Model save
Browse files- README.md +2 -4
- all_results.json +4 -4
- train_results.json +4 -4
- trainer_state.json +886 -886
README.md
CHANGED
|
@@ -1,11 +1,9 @@
|
|
| 1 |
---
|
| 2 |
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
|
| 3 |
-
datasets: knoveleng/open-rs
|
| 4 |
library_name: transformers
|
| 5 |
model_name: OpenRS-GRPO
|
| 6 |
tags:
|
| 7 |
- generated_from_trainer
|
| 8 |
-
- open-r1
|
| 9 |
- trl
|
| 10 |
- grpo
|
| 11 |
licence: license
|
|
@@ -13,7 +11,7 @@ licence: license
|
|
| 13 |
|
| 14 |
# Model Card for OpenRS-GRPO
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B)
|
| 17 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 18 |
|
| 19 |
## Quick start
|
|
@@ -29,7 +27,7 @@ print(output["generated_text"])
|
|
| 29 |
|
| 30 |
## Training procedure
|
| 31 |
|
| 32 |
-
[<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/sonicdog00-korea-university/openrs/runs/
|
| 33 |
|
| 34 |
|
| 35 |
This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
|
|
|
|
| 1 |
---
|
| 2 |
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
|
|
|
|
| 3 |
library_name: transformers
|
| 4 |
model_name: OpenRS-GRPO
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
| 7 |
- trl
|
| 8 |
- grpo
|
| 9 |
licence: license
|
|
|
|
| 11 |
|
| 12 |
# Model Card for OpenRS-GRPO
|
| 13 |
|
| 14 |
+
This model is a fine-tuned version of [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B).
|
| 15 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
|
| 17 |
## Quick start
|
|
|
|
| 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/sonicdog00-korea-university/openrs/runs/ez1iu4d5)
|
| 31 |
|
| 32 |
|
| 33 |
This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 7000,
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.0719883194193244,
|
| 4 |
+
"train_runtime": 18385.1788,
|
| 5 |
"train_samples": 7000,
|
| 6 |
+
"train_samples_per_second": 0.392,
|
| 7 |
+
"train_steps_per_second": 0.005
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 7000,
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.0719883194193244,
|
| 4 |
+
"train_runtime": 18385.1788,
|
| 5 |
"train_samples": 7000,
|
| 6 |
+
"train_samples_per_second": 0.392,
|
| 7 |
+
"train_steps_per_second": 0.005
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -10,1612 +10,1612 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"clip_ratio": 0.0,
|
| 13 |
-
"completion_length":
|
| 14 |
"epoch": 0.001713796058269066,
|
| 15 |
-
"grad_norm": 0.
|
| 16 |
"kl": 0.0,
|
| 17 |
"learning_rate": 1e-07,
|
| 18 |
-
"loss": 0.
|
| 19 |
-
"reward": 0.
|
| 20 |
-
"reward_std": 0.
|
| 21 |
-
"rewards/accuracy_reward": 0.
|
| 22 |
"rewards/difficulty_following_reward": 0.0,
|
| 23 |
-
"rewards/format_reward": 0.
|
| 24 |
"step": 1,
|
| 25 |
-
"vanishing_advantage_ratio": 0.
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"clip_ratio": 0.0,
|
| 29 |
-
"completion_length":
|
| 30 |
"epoch": 0.003427592116538132,
|
| 31 |
-
"grad_norm": 0.
|
| 32 |
"kl": 0.0,
|
| 33 |
"learning_rate": 2e-07,
|
| 34 |
-
"loss": 0.
|
| 35 |
-
"reward": 0.
|
| 36 |
-
"reward_std": 0.
|
| 37 |
-
"rewards/accuracy_reward": 0.
|
| 38 |
"rewards/difficulty_following_reward": 0.0,
|
| 39 |
-
"rewards/format_reward": 0.
|
| 40 |
"step": 2,
|
| 41 |
-
"vanishing_advantage_ratio": 0.
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"clip_ratio": 0.0,
|
| 45 |
-
"completion_length":
|
| 46 |
"epoch": 0.005141388174807198,
|
| 47 |
-
"grad_norm": 0.
|
| 48 |
-
"kl": 4.
|
| 49 |
"learning_rate": 3e-07,
|
| 50 |
-
"loss": 0.
|
| 51 |
-
"reward": 0.
|
| 52 |
-
"reward_std": 0.
|
| 53 |
-
"rewards/accuracy_reward": 0.
|
| 54 |
"rewards/difficulty_following_reward": 0.0,
|
| 55 |
-
"rewards/format_reward": 0.
|
| 56 |
"step": 3,
|
| 57 |
-
"vanishing_advantage_ratio": 0.
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"clip_ratio": 0.0,
|
| 61 |
-
"completion_length":
|
| 62 |
"epoch": 0.006855184233076264,
|
| 63 |
-
"grad_norm": 0.
|
| 64 |
-
"kl":
|
| 65 |
"learning_rate": 4e-07,
|
| 66 |
-
"loss": 0.
|
| 67 |
-
"reward": 0.
|
| 68 |
-
"reward_std": 0.
|
| 69 |
-
"rewards/accuracy_reward": 0.
|
| 70 |
"rewards/difficulty_following_reward": 0.0,
|
| 71 |
-
"rewards/format_reward": 0.
|
| 72 |
"step": 4,
|
| 73 |
-
"vanishing_advantage_ratio": 0.
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"clip_ratio": 0.0,
|
| 77 |
-
"completion_length":
|
| 78 |
"epoch": 0.00856898029134533,
|
| 79 |
-
"grad_norm": 0.
|
| 80 |
-
"kl":
|
| 81 |
"learning_rate": 5e-07,
|
| 82 |
-
"loss": 0.
|
| 83 |
-
"reward": 0.
|
| 84 |
-
"reward_std": 0.
|
| 85 |
-
"rewards/accuracy_reward": 0.
|
| 86 |
"rewards/difficulty_following_reward": 0.0,
|
| 87 |
-
"rewards/format_reward": 0.
|
| 88 |
"step": 5,
|
| 89 |
-
"vanishing_advantage_ratio": 0.
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"clip_ratio": 0.0,
|
| 93 |
-
"completion_length":
|
| 94 |
"epoch": 0.010282776349614395,
|
| 95 |
-
"grad_norm": 0.
|
| 96 |
-
"kl": 5.
|
| 97 |
"learning_rate": 6e-07,
|
| 98 |
-
"loss": 0.
|
| 99 |
-
"reward": 0.
|
| 100 |
-
"reward_std": 0.
|
| 101 |
-
"rewards/accuracy_reward": 0.
|
| 102 |
"rewards/difficulty_following_reward": 0.0,
|
| 103 |
-
"rewards/format_reward": 0.
|
| 104 |
"step": 6,
|
| 105 |
-
"vanishing_advantage_ratio": 0.
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"clip_ratio": 0.0,
|
| 109 |
-
"completion_length":
|
| 110 |
"epoch": 0.011996572407883462,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
-
"kl": 5.
|
| 113 |
"learning_rate": 7e-07,
|
| 114 |
-
"loss": 0.
|
| 115 |
-
"reward": 0.
|
| 116 |
-
"reward_std": 0.
|
| 117 |
-
"rewards/accuracy_reward": 0.
|
| 118 |
"rewards/difficulty_following_reward": 0.0,
|
| 119 |
-
"rewards/format_reward": 0.
|
| 120 |
"step": 7,
|
| 121 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"clip_ratio": 0.0,
|
| 125 |
-
"completion_length":
|
| 126 |
"epoch": 0.013710368466152529,
|
| 127 |
-
"grad_norm": 0.
|
| 128 |
-
"kl": 3.
|
| 129 |
"learning_rate": 8e-07,
|
| 130 |
-
"loss": 0.
|
| 131 |
-
"reward": 0.
|
| 132 |
-
"reward_std": 0.
|
| 133 |
-
"rewards/accuracy_reward": 0.
|
| 134 |
"rewards/difficulty_following_reward": 0.0,
|
| 135 |
-
"rewards/format_reward": 0.
|
| 136 |
"step": 8,
|
| 137 |
-
"vanishing_advantage_ratio": 0.
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"clip_ratio": 0.0,
|
| 141 |
-
"completion_length":
|
| 142 |
"epoch": 0.015424164524421594,
|
| 143 |
-
"grad_norm": 0.
|
| 144 |
-
"kl":
|
| 145 |
"learning_rate": 9e-07,
|
| 146 |
-
"loss": 0.
|
| 147 |
-
"reward": 0.
|
| 148 |
-
"reward_std": 0.
|
| 149 |
-
"rewards/accuracy_reward": 0.
|
| 150 |
"rewards/difficulty_following_reward": 0.0,
|
| 151 |
-
"rewards/format_reward": 0.
|
| 152 |
"step": 9,
|
| 153 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"clip_ratio": 0.0,
|
| 157 |
-
"completion_length":
|
| 158 |
"epoch": 0.01713796058269066,
|
| 159 |
-
"grad_norm": 0.
|
| 160 |
-
"kl": 3.
|
| 161 |
"learning_rate": 1e-06,
|
| 162 |
-
"loss": 0.
|
| 163 |
-
"reward": 0.
|
| 164 |
-
"reward_std": 0.
|
| 165 |
-
"rewards/accuracy_reward": 0.
|
| 166 |
"rewards/difficulty_following_reward": 0.0,
|
| 167 |
-
"rewards/format_reward": 0.
|
| 168 |
"step": 10,
|
| 169 |
-
"vanishing_advantage_ratio": 0.
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"clip_ratio": 0.0,
|
| 173 |
-
"completion_length":
|
| 174 |
"epoch": 0.018851756640959727,
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
-
"kl":
|
| 177 |
"learning_rate": 9.997258721585931e-07,
|
| 178 |
-
"loss": 0.
|
| 179 |
-
"reward": 0.
|
| 180 |
-
"reward_std": 0.
|
| 181 |
-
"rewards/accuracy_reward": 0.
|
| 182 |
"rewards/difficulty_following_reward": 0.0,
|
| 183 |
-
"rewards/format_reward": 0.
|
| 184 |
"step": 11,
|
| 185 |
-
"vanishing_advantage_ratio": 0.
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"clip_ratio": 0.0,
|
| 189 |
-
"completion_length":
|
| 190 |
"epoch": 0.02056555269922879,
|
| 191 |
-
"grad_norm": 0.
|
| 192 |
-
"kl": 3.
|
| 193 |
"learning_rate": 9.989038226169207e-07,
|
| 194 |
-
"loss": 0.
|
| 195 |
-
"reward": 0.
|
| 196 |
-
"reward_std": 0.
|
| 197 |
-
"rewards/accuracy_reward": 0.
|
| 198 |
"rewards/difficulty_following_reward": 0.0,
|
| 199 |
-
"rewards/format_reward": 0.
|
| 200 |
"step": 12,
|
| 201 |
-
"vanishing_advantage_ratio": 0.
