
"eval_runtime": 12.435,
|
| 318 |
-
"eval_samples_per_second": 90.39,
|
| 319 |
-
"eval_steps_per_second": 15.119,
|
| 320 |
-
"step": 5812
|
| 321 |
-
},
|
| 322 |
-
{
|
| 323 |
-
"epoch": 32.0,
|
| 324 |
-
"eval_loss": 2.070535898208618,
|
| 325 |
-
"eval_runtime": 12.426,
|
| 326 |
-
"eval_samples_per_second": 90.455,
|
| 327 |
-
"eval_steps_per_second": 15.13,
|
| 328 |
-
"step": 6000
|
| 329 |
-
},
|
| 330 |
-
{
|
| 331 |
-
"epoch": 32.99733333333333,
|
| 332 |
-
"eval_loss": 2.087083339691162,
|
| 333 |
-
"eval_runtime": 12.428,
|
| 334 |
-
"eval_samples_per_second": 90.441,
|
| 335 |
-
"eval_steps_per_second": 15.127,
|
| 336 |
-
"step": 6187
|
| 337 |
-
},
|
| 338 |
-
{
|
| 339 |
-
"epoch": 33.6,
|
| 340 |
-
"grad_norm": 0.0002252462727483362,
|
| 341 |
-
"learning_rate": 1.8583028637387568e-05,
|
| 342 |
-
"loss": 0.0001,
|
| 343 |
-
"step": 6300
|
| 344 |
-
},
|
| 345 |
-
{
|
| 346 |
-
"epoch": 34.0,
|
| 347 |
-
"eval_loss": 2.099353551864624,
|
| 348 |
-
"eval_runtime": 12.4444,
|
| 349 |
-
"eval_samples_per_second": 90.322,
|
| 350 |
-
"eval_steps_per_second": 15.107,
|
| 351 |
-
"step": 6375
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 34.99733333333333,
|
| 355 |
-
"eval_loss": 2.1083507537841797,
|
| 356 |
-
"eval_runtime": 12.431,
|
| 357 |
-
"eval_samples_per_second": 90.419,
|
| 358 |
-
"eval_steps_per_second": 15.123,
|
| 359 |
-
"step": 6562
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 36.0,
|
| 363 |
-
"eval_loss": 2.114396333694458,
|
| 364 |
-
"eval_runtime": 12.8335,
|
| 365 |
-
"eval_samples_per_second": 87.583,
|
| 366 |
-
"eval_steps_per_second": 14.649,
|
| 367 |
-
"step": 6750
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 36.99733333333333,
|
| 371 |
-
"eval_loss": 2.1184046268463135,
|
| 372 |
-
"eval_runtime": 12.4403,
|
| 373 |
-
"eval_samples_per_second": 90.352,
|
| 374 |
-
"eval_steps_per_second": 15.112,
|
| 375 |
-
"step": 6937
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 37.333333333333336,
|
| 379 |
-
"grad_norm": 0.00023204727040138096,
|
| 380 |
-
"learning_rate": 3.1584176769466346e-06,
|
| 381 |
-
"loss": 0.0,
|
| 382 |
-
"step": 7000
|
| 383 |
-
},
|
| 384 |
-
{
|
| 385 |
-
"epoch": 38.0,
|
| 386 |
-
"eval_loss": 2.12043833732605,
|
| 387 |
-
"eval_runtime": 12.6027,
|
| 388 |
-
"eval_samples_per_second": 89.187,
|
| 389 |
-
"eval_steps_per_second": 14.917,
|
| 390 |
-
"step": 7125
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 38.99733333333333,
|
| 394 |
-
"eval_loss": 2.121260643005371,
|
| 395 |
-
"eval_runtime": 12.4106,
|
| 396 |
-
"eval_samples_per_second": 90.568,
|
| 397 |
-
"eval_steps_per_second": 15.148,
|
| 398 |
-
"step": 7312
|
| 399 |
-
},
|
| 400 |
-
{
|
| 401 |
