End of training
Browse files- all_results.json +5 -5
- model.safetensors +1 -1
- train_results.json +5 -5
- trainer_state.json +138 -898
- training_args.bin +1 -1
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 4.617630838765763e-06,
|
| 4 |
+
"train_runtime": 463.4946,
|
| 5 |
+
"train_samples": 599,
|
| 6 |
+
"train_samples_per_second": 0.69,
|
| 7 |
+
"train_steps_per_second": 0.043
|
| 8 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1976163472
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a57fd08d5b5d0ca9650c6d1a6a1f2bdab037daed303c9a4ec056d64690a2823
|
| 3 |
size 1976163472
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 4.617630838765763e-06,
|
| 4 |
+
"train_runtime": 463.4946,
|
| 5 |
+
"train_samples": 599,
|
| 6 |
+
"train_samples_per_second": 0.69,
|
| 7 |
+
"train_steps_per_second": 0.043
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -2,975 +2,215 @@
|
|
| 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 |
-
"completion_length":
|
| 14 |
-
"epoch": 0.
|
| 15 |
-
"grad_norm":
|
| 16 |
"kl": 0.0,
|
| 17 |
-
"learning_rate":
|
| 18 |
"loss": 0.0,
|
| 19 |
-
"reward":
|
| 20 |
-
"reward_std": 0.
|
| 21 |
-
"rewards/concensus_correctness_reward_func": 0.
|
| 22 |
-
"rewards/consensus_reward_func":
|
| 23 |
"rewards/cumulative_reward_2": 0.0,
|
| 24 |
-
"rewards/final_correctness_reward_func": 0.
|
| 25 |
-
"rewards/question_recreation_reward_func": 0.
|
| 26 |
"rewards/soft_format_reward_func": 0.0,
|
| 27 |
-
"rewards/strict_format_reward_func": 0.
|
| 28 |
-
"rewards/xmlcount_reward_func": 0.
|
| 29 |
"step": 2
|
| 30 |
},
|
| 31 |
{
|
| 32 |
-
"completion_length":
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"kl":
|
| 36 |
-
"learning_rate":
|
| 37 |
"loss": 0.0,
|
| 38 |
-
"reward":
|
| 39 |
-
"reward_std":
|
| 40 |
-
"rewards/concensus_correctness_reward_func":
|
| 41 |
-
"rewards/consensus_reward_func":
|
| 42 |
"rewards/cumulative_reward_2": 0.0,
|
| 43 |
-
"rewards/final_correctness_reward_func": 0.
|
| 44 |
-
"rewards/question_recreation_reward_func":
|
| 45 |
"rewards/soft_format_reward_func": 0.0,
|
| 46 |
-
"rewards/strict_format_reward_func": 0.
|
| 47 |
-
"rewards/xmlcount_reward_func":
|
| 48 |
"step": 4
|
| 49 |
},
|
| 50 |
{
|
| 51 |
-
"completion_length":
|
| 52 |
-
"epoch": 0.
|
| 53 |
-
"grad_norm":
|
| 54 |
-
"kl": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
"loss": 0.0,
|
| 57 |
-
"reward":
|
| 58 |
-
"reward_std": 0.
|
| 59 |
-
"rewards/concensus_correctness_reward_func":
|
| 60 |
-
"rewards/consensus_reward_func":
|
| 61 |
"rewards/cumulative_reward_2": 0.0,
|
| 62 |
-
"rewards/final_correctness_reward_func": 0.
|
| 63 |
-
"rewards/question_recreation_reward_func":
|
| 64 |
"rewards/soft_format_reward_func": 0.0,
|
| 65 |
-
"rewards/strict_format_reward_func": 0.
|
| 66 |
-
"rewards/xmlcount_reward_func":
|
| 67 |
"step": 6
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"completion_length":
|
| 71 |
-
"epoch": 0.
|
| 72 |
-
"grad_norm":
|
| 73 |
-
"kl": 0.
|
| 74 |
-
"learning_rate":
|
| 75 |
"loss": 0.0,
|
| 76 |
-
"reward":
|
| 77 |
-
"reward_std": 0.
|
| 78 |
-
"rewards/concensus_correctness_reward_func":
|
| 79 |
-
"rewards/consensus_reward_func":
|
| 80 |
"rewards/cumulative_reward_2": 0.0,
|
| 81 |
"rewards/final_correctness_reward_func": 0.125,
|
| 82 |
-
"rewards/question_recreation_reward_func":
|
| 83 |
"rewards/soft_format_reward_func": 0.0,
|
| 84 |
-
"rewards/strict_format_reward_func": 0.
|
| 85 |
-
"rewards/xmlcount_reward_func":
|
| 86 |
"step": 8
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"completion_length":
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm":
|
| 92 |
-
"kl": 0.
|
| 93 |
-
"learning_rate":
|
| 94 |
"loss": 0.0,
|
| 95 |
-
"reward":
|
| 96 |
-
"reward_std": 0.
|
| 97 |
-
"rewards/concensus_correctness_reward_func":
|
| 98 |
-
"rewards/consensus_reward_func":
|
| 99 |
"rewards/cumulative_reward_2": 0.0,
|
| 100 |
-
"rewards/final_correctness_reward_func": 0.
|
| 101 |
-
"rewards/question_recreation_reward_func":
|
| 102 |
"rewards/soft_format_reward_func": 0.0,
|
| 103 |
-
"rewards/strict_format_reward_func": 0.
|
| 104 |
-
"rewards/xmlcount_reward_func":
|
| 105 |
"step": 10
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"completion_length":
|
| 109 |
-
"epoch": 0.
|
| 110 |
-
"grad_norm":
|
| 111 |
-
"kl": 0.
|
| 112 |
-
"learning_rate": 1.
|
| 113 |
-
"loss": 0.
|
| 114 |
-
"reward":
|
| 115 |
-
"reward_std": 0.
|
| 116 |
-
"rewards/concensus_correctness_reward_func":
|
| 117 |
-
"rewards/consensus_reward_func":
|
| 118 |
"rewards/cumulative_reward_2": 0.0,
|
| 119 |
-
"rewards/final_correctness_reward_func": 0.
|
| 120 |
-
"rewards/question_recreation_reward_func":
|
| 121 |
"rewards/soft_format_reward_func": 0.0,
|
| 122 |
-
"rewards/strict_format_reward_func": 0.
|
| 123 |
-
"rewards/xmlcount_reward_func":
|
| 124 |
"step": 12
|
| 125 |
},
|
| 126 |
{
|
| 127 |
-
"completion_length":
|
| 128 |
-
"epoch": 0.
|
| 129 |
-
"grad_norm":
|
| 130 |
-
"kl": 0.
|
| 131 |
-
"learning_rate": 1.
|
| 132 |
"loss": 0.0,
|
| 133 |
-
"reward":
|
| 134 |
-
"reward_std":
|
| 135 |
-
"rewards/concensus_correctness_reward_func":
|
| 136 |
-
"rewards/consensus_reward_func":
|
| 137 |
"rewards/cumulative_reward_2": 0.0,
|
| 138 |
-
"rewards/final_correctness_reward_func": 0.