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"clip_ratio": 0.0,
|
| 205 |
-
"completion_length":
|
| 206 |
"epoch": 0.022279348757497857,
|
| 207 |
-
"grad_norm": 0.
|
| 208 |
-
"kl":
|
| 209 |
"learning_rate": 9.975348529157229e-07,
|
| 210 |
-
"loss": 0.
|
| 211 |
-
"reward": 0.
|
| 212 |
-
"reward_std": 0.
|
| 213 |
-
"rewards/accuracy_reward": 0.
|
| 214 |
"rewards/difficulty_following_reward": 0.0,
|
| 215 |
-
"rewards/format_reward": 0.
|
| 216 |
"step": 13,
|
| 217 |
-
"vanishing_advantage_ratio": 0.
|
| 218 |
},
|
| 219 |
{
|
| 220 |
"clip_ratio": 0.0,
|
| 221 |
-
"completion_length":
|
| 222 |
"epoch": 0.023993144815766924,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
-
"kl":
|
| 225 |
"learning_rate": 9.956206309337066e-07,
|
| 226 |
-
"loss": 0.
|
| 227 |
-
"reward": 0.
|
| 228 |
-
"reward_std": 0.
|
| 229 |
-
"rewards/accuracy_reward": 0.
|
| 230 |
"rewards/difficulty_following_reward": 0.0,
|
| 231 |
-
"rewards/format_reward": 0.
|
| 232 |
"step": 14,
|
| 233 |
-
"vanishing_advantage_ratio": 0.
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"clip_ratio": 0.0,
|
| 237 |
-
"completion_length":
|
| 238 |
"epoch": 0.02570694087403599,
|
| 239 |
-
"grad_norm": 0.
|
| 240 |
-
"kl":
|
| 241 |
"learning_rate": 9.931634888554935e-07,
|
| 242 |
-
"loss": 0.
|
| 243 |
-
"reward": 0.
|
| 244 |
-
"reward_std": 0.
|
| 245 |
-
"rewards/accuracy_reward": 0.
|
| 246 |
"rewards/difficulty_following_reward": 0.0,
|
| 247 |
-
"rewards/format_reward": 0.
|
| 248 |
"step": 15,
|
| 249 |
-
"vanishing_advantage_ratio": 0.
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"clip_ratio": 0.0,
|
| 253 |
-
"completion_length":
|
| 254 |
"epoch": 0.027420736932305057,
|
| 255 |
-
"grad_norm": 0.
|
| 256 |
-
"kl":
|
| 257 |
"learning_rate": 9.901664203302124e-07,
|
| 258 |
-
"loss": 0.
|
| 259 |
-
"reward": 0.
|
| 260 |
-
"reward_std": 0.
|
| 261 |
-
"rewards/accuracy_reward": 0.
|
| 262 |
"rewards/difficulty_following_reward": 0.0,
|
| 263 |
-
"rewards/format_reward": 0.
|
| 264 |
"step": 16,
|
| 265 |
-
"vanishing_advantage_ratio": 0.
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"clip_ratio": 0.0,
|
| 269 |
-
"completion_length":
|
| 270 |
"epoch": 0.02913453299057412,
|
| 271 |
-
"grad_norm": 0.
|
| 272 |
-
"kl":
|
| 273 |
"learning_rate": 9.866330768241983e-07,
|
| 274 |
-
"loss": 0.
|
| 275 |
-
"reward": 0.
|
| 276 |
-
"reward_std": 0.
|
| 277 |
-
"rewards/accuracy_reward": 0.
|
| 278 |
"rewards/difficulty_following_reward": 0.0,
|
| 279 |
-
"rewards/format_reward": 0.
|
| 280 |
"step": 17,
|
| 281 |
-
"vanishing_advantage_ratio": 0.
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"clip_ratio": 0.0,
|
| 285 |
-
"completion_length":
|
| 286 |
"epoch": 0.030848329048843187,
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"kl":
|
| 289 |
"learning_rate": 9.825677631722435e-07,
|
| 290 |
-
"loss": 0.
|
| 291 |
-
"reward": 0.
|
| 292 |
-
"reward_std": 0.
|
| 293 |
-
"rewards/accuracy_reward": 0.
|
| 294 |
"rewards/difficulty_following_reward": 0.0,
|
| 295 |
-
"rewards/format_reward": 0.
|
| 296 |
"step": 18,
|
| 297 |
-
"vanishing_advantage_ratio": 0.
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"clip_ratio": 0.0,
|
| 301 |
-
"completion_length":
|
| 302 |
"epoch": 0.032562125107112254,
|
| 303 |
-
"grad_norm": 0.
|
| 304 |
-
"kl":
|
| 305 |
"learning_rate": 9.779754323328192e-07,
|
| 306 |
-
"loss": 0.
|
| 307 |
-
"reward": 0.
|
| 308 |
-
"reward_std": 0.
|
| 309 |
-
"rewards/accuracy_reward": 0.
|
| 310 |
"rewards/difficulty_following_reward": 0.0,
|
| 311 |
-
"rewards/format_reward": 0.
|
| 312 |
"step": 19,
|
| 313 |
-
"vanishing_advantage_ratio": 0.
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"clip_ratio": 0.0,
|
| 317 |
-
"completion_length":
|
| 318 |
"epoch": 0.03427592116538132,
|
| 319 |
-
"grad_norm": 0.
|
| 320 |
-
"kl":
|
| 321 |
"learning_rate": 9.728616793536587e-07,
|
| 322 |
-
"loss": 0.
|
| 323 |
-
"reward": 0.
|
| 324 |
-
"reward_std": 0.
|
| 325 |
-
"rewards/accuracy_reward": 0.
|
| 326 |
"rewards/difficulty_following_reward": 0.0,
|
| 327 |
-
"rewards/format_reward": 0.
|
| 328 |
"step": 20,
|
| 329 |
-
"vanishing_advantage_ratio": 0.
|
| 330 |
},
|
| 331 |
{
|
| 332 |
"clip_ratio": 0.0,
|
| 333 |
-
"completion_length":
|
| 334 |
"epoch": 0.03598971722365039,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"kl":
|
| 337 |
"learning_rate": 9.672327345550543e-07,
|
| 338 |
-
"loss": 0.
|
| 339 |
-
"reward": 0.
|
| 340 |
-
"reward_std": 0.
|
| 341 |
-
"rewards/accuracy_reward": 0.
|
| 342 |
"rewards/difficulty_following_reward": 0.0,
|
| 343 |
-
"rewards/format_reward": 0.
|
| 344 |
"step": 21,
|
| 345 |
-
"vanishing_advantage_ratio": 0.
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"clip_ratio": 0.0,
|
| 349 |
-
"completion_length":
|
| 350 |
"epoch": 0.037703513281919454,
|
| 351 |
-
"grad_norm": 0.
|
| 352 |
-
"kl":
|
| 353 |
"learning_rate": 9.610954559391704e-07,
|
| 354 |
-
"loss": 0.
|
| 355 |
-
"reward": 0.
|
| 356 |
-
"reward_std": 0.
|
| 357 |
-
"rewards/accuracy_reward": 0.
|
| 358 |
"rewards/difficulty_following_reward": 0.0,
|
| 359 |
-
"rewards/format_reward": 0.
|
| 360 |
"step": 22,
|
| 361 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"clip_ratio": 0.0,
|
| 365 |
-
"completion_length":
|
| 366 |
"epoch": 0.03941730934018852,
|
| 367 |
-
"grad_norm": 0.
|
| 368 |
-
"kl":
|
| 369 |
"learning_rate": 9.54457320834625e-07,
|
| 370 |
-
"loss": 0.
|
| 371 |
-
"reward": 0.
|
| 372 |
-
"reward_std": 0.
|
| 373 |
-
"rewards/accuracy_reward": 0.
|
| 374 |
"rewards/difficulty_following_reward": 0.0,
|
| 375 |
-
"rewards/format_reward": 0.
|
| 376 |
"step": 23,
|
| 377 |
-
"vanishing_advantage_ratio": 0.
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"clip_ratio": 0.0,
|
| 381 |
-
"completion_length":
|
| 382 |
"epoch": 0.04113110539845758,
|
| 383 |
-
"grad_norm": 0.
|
| 384 |
-
"kl": 0.
|
| 385 |
"learning_rate": 9.473264167865171e-07,
|
| 386 |
-
"loss": 0.
|
| 387 |
-
"reward": 0.
|
| 388 |
-
"reward_std": 0.
|
| 389 |
-
"rewards/accuracy_reward": 0.
|
| 390 |
"rewards/difficulty_following_reward": 0.0,
|
| 391 |
-
"rewards/format_reward": 0.
|
| 392 |
"step": 24,
|
| 393 |
-
"vanishing_advantage_ratio": 0.
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"clip_ratio": 0.0,
|
| 397 |
-
"completion_length":
|
| 398 |
"epoch": 0.04284490145672665,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
-
"kl": 0.
|
| 401 |
"learning_rate": 9.397114317029974e-07,
|
| 402 |
-
"loss": 0.
|
| 403 |
-
"reward": 0.
|
| 404 |
-
"reward_std": 0.
|
| 405 |
-
"rewards/accuracy_reward": 0.
|
| 406 |
"rewards/difficulty_following_reward": 0.0,
|
| 407 |
-
"rewards/format_reward": 0.
|
| 408 |
"step": 25,
|
| 409 |
-
"vanishing_advantage_ratio": 0.
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"clip_ratio": 0.0,
|
| 413 |
-
"completion_length":
|
| 414 |
"epoch": 0.044558697514995714,
|
| 415 |
-
"grad_norm": 0.
|
| 416 |
-
"kl": 0.
|
| 417 |
"learning_rate": 9.316216432703916e-07,
|
| 418 |
-
"loss": 0.
|
| 419 |
-
"reward": 0.
|
| 420 |
-
"reward_std": 0.
|
| 421 |
-
"rewards/accuracy_reward": 0.
|
| 422 |
"rewards/difficulty_following_reward": 0.0,
|
| 423 |
-
"rewards/format_reward": 0.
|
| 424 |
"step": 26,
|
| 425 |
-
"vanishing_advantage_ratio": 0.
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"clip_ratio": 0.0,
|
| 429 |
-
"completion_length":
|
| 430 |
"epoch": 0.04627249357326478,
|
| 431 |
-
"grad_norm": 0.
|
| 432 |
-
"kl": 0.
|
| 433 |
"learning_rate": 9.230669076497687e-07,
|
| 434 |
-
"loss": 0.
|
| 435 |
-
"reward": 0.
|
| 436 |
-
"reward_std": 0.
|
| 437 |
-
"rewards/accuracy_reward": 0.
|
| 438 |
"rewards/difficulty_following_reward": 0.0,
|
| 439 |
-
"rewards/format_reward": 0.
|
| 440 |
"step": 27,
|
| 441 |
-
"vanishing_advantage_ratio": 0.
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"clip_ratio": 0.0,
|
| 445 |
-
"completion_length":
|
| 446 |
"epoch": 0.04798628963153385,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
-
"kl": 0.
|
| 449 |
"learning_rate": 9.140576474687263e-07,
|
| 450 |
-
"loss": 0.
|
| 451 |
-
"reward": 0.
|
| 452 |
-
"reward_std": 0.
|
| 453 |
-
"rewards/accuracy_reward": 0.
|
| 454 |
"rewards/difficulty_following_reward": 0.0,
|
| 455 |
-
"rewards/format_reward": 0.
|
| 456 |
"step": 28,
|
| 457 |
-
"vanishing_advantage_ratio": 0.