-
"epoch": 39.89333333333333,
|
| 402 |
-
"eval_loss": 2.1213064193725586,
|
| 403 |
-
"eval_runtime": 12.42,
|
| 404 |
-
"eval_samples_per_second": 90.499,
|
| 405 |
-
"eval_steps_per_second": 15.137,
|
| 406 |
-
"step": 7480
|
| 407 |
-
},
|
| 408 |
-
{
|
| 409 |
-
"epoch": 39.89333333333333,
|
| 410 |
-
"step": 7480,
|
| 411 |
-
"total_flos": 1.743333613371392e+18,
|
| 412 |
-
"train_loss": 0.10288844243324496,
|
| 413 |
-
"train_runtime": 9416.1259,
|
| 414 |
-
"train_samples_per_second": 19.108,
|
| 415 |
-
"train_steps_per_second": 0.794
|
| 416 |
}
|
| 417 |
],
|
| 418 |
"logging_steps": 700,
|
| 419 |
-
"max_steps":
|
| 420 |
"num_input_tokens_seen": 0,
|
| 421 |
-
"num_train_epochs":
|
| 422 |
"save_steps": 500,
|
| 423 |
"stateful_callbacks": {
|
| 424 |
"TrainerControl": {
|
|
@@ -432,7 +222,7 @@
|
|
| 432 |
"attributes": {}
|
| 433 |
}
|
| 434 |
},
|
| 435 |
-
"total_flos":
|
| 436 |
"train_batch_size": 4,
|
| 437 |
"trial_name": null,
|
| 438 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 240,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.08333333333333333,
|
| 13 |
+
"grad_norm": 5.815704345703125,
|
| 14 |
+
"learning_rate": 4.166666666666667e-06,
|
| 15 |
+
"loss": 0.9944,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy_fol": 1.0,
|
| 21 |
+
"eval_loss": 0.37742123007774353,
|
| 22 |
+
"eval_runtime": 10.6636,
|
| 23 |
+
"eval_samples_per_second": 6.471,
|
| 24 |
+
"eval_steps_per_second": 1.125,
|
| 25 |
+
"step": 12
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 2.0,
|
| 29 |
+
"eval_accuracy_fol": 0.9855072463768116,
|
| 30 |
+
"eval_loss": 0.36990299820899963,
|
| 31 |
+
"eval_runtime": 10.0571,
|
| 32 |
+
"eval_samples_per_second": 6.861,
|
| 33 |
+
"eval_steps_per_second": 1.193,
|
| 34 |
+
"step": 24
|
| 35 |
},
|
| 36 |
{
|
| 37 |
+
"epoch": 3.0,
|
| 38 |
+
"eval_accuracy_fol": 1.0,
|
| 39 |
+
"eval_loss": 0.4505714774131775,
|
| 40 |
+
"eval_runtime": 10.0403,
|
| 41 |
+
"eval_samples_per_second": 6.872,
|
| 42 |
+
"eval_steps_per_second": 1.195,
|
| 43 |
+
"step": 36
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
},
|
| 45 |
{
|
| 46 |
"epoch": 4.0,
|
| 47 |
+
"eval_accuracy_fol": 1.0,
|
| 48 |
+
"eval_loss": 0.49083924293518066,
|
| 49 |
+
"eval_runtime": 10.1223,
|
| 50 |
+
"eval_samples_per_second": 6.817,
|
| 51 |
+
"eval_steps_per_second": 1.186,
|
| 52 |
+
"step": 48
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 5.0,
|
| 56 |
+
"eval_accuracy_fol": 1.0,
|
| 57 |
+
"eval_loss": 0.551354706287384,
|
| 58 |
+
"eval_runtime": 9.943,
|
| 59 |
+
"eval_samples_per_second": 6.94,
|
| 60 |
+
"eval_steps_per_second": 1.207,
|
| 61 |
+
"step": 60
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 6.0,