|
| 139 |
-
"rewards/question_recreation_reward_func":
|
| 140 |
"rewards/soft_format_reward_func": 0.0,
|
| 141 |
-
"rewards/strict_format_reward_func": 0.
|
| 142 |
-
"rewards/xmlcount_reward_func":
|
| 143 |
"step": 14
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"completion_length":
|
| 147 |
-
"epoch": 0.
|
| 148 |
-
"grad_norm":
|
| 149 |
-
"kl": 0.
|
| 150 |
-
"learning_rate":
|
| 151 |
"loss": 0.0,
|
| 152 |
-
"reward":
|
| 153 |
-
"reward_std": 0.
|
| 154 |
-
"rewards/concensus_correctness_reward_func":
|
| 155 |
-
"rewards/consensus_reward_func":
|
| 156 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 157 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 158 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 159 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 160 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 161 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 162 |
-
"step": 16
|
| 163 |
-
},
|
| 164 |
-
{
|
| 165 |
-
"completion_length": 107.6875,
|
| 166 |
-
"epoch": 0.18,
|
| 167 |
-
"grad_norm": 0.608063280582428,
|
| 168 |
-
"kl": 0.012466257388950908,
|
| 169 |
-
"learning_rate": 1.1305772944380141e-06,
|
| 170 |
-
"loss": 0.0002,
|
| 171 |
-
"reward": 7.613874942064285,
|
| 172 |
-
"reward_std": 0.022097086533904076,
|
| 173 |
-
"rewards/concensus_correctness_reward_func": 2.3795000091195107,
|
| 174 |
-
"rewards/consensus_reward_func": 2.0,
|
| 175 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 176 |
-
"rewards/final_correctness_reward_func": 0.5,
|
| 177 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 178 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 179 |
-
"rewards/strict_format_reward_func": 0.484375,
|
| 180 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 181 |
-
"step": 18
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"completion_length": 113.71875,
|
| 185 |
-
"epoch": 0.2,
|
| 186 |
-
"grad_norm": 0.03589734435081482,
|
| 187 |
-
"kl": 0.002400611226789806,
|
| 188 |
-
"learning_rate": 1.1113300941935815e-06,
|
| 189 |
-
"loss": 0.0,
|
| 190 |
-
"reward": 6.913101881742477,
|
| 191 |
-
"reward_std": 0.0002745024103205651,
|
| 192 |
-
"rewards/concensus_correctness_reward_func": 2.038499988615513,
|
| 193 |
-
"rewards/consensus_reward_func": 2.0,
|
| 194 |
"rewards/cumulative_reward_2": 0.0,
|
| 195 |
"rewards/final_correctness_reward_func": 0.125,
|
| 196 |
-
"rewards/question_recreation_reward_func": 0.
|
| 197 |
"rewards/soft_format_reward_func": 0.0,
|
| 198 |
-
"rewards/strict_format_reward_func": 0.
|
| 199 |
-
"rewards/xmlcount_reward_func":
|
| 200 |
-
"step":
|
| 201 |
-
},
|
| 202 |
-
{
|
| 203 |
-
"completion_length": 128.1875,
|
| 204 |
-
"epoch": 0.22,
|
| 205 |
-
"grad_norm": 0.017112350091338158,
|
| 206 |
-
"kl": 0.0014598190628021257,
|
| 207 |
-
"learning_rate": 1.0899381978343265e-06,
|
| 208 |
-
"loss": 0.0,
|
| 209 |
-
"reward": 6.634874999523163,
|
| 210 |
-
"reward_std": 0.0,
|
| 211 |
-
"rewards/concensus_correctness_reward_func": 1.884875014424324,
|
| 212 |
-
"rewards/consensus_reward_func": 2.0,
|
| 213 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 214 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 215 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 216 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 217 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 218 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 219 |
-
"step": 22
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"completion_length": 112.6875,
|
| 223 |
-
"epoch": 0.24,
|
| 224 |
-
"grad_norm": 0.048251923173666,
|
| 225 |
-
"kl": 0.0009265093149224413,
|
| 226 |
-
"learning_rate": 1.0664913304077896e-06,
|
| 227 |
-
"loss": 0.0,
|
| 228 |
-
"reward": 6.918624967336655,
|
| 229 |
-
"reward_std": 0.0,
|
| 230 |
-
"rewards/concensus_correctness_reward_func": 2.0436250120401382,
|
| 231 |
-
"rewards/consensus_reward_func": 2.0,
|
| 232 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 233 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 234 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 235 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 236 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 237 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 238 |
-
"step": 24
|
| 239 |
},
|
| 240 |
{
|
| 241 |
-
"completion_length":
|
| 242 |
-
"epoch": 0.
|
| 243 |
-
"grad_norm":
|
| 244 |
-
"kl": 0.
|
| 245 |
-
"learning_rate": 1.
|
| 246 |
"loss": 0.0,
|
| 247 |
-
"reward":
|
| 248 |
-
"reward_std": 0.
|
| 249 |
-
"rewards/concensus_correctness_reward_func":
|
| 250 |
-
"rewards/consensus_reward_func":
|
| 251 |
"rewards/cumulative_reward_2": 0.0,
|
| 252 |
"rewards/final_correctness_reward_func": 0.0625,
|
| 253 |
-
"rewards/question_recreation_reward_func": 0.
|
| 254 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 255 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 256 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 257 |
-
"step": 26
|
| 258 |
-
},
|
| 259 |
-
{
|
| 260 |
-
"completion_length": 105.0625,
|
| 261 |
-
"epoch": 0.28,
|
| 262 |
-
"grad_norm": 0.02676808461546898,
|
| 263 |
-
"kl": 0.0018403055837552529,
|
| 264 |
-
"learning_rate": 1.0138342663668626e-06,
|
| 265 |
-
"loss": 0.0,
|
| 266 |
-
"reward": 6.66587495803833,
|
| 267 |
-
"reward_std": 0.0,
|
| 268 |
-
"rewards/concensus_correctness_reward_func": 1.9158750250935555,
|
| 269 |
-
"rewards/consensus_reward_func": 2.0,
|
| 270 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 271 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 272 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 273 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 274 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 275 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 276 |
-
"step": 28
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"completion_length": 113.34375,
|
| 280 |
-
"epoch": 0.3,
|
| 281 |
-
"grad_norm": 0.02248120680451393,
|
| 282 |
-
"kl": 0.001570980580709147,
|
| 283 |
-
"learning_rate": 9.848449317786839e-07,
|
| 284 |
-
"loss": 0.0,
|
| 285 |
-
"reward": 6.9081249833106995,
|
| 286 |
-
"reward_std": 0.0,
|
| 287 |
-
"rewards/concensus_correctness_reward_func": 2.0331250056624413,
|
| 288 |
-
"rewards/consensus_reward_func": 2.0,
|
| 289 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 290 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 291 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 292 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 293 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 294 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 295 |
-
"step": 30
|
| 296 |
-
},
|
| 297 |
-
{
|
| 298 |
-
"completion_length": 111.875,
|
| 299 |
-
"epoch": 0.32,
|
| 300 |
-
"grad_norm": 0.06672874093055725,
|
| 301 |
-
"kl": 0.002423093282232003,
|
| 302 |
-
"learning_rate": 9.542414238035536e-07,
|
| 303 |
-
"loss": 0.0,
|
| 304 |
-
"reward": 7.160874992609024,
|
| 305 |
-
"reward_std": 0.0,
|
| 306 |
-
"rewards/concensus_correctness_reward_func": 2.1608749851584435,
|
| 307 |
-
"rewards/consensus_reward_func": 2.0,
|
| 308 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 309 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 310 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 311 |
"rewards/soft_format_reward_func": 0.0,
|
| 312 |
-
"rewards/strict_format_reward_func": 0.