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"clip_ratio": 0.0,
|
| 461 |
-
"completion_length":
|
| 462 |
"epoch": 0.049700085689802914,
|
| 463 |
-
"grad_norm": 0.
|
| 464 |
-
"kl": 0.
|
| 465 |
"learning_rate": 9.046048391230247e-07,
|
| 466 |
-
"loss": 0.
|
| 467 |
-
"reward": 0.
|
| 468 |
-
"reward_std": 0.
|
| 469 |
-
"rewards/accuracy_reward": 0.
|
| 470 |
"rewards/difficulty_following_reward": 0.0,
|
| 471 |
-
"rewards/format_reward": 0.
|
| 472 |
"step": 29,
|
| 473 |
-
"vanishing_advantage_ratio": 0.
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"clip_ratio": 0.0,
|
| 477 |
-
"completion_length":
|
| 478 |
"epoch": 0.05141388174807198,
|
| 479 |
-
"grad_norm": 0.
|
| 480 |
-
"kl": 0.
|
| 481 |
"learning_rate": 8.9471999940354e-07,
|
| 482 |
-
"loss": 0.
|
| 483 |
-
"reward": 0.
|
| 484 |
-
"reward_std": 0.
|
| 485 |
-
"rewards/accuracy_reward": 0.
|
| 486 |
"rewards/difficulty_following_reward": 0.0,
|
| 487 |
-
"rewards/format_reward": 0.
|
| 488 |
"step": 30,
|
| 489 |
-
"vanishing_advantage_ratio": 0.
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"clip_ratio": 0.0,
|
| 493 |
-
"completion_length":
|
| 494 |
"epoch": 0.05312767780634105,
|
| 495 |
-
"grad_norm": 0.
|
| 496 |
-
"kl": 0.
|
| 497 |
"learning_rate": 8.844151714648274e-07,
|
| 498 |
-
"loss": 0.
|
| 499 |
-
"reward": 0.
|
| 500 |
-
"reward_std": 0.
|
| 501 |
-
"rewards/accuracy_reward": 0.
|
| 502 |
"rewards/difficulty_following_reward": 0.0,
|
| 503 |
-
"rewards/format_reward": 0.
|
| 504 |
"step": 31,
|
| 505 |
-
"vanishing_advantage_ratio": 0.
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"clip_ratio": 0.0,
|
| 509 |
-
"completion_length":
|
| 510 |
"epoch": 0.054841473864610114,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
-
"kl": 0.
|
| 513 |
"learning_rate": 8.737029101523929e-07,
|
| 514 |
-
"loss": 0.
|
| 515 |
-
"reward": 0.
|
| 516 |
-
"reward_std": 0.
|
| 517 |
-
"rewards/accuracy_reward": 0.
|
| 518 |
"rewards/difficulty_following_reward": 0.0,
|
| 519 |
-
"rewards/format_reward": 0.
|
| 520 |
"step": 32,
|
| 521 |
-
"vanishing_advantage_ratio": 0.
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"clip_ratio": 0.0,
|
| 525 |
-
"completion_length":
|
| 526 |
"epoch": 0.056555269922879174,
|
| 527 |
-
"grad_norm": 0.
|
| 528 |
-
"kl": 0.
|
| 529 |
"learning_rate": 8.625962667065487e-07,
|
| 530 |
-
"loss": 0.
|
| 531 |
-
"reward": 0.
|
| 532 |
-
"reward_std": 0.
|
| 533 |
-
"rewards/accuracy_reward": 0.
|
| 534 |
"rewards/difficulty_following_reward": 0.0,
|
| 535 |
-
"rewards/format_reward": 0.
|
| 536 |
"step": 33,
|
| 537 |
-
"vanishing_advantage_ratio": 0.
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"clip_ratio": 0.0,
|
| 541 |
-
"completion_length":
|
| 542 |
"epoch": 0.05826906598114824,
|
| 543 |
-
"grad_norm": 0.
|
| 544 |
-
"kl": 0.
|
| 545 |
"learning_rate": 8.511087728614862e-07,
|
| 546 |
-
"loss": 0.
|
| 547 |
-
"reward": 0.
|
| 548 |
-
"reward_std": 0.
|
| 549 |
-
"rewards/accuracy_reward": 0.
|
| 550 |
"rewards/difficulty_following_reward": 0.0,
|
| 551 |
-
"rewards/format_reward": 0.
|
| 552 |
"step": 34,
|
| 553 |
-
"vanishing_advantage_ratio": 0.
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"clip_ratio": 0.0,
|
| 557 |
-
"completion_length":
|
| 558 |
"epoch": 0.05998286203941731,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
-
"kl": 0.
|
| 561 |
"learning_rate": 8.392544243589427e-07,
|
| 562 |
-
"loss": 0.
|
| 563 |
-
"reward": 0.
|
| 564 |
-
"reward_std": 0.
|
| 565 |
-
"rewards/accuracy_reward": 0.
|
| 566 |
"rewards/difficulty_following_reward": 0.0,
|
| 567 |
-
"rewards/format_reward": 0.
|
| 568 |
"step": 35,
|
| 569 |
-
"vanishing_advantage_ratio": 0.
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"clip_ratio": 0.0,
|
| 573 |
-
"completion_length":
|
| 574 |
"epoch": 0.061696658097686374,
|
| 575 |
-
"grad_norm": 0.
|
| 576 |
-
"kl": 0.
|
| 577 |
"learning_rate": 8.270476638965461e-07,
|
| 578 |
-
"loss": 0.
|
| 579 |
-
"reward": 0.
|
| 580 |
-
"reward_std": 0.
|
| 581 |
-
"rewards/accuracy_reward": 0.
|
| 582 |
"rewards/difficulty_following_reward": 0.0,
|
| 583 |
-
"rewards/format_reward": 0.
|
| 584 |
"step": 36,
|
| 585 |
"vanishing_advantage_ratio": 0.5000000149011612
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"clip_ratio": 0.0,
|
| 589 |
-
"completion_length":
|
| 590 |
"epoch": 0.06341045415595545,
|
| 591 |
-
"grad_norm": 0.
|
| 592 |
-
"kl": 0.
|
| 593 |
"learning_rate": 8.145033635316128e-07,
|
| 594 |
-
"loss": 0.
|
| 595 |
-
"reward": 0.
|
| 596 |
-
"reward_std": 0.
|
| 597 |
-
"rewards/accuracy_reward": 0.
|
| 598 |
"rewards/difficulty_following_reward": 0.0,
|
| 599 |
-
"rewards/format_reward": 0.
|
| 600 |
"step": 37,
|
| 601 |
-
"vanishing_advantage_ratio": 0.
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"clip_ratio": 0.0,
|
| 605 |
-
"completion_length":
|
| 606 |
"epoch": 0.06512425021422451,
|
| 607 |
-
"grad_norm": 0.
|
| 608 |
-
"kl": 0.
|
| 609 |
"learning_rate": 8.01636806561836e-07,
|
| 610 |
-
"loss": 0.
|
| 611 |
-
"reward": 0.
|
| 612 |
-
"reward_std": 0.
|
| 613 |
-
"rewards/accuracy_reward": 0.
|
| 614 |
"rewards/difficulty_following_reward": 0.0,
|
| 615 |
-
"rewards/format_reward": 0.
|
| 616 |
"step": 38,
|
| 617 |
-
"vanishing_advantage_ratio": 0.
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"clip_ratio": 0.0,
|
| 621 |
-
"completion_length":
|
| 622 |
"epoch": 0.06683804627249357,
|
| 623 |
-
"grad_norm": 0.
|
| 624 |
-
"kl": 0.
|
| 625 |
"learning_rate": 7.884636689049422e-07,
|
| 626 |
-
"loss": 0.
|
| 627 |
-
"reward": 0.
|
| 628 |
-
"reward_std": 0.
|
| 629 |
-
"rewards/accuracy_reward": 0.
|
| 630 |
"rewards/difficulty_following_reward": 0.0,
|
| 631 |
-
"rewards/format_reward": 0.
|
| 632 |
"step": 39,
|
| 633 |
"vanishing_advantage_ratio": 0.2500000074505806
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"clip_ratio": 0.0,
|
| 637 |
-
"completion_length":
|
| 638 |
"epoch": 0.06855184233076264,
|
| 639 |
-
"grad_norm": 0.
|
| 640 |
-
"kl": 0.
|
| 641 |
"learning_rate": 7.75e-07,
|
| 642 |
-
"loss": 0.
|
| 643 |
-
"reward": 0.
|
| 644 |
-
"reward_std": 0.
|
| 645 |
-
"rewards/accuracy_reward": 0.
|
| 646 |
"rewards/difficulty_following_reward": 0.0,
|
| 647 |
-
"rewards/format_reward": 0.
|
| 648 |
"step": 40,
|
| 649 |
"vanishing_advantage_ratio": 0.1944444552063942
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"clip_ratio": 0.0,
|
| 653 |
-
"completion_length":
|
| 654 |
"epoch": 0.0702656383890317,
|
| 655 |
-
"grad_norm": 0.
|
| 656 |
-
"kl": 0.
|
| 657 |
"learning_rate": 7.612622032536507e-07,
|
| 658 |
-
"loss": 0.
|
| 659 |
-
"reward": 0.
|
| 660 |
-
"reward_std": 0.
|
| 661 |
-
"rewards/accuracy_reward": 0.
|
| 662 |
"rewards/difficulty_following_reward": 0.0,
|
| 663 |
"rewards/format_reward": 0.18055556062608957,
|
| 664 |
"step": 41,
|
| 665 |
-
"vanishing_advantage_ratio": 0.
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"clip_ratio": 0.0,
|
| 669 |
-
"completion_length":
|
| 670 |
"epoch": 0.07197943444730077,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
-
"kl": 0.
|
| 673 |
"learning_rate": 7.472670160550848e-07,
|
| 674 |
-
"loss": 0.
|
| 675 |
-
"reward": 0.
|
| 676 |
-
"reward_std": 0.
|
| 677 |
"rewards/accuracy_reward": 0.09722222294658422,
|
| 678 |
"rewards/difficulty_following_reward": 0.0,
|
| 679 |
-
"rewards/format_reward": 0.
|
| 680 |
"step": 42,
|
| 681 |
-
"vanishing_advantage_ratio": 0.
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"clip_ratio": 0.0,
|
| 685 |
-
"completion_length":
|
| 686 |
"epoch": 0.07369323050556983,
|
| 687 |
-
"grad_norm": 0.
|
| 688 |
-
"kl": 0.
|
| 689 |
"learning_rate": 7.330314893841101e-07,
|
| 690 |
-
"loss": 0.
|
| 691 |
-
"reward": 0.
|
| 692 |
-
"reward_std": 0.
|
| 693 |
-
"rewards/accuracy_reward": 0.
|
| 694 |
"rewards/difficulty_following_reward": 0.0,
|
| 695 |
-
"rewards/format_reward": 0.
|
| 696 |
"step": 43,
|
| 697 |
-
"vanishing_advantage_ratio": 0.
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"clip_ratio": 0.0,
|
| 701 |
-
"completion_length":
|
| 702 |
"epoch": 0.07540702656383891,
|
| 703 |
-
"grad_norm": 0.
|
| 704 |
-
"kl": 0.
|
| 705 |
"learning_rate": 7.185729670371604e-07,
|
| 706 |
-
"loss": 0.
|
| 707 |
-
"reward": 0.
|
| 708 |
-
"reward_std": 0.
|
| 709 |
-
"rewards/accuracy_reward": 0.
|
| 710 |
"rewards/difficulty_following_reward": 0.0,
|
| 711 |
-
"rewards/format_reward": 0.
|
| 712 |
"step": 44,
|
| 713 |
-
"vanishing_advantage_ratio": 0.