|
| 65 |
+
"eval_accuracy_fol": 0.9565217391304348,
|
| 66 |
+
"eval_loss": 0.5758901834487915,
|
| 67 |
+
"eval_runtime": 10.0414,
|
| 68 |
+
"eval_samples_per_second": 6.872,
|
| 69 |
+
"eval_steps_per_second": 1.195,
|
| 70 |
+
"step": 72
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
},
|
| 72 |
{
|
| 73 |
+
"epoch": 7.0,
|
| 74 |
+
"eval_accuracy_fol": 1.0,
|
| 75 |
+
"eval_loss": 0.6346259117126465,
|
| 76 |
+
"eval_runtime": 10.0121,
|
| 77 |
+
"eval_samples_per_second": 6.892,
|
| 78 |
+
"eval_steps_per_second": 1.199,
|
| 79 |
+
"step": 84
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 8.0,
|
| 83 |
+
"eval_accuracy_fol": 0.9855072463768116,
|
| 84 |
+
"eval_loss": 0.6786300539970398,
|
| 85 |
+
"eval_runtime": 9.853,
|
| 86 |
+
"eval_samples_per_second": 7.003,
|
| 87 |
+
"eval_steps_per_second": 1.218,
|
| 88 |
+
"step": 96
|
| 89 |
},
|
| 90 |
{
|
| 91 |
+
"epoch": 9.0,
|
| 92 |
+
"eval_accuracy_fol": 0.9855072463768116,
|
| 93 |
+
"eval_loss": 0.7070556282997131,
|
| 94 |
+
"eval_runtime": 10.0666,
|
| 95 |
+
"eval_samples_per_second": 6.854,
|
| 96 |
+
"eval_steps_per_second": 1.192,
|
| 97 |
+
"step": 108
|
| 98 |
},
|
| 99 |
{
|
| 100 |
"epoch": 10.0,
|
| 101 |
+
"eval_accuracy_fol": 1.0,
|
| 102 |
+
"eval_loss": 0.7301931381225586,
|
| 103 |
+
"eval_runtime": 9.8811,
|
| 104 |
+
"eval_samples_per_second": 6.983,
|
| 105 |
+
"eval_steps_per_second": 1.214,
|
| 106 |
+
"step": 120
|
| 107 |
},
|
| 108 |
{
|
| 109 |
+
"epoch": 11.0,
|
| 110 |
+
"eval_accuracy_fol": 1.0,
|
| 111 |
+
"eval_loss": 0.7777052521705627,
|
| 112 |
+
"eval_runtime": 10.0443,
|
| 113 |
+
"eval_samples_per_second": 6.87,
|
| 114 |
+
"eval_steps_per_second": 1.195,
|
| 115 |
+
"step": 132
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 12.0,
|
| 119 |
+
"eval_accuracy_fol": 1.0,
|
| 120 |
+
"eval_loss": 0.7906161546707153,
|
| 121 |
+
"eval_runtime": 9.9021,
|
| 122 |
+
"eval_samples_per_second": 6.968,
|
| 123 |
+
"eval_steps_per_second": 1.212,
|
| 124 |
+
"step": 144
|
| 125 |
},
|
| 126 |
{
|
| 127 |
+
"epoch": 13.0,
|
| 128 |
+
"eval_accuracy_fol": 1.0,
|
| 129 |
+
"eval_loss": 0.7852453589439392,
|
| 130 |
+
"eval_runtime": 10.0952,
|
| 131 |
+
"eval_samples_per_second": 6.835,
|
| 132 |
+
"eval_steps_per_second": 1.189,
|
| 133 |
+
"step": 156
|
| 134 |
},
|
| 135 |
{
|
| 136 |
"epoch": 14.0,
|
| 137 |
+
"eval_accuracy_fol": 1.0,
|
| 138 |
+
"eval_loss": 0.8031890392303467,
|
| 139 |
+
"eval_runtime": 9.9012,
|
| 140 |
+
"eval_samples_per_second": 6.969,
|
| 141 |
+
"eval_steps_per_second": 1.212,
|
| 142 |
+
"step": 168
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 15.0,
|
| 146 |
+
"eval_accuracy_fol": 1.0,
|
| 147 |
+
"eval_loss": 0.8141315579414368,