|
| 313 |
-
"rewards/xmlcount_reward_func":
|
| 314 |
-
"step":
|
| 315 |
-
},
|
| 316 |
-
{
|
| 317 |
-
"completion_length": 124.71875,
|
| 318 |
-
"epoch": 0.34,
|
| 319 |
-
"grad_norm": 0.15347696840763092,
|
| 320 |
-
"kl": 0.0032288330085066264,
|
| 321 |
-
"learning_rate": 9.221521041893904e-07,
|
| 322 |
-
"loss": 0.0,
|
| 323 |
-
"reward": 7.154421359300613,
|
| 324 |
-
"reward_std": 0.004177047871053219,
|
| 325 |
-
"rewards/concensus_correctness_reward_func": 2.157375007867813,
|
| 326 |
-
"rewards/consensus_reward_func": 2.0,
|
| 327 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 328 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 329 |
-
"rewards/question_recreation_reward_func": 0.9970463812351227,
|
| 330 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 331 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 332 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 333 |
-
"step": 34
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"completion_length": 124.125,
|
| 337 |
-
"epoch": 0.36,
|
| 338 |
-
"grad_norm": 0.20969295501708984,
|
| 339 |
-
"kl": 0.006022399147695978,
|
| 340 |
-
"learning_rate": 8.88711566694483e-07,
|
| 341 |
-
"loss": 0.0001,
|
| 342 |
-
"reward": 7.163678020238876,
|
| 343 |
-
"reward_std": 0.00010184785787714645,
|
| 344 |
-
"rewards/concensus_correctness_reward_func": 2.163750007748604,
|
| 345 |
-
"rewards/consensus_reward_func": 2.0,
|
| 346 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 347 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 348 |
-
"rewards/question_recreation_reward_func": 0.9999279975891113,
|
| 349 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 350 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 351 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 352 |
-
"step": 36
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"completion_length": 113.40625,
|
| 356 |
-
"epoch": 0.38,
|
| 357 |
-
"grad_norm": 0.016258783638477325,
|
| 358 |
-
"kl": 0.0014696058724439354,
|
| 359 |
-
"learning_rate": 8.540600725544942e-07,
|
| 360 |
-
"loss": 0.0,
|
| 361 |
-
"reward": 6.662374943494797,
|
| 362 |
-
"reward_std": 0.0,
|
| 363 |
-
"rewards/concensus_correctness_reward_func": 1.9123750254511833,
|
| 364 |
-
"rewards/consensus_reward_func": 2.0,
|
| 365 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 366 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 367 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 368 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 369 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 370 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 371 |
-
"step": 38
|
| 372 |
-
},
|
| 373 |
-
{
|
| 374 |
-
"completion_length": 105.71875,
|
| 375 |
-
"epoch": 0.4,
|
| 376 |
-
"grad_norm": 0.3598341643810272,
|
| 377 |
-
"kl": 0.011058556609668813,
|
| 378 |
-
"learning_rate": 8.183429621780925e-07,
|
| 379 |
-
"loss": 0.0002,
|
| 380 |
-
"reward": 7.289874881505966,
|
| 381 |
-
"reward_std": 0.0,
|
| 382 |
-
"rewards/concensus_correctness_reward_func": 2.289875015616417,
|
| 383 |
-
"rewards/consensus_reward_func": 2.0,
|
| 384 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 385 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 386 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 387 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 388 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 389 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 390 |
-
"step": 40
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"completion_length": 105.59375,
|
| 394 |
-
"epoch": 0.42,
|
| 395 |
-
"grad_norm": 0.01676054112613201,
|
| 396 |
-
"kl": 0.001547790251834158,
|
| 397 |
-
"learning_rate": 7.817100455387703e-07,
|
| 398 |
-
"loss": 0.0,
|
| 399 |
-
"reward": 7.168750017881393,
|
| 400 |
-
"reward_std": 0.0,
|
| 401 |
-
"rewards/concensus_correctness_reward_func": 2.1687499955296516,
|
| 402 |
-
"rewards/consensus_reward_func": 2.0,
|
| 403 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 404 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 405 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 406 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 407 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 408 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 409 |
-
"step": 42
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"completion_length": 116.09375,
|
| 413 |
-
"epoch": 0.44,
|
| 414 |
-
"grad_norm": 0.040649134665727615,
|
| 415 |
-
"kl": 0.0027483347312227124,
|
| 416 |
-
"learning_rate": 7.443149738197521e-07,
|
| 417 |
-
"loss": 0.0,
|
| 418 |
-
"reward": 7.121432453393936,
|
| 419 |
-
"reward_std": 0.008820239454507828,
|
| 420 |
-
"rewards/concensus_correctness_reward_func": 2.137750022113323,
|
| 421 |
-
"rewards/consensus_reward_func": 2.0,
|
| 422 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 423 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 424 |
-
"rewards/question_recreation_reward_func": 0.9836825020611286,
|
| 425 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 426 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 427 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 428 |
-
"step": 44
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"completion_length": 114.40625,
|
| 432 |
-
"epoch": 0.46,
|
| 433 |
-
"grad_norm": 0.01834481954574585,
|
| 434 |
-
"kl": 0.011484232741167943,
|
| 435 |
-
"learning_rate": 7.063145949475337e-07,
|
| 436 |
-
"loss": 0.0002,
|
| 437 |
-
"reward": 7.043624967336655,
|
| 438 |
-
"reward_std": 0.0,
|
| 439 |
-
"rewards/concensus_correctness_reward_func": 2.168625019490719,
|
| 440 |
-
"rewards/consensus_reward_func": 1.875,
|
| 441 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 442 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 443 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 444 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 445 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 446 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 447 |
-
"step": 46
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"completion_length": 120.0,
|
| 451 |
-
"epoch": 0.48,
|
| 452 |
-
"grad_norm": 0.04231150075793266,
|
| 453 |
-
"kl": 0.017316492816462414,
|
| 454 |
-
"learning_rate": 6.678682957171706e-07,
|
| 455 |
-
"loss": 0.0003,
|
| 456 |
-
"reward": 6.907124996185303,
|
| 457 |
-
"reward_std": 0.0,
|
| 458 |
-
"rewards/concensus_correctness_reward_func": 2.0321250036358833,
|
| 459 |
-
"rewards/consensus_reward_func": 2.0,
|
| 460 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 461 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 462 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 463 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 464 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 465 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 466 |
-
"step": 48
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"completion_length": 105.71875,
|
| 470 |
-
"epoch": 0.5,
|
| 471 |
-
"grad_norm": 0.9996789693832397,
|
| 472 |
-