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"clip_ratio": 0.0,
|
| 717 |
-
"completion_length":
|
| 718 |
"epoch": 0.07712082262210797,
|
| 719 |
-
"grad_norm": 0.
|
| 720 |
-
"kl": 0.
|
| 721 |
"learning_rate": 7.039090644965509e-07,
|
| 722 |
-
"loss": 0.
|
| 723 |
-
"reward": 0.
|
| 724 |
-
"reward_std": 0.
|
| 725 |
-
"rewards/accuracy_reward": 0.
|
| 726 |
"rewards/difficulty_following_reward": 0.0,
|
| 727 |
-
"rewards/format_reward": 0.
|
| 728 |
"step": 45,
|
| 729 |
-
"vanishing_advantage_ratio": 0.
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"clip_ratio": 0.0,
|
| 733 |
-
"completion_length":
|
| 734 |
"epoch": 0.07883461868037704,
|
| 735 |
-
"grad_norm": 0.
|
| 736 |
-
"kl": 0.
|
| 737 |
"learning_rate": 6.890576474687263e-07,
|
| 738 |
-
"loss": 0.
|
| 739 |
-
"reward": 0.
|
| 740 |
-
"reward_std": 0.
|
| 741 |
-
"rewards/accuracy_reward": 0.
|
| 742 |
"rewards/difficulty_following_reward": 0.0,
|
| 743 |
-
"rewards/format_reward": 0.
|
| 744 |
"step": 46,
|
| 745 |
-
"vanishing_advantage_ratio": 0.
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"clip_ratio": 0.0,
|
| 749 |
-
"completion_length":
|
| 750 |
"epoch": 0.0805484147386461,
|
| 751 |
-
"grad_norm": 0.
|
| 752 |
-
"kl": 0.
|
| 753 |
"learning_rate": 6.740368101176495e-07,
|
| 754 |
-
"loss": 0.
|
| 755 |
-
"reward": 0.
|
| 756 |
-
"reward_std": 0.
|
| 757 |
-
"rewards/accuracy_reward": 0.
|
| 758 |
"rewards/difficulty_following_reward": 0.0,
|
| 759 |
-
"rewards/format_reward": 0.
|
| 760 |
"step": 47,
|
| 761 |
-
"vanishing_advantage_ratio": 0.
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"clip_ratio": 0.0,
|
| 765 |
-
"completion_length":
|
| 766 |
"epoch": 0.08226221079691516,
|
| 767 |
-
"grad_norm": 0.
|
| 768 |
-
"kl": 0.
|
| 769 |
"learning_rate": 6.588648530198504e-07,
|
| 770 |
-
"loss": 0.
|
| 771 |
-
"reward": 0.
|
| 772 |
-
"reward_std": 0.
|
| 773 |
-
"rewards/accuracy_reward": 0.
|
| 774 |
"rewards/difficulty_following_reward": 0.0,
|
| 775 |
-
"rewards/format_reward": 0.
|
| 776 |
"step": 48,
|
| 777 |
-
"vanishing_advantage_ratio": 0.
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"clip_ratio": 0.0,
|
| 781 |
-
"completion_length":
|
| 782 |
"epoch": 0.08397600685518423,
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
-
"kl": 0.
|
| 785 |
"learning_rate": 6.435602608679916e-07,
|
| 786 |
-
"loss": 0.
|
| 787 |
-
"reward": 0.
|
| 788 |
-
"reward_std": 0.
|
| 789 |
-
"rewards/accuracy_reward": 0.
|
| 790 |
"rewards/difficulty_following_reward": 0.0,
|
| 791 |
-
"rewards/format_reward": 0.
|
| 792 |
"step": 49,
|
| 793 |
-
"vanishing_advantage_ratio": 0.
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"clip_ratio": 0.0,
|
| 797 |
-
"completion_length":
|
| 798 |
"epoch": 0.0856898029134533,
|
| 799 |
-
"grad_norm": 0.
|
| 800 |
-
"kl": 0.
|
| 801 |
"learning_rate": 6.281416799501187e-07,
|
| 802 |
-
"loss": 0.
|
| 803 |
-
"reward": 0.
|
| 804 |
-
"reward_std": 0.
|
| 805 |
-
"rewards/accuracy_reward": 0.
|
| 806 |
"rewards/difficulty_following_reward": 0.0,
|
| 807 |
-
"rewards/format_reward": 0.
|
| 808 |
"step": 50,
|
| 809 |
-
"vanishing_advantage_ratio": 0.
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"clip_ratio": 0.0,
|
| 813 |
-
"completion_length":
|
| 814 |
"epoch": 0.08740359897172237,
|
| 815 |
-
"grad_norm": 0.
|
| 816 |
-
"kl": 0.
|
| 817 |
"learning_rate": 6.126278954320294e-07,
|
| 818 |
-
"loss": 0.
|
| 819 |
-
"reward": 0.
|
| 820 |
-
"reward_std": 0.
|
| 821 |
-
"rewards/accuracy_reward": 0.
|
| 822 |
"rewards/difficulty_following_reward": 0.0,
|
| 823 |
-
"rewards/format_reward": 0.
|
| 824 |
"step": 51,
|
| 825 |
-
"vanishing_advantage_ratio": 0.
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"clip_ratio": 0.0,
|
| 829 |
-
"completion_length":
|
| 830 |
"epoch": 0.08911739502999143,
|
| 831 |
-
"grad_norm": 0.
|
| 832 |
-
"kl": 0.
|
| 833 |
"learning_rate": 5.97037808470444e-07,
|
| 834 |
-
"loss": 0.
|
| 835 |
-
"reward": 0.
|
| 836 |
-
"reward_std": 0.
|
| 837 |
-
"rewards/accuracy_reward": 0.
|
| 838 |
"rewards/difficulty_following_reward": 0.0,
|
| 839 |
-
"rewards/format_reward": 0.
|
| 840 |
"step": 52,
|
| 841 |
-
"vanishing_advantage_ratio": 0.
|
| 842 |
},
|
| 843 |
{
|
| 844 |
"clip_ratio": 0.0,
|
| 845 |
-
"completion_length":
|
| 846 |
"epoch": 0.0908311910882605,
|
| 847 |
-
"grad_norm": 0.
|
| 848 |
-
"kl": 0.
|
| 849 |
"learning_rate": 5.813904131848564e-07,
|
| 850 |
-
"loss": 0.
|
| 851 |
-
"reward": 0.
|
| 852 |
-
"reward_std": 0.
|
| 853 |
-
"rewards/accuracy_reward": 0.
|
| 854 |
"rewards/difficulty_following_reward": 0.0,
|
| 855 |
-
"rewards/format_reward": 0.
|
| 856 |
"step": 53,
|
| 857 |
-
"vanishing_advantage_ratio": 0.
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"clip_ratio": 0.0,
|
| 861 |
-
"completion_length":
|
| 862 |
"epoch": 0.09254498714652956,
|
| 863 |
-
"grad_norm": 0.
|
| 864 |
-
"kl": 0.
|
| 865 |
"learning_rate": 5.657047735161255e-07,
|
| 866 |
-
"loss": 0.
|
| 867 |
-
"reward": 0.
|
| 868 |
-
"reward_std": 0.
|
| 869 |
-
"rewards/accuracy_reward": 0.
|
| 870 |
"rewards/difficulty_following_reward": 0.0,
|
| 871 |
-
"rewards/format_reward": 0.
|
| 872 |
"step": 54,
|
| 873 |
-
"vanishing_advantage_ratio": 0.
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"clip_ratio": 0.0,
|
| 877 |
-
"completion_length":
|
| 878 |
"epoch": 0.09425878320479864,
|
| 879 |
-
"grad_norm": 0.
|
| 880 |
-
"kl": 0.
|
| 881 |
"learning_rate": 5.5e-07,
|
| 882 |
-
"loss": 0.
|
| 883 |
-
"reward": 0.
|
| 884 |
-
"reward_std": 0.
|
| 885 |
-
"rewards/accuracy_reward": 0.
|
| 886 |
"rewards/difficulty_following_reward": 0.0,
|
| 887 |
-
"rewards/format_reward": 0.
|
| 888 |
"step": 55,
|
| 889 |
-
"vanishing_advantage_ratio": 0.
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"clip_ratio": 0.0,
|
| 893 |
-
"completion_length":
|
| 894 |
"epoch": 0.0959725792630677,
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
-
"kl": 0.
|
| 897 |
"learning_rate": 5.342952264838747e-07,
|
| 898 |
-
"loss": 0.
|
| 899 |
-
"reward": 0.
|
| 900 |
-
"reward_std": 0.
|
| 901 |
-
"rewards/accuracy_reward": 0.
|
| 902 |
"rewards/difficulty_following_reward": 0.0,
|
| 903 |
-
"rewards/format_reward": 0.
|
| 904 |
"step": 56,
|
| 905 |
-
"vanishing_advantage_ratio": 0.
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"clip_ratio": 0.0,
|
| 909 |
-
"completion_length":
|
| 910 |
"epoch": 0.09768637532133675,
|
| 911 |
-
"grad_norm": 0.
|
| 912 |
-
"kl": 0.
|
| 913 |
"learning_rate": 5.186095868151436e-07,
|
| 914 |
-
"loss": 0.
|
| 915 |
-
"reward": 0.
|
| 916 |
-
"reward_std": 0.
|
| 917 |
-
"rewards/accuracy_reward": 0.
|
| 918 |
"rewards/difficulty_following_reward": 0.0,
|
| 919 |
-
"rewards/format_reward": 0.
|
| 920 |
"step": 57,
|
| 921 |
-
"vanishing_advantage_ratio": 0.
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"clip_ratio": 0.0,
|
| 925 |
-
"completion_length":
|
| 926 |
"epoch": 0.09940017137960583,
|
| 927 |
-
"grad_norm": 0.
|
| 928 |
-
"kl": 0.
|
| 929 |
"learning_rate": 5.02962191529556e-07,
|
| 930 |
-
"loss": 0.
|
| 931 |
-
"reward": 0.
|
| 932 |
-
"reward_std": 0.
|
| 933 |
-
"rewards/accuracy_reward": 0.
|
| 934 |
"rewards/difficulty_following_reward": 0.0,
|
| 935 |
-
"rewards/format_reward": 0.
|
| 936 |
"step": 58,
|
| 937 |
-
"vanishing_advantage_ratio": 0.
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"clip_ratio": 0.0,
|
| 941 |
-
"completion_length":
|
| 942 |
"epoch": 0.10111396743787489,
|
| 943 |
-
"grad_norm": 0.
|
| 944 |
-
"kl": 0.
|
| 945 |
"learning_rate": 4.873721045679706e-07,
|
| 946 |
-
"loss": 0.
|
| 947 |
-
"reward": 0.
|
| 948 |
-
"reward_std": 0.
|
| 949 |
-
"rewards/accuracy_reward": 0.
|
| 950 |
"rewards/difficulty_following_reward": 0.0,
|
| 951 |
-
"rewards/format_reward": 0.
|
| 952 |
"step": 59,
|
| 953 |
"vanishing_advantage_ratio": 0.1666666716337204
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"clip_ratio": 0.0,
|
| 957 |
-
"completion_length":
|
| 958 |
"epoch": 0.10282776349614396,
|
| 959 |
-
"grad_norm": 0.
|
| 960 |
-
"kl": 0.
|
| 961 |
"learning_rate": 4.7185832004988133e-07,
|
| 962 |
-
"loss": 0.
|
| 963 |
-
"reward": 0.
|
| 964 |
-
"reward_std": 0.
|
| 965 |
-
"rewards/accuracy_reward": 0.
|
| 966 |
"rewards/difficulty_following_reward": 0.0,
|
| 967 |
-
"rewards/format_reward": 0.
|
| 968 |
"step": 60,
|
| 969 |
-
"vanishing_advantage_ratio": 0.