|
| 148 |
+
"eval_runtime": 10.0811,
|
| 149 |
+
"eval_samples_per_second": 6.845,
|
| 150 |
+
"eval_steps_per_second": 1.19,
|
| 151 |
+
"step": 180
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 16.0,
|
| 155 |
+
"eval_accuracy_fol": 1.0,
|
| 156 |
+
"eval_loss": 0.8203307390213013,
|
| 157 |
+
"eval_runtime": 10.0556,
|
| 158 |
+
"eval_samples_per_second": 6.862,
|
| 159 |
+
"eval_steps_per_second": 1.193,
|
| 160 |
+
"step": 192
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"epoch": 17.0,
|
| 164 |
+
"eval_accuracy_fol": 1.0,
|
| 165 |
+
"eval_loss": 0.8233786225318909,
|
| 166 |
+
"eval_runtime": 10.038,
|
| 167 |
+
"eval_samples_per_second": 6.874,
|
| 168 |
+
"eval_steps_per_second": 1.195,
|
| 169 |
+
"step": 204
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 18.0,
|
| 173 |
+
"eval_accuracy_fol": 1.0,
|
| 174 |
+
"eval_loss": 0.8255409002304077,
|
| 175 |
+
"eval_runtime": 9.8663,
|
| 176 |
+
"eval_samples_per_second": 6.993,
|
| 177 |
+
"eval_steps_per_second": 1.216,
|
| 178 |
+
"step": 216
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 19.0,
|
| 182 |
+
"eval_accuracy_fol": 1.0,
|
| 183 |
+
"eval_loss": 0.8259706497192383,
|
| 184 |
+
"eval_runtime": 9.9818,
|
| 185 |
+
"eval_samples_per_second": 6.913,
|
| 186 |
+
"eval_steps_per_second": 1.202,
|
| 187 |
+
"step": 228
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 188 |
},
|
| 189 |
{
|
| 190 |
"epoch": 20.0,
|
| 191 |
+
"eval_accuracy_fol": 1.0,
|
| 192 |
+
"eval_loss": 0.8260682821273804,
|
| 193 |
+
"eval_runtime": 9.8198,
|
| 194 |
+
"eval_samples_per_second": 7.027,
|
| 195 |
+
"eval_steps_per_second": 1.222,
|
| 196 |
+
"step": 240
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 197 |
},
|
| 198 |
{
|
| 199 |
+
"epoch": 20.0,
|
| 200 |
+
"step": 240,
|
| 201 |
+
"total_flos": 4.93214716919808e+16,
|
| 202 |
+
"train_loss": 0.09651691963275273,
|
| 203 |
+
"train_runtime": 472.7445,
|
| 204 |
+
"train_samples_per_second": 11.846,
|
| 205 |
+
"train_steps_per_second": 0.508
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 206 |
}
|
| 207 |
],
|
| 208 |
"logging_steps": 700,
|
| 209 |
+
"max_steps": 240,
|
| 210 |
"num_input_tokens_seen": 0,
|
| 211 |
+
"num_train_epochs": 20,
|
| 212 |
"save_steps": 500,
|
| 213 |
"stateful_callbacks": {
|
| 214 |
"TrainerControl": {
|
|
|
|
| 222 |
"attributes": {}
|
| 223 |
}
|
| 224 |
},
|
| 225 |
+
"total_flos": 4.93214716919808e+16,
|
| 226 |
"train_batch_size": 4,
|
| 227 |
"trial_name": null,
|
| 228 |
"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 6968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f52ac1e5f794b25e82a6d31d954a349d1b11035ec6c6fd322a263d592362a70f
|
| 3 |
size 6968
|