"kl": 0.0018611156865517842,
|
| 473 |
-
"learning_rate": 6.29137333268669e-07,
|
| 474 |
-
"loss": 0.0,
|
| 475 |
-
"reward": 7.403761506080627,
|
| 476 |
-
"reward_std": 0.00033736444311216474,
|
| 477 |
-
"rewards/concensus_correctness_reward_func": 2.2790000066161156,
|
| 478 |
-
"rewards/consensus_reward_func": 2.0,
|
| 479 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 480 |
-
"rewards/final_correctness_reward_func": 0.375,
|
| 481 |
-
"rewards/question_recreation_reward_func": 0.999761451035738,
|
| 482 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 483 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 484 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 485 |
-
"step": 50
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"completion_length": 111.34375,
|
| 489 |
-
"epoch": 0.52,
|
| 490 |
-
"grad_norm": 0.04733134061098099,
|
| 491 |
-
"kl": 0.0036280953381719883,
|
| 492 |
-
"learning_rate": 5.902841587184991e-07,
|
| 493 |
-
"loss": 0.0001,
|
| 494 |
-
"reward": 6.6659999787807465,
|
| 495 |
-
"reward_std": 0.0,
|
| 496 |
-
"rewards/concensus_correctness_reward_func": 1.9160000085830688,
|
| 497 |
-
"rewards/consensus_reward_func": 2.0,
|
| 498 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 499 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 500 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 501 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 502 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 503 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 504 |
-
"step": 52
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"completion_length": 115.375,
|
| 508 |
-
"epoch": 0.54,
|
| 509 |
-
"grad_norm": 0.0800234004855156,
|
| 510 |
-
"kl": 0.014154426544337184,
|
| 511 |
-
"learning_rate": 5.514717357831576e-07,
|
| 512 |
-
"loss": 0.0002,
|
| 513 |
-
"reward": 7.356687515974045,
|
| 514 |
-
"reward_std": 0.26171789318323135,
|
| 515 |
-
"rewards/concensus_correctness_reward_func": 2.294187508523464,
|
| 516 |
-
"rewards/consensus_reward_func": 1.875,
|
| 517 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 518 |
-
"rewards/final_correctness_reward_func": 0.4375,
|
| 519 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 520 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 521 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 522 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 523 |
-
"step": 54
|
| 524 |
-
},
|
| 525 |
-
{
|
| 526 |
-
"completion_length": 116.65625,
|
| 527 |
-
"epoch": 0.56,
|
| 528 |
-
"grad_norm": 0.05929900333285332,
|
| 529 |
-
"kl": 0.010289387573720887,
|
| 530 |
-
"learning_rate": 5.12862857252706e-07,
|
| 531 |
-
"loss": 0.0002,
|
| 532 |
-
"reward": 7.364749997854233,
|
| 533 |
-
"reward_std": 0.0,
|
| 534 |
-
"rewards/concensus_correctness_reward_func": 2.239749997854233,
|
| 535 |
-
"rewards/consensus_reward_func": 2.0,
|
| 536 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 537 |
-
"rewards/final_correctness_reward_func": 0.375,
|
| 538 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 539 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 540 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 541 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 542 |
-
"step": 56
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"completion_length": 113.65625,
|
| 546 |
-
"epoch": 0.58,
|
| 547 |
-
"grad_norm": 0.08536501228809357,
|
| 548 |
-
"kl": 0.0039269381459234864,
|
| 549 |
-
"learning_rate": 4.7461946218123493e-07,
|
| 550 |
-
"loss": 0.0001,
|
| 551 |
-
"reward": 6.906499952077866,
|
| 552 |
-
"reward_std": 0.0,
|
| 553 |
-
"rewards/concensus_correctness_reward_func": 2.031499996781349,
|
| 554 |
-
"rewards/consensus_reward_func": 2.0,
|
| 555 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 556 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 557 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 558 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 559 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 560 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 561 |
-
"step": 58
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"completion_length": 112.6875,
|
| 565 |
-
"epoch": 0.6,
|
| 566 |
-
"grad_norm": 0.0589258149266243,
|
| 567 |
-
"kl": 0.005026609829656081,
|
| 568 |
-
"learning_rate": 4.369019566581942e-07,
|
| 569 |
-
"loss": 0.0001,
|
| 570 |
-
"reward": 7.163875013589859,
|
| 571 |
-
"reward_std": 0.0,
|
| 572 |
-
"rewards/concensus_correctness_reward_func": 2.163875013589859,
|
| 573 |
-
"rewards/consensus_reward_func": 2.0,
|
| 574 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 575 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 576 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 577 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 578 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 579 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 580 |
-
"step": 60
|
| 581 |
-
},
|
| 582 |
-
{
|
| 583 |
-
"completion_length": 115.5,
|
| 584 |
-
"epoch": 0.62,
|
| 585 |
-
"grad_norm": 1.4010002613067627,
|
| 586 |
-
"kl": 0.01616358994942857,
|
| 587 |
-
"learning_rate": 3.9986854100950773e-07,
|
| 588 |
-
"loss": 0.0002,
|
| 589 |
-
"reward": 6.842124968767166,
|
| 590 |
-
"reward_std": 0.0,
|
| 591 |
-
"rewards/concensus_correctness_reward_func": 1.9671249985694885,
|
| 592 |
-
"rewards/consensus_reward_func": 2.0,
|
| 593 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 594 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 595 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 596 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 597 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 598 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 599 |
-
"step": 62
|
| 600 |
-
},
|
| 601 |
-
{
|
| 602 |
-
"completion_length": 110.21875,
|
| 603 |
-
"epoch": 0.64,
|
| 604 |
-
"grad_norm": 0.04362049698829651,
|
| 605 |
-
"kl": 0.009225194641658163,
|
| 606 |
-
"learning_rate": 3.6367454625042783e-07,
|
| 607 |
-
"loss": 0.0001,
|
| 608 |
-
"reward": 7.585927993059158,
|
| 609 |
-
"reward_std": 0.0,
|
| 610 |
-
"rewards/concensus_correctness_reward_func": 2.3382499888539314,
|
| 611 |
-
"rewards/consensus_reward_func": 2.0,
|
| 612 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 613 |
-
"rewards/final_correctness_reward_func": 0.5,
|
| 614 |
-
"rewards/question_recreation_reward_func": 0.9976780191063881,
|
| 615 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 616 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 617 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 618 |
-
"step": 64
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"completion_length": 117.0625,
|
| 622 |
-
"epoch": 0.66,
|
| 623 |
-
"grad_norm": 0.04796240106225014,
|
| 624 |
-
"kl": 0.0031241076030710246,
|
| 625 |
-
"learning_rate": 3.284717825732767e-07,
|
| 626 |
-
"loss": 0.0,
|
| 627 |
-
"reward": 6.894343733787537,
|
| 628 |
-
"reward_std": 0.027621358633041382,
|
| 629 |
-
"rewards/concensus_correctness_reward_func": 2.038875013589859,
|
| 630 |
-