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"clip_ratio": 0.0,
|
| 973 |
-
"completion_length":
|
| 974 |
"epoch": 0.10454155955441302,
|
| 975 |
-
"grad_norm": 0.
|
| 976 |
-
"kl": 0.
|
| 977 |
"learning_rate": 4.5643973913200837e-07,
|
| 978 |
-
"loss": 0.
|
| 979 |
-
"reward": 0.
|
| 980 |
-
"reward_std": 0.
|
| 981 |
-
"rewards/accuracy_reward": 0.
|
| 982 |
"rewards/difficulty_following_reward": 0.0,
|
| 983 |
-
"rewards/format_reward": 0.
|
| 984 |
"step": 61,
|
| 985 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"clip_ratio": 0.0,
|
| 989 |
-
"completion_length":
|
| 990 |
"epoch": 0.1062553556126821,
|
| 991 |
-
"grad_norm": 0.
|
| 992 |
-
"kl": 0.
|
| 993 |
"learning_rate": 4.4113514698014953e-07,
|
| 994 |
-
"loss": 0.
|
| 995 |
-
"reward": 0.
|
| 996 |
-
"reward_std": 0.
|
| 997 |
-
"rewards/accuracy_reward": 0.
|
| 998 |
"rewards/difficulty_following_reward": 0.0,
|
| 999 |
-
"rewards/format_reward": 0.
|
| 1000 |
"step": 62,
|
| 1001 |
-
"vanishing_advantage_ratio": 0.
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"clip_ratio": 0.0,
|
| 1005 |
-
"completion_length":
|
| 1006 |
"epoch": 0.10796915167095116,
|
| 1007 |
-
"grad_norm": 0.
|
| 1008 |
-
"kl": 0.
|
| 1009 |
"learning_rate": 4.2596318988235037e-07,
|
| 1010 |
-
"loss": 0.
|
| 1011 |
-
"reward": 0.
|
| 1012 |
-
"reward_std": 0.
|
| 1013 |
-
"rewards/accuracy_reward": 0.
|
| 1014 |
"rewards/difficulty_following_reward": 0.0,
|
| 1015 |
-
"rewards/format_reward": 0.
|
| 1016 |
"step": 63,
|
| 1017 |
-
"vanishing_advantage_ratio": 0.
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"clip_ratio": 0.0,
|
| 1021 |
-
"completion_length":
|
| 1022 |
"epoch": 0.10968294772922023,
|
| 1023 |
-
"grad_norm": 0.
|
| 1024 |
-
"kl": 0.
|
| 1025 |
"learning_rate": 4.1094235253127374e-07,
|
| 1026 |
-
"loss": 0.
|
| 1027 |
-
"reward": 0.
|
| 1028 |
-
"reward_std": 0.
|
| 1029 |
-
"rewards/accuracy_reward": 0.
|
| 1030 |
"rewards/difficulty_following_reward": 0.0,
|
| 1031 |
-
"rewards/format_reward": 0.
|
| 1032 |
"step": 64,
|
| 1033 |
-
"vanishing_advantage_ratio": 0.
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"clip_ratio": 0.0,
|
| 1037 |
-
"completion_length":
|
| 1038 |
"epoch": 0.11139674378748929,
|
| 1039 |
-
"grad_norm": 0.
|
| 1040 |
-
"kl": 0.
|
| 1041 |
"learning_rate": 3.9609093550344907e-07,
|
| 1042 |
-
"loss": 0.
|
| 1043 |
-
"reward": 0.
|
| 1044 |
-
"reward_std": 0.
|
| 1045 |
-
"rewards/accuracy_reward": 0.
|
| 1046 |
"rewards/difficulty_following_reward": 0.0,
|
| 1047 |
-
"rewards/format_reward": 0.
|
| 1048 |
"step": 65,
|
| 1049 |
-
"vanishing_advantage_ratio": 0.
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
"clip_ratio": 0.0,
|
| 1053 |
-
"completion_length":
|
| 1054 |
"epoch": 0.11311053984575835,
|
| 1055 |
-
"grad_norm": 0.
|
| 1056 |
-
"kl": 0.
|
| 1057 |
"learning_rate": 3.8142703296283953e-07,
|
| 1058 |
-
"loss": 0.
|
| 1059 |
-
"reward": 0.
|
| 1060 |
-
"reward_std": 0.
|
| 1061 |
-
"rewards/accuracy_reward": 0.
|
| 1062 |
"rewards/difficulty_following_reward": 0.0,
|
| 1063 |
-
"rewards/format_reward": 0.
|
| 1064 |
"step": 66,
|
| 1065 |
-
"vanishing_advantage_ratio": 0.
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"clip_ratio": 0.0,
|
| 1069 |
-
"completion_length":
|
| 1070 |
"epoch": 0.11482433590402742,
|
| 1071 |
-
"grad_norm": 0.
|
| 1072 |
-
"kl": 0.
|
| 1073 |
"learning_rate": 3.6696851061588994e-07,
|
| 1074 |
-
"loss": 0.
|
| 1075 |
-
"reward": 0.
|
| 1076 |
-
"reward_std": 0.
|
| 1077 |
-
"rewards/accuracy_reward": 0.
|
| 1078 |
"rewards/difficulty_following_reward": 0.0,
|
| 1079 |
-
"rewards/format_reward": 0.
|
| 1080 |
"step": 67,
|
| 1081 |
-
"vanishing_advantage_ratio": 0.
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"clip_ratio": 0.0,
|
| 1085 |
-
"completion_length":
|
| 1086 |
"epoch": 0.11653813196229648,
|
| 1087 |
-
"grad_norm": 0.
|
| 1088 |
-
"kl": 0.
|
| 1089 |
"learning_rate": 3.5273298394491515e-07,
|
| 1090 |
-
"loss": 0.
|
| 1091 |
-
"reward": 0.
|
| 1092 |
-
"reward_std": 0.
|
| 1093 |
-
"rewards/accuracy_reward": 0.
|
| 1094 |
"rewards/difficulty_following_reward": 0.0,
|
| 1095 |
-
"rewards/format_reward": 0.
|
| 1096 |
"step": 68,
|
| 1097 |
-
"vanishing_advantage_ratio": 0.
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"clip_ratio": 0.0,
|
| 1101 |
-
"completion_length":
|
| 1102 |
"epoch": 0.11825192802056556,
|
| 1103 |
-
"grad_norm": 0.
|
| 1104 |
-
"kl": 0.
|
| 1105 |
"learning_rate": 3.387377967463493e-07,
|
| 1106 |
-
"loss": 0.
|
| 1107 |
-
"reward": 0.
|
| 1108 |
-
"reward_std": 0.
|
| 1109 |
-
"rewards/accuracy_reward": 0.
|
| 1110 |
"rewards/difficulty_following_reward": 0.0,
|
| 1111 |
-
"rewards/format_reward": 0.
|
| 1112 |
"step": 69,
|
| 1113 |
-
"vanishing_advantage_ratio": 0.
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"clip_ratio": 0.0,
|
| 1117 |
-
"completion_length":
|
| 1118 |
"epoch": 0.11996572407883462,
|
| 1119 |
-
"grad_norm": 0.
|
| 1120 |
-
"kl": 0.
|
| 1121 |
"learning_rate": 3.250000000000001e-07,
|
| 1122 |
-
"loss": 0.
|
| 1123 |
-
"reward": 0.
|
| 1124 |
-
"reward_std": 0.
|
| 1125 |
-
"rewards/accuracy_reward": 0.
|
| 1126 |
"rewards/difficulty_following_reward": 0.0,
|
| 1127 |
-
"rewards/format_reward": 0.
|
| 1128 |
"step": 70,
|
| 1129 |
-
"vanishing_advantage_ratio": 0.
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"clip_ratio": 0.0,
|
| 1133 |
-
"completion_length":
|
| 1134 |
"epoch": 0.12167952013710369,
|
| 1135 |
-
"grad_norm": 0.
|
| 1136 |
-
"kl": 0.
|
| 1137 |
"learning_rate": 3.115363310950578e-07,
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"reward": 0.
|
| 1140 |
-
"reward_std": 0.
|
| 1141 |
-
"rewards/accuracy_reward": 0.
|
| 1142 |
"rewards/difficulty_following_reward": 0.0,
|
| 1143 |
-
"rewards/format_reward": 0.
|
| 1144 |
"step": 71,
|
| 1145 |
-
"vanishing_advantage_ratio": 0.
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"clip_ratio": 0.0,
|
| 1149 |
-
"completion_length":
|
| 1150 |
"epoch": 0.12339331619537275,
|
| 1151 |
-
"grad_norm": 0.
|
| 1152 |
-
"kl": 0.
|
| 1153 |
"learning_rate": 2.9836319343816397e-07,
|
| 1154 |
-
"loss": 0.
|
| 1155 |
-
"reward": 0.
|
| 1156 |
-
"reward_std": 0.
|
| 1157 |
-
"rewards/accuracy_reward": 0.
|
| 1158 |
"rewards/difficulty_following_reward": 0.0,
|
| 1159 |
-
"rewards/format_reward": 0.
|
| 1160 |
"step": 72,
|
| 1161 |
-
"vanishing_advantage_ratio": 0.
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"clip_ratio": 0.0,
|
| 1165 |
-
"completion_length":
|
| 1166 |
"epoch": 0.12510711225364182,
|
| 1167 |
-
"grad_norm": 0.
|
| 1168 |
-
"kl": 0.
|
| 1169 |
"learning_rate": 2.854966364683872e-07,
|
| 1170 |
-
"loss": 0.
|
| 1171 |
-
"reward": 0.
|
| 1172 |
-
"reward_std": 0.
|
| 1173 |
-
"rewards/accuracy_reward": 0.
|
| 1174 |
"rewards/difficulty_following_reward": 0.0,
|
| 1175 |
-
"rewards/format_reward": 0.
|
| 1176 |
"step": 73,
|
| 1177 |
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"clip_ratio": 0.0,
|
| 1181 |
-
"completion_length":
|
| 1182 |
"epoch": 0.1268209083119109,
|
| 1183 |
-
"grad_norm": 0.
|
| 1184 |
-
"kl": 0.
|
| 1185 |
"learning_rate": 2.729523361034538e-07,
|
| 1186 |
-
"loss": 0.
|
| 1187 |
-
"reward": 0.
|
| 1188 |
-
"reward_std": 0.
|
| 1189 |
-
"rewards/accuracy_reward": 0.
|
| 1190 |
"rewards/difficulty_following_reward": 0.0,
|
| 1191 |
-
"rewards/format_reward": 0.
|
| 1192 |
"step": 74,
|
| 1193 |
-
"vanishing_advantage_ratio": 0.
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"clip_ratio": 0.0,
|
| 1197 |
-
"completion_length":
|
| 1198 |
"epoch": 0.12853470437017994,
|
| 1199 |
-
"grad_norm": 0.
|
| 1200 |
-
"kl": 0.
|
| 1201 |
"learning_rate": 2.6074557564105724e-07,
|
| 1202 |
-
"loss": 0.
|
| 1203 |
-
"reward": 0.
|
| 1204 |
-
"reward_std": 0.
|
| 1205 |
-
"rewards/accuracy_reward": 0.
|
| 1206 |
"rewards/difficulty_following_reward": 0.0,
|
| 1207 |
-
"rewards/format_reward": 0.
|
| 1208 |
"step": 75,
|
| 1209 |
-
"vanishing_advantage_ratio": 0.
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"clip_ratio": 0.0,
|
| 1213 |
-
"completion_length":
|
| 1214 |
"epoch": 0.13024850042844902,
|
| 1215 |
-
"grad_norm": 0.
|
| 1216 |
-
"kl": 0.
|
| 1217 |
"learning_rate": 2.488912271385139e-07,
|
| 1218 |
-
"loss": 0.
|
| 1219 |
-
"reward": 0.
|
| 1220 |
-
"reward_std": 0.
|
| 1221 |
-
"rewards/accuracy_reward": 0.
|
| 1222 |
"rewards/difficulty_following_reward": 0.0,
|
| 1223 |
-
"rewards/format_reward": 0.
|
| 1224 |
"step": 76,
|
| 1225 |
-
"vanishing_advantage_ratio": 0.
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"clip_ratio": 0.0,
|
| 1229 |
-
"completion_length":
|
| 1230 |
"epoch": 0.1319622964867181,
|
| 1231 |
-
"grad_norm": 0.
|
| 1232 |
-
"kl": 0.
|
| 1233 |
"learning_rate": 2.374037332934512e-07,
|
| 1234 |
-
"loss": 0.
|
| 1235 |
-
"reward": 0.
|
| 1236 |
-
"reward_std": 0.
|
| 1237 |
-
"rewards/accuracy_reward": 0.
|
| 1238 |
"rewards/difficulty_following_reward": 0.0,
|
| 1239 |
-
"rewards/format_reward": 0.
|
| 1240 |
"step": 77,
|
| 1241 |
-
"vanishing_advantage_ratio": 0.
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"clip_ratio": 0.0,
|
| 1245 |
-
"completion_length":
|
| 1246 |
"epoch": 0.13367609254498714,
|
| 1247 |
-
"grad_norm": 0.
|
| 1248 |
-
"kl": 0.
|
| 1249 |
"learning_rate": 2.2629708984760706e-07,
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"reward": 0.
|
| 1252 |
-
"reward_std": 0.
|
| 1253 |
-
"rewards/accuracy_reward": 0.
|
| 1254 |
"rewards/difficulty_following_reward": 0.0,
|
| 1255 |
-
"rewards/format_reward": 0.
|
| 1256 |
"step": 78,
|
| 1257 |
-
"vanishing_advantage_ratio": 0.
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"clip_ratio": 0.0,
|
| 1261 |
-
"completion_length":
|
| 1262 |
"epoch": 0.1353898886032562,
|
| 1263 |
-
"grad_norm": 0.
|
| 1264 |
-
"kl": 0.
|
| 1265 |
"learning_rate": 2.1558482853517253e-07,
|
| 1266 |
-
"loss": 0.
|
| 1267 |
-
"reward": 0.
|
| 1268 |
-
"reward_std": 0.
|
| 1269 |
-
"rewards/accuracy_reward": 0.
|
| 1270 |
"rewards/difficulty_following_reward": 0.0,
|
| 1271 |
-
"rewards/format_reward": 0.
|
| 1272 |
"step": 79,
|
| 1273 |
-
"vanishing_advantage_ratio": 0.
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"clip_ratio": 0.0,
|
| 1277 |
-
"completion_length":
|
| 1278 |
"epoch": 0.13710368466152528,
|
| 1279 |
-
"grad_norm": 0.
|
| 1280 |
-
"kl": 0.
|
| 1281 |
"learning_rate": 2.0528000059645995e-07,
|
| 1282 |
-
"loss": 0.
|
| 1283 |
-
"reward": 0.
|
| 1284 |
-
"reward_std": 0.
|
| 1285 |
-
"rewards/accuracy_reward": 0.
|
| 1286 |
"rewards/difficulty_following_reward": 0.0,
|
| 1287 |
-
"rewards/format_reward": 0.
|
| 1288 |
"step": 80,
|
| 1289 |
-
"vanishing_advantage_ratio": 0.
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
"clip_ratio": 0.0,
|
| 1293 |
-
"completion_length":
|
| 1294 |
"epoch": 0.13881748071979436,
|
| 1295 |
-
"grad_norm": 0.
|
| 1296 |
-
"kl": 0.
|
| 1297 |
"learning_rate": 1.9539516087697517e-07,
|
| 1298 |
-
"loss": 0.
|
| 1299 |
-
"reward": 0.
|
| 1300 |
-
"reward_std": 0.
|
| 1301 |
-
"rewards/accuracy_reward": 0.
|
| 1302 |
"rewards/difficulty_following_reward": 0.0,
|
| 1303 |
-
"rewards/format_reward": 0.
|
| 1304 |
"step": 81,
|
| 1305 |
-
"vanishing_advantage_ratio": 0.
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"clip_ratio": 0.0,
|
| 1309 |
-
"completion_length":
|
| 1310 |
"epoch": 0.1405312767780634,
|
| 1311 |
-
"grad_norm": 0.
|
| 1312 |
-
"kl": 0.
|
| 1313 |
"learning_rate": 1.8594235253127372e-07,
|
| 1314 |
-
"loss": 0.
|
| 1315 |
-
"reward": 0.
|
| 1316 |
-
"reward_std": 0.
|
| 1317 |
-
"rewards/accuracy_reward": 0.
|
| 1318 |
"rewards/difficulty_following_reward": 0.0,
|
| 1319 |
-
"rewards/format_reward": 0.
|
| 1320 |
"step": 82,
|
| 1321 |
-
"vanishing_advantage_ratio": 0.
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"clip_ratio": 0.0,
|
| 1325 |
-
"completion_length":
|
| 1326 |
"epoch": 0.14224507283633248,
|
| 1327 |
-
"grad_norm": 0.
|
| 1328 |
-
"kl": 0.
|
| 1329 |
"learning_rate": 1.7693309235023127e-07,
|
| 1330 |
-
"loss": 0.
|
| 1331 |
-
"reward": 0.
|
| 1332 |
-
"reward_std": 0.
|
| 1333 |
-
"rewards/accuracy_reward": 0.
|
| 1334 |
"rewards/difficulty_following_reward": 0.0,
|
| 1335 |
-
"rewards/format_reward": 0.
|
| 1336 |
"step": 83,
|
| 1337 |
-
"vanishing_advantage_ratio": 0.
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"clip_ratio": 0.0,
|
| 1341 |
-
"completion_length":
|
| 1342 |
"epoch": 0.14395886889460155,
|
| 1343 |
-
"grad_norm": 0.
|
| 1344 |
-
"kl": 0.
|
| 1345 |
"learning_rate": 1.6837835672960831e-07,
|
| 1346 |
-
"loss": 0.
|
| 1347 |
-
"reward": 0.
|
| 1348 |
-
"reward_std": 0.
|
| 1349 |
-
"rewards/accuracy_reward": 0.
|
| 1350 |
"rewards/difficulty_following_reward": 0.0,
|
| 1351 |
-
"rewards/format_reward": 0.
|
| 1352 |
"step": 84,
|
| 1353 |
-
"vanishing_advantage_ratio": 0.
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"clip_ratio": 0.0,
|
| 1357 |
-
"completion_length":
|
| 1358 |
"epoch": 0.1456726649528706,
|
| 1359 |
-
"grad_norm": 0.
|
| 1360 |
-
"kl": 0.
|
| 1361 |
"learning_rate": 1.6028856829700258e-07,
|
| 1362 |
-
"loss": 0.
|
| 1363 |
-
"reward": 0.
|
| 1364 |
-
"reward_std": 0.
|
| 1365 |
-
"rewards/accuracy_reward": 0.
|
| 1366 |
"rewards/difficulty_following_reward": 0.0,
|
| 1367 |
-
"rewards/format_reward": 0.
|
| 1368 |
"step": 85,
|
| 1369 |
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"clip_ratio": 0.0,
|
| 1373 |
-
"completion_length":
|
| 1374 |
"epoch": 0.14738646101113967,
|
| 1375 |
-
"grad_norm": 0.
|
| 1376 |
-
"kl": 0.
|
| 1377 |
"learning_rate": 1.5267358321348285e-07,
|
| 1378 |
-
"loss": 0.
|
| 1379 |
-
"reward": 0.
|
| 1380 |
-
"reward_std": 0.
|
| 1381 |
-
"rewards/accuracy_reward": 0.
|
| 1382 |
"rewards/difficulty_following_reward": 0.0,
|
| 1383 |
-
"rewards/format_reward": 0.
|
| 1384 |
"step": 86,
|
| 1385 |
-
"vanishing_advantage_ratio": 0.
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"clip_ratio": 0.0,
|
| 1389 |
-
"completion_length":
|
| 1390 |
"epoch": 0.14910025706940874,
|
| 1391 |
-
"grad_norm": 0.
|
| 1392 |
-
"kl": 0.
|
| 1393 |
"learning_rate": 1.4554267916537495e-07,
|
| 1394 |
-
"loss": 0.
|
| 1395 |
-
"reward": 0.
|
| 1396 |
-
"reward_std": 0.
|
| 1397 |
-
"rewards/accuracy_reward": 0.
|
| 1398 |
"rewards/difficulty_following_reward": 0.0,
|
| 1399 |
-
"rewards/format_reward": 0.
|
| 1400 |
"step": 87,
|
| 1401 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1402 |
},
|
| 1403 |
{
|
| 1404 |
"clip_ratio": 0.0,
|
| 1405 |
-
"completion_length":
|
| 1406 |
"epoch": 0.15081405312767782,
|
| 1407 |
-
"grad_norm": 0.
|
| 1408 |
-
"kl": 0.
|
| 1409 |
"learning_rate": 1.3890454406082956e-07,
|
| 1410 |
-
"loss": 0.
|
| 1411 |
-
"reward": 0.
|
| 1412 |
-
"reward_std": 0.
|
| 1413 |
-
"rewards/accuracy_reward": 0.
|
| 1414 |
"rewards/difficulty_following_reward": 0.0,
|
| 1415 |
-
"rewards/format_reward": 0.
|
| 1416 |
"step": 88,
|
| 1417 |
-
"vanishing_advantage_ratio": 0.
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"clip_ratio": 0.0,
|
| 1421 |
-
"completion_length":
|
| 1422 |
"epoch": 0.15252784918594686,
|
| 1423 |
-
"grad_norm": 0.
|
| 1424 |
-
"kl": 0.
|
| 1425 |
"learning_rate": 1.3276726544494571e-07,
|
| 1426 |
-
"loss": 0.
|
| 1427 |
-
"reward": 0.
|
| 1428 |
-
"reward_std": 0.
|
| 1429 |
-
"rewards/accuracy_reward": 0.
|
| 1430 |
"rewards/difficulty_following_reward": 0.0,
|
| 1431 |
-
"rewards/format_reward": 0.
|
| 1432 |
"step": 89,
|
| 1433 |
-
"vanishing_advantage_ratio": 0.