"rewards/consensus_reward_func": 2.0,
|
| 631 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 632 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 633 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 634 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 635 |
-
"rewards/strict_format_reward_func": 0.484375,
|
| 636 |
-
"rewards/xmlcount_reward_func": 1.24609375,
|
| 637 |
-
"step": 66
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"completion_length": 120.53125,
|
| 641 |
-
"epoch": 0.68,
|
| 642 |
-
"grad_norm": 0.6480430960655212,
|
| 643 |
-
"kl": 0.06253671760896395,
|
| 644 |
-
"learning_rate": 2.944079026027424e-07,
|
| 645 |
-
"loss": 0.0009,
|
| 646 |
-
"reward": 7.561367034912109,
|
| 647 |
-
"reward_std": 0.08847127005719813,
|
| 648 |
-
"rewards/concensus_correctness_reward_func": 2.374874994158745,
|
| 649 |
-
"rewards/consensus_reward_func": 2.0,
|
| 650 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 651 |
-
"rewards/final_correctness_reward_func": 0.4375,
|
| 652 |
-
"rewards/question_recreation_reward_func": 0.9989920035004616,
|
| 653 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 654 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 655 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 656 |
-
"step": 68
|
| 657 |
-
},
|
| 658 |
-
{
|
| 659 |
-
"completion_length": 109.6875,
|
| 660 |
-
"epoch": 0.7,
|
| 661 |
-
"grad_norm": 0.05288118124008179,
|
| 662 |
-
"kl": 0.10043014965231123,
|
| 663 |
-
"learning_rate": 2.6162578208945604e-07,
|
| 664 |
-
"loss": 0.0015,
|
| 665 |
-
"reward": 7.849874943494797,
|
| 666 |
-
"reward_std": 0.0,
|
| 667 |
-
"rewards/concensus_correctness_reward_func": 2.474875002168119,
|
| 668 |
-
"rewards/consensus_reward_func": 1.875,
|
| 669 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 670 |
-
"rewards/final_correctness_reward_func": 0.75,
|
| 671 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 672 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 673 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 674 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 675 |
-
"step": 70
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"completion_length": 110.3125,
|
| 679 |
-
"epoch": 0.72,
|
| 680 |
-
"grad_norm": 0.016276629641652107,
|
| 681 |
-
"kl": 0.0029021119621575053,
|
| 682 |
-
"learning_rate": 2.3026292063944357e-07,
|
| 683 |
-
"loss": 0.0,
|
| 684 |
-
"reward": 7.09412494301796,
|
| 685 |
-
"reward_std": 0.0883883461356163,
|
| 686 |
-
"rewards/concensus_correctness_reward_func": 2.156625010073185,
|
| 687 |
-
"rewards/consensus_reward_func": 2.0,
|
| 688 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 689 |
-
"rewards/final_correctness_reward_func": 0.1875,
|
| 690 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 691 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 692 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 693 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 694 |
-
"step": 72
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"completion_length": 116.0625,
|
| 698 |
-
"epoch": 0.74,
|
| 699 |
-
"grad_norm": 0.0852767825126648,
|
| 700 |
-
"kl": 0.005175539316041977,
|
| 701 |
-
"learning_rate": 2.0045086499299148e-07,
|
| 702 |
-
"loss": 0.0001,
|
| 703 |
-
"reward": 7.029082745313644,
|
| 704 |
-
"reward_std": 0.0,
|
| 705 |
-
"rewards/concensus_correctness_reward_func": 2.1545000076293945,
|
| 706 |
-
"rewards/consensus_reward_func": 1.875,
|
| 707 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 708 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 709 |
-
"rewards/question_recreation_reward_func": 0.9995827786624432,
|
| 710 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 711 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 712 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 713 |
-
"step": 74
|
| 714 |
-
},
|
| 715 |
-
{
|
| 716 |
-
"completion_length": 118.71875,
|
| 717 |
-
"epoch": 0.76,
|
| 718 |
-
"grad_norm": 0.11004471033811569,
|
| 719 |
-
"kl": 0.006573738556653552,
|
| 720 |
-
"learning_rate": 1.723146572719026e-07,
|
| 721 |
-
"loss": 0.0001,
|
| 722 |
-
"reward": 6.78612494468689,
|
| 723 |
-
"reward_std": 0.0,
|
| 724 |
-
"rewards/concensus_correctness_reward_func": 1.9111249893903732,
|
| 725 |
-
"rewards/consensus_reward_func": 2.0,
|
| 726 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 727 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 728 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 729 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 730 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 731 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 732 |
-
"step": 76
|
| 733 |
-
},
|
| 734 |
-
{
|
| 735 |
-
"completion_length": 109.6875,
|
| 736 |
-
"epoch": 0.78,
|
| 737 |
-
"grad_norm": 0.012919344939291477,
|
| 738 |
-
"kl": 0.005183771483643795,
|
| 739 |
-
"learning_rate": 1.459723105093828e-07,
|
| 740 |
-
"loss": 0.0001,
|
| 741 |
-
"reward": 6.658572971820831,
|
| 742 |
-
"reward_std": 7.354625267907977e-05,
|
| 743 |
-
"rewards/concensus_correctness_reward_func": 1.9086250066757202,
|
| 744 |
-
"rewards/consensus_reward_func": 2.0,
|
| 745 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 746 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 747 |
-
"rewards/question_recreation_reward_func": 0.9999480023980141,
|
| 748 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 749 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 750 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 751 |
-
"step": 78
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"completion_length": 112.875,
|
| 755 |
-
"epoch": 0.8,
|
| 756 |
-
"grad_norm": 0.021166132763028145,
|
| 757 |
-
"kl": 0.007018942174909171,
|
| 758 |
-
"learning_rate": 1.2153431366236765e-07,
|
| 759 |
-
"loss": 0.0001,
|
| 760 |
-
"reward": 6.880374938249588,
|
| 761 |
-
"reward_std": 0.0,
|
| 762 |
-
"rewards/concensus_correctness_reward_func": 2.0053750053048134,
|
| 763 |
-
"rewards/consensus_reward_func": 2.0,
|
| 764 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 765 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 766 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 767 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 768 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 769 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 770 |
-
"step": 80
|
| 771 |
-
},
|
| 772 |
-
{
|
| 773 |
-
"completion_length": 118.28125,
|
| 774 |
-
"epoch": 0.82,
|
| 775 |
-
"grad_norm": 0.7134923338890076,
|
| 776 |
-
"kl": 0.01929802937729619,
|
| 777 |
-
"learning_rate": 9.910316818243741e-08,
|
| 778 |
-
"loss": 0.0003,
|
| 779 |
-
"reward": 7.154624938964844,
|
| 780 |
-
"reward_std": 0.0,
|
| 781 |
-
"rewards/concensus_correctness_reward_func": 2.154625006020069,
|
| 782 |
-
"rewards/consensus_reward_func": 2.0,
|
| 783 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 784 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 785 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 786 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 787 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 788 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 789 |