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
"clip_ratio": 0.0,
|
| 1437 |
-
"completion_length":
|
| 1438 |
"epoch": 0.15424164524421594,
|
| 1439 |
-
"grad_norm": 0.
|
| 1440 |
-
"kl": 0.
|
| 1441 |
"learning_rate": 1.2713832064634125e-07,
|
| 1442 |
-
"loss": 0.
|
| 1443 |
-
"reward": 0.
|
| 1444 |
-
"reward_std": 0.
|
| 1445 |
"rewards/accuracy_reward": 0.055555556900799274,
|
| 1446 |
"rewards/difficulty_following_reward": 0.0,
|
| 1447 |
-
"rewards/format_reward": 0.
|
| 1448 |
"step": 90,
|
| 1449 |
-
"vanishing_advantage_ratio": 0.
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"clip_ratio": 0.0,
|
| 1453 |
-
"completion_length":
|
| 1454 |
"epoch": 0.155955441302485,
|
| 1455 |
-
"grad_norm": 0.
|
| 1456 |
-
"kl": 0.
|
| 1457 |
"learning_rate": 1.220245676671809e-07,
|
| 1458 |
-
"loss": 0.
|
| 1459 |
-
"reward": 0.
|
| 1460 |
-
"reward_std": 0.
|
| 1461 |
-
"rewards/accuracy_reward": 0.
|
| 1462 |
"rewards/difficulty_following_reward": 0.0,
|
| 1463 |
-
"rewards/format_reward": 0.
|
| 1464 |
"step": 91,
|
| 1465 |
-
"vanishing_advantage_ratio": 0.
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"clip_ratio": 0.0,
|
| 1469 |
-
"completion_length":
|
| 1470 |
"epoch": 0.15766923736075408,
|
| 1471 |
-
"grad_norm": 0.
|
| 1472 |
-
"kl": 0.
|
| 1473 |
"learning_rate": 1.1743223682775649e-07,
|
| 1474 |
-
"loss": 0.
|
| 1475 |
-
"reward": 0.
|
| 1476 |
-
"reward_std": 0.
|
| 1477 |
-
"rewards/accuracy_reward": 0.
|
| 1478 |
"rewards/difficulty_following_reward": 0.0,
|
| 1479 |
-
"rewards/format_reward": 0.
|
| 1480 |
"step": 92,
|
| 1481 |
-
"vanishing_advantage_ratio": 0.
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"clip_ratio": 0.0,
|
| 1485 |
-
"completion_length":
|
| 1486 |
"epoch": 0.15938303341902313,
|
| 1487 |
-
"grad_norm": 0.
|
| 1488 |
-
"kl": 0.
|
| 1489 |
"learning_rate": 1.1336692317580158e-07,
|
| 1490 |
-
"loss": 0.
|
| 1491 |
-
"reward": 0.
|
| 1492 |
-
"reward_std": 0.
|
| 1493 |
-
"rewards/accuracy_reward": 0.
|
| 1494 |
"rewards/difficulty_following_reward": 0.0,
|
| 1495 |
-
"rewards/format_reward": 0.
|
| 1496 |
"step": 93,
|
| 1497 |
-
"vanishing_advantage_ratio": 0.
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"clip_ratio": 0.0,
|
| 1501 |
-
"completion_length":
|
| 1502 |
"epoch": 0.1610968294772922,
|
| 1503 |
-
"grad_norm": 0.
|
| 1504 |
-
"kl": 0.
|
| 1505 |
"learning_rate": 1.0983357966978745e-07,
|
| 1506 |
-
"loss": 0.
|
| 1507 |
-
"reward": 0.
|
| 1508 |
-
"reward_std": 0.
|
| 1509 |
-
"rewards/accuracy_reward": 0.
|
| 1510 |
"rewards/difficulty_following_reward": 0.0,
|
| 1511 |
-
"rewards/format_reward": 0.
|
| 1512 |
"step": 94,
|
| 1513 |
"vanishing_advantage_ratio": 0.4166666716337204
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
"clip_ratio": 0.0,
|
| 1517 |
-
"completion_length":
|
| 1518 |
"epoch": 0.16281062553556128,
|
| 1519 |
-
"grad_norm": 0.
|
| 1520 |
-
"kl": 0.
|
| 1521 |
"learning_rate": 1.068365111445064e-07,
|
| 1522 |
-
"loss": 0.
|
| 1523 |
-
"reward": 0.
|
| 1524 |
-
"reward_std": 0.
|
| 1525 |
-
"rewards/accuracy_reward": 0.
|
| 1526 |
"rewards/difficulty_following_reward": 0.0,
|
| 1527 |
-
"rewards/format_reward": 0.
|
| 1528 |
"step": 95,
|
| 1529 |
"vanishing_advantage_ratio": 0.0833333358168602
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"clip_ratio": 0.0,
|
| 1533 |
-
"completion_length":
|
| 1534 |
"epoch": 0.16452442159383032,
|
| 1535 |
-
"grad_norm": 0.
|
| 1536 |
-
"kl": 0.
|
| 1537 |
"learning_rate": 1.0437936906629334e-07,
|
| 1538 |
-
"loss": 0.
|
| 1539 |
-
"reward": 0.
|
| 1540 |
-
"reward_std": 0.
|
| 1541 |
-
"rewards/accuracy_reward": 0.
|
| 1542 |
"rewards/difficulty_following_reward": 0.0,
|
| 1543 |
-
"rewards/format_reward": 0.
|
| 1544 |
"step": 96,
|
| 1545 |
-
"vanishing_advantage_ratio": 0.
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"clip_ratio": 0.0,
|
| 1549 |
-
"completion_length":
|
| 1550 |
"epoch": 0.1662382176520994,
|
| 1551 |
-
"grad_norm": 0.
|
| 1552 |
-
"kl": 0.
|
| 1553 |
"learning_rate": 1.0246514708427701e-07,
|
| 1554 |
-
"loss": 0.
|
| 1555 |
-
"reward": 0.
|
| 1556 |
-
"reward_std": 0.
|
| 1557 |
-
"rewards/accuracy_reward": 0.
|
| 1558 |
"rewards/difficulty_following_reward": 0.0,
|
| 1559 |
-
"rewards/format_reward": 0.
|
| 1560 |
"step": 97,
|
| 1561 |
-
"vanishing_advantage_ratio": 0.
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"clip_ratio": 0.0,
|
| 1565 |
-
"completion_length":
|
| 1566 |
"epoch": 0.16795201371036847,
|
| 1567 |
-
"grad_norm": 0.
|
| 1568 |
-
"kl": 0.
|
| 1569 |
"learning_rate": 1.0109617738307911e-07,
|
| 1570 |
-
"loss": 0.
|
| 1571 |
-
"reward": 0.
|
| 1572 |
-
"reward_std": 0.
|
| 1573 |
-
"rewards/accuracy_reward": 0.
|
| 1574 |
"rewards/difficulty_following_reward": 0.0,
|
| 1575 |
-
"rewards/format_reward": 0.
|
| 1576 |
"step": 98,
|
| 1577 |
-
"vanishing_advantage_ratio": 0.
|
| 1578 |
},
|
| 1579 |
{
|
| 1580 |
"clip_ratio": 0.0,
|
| 1581 |
-
"completion_length":
|
| 1582 |
"epoch": 0.16966580976863754,
|
| 1583 |
-
"grad_norm": 0.
|
| 1584 |
-
"kl": 0.
|
| 1585 |
"learning_rate": 1.002741278414069e-07,
|
| 1586 |
-
"loss": 0.
|
| 1587 |
-
"reward": 0.
|
| 1588 |
-
"reward_std": 0.
|
| 1589 |
-
"rewards/accuracy_reward": 0.
|
| 1590 |
"rewards/difficulty_following_reward": 0.0,
|
| 1591 |
-
"rewards/format_reward": 0.
|
| 1592 |
"step": 99,
|
| 1593 |
-
"vanishing_advantage_ratio": 0.
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"clip_ratio": 0.0,
|
| 1597 |
-
"completion_length":
|
| 1598 |
"epoch": 0.1713796058269066,
|
| 1599 |
-
"grad_norm": 0.
|
| 1600 |
-
"kl": 0.
|
| 1601 |
"learning_rate": 1e-07,
|
| 1602 |
-
"loss": 0.
|
| 1603 |
-
"reward": 0.
|
| 1604 |
-
"reward_std": 0.
|
| 1605 |
-
"rewards/accuracy_reward": 0.
|
| 1606 |
"rewards/difficulty_following_reward": 0.0,
|
| 1607 |
-
"rewards/format_reward": 0.
|
| 1608 |
"step": 100,
|
| 1609 |
-
"vanishing_advantage_ratio": 0.
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
"epoch": 0.1713796058269066,
|
| 1613 |
"step": 100,
|
| 1614 |
"total_flos": 0.0,
|
| 1615 |
-
"train_loss": 0.
|
| 1616 |
-
"train_runtime":
|
| 1617 |
-
"train_samples_per_second": 0.
|
| 1618 |
-
"train_steps_per_second": 0.