-
"step": 82
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"completion_length": 115.1875,
|
| 793 |
-
"epoch": 0.84,
|
| 794 |
-
"grad_norm": 1.3774880170822144,
|
| 795 |
-
"kl": 0.005105306326640857,
|
| 796 |
-
"learning_rate": 7.877295808910932e-08,
|
| 797 |
-
"loss": 0.0001,
|
| 798 |
-
"reward": 6.884448528289795,
|
| 799 |
-
"reward_std": 0.022340359166264534,
|
| 800 |
-
"rewards/concensus_correctness_reward_func": 2.025500014424324,
|
| 801 |
-
"rewards/consensus_reward_func": 2.0,
|
| 802 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 803 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 804 |
-
"rewards/question_recreation_reward_func": 0.9995735697448254,
|
| 805 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 806 |
-
"rewards/strict_format_reward_func": 0.484375,
|
| 807 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 808 |
-
"step": 84
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"completion_length": 113.625,
|
| 812 |
-
"epoch": 0.86,
|
| 813 |
-
"grad_norm": 0.03088958002626896,
|
| 814 |
-
"kl": 0.0026084017076755117,
|
| 815 |
-
"learning_rate": 6.062895534876287e-08,
|
| 816 |
-
"loss": 0.0,
|
| 817 |
-
"reward": 6.6597499549388885,
|
| 818 |
-
"reward_std": 0.0,
|
| 819 |
-
"rewards/concensus_correctness_reward_func": 1.9097500145435333,
|
| 820 |
-
"rewards/consensus_reward_func": 2.0,
|
| 821 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 822 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 823 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 824 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 825 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 826 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 827 |
-
"step": 86
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"completion_length": 117.15625,
|
| 831 |
-
"epoch": 0.88,
|
| 832 |
-
"grad_norm": 0.06499552726745605,
|
| 833 |
-
"kl": 0.004322356182001386,
|
| 834 |
-
"learning_rate": 4.474726221437752e-08,
|
| 835 |
-
"loss": 0.0001,
|
| 836 |
-
"reward": 7.145511776208878,
|
| 837 |
-
"reward_std": 0.0,
|
| 838 |
-
"rewards/concensus_correctness_reward_func": 2.1503750011324883,
|
| 839 |
-
"rewards/consensus_reward_func": 2.0,
|
| 840 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 841 |
-
"rewards/final_correctness_reward_func": 0.25,
|
| 842 |
-
"rewards/question_recreation_reward_func": 0.995136771351099,
|
| 843 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 844 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 845 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 846 |
-
"step": 88
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"completion_length": 119.4375,
|
| 850 |
-
"epoch": 0.9,
|
| 851 |
-
"grad_norm": 0.09833624958992004,
|
| 852 |
-
"kl": 0.006649176628343412,
|
| 853 |
-
"learning_rate": 3.119449202623774e-08,
|
| 854 |
-
"loss": 0.0001,
|
| 855 |
-
"reward": 6.901250004768372,
|
| 856 |
-
"reward_std": 0.0,
|
| 857 |
-
"rewards/concensus_correctness_reward_func": 2.0262500047683716,
|
| 858 |
-
"rewards/consensus_reward_func": 2.0,
|
| 859 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 860 |
-
"rewards/final_correctness_reward_func": 0.125,
|
| 861 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 862 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 863 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 864 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 865 |
-
"step": 90
|
| 866 |
-
},
|
| 867 |
-
{
|
| 868 |
-
"completion_length": 111.125,
|
| 869 |
-
"epoch": 0.92,
|
| 870 |
-
"grad_norm": 0.029738876968622208,
|
| 871 |
-
"kl": 0.003505616401525913,
|
| 872 |
-
"learning_rate": 2.0027489812433317e-08,
|
| 873 |
-
"loss": 0.0001,
|
| 874 |
-
"reward": 6.53612494468689,
|
| 875 |
-
"reward_std": 0.0,
|
| 876 |
-
"rewards/concensus_correctness_reward_func": 1.9111249893903732,
|
| 877 |
-
"rewards/consensus_reward_func": 1.875,
|
| 878 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 879 |
-
"rewards/final_correctness_reward_func": 0.0,
|
| 880 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 881 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 882 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 883 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 884 |
-
"step": 92
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"completion_length": 123.40625,
|
| 888 |
-
"epoch": 0.94,
|
| 889 |
-
"grad_norm": 0.026749616488814354,
|
| 890 |
-
"kl": 0.0031833851862757,
|
| 891 |
-
"learning_rate": 1.1293093861056569e-08,
|
| 892 |
-
"loss": 0.0,
|
| 893 |
-
"reward": 7.906374961137772,
|
| 894 |
-
"reward_std": 0.0,
|
| 895 |
-
"rewards/concensus_correctness_reward_func": 2.5313749983906746,
|
| 896 |
-
"rewards/consensus_reward_func": 2.0,
|
| 897 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 898 |
-
"rewards/final_correctness_reward_func": 0.625,
|
| 899 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 900 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 901 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 902 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 903 |
-
"step": 94
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"completion_length": 124.125,
|
| 907 |
-
"epoch": 0.96,
|
| 908 |
-
"grad_norm": 0.7131932973861694,
|
| 909 |
-
"kl": 0.0020799475960302516,
|
| 910 |
-
"learning_rate": 5.027939264144954e-09,
|
| 911 |
-
"loss": 0.0,
|
| 912 |
-
"reward": 7.410148322582245,
|
| 913 |
-
"reward_std": 0.0004973328104824759,
|
| 914 |
-
"rewards/concensus_correctness_reward_func": 2.285500004887581,
|
| 915 |
-
"rewards/consensus_reward_func": 2.0,
|
| 916 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 917 |
-
"rewards/final_correctness_reward_func": 0.375,
|
| 918 |
-
"rewards/question_recreation_reward_func": 0.9996483325958252,
|
| 919 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 920 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 921 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 922 |
-
"step": 96
|
| 923 |
-
},
|
| 924 |
-
{
|
| 925 |
-
"completion_length": 118.875,
|
| 926 |
-
"epoch": 0.98,
|
| 927 |
-
"grad_norm": 0.04078225791454315,
|
| 928 |
-
"kl": 0.008958728666584648,
|
| 929 |
-
"learning_rate": 1.2583042573730107e-09,
|
| 930 |
-
"loss": 0.0001,
|
| 931 |
-
"reward": 7.291375011205673,
|
| 932 |
-
"reward_std": 0.0,
|
| 933 |
-
"rewards/concensus_correctness_reward_func": 2.2913750037550926,
|
| 934 |
-
"rewards/consensus_reward_func": 1.875,
|
| 935 |
-
"rewards/cumulative_reward_2": 0.0,
|
| 936 |
-
"rewards/final_correctness_reward_func": 0.375,
|
| 937 |
-
"rewards/question_recreation_reward_func": 1.0,
|
| 938 |
-
"rewards/soft_format_reward_func": 0.0,
|
| 939 |
-
"rewards/strict_format_reward_func": 0.5,
|
| 940 |
-
"rewards/xmlcount_reward_func": 1.25,
|
| 941 |
-
"step": 98
|
| 942 |
},
|
| 943 |
{
|
| 944 |
-
"completion_length":
|
| 945 |
-
"epoch":
|
| 946 |
-
"grad_norm":
|
| 947 |
-
"kl": 0.
|
| 948 |
"learning_rate": 0.0,
|
| 949 |
"loss": 0.0,
|
| 950 |
-
"reward":
|
| 951 |
-
"reward_std": 0.
|
| 952 |
-
"rewards/concensus_correctness_reward_func":
|
| 953 |
-
"rewards/consensus_reward_func":
|
| 954 |
"rewards/cumulative_reward_2": 0.0,
|
| 955 |
-
"rewards/final_correctness_reward_func": 0.
|
| 956 |
-
"rewards/question_recreation_reward_func":
|
| 957 |
"rewards/soft_format_reward_func": 0.0,
|
| 958 |
-
"rewards/strict_format_reward_func": 0.
|
| 959 |
-
"rewards/xmlcount_reward_func":
|
| 960 |
-
"step":
|
| 961 |
},
|
| 962 |
{
|
| 963 |
-
"epoch":
|
| 964 |
-
"step":
|
| 965 |
"total_flos": 0.0,
|
| 966 |
-
"train_loss":
|
| 967 |
-
"train_runtime":
|
| 968 |
-
"train_samples_per_second":
|
| 969 |
-
"train_steps_per_second": 0.
|
| 970 |
}
|
| 971 |
],
|
| 972 |
"logging_steps": 2,
|
| 973 |
-
"max_steps":
|
| 974 |
"num_input_tokens_seen": 0,
|
| 975 |
"num_train_epochs": 1,
|
| 976 |
"save_steps": 25,
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.2671118530884808,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 20,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"completion_length": 299.34375,
|
| 14 |
+
"epoch": 0.02671118530884808,
|
| 15 |
+
"grad_norm": 94.53134155273438,
|
| 16 |
"kl": 0.0,
|
| 17 |
+
"learning_rate": 4.965903258506806e-07,
|
| 18 |
"loss": 0.0,
|
| 19 |
+
"reward": 0.4095814856700599,
|
| 20 |
+
"reward_std": 0.24939912266563624,
|
| 21 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 22 |
+
"rewards/consensus_reward_func": 0.0625,
|
| 23 |
"rewards/cumulative_reward_2": 0.0,
|
| 24 |
+
"rewards/final_correctness_reward_func": 0.25,
|
| 25 |
+
"rewards/question_recreation_reward_func": 0.17958148289471865,
|
| 26 |
"rewards/soft_format_reward_func": 0.0,
|
| 27 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 28 |
+
"rewards/xmlcount_reward_func": -0.08249999745748937,
|
| 29 |
"step": 2
|
| 30 |
},
|
| 31 |
{
|
| 32 |
+
"completion_length": 316.9375,
|
| 33 |
+
"epoch": 0.05342237061769616,
|
| 34 |
+
"grad_norm": 9.352435111999512,
|
| 35 |
+
"kl": 0.0014936136085452745,
|
| 36 |
+
"learning_rate": 4.698684378016222e-07,
|
| 37 |
"loss": 0.0,
|
| 38 |
+
"reward": 1.0824220394715667,
|
| 39 |
+
"reward_std": 1.2490277905308176,
|
| 40 |
+
"rewards/concensus_correctness_reward_func": 0.625,
|
| 41 |
+
"rewards/consensus_reward_func": 0.0625,
|
| 42 |
"rewards/cumulative_reward_2": 0.0,
|
| 43 |
+
"rewards/final_correctness_reward_func": 0.0625,
|
| 44 |
+
"rewards/question_recreation_reward_func": 0.1588594950735569,
|
| 45 |
"rewards/soft_format_reward_func": 0.0,
|
| 46 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 47 |
+
"rewards/xmlcount_reward_func": 0.1735625034198165,
|
| 48 |
"step": 4
|
| 49 |
},
|
| 50 |
{
|
| 51 |
+
"completion_length": 432.5,
|
| 52 |
+
"epoch": 0.08013355592654424,
|
| 53 |
+
"grad_norm": 72.27102661132812,
|
| 54 |
+
"kl": 0.0026759140364447376,
|
| 55 |
+
"learning_rate": 4.193203929064353e-07,
|
| 56 |
"loss": 0.0,
|
| 57 |
+
"reward": 0.26975517414393835,
|
| 58 |
+
"reward_std": 0.23289441590895876,
|
| 59 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 60 |
+
"rewards/consensus_reward_func": 0.0,
|
| 61 |
"rewards/cumulative_reward_2": 0.0,
|
| 62 |
+
"rewards/final_correctness_reward_func": 0.0,
|
| 63 |
+
"rewards/question_recreation_reward_func": 0.16150517805363052,
|
| 64 |
"rewards/soft_format_reward_func": 0.0,
|
| 65 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 66 |
+
"rewards/xmlcount_reward_func": 0.10824999865144491,
|
| 67 |
"step": 6
|
| 68 |
},
|
| 69 |
{
|
| 70 |
+
"completion_length": 424.3125,
|
| 71 |
+
"epoch": 0.10684474123539232,
|
| 72 |
+
"grad_norm": 57.67626190185547,
|
| 73 |
+
"kl": 0.003170746138493996,
|
| 74 |
+
"learning_rate": 3.5042385616324236e-07,
|
| 75 |
"loss": 0.0,
|
| 76 |
+
"reward": 0.38037113891914487,
|
| 77 |
+
"reward_std": 0.40154265408637,
|
| 78 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 79 |
+
"rewards/consensus_reward_func": 0.0,
|
| 80 |
"rewards/cumulative_reward_2": 0.0,
|
| 81 |
"rewards/final_correctness_reward_func": 0.125,
|
| 82 |
+
"rewards/question_recreation_reward_func": 0.14277737913653255,
|
| 83 |
"rewards/soft_format_reward_func": 0.0,
|
| 84 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 85 |
+
"rewards/xmlcount_reward_func": 0.11259375081135659,
|
| 86 |
"step": 8
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"completion_length": 388.5,
|
| 90 |
+
"epoch": 0.1335559265442404,
|
| 91 |
+
"grad_norm": 5.881807804107666,
|
| 92 |
+
"kl": 0.008150014091370394,
|
| 93 |
+
"learning_rate": 2.706448363680831e-07,
|
| 94 |
"loss": 0.0,
|
| 95 |
+
"reward": 0.009647406113799661,
|
| 96 |
+
"reward_std": 0.48192066283809254,
|
| 97 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 98 |
+
"rewards/consensus_reward_func": 0.0,
|
| 99 |
"rewards/cumulative_reward_2": 0.0,
|
| 100 |
+
"rewards/final_correctness_reward_func": 0.0,
|
| 101 |
+
"rewards/question_recreation_reward_func": 0.11724113661330193,
|
| 102 |
"rewards/soft_format_reward_func": 0.0,
|
| 103 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 104 |
+
"rewards/xmlcount_reward_func": -0.10759374493500218,
|
| 105 |
"step": 10
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"completion_length": 329.3125,
|
| 109 |
+
"epoch": 0.16026711185308848,
|
| 110 |
+
"grad_norm": 32.46157455444336,
|
| 111 |
+
"kl": 0.0016044293206505245,
|
| 112 |
+
"learning_rate": 1.886286282148002e-07,
|
| 113 |
+
"loss": 0.0,
|
| 114 |
+
"reward": 0.3704580693738535,
|
| 115 |
+
"reward_std": 0.46280903373553883,
|
| 116 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 117 |
+
"rewards/consensus_reward_func": 0.0,
|
| 118 |
"rewards/cumulative_reward_2": 0.0,
|
| 119 |
+
"rewards/final_correctness_reward_func": 0.0625,
|
| 120 |
+
"rewards/question_recreation_reward_func": 0.19455181124794763,
|
| 121 |
"rewards/soft_format_reward_func": 0.0,
|
| 122 |
+
"rewards/strict_format_reward_func": 0.015625,
|
| 123 |
+
"rewards/xmlcount_reward_func": 0.09778124839067459,
|
| 124 |
"step": 12
|
| 125 |
},
|
| 126 |
{
|
| 127 |
+
"completion_length": 385.28125,
|
| 128 |
+
"epoch": 0.18697829716193656,
|
| 129 |
+
"grad_norm": 63.96397018432617,
|
| 130 |
+
"kl": 0.019191041095837136,
|
| 131 |
+
"learning_rate": 1.1326296046939333e-07,
|
| 132 |
"loss": 0.0,
|
| 133 |
+
"reward": 0.7466699600918218,
|
| 134 |
+
"reward_std": 1.2343180573134873,
|
| 135 |
+
"rewards/concensus_correctness_reward_func": 0.625,
|
| 136 |
+
"rewards/consensus_reward_func": 0.0,
|
| 137 |
"rewards/cumulative_reward_2": 0.0,
|
| 138 |
+
"rewards/final_correctness_reward_func": 0.0625,
|
| 139 |
+
"rewards/question_recreation_reward_func": 0.1140449601225555,
|
| 140 |
"rewards/soft_format_reward_func": 0.0,
|
| 141 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 142 |
+
"rewards/xmlcount_reward_func": -0.05487499665468931,
|
| 143 |
"step": 14
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"completion_length": 381.03125,
|
| 147 |
+
"epoch": 0.21368948247078465,
|
| 148 |
+
"grad_norm": 12.218117713928223,
|
| 149 |
+
"kl": 0.003359480255312519,
|
| 150 |
+
"learning_rate": 5.271487265090163e-08,
|
| 151 |
"loss": 0.0,
|
| 152 |
+
"reward": 0.4025605304632336,
|
| 153 |
+
"reward_std": 0.7095250401762314,
|
| 154 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 155 |
+
"rewards/consensus_reward_func": 0.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 156 |
"rewards/cumulative_reward_2": 0.0,
|
| 157 |
"rewards/final_correctness_reward_func": 0.125,
|
| 158 |
+
"rewards/question_recreation_reward_func": 0.23962301563005894,
|
| 159 |
"rewards/soft_format_reward_func": 0.0,
|
| 160 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 161 |
+
"rewards/xmlcount_reward_func": 0.037937495624646544,
|
| 162 |
+
"step": 16
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 163 |
},
|
| 164 |
{
|
| 165 |
+
"completion_length": 261.03125,
|
| 166 |
+
"epoch": 0.24040066777963273,
|
| 167 |
+
"grad_norm": 9.750836372375488,
|
| 168 |
+
"kl": 0.0017970694570976775,
|
| 169 |
+
"learning_rate": 1.3545689574841341e-08,
|
| 170 |
"loss": 0.0,
|
| 171 |
+
"reward": 0.4833010680740699,
|
| 172 |
+
"reward_std": 0.5299922852864256,
|
| 173 |
+
"rewards/concensus_correctness_reward_func": 0.0,
|
| 174 |
+
"rewards/consensus_reward_func": 0.0625,
|
| 175 |
"rewards/cumulative_reward_2": 0.0,
|
| 176 |
"rewards/final_correctness_reward_func": 0.0625,
|
| 177 |
+
"rewards/question_recreation_reward_func": 0.2038323215674609,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 178 |
"rewards/soft_format_reward_func": 0.0,
|
| 179 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 180 |
+
"rewards/xmlcount_reward_func": 0.15446875197812915,
|
| 181 |
+
"step": 18
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 182 |
},
|
| 183 |
{
|
| 184 |
+
"completion_length": 358.25,
|
| 185 |
+
"epoch": 0.2671118530884808,
|
| 186 |
+
"grad_norm": 30.307376861572266,
|
| 187 |
+
"kl": 0.0023762249620631337,
|
| 188 |
"learning_rate": 0.0,
|
| 189 |
"loss": 0.0,
|
| 190 |
+
"reward": 0.5694868991849944,
|
| 191 |
+
"reward_std": 0.5472512795531657,
|
| 192 |
+
"rewards/concensus_correctness_reward_func": 0.05999999865889549,
|
| 193 |
+
"rewards/consensus_reward_func": 0.0625,
|
| 194 |
"rewards/cumulative_reward_2": 0.0,
|
| 195 |
+
"rewards/final_correctness_reward_func": 0.0625,
|
| 196 |
+
"rewards/question_recreation_reward_func": 0.25226815952919424,
|
| 197 |
"rewards/soft_format_reward_func": 0.0,
|
| 198 |
+
"rewards/strict_format_reward_func": 0.0,
|
| 199 |
+
"rewards/xmlcount_reward_func": 0.13221875578165054,
|
| 200 |
+
"step": 20
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"epoch": 0.2671118530884808,
|
| 204 |
+
"step": 20,
|
| 205 |
"total_flos": 0.0,
|
| 206 |
+
"train_loss": 4.617630838765763e-06,
|
| 207 |
+
"train_runtime": 463.4946,
|
| 208 |
+
"train_samples_per_second": 0.69,
|
| 209 |
+
"train_steps_per_second": 0.043
|
| 210 |
}
|
| 211 |
],
|
| 212 |
"logging_steps": 2,
|
| 213 |
+
"max_steps": 20,
|
| 214 |
"num_input_tokens_seen": 0,
|
| 215 |
"num_train_epochs": 1,
|
| 216 |
"save_steps": 25,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6008
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76acb6803d24a05a0407ebe59d186d61c08a8dec84cd1a5c3e957bf419604560
|
| 3 |
size 6008
|