|
| 1619 |
}
|
| 1620 |
],
|
| 1621 |
"logging_steps": 1,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"clip_ratio": 0.0,
|
| 13 |
+
"completion_length": 2910.513916015625,
|
| 14 |
"epoch": 0.001713796058269066,
|
| 15 |
+
"grad_norm": 0.07309982925653458,
|
| 16 |
"kl": 0.0,
|
| 17 |
"learning_rate": 1e-07,
|
| 18 |
+
"loss": 0.0596,
|
| 19 |
+
"reward": 0.11805555876344442,
|
| 20 |
+
"reward_std": 0.060043493285775185,
|
| 21 |
+
"rewards/accuracy_reward": 0.0,
|
| 22 |
"rewards/difficulty_following_reward": 0.0,
|
| 23 |
+
"rewards/format_reward": 0.11805555876344442,
|
| 24 |
"step": 1,
|
| 25 |
+
"vanishing_advantage_ratio": 0.7500000074505806
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"clip_ratio": 0.0,
|
| 29 |
+
"completion_length": 2873.4722900390625,
|
| 30 |
"epoch": 0.003427592116538132,
|
| 31 |
+
"grad_norm": 0.13450193405151367,
|
| 32 |
"kl": 0.0,
|
| 33 |
"learning_rate": 2e-07,
|
| 34 |
+
"loss": 0.0299,
|
| 35 |
+
"reward": 0.17361111612990499,
|
| 36 |
+
"reward_std": 0.13970773108303547,
|
| 37 |
+
"rewards/accuracy_reward": 0.0,
|
| 38 |
"rewards/difficulty_following_reward": 0.0,
|
| 39 |
+
"rewards/format_reward": 0.17361111612990499,
|
| 40 |
"step": 2,
|
| 41 |
+
"vanishing_advantage_ratio": 0.5416666865348816
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"clip_ratio": 0.0,
|
| 45 |
+
"completion_length": 2874.8056030273438,
|
| 46 |
"epoch": 0.005141388174807198,
|
| 47 |
+
"grad_norm": 0.1418725550174713,
|
| 48 |
+
"kl": 4.89354133605957e-05,
|
| 49 |
"learning_rate": 3e-07,
|
| 50 |
+
"loss": 0.0777,
|
| 51 |
+
"reward": 0.1527777835726738,
|
| 52 |
+
"reward_std": 0.176795095205307,
|
| 53 |
+
"rewards/accuracy_reward": 0.0,
|
| 54 |
"rewards/difficulty_following_reward": 0.0,
|
| 55 |
+
"rewards/format_reward": 0.1527777835726738,
|
| 56 |
"step": 3,
|
| 57 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"clip_ratio": 0.0,
|
| 61 |
+
"completion_length": 2538.0694580078125,
|
| 62 |
"epoch": 0.006855184233076264,
|
| 63 |
+
"grad_norm": 0.23347292840480804,
|
| 64 |
+
"kl": 6.121397018432617e-05,
|
| 65 |
"learning_rate": 4e-07,
|
| 66 |
+
"loss": 0.1711,
|
| 67 |
+
"reward": 0.2500000074505806,
|
| 68 |
+
"reward_std": 0.23697294667363167,
|
| 69 |
+
"rewards/accuracy_reward": 0.0,
|
| 70 |
"rewards/difficulty_following_reward": 0.0,
|
| 71 |
+
"rewards/format_reward": 0.2500000074505806,
|
| 72 |
"step": 4,
|
| 73 |
+
"vanishing_advantage_ratio": 0.0833333358168602
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"clip_ratio": 0.0,
|
| 77 |
+
"completion_length": 2422.1666259765625,
|
| 78 |
"epoch": 0.00856898029134533,
|
| 79 |
+
"grad_norm": 0.25554895401000977,
|
| 80 |
+
"kl": 6.949901580810547e-05,
|
| 81 |
"learning_rate": 5e-07,
|
| 82 |
+
"loss": 0.1879,
|
| 83 |
+
"reward": 0.22916666977107525,
|
| 84 |
+
"reward_std": 0.21095014363527298,
|
| 85 |
+
"rewards/accuracy_reward": 0.0,
|
| 86 |
"rewards/difficulty_following_reward": 0.0,
|
| 87 |
+
"rewards/format_reward": 0.22916666977107525,
|
| 88 |
"step": 5,
|
| 89 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"clip_ratio": 0.0,
|
| 93 |
+
"completion_length": 2720.361083984375,
|
| 94 |
"epoch": 0.010282776349614395,
|
| 95 |
+
"grad_norm": 0.1855655461549759,
|
| 96 |
+
"kl": 5.7816505432128906e-05,
|
| 97 |
"learning_rate": 6e-07,
|
| 98 |
+
"loss": 0.1443,
|
| 99 |
+
"reward": 0.1666666716337204,
|
| 100 |
+
"reward_std": 0.2220027968287468,
|
| 101 |
+
"rewards/accuracy_reward": 0.0,
|
| 102 |
"rewards/difficulty_following_reward": 0.0,
|
| 103 |
+
"rewards/format_reward": 0.1666666716337204,
|
| 104 |
"step": 6,
|
| 105 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"clip_ratio": 0.0,
|
| 109 |
+
"completion_length": 2124.763885498047,
|
| 110 |
"epoch": 0.011996572407883462,
|
| 111 |
+
"grad_norm": 0.19423389434814453,
|
| 112 |
+
"kl": 5.745887756347656e-05,
|
| 113 |
"learning_rate": 7e-07,
|
| 114 |
+
"loss": 0.0686,
|
| 115 |
+
"reward": 0.25,
|
| 116 |
+
"reward_std": 0.17486263997852802,
|
| 117 |
+
"rewards/accuracy_reward": 0.0,
|
| 118 |
"rewards/difficulty_following_reward": 0.0,
|
| 119 |
+
"rewards/format_reward": 0.25,
|
| 120 |
"step": 7,
|
| 121 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"clip_ratio": 0.0,
|
| 125 |
+
"completion_length": 2672.6527709960938,
|
| 126 |
"epoch": 0.013710368466152529,
|
| 127 |
+
"grad_norm": 0.1445378065109253,
|
| 128 |
+
"kl": 3.981590270996094e-05,
|
| 129 |
"learning_rate": 8e-07,
|
| 130 |
+
"loss": 0.0707,
|
| 131 |
+
"reward": 0.18055555783212185,
|
| 132 |
+
"reward_std": 0.20800206437706947,
|
| 133 |
+
"rewards/accuracy_reward": 0.0,
|
| 134 |
"rewards/difficulty_following_reward": 0.0,
|
| 135 |
+
"rewards/format_reward": 0.18055555783212185,
|
| 136 |
"step": 8,
|
| 137 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"clip_ratio": 0.0,
|
| 141 |
+
"completion_length": 2435.027801513672,
|
| 142 |
"epoch": 0.015424164524421594,
|
| 143 |
+
"grad_norm": 0.2157757580280304,
|
| 144 |
+
"kl": 5.0008296966552734e-05,
|
| 145 |
"learning_rate": 9e-07,
|
| 146 |
+
"loss": 0.1989,
|
| 147 |
+
"reward": 0.2152777798473835,
|
| 148 |
+
"reward_std": 0.16122430190443993,
|
| 149 |
+
"rewards/accuracy_reward": 0.0,
|
| 150 |
"rewards/difficulty_following_reward": 0.0,
|
| 151 |
+
"rewards/format_reward": 0.2152777798473835,
|
| 152 |
"step": 9,
|
| 153 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"clip_ratio": 0.0,
|
| 157 |
+
"completion_length": 2983.874969482422,
|
| 158 |
"epoch": 0.01713796058269066,
|
| 159 |
+
"grad_norm": 0.1537931114435196,
|
| 160 |
+
"kl": 3.7670135498046875e-05,
|
| 161 |
"learning_rate": 1e-06,
|
| 162 |
+
"loss": 0.0572,
|
| 163 |
+
"reward": 0.1388888917863369,
|
| 164 |
+
"reward_std": 0.1325911059975624,
|
| 165 |
+
"rewards/accuracy_reward": 0.0,
|
| 166 |
"rewards/difficulty_following_reward": 0.0,
|
| 167 |
+
"rewards/format_reward": 0.1388888917863369,
|
| 168 |
"step": 10,
|
| 169 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"clip_ratio": 0.0,
|
| 173 |
+
"completion_length": 2714.6666870117188,
|
| 174 |
"epoch": 0.018851756640959727,
|
| 175 |
+
"grad_norm": 0.14581632614135742,
|
| 176 |
+
"kl": 4.884600639343262e-05,
|
| 177 |
"learning_rate": 9.997258721585931e-07,
|
| 178 |
+
"loss": 0.0249,
|
| 179 |
+
"reward": 0.4097222276031971,
|
| 180 |
+
"reward_std": 0.5287230610847473,
|
| 181 |
+
"rewards/accuracy_reward": 0.13888889085501432,
|
| 182 |
"rewards/difficulty_following_reward": 0.0,
|
| 183 |
+
"rewards/format_reward": 0.1319444477558136,
|
| 184 |
"step": 11,
|
| 185 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"clip_ratio": 0.0,
|
| 189 |
+
"completion_length": 3190.5555419921875,
|
| 190 |
"epoch": 0.02056555269922879,
|
| 191 |
+
"grad_norm": 0.12614919245243073,
|
| 192 |
+
"kl": 3.2901763916015625e-05,
|
| 193 |
"learning_rate": 9.989038226169207e-07,
|
| 194 |
+
"loss": 0.0385,
|
| 195 |
+
"reward": 0.305555556435138,
|
| 196 |
+
"reward_std": 0.44778930954635143,
|
| 197 |
+
"rewards/accuracy_reward": 0.08333333488553762,
|
| 198 |
"rewards/difficulty_following_reward": 0.0,
|
| 199 |
+
"rewards/format_reward": 0.13888889411464334,
|
| 200 |
"step": 12,
|
| 201 |
+
"vanishing_advantage_ratio": 0.2638889029622078
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"clip_ratio": 0.0,
|
| 205 |
+
"completion_length": 2701.9583129882812,
|
| 206 |
"epoch": 0.022279348757497857,
|
| 207 |
+
"grad_norm": 0.17424772679805756,
|
| 208 |
+
"kl": 0.00010010600090026855,
|
| 209 |
"learning_rate": 9.975348529157229e-07,
|
| 210 |
+
"loss": 0.0723,
|
| 211 |
+
"reward": 0.486111119389534,
|
| 212 |
+
"reward_std": 0.6361183077096939,
|
| 213 |
+
"rewards/accuracy_reward": 0.15277778264135122,
|
| 214 |
"rewards/difficulty_following_reward": 0.0,
|
| 215 |
+
"rewards/format_reward": 0.18055556016042829,
|
| 216 |
"step": 13,
|
| 217 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 218 |
},
|
| 219 |
{
|
| 220 |
"clip_ratio": 0.0,
|
| 221 |
+
"completion_length": 2698.0972595214844,
|
| 222 |
"epoch": 0.023993144815766924,
|
| 223 |
+
"grad_norm": 0.13479822874069214,
|
| 224 |
+
"kl": 7.367134094238281e-05,
|
| 225 |
"learning_rate": 9.956206309337066e-07,
|
| 226 |
+
"loss": 0.0696,
|
| 227 |
+
"reward": 0.38194445334374905,
|
| 228 |
+
"reward_std": 0.4227895326912403,
|
| 229 |
+
"rewards/accuracy_reward": 0.09722222294658422,
|
| 230 |
"rewards/difficulty_following_reward": 0.0,
|
| 231 |
+
"rewards/format_reward": 0.1875,
|
| 232 |
"step": 14,
|
| 233 |
+
"vanishing_advantage_ratio": 0.3611111268401146
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"clip_ratio": 0.0,
|
| 237 |
+
"completion_length": 3156.9722290039062,
|
| 238 |
"epoch": 0.02570694087403599,
|
| 239 |
+
"grad_norm": 0.12347124516963959,
|
| 240 |
+
"kl": 4.169344902038574e-05,
|
| 241 |
"learning_rate": 9.931634888554935e-07,
|
| 242 |
+
"loss": 0.0423,
|
| 243 |
+
"reward": 0.4791666679084301,
|
| 244 |
+
"reward_std": 0.5301840528845787,
|
| 245 |
+
"rewards/accuracy_reward": 0.16666666883975267,
|
| 246 |
"rewards/difficulty_following_reward": 0.0,
|
| 247 |
+
"rewards/format_reward": 0.14583333488553762,
|
| 248 |
"step": 15,
|
| 249 |
+
"vanishing_advantage_ratio": 0.2916666753590107
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"clip_ratio": 0.0,
|
| 253 |
+
"completion_length": 2310.138916015625,
|
| 254 |
"epoch": 0.027420736932305057,
|
| 255 |
+
"grad_norm": 0.20043165981769562,
|
| 256 |
+
"kl": 0.0002892017364501953,
|
| 257 |
"learning_rate": 9.901664203302124e-07,
|
| 258 |
+
"loss": 0.0751,
|
| 259 |
+
"reward": 0.5208333283662796,
|
| 260 |
+
"reward_std": 0.6124575287103653,
|
| 261 |
+
"rewards/accuracy_reward": 0.12500000279396772,
|
| 262 |
"rewards/difficulty_following_reward": 0.0,
|
| 263 |
+
"rewards/format_reward": 0.27083333767950535,
|
| 264 |
"step": 16,
|
| 265 |
+
"vanishing_advantage_ratio": 0.1944444552063942
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"clip_ratio": 0.0,
|
| 269 |
+
"completion_length": 2492.9583435058594,
|
| 270 |
"epoch": 0.02913453299057412,
|
| 271 |
+
"grad_norm": 0.17674975097179413,
|
| 272 |
+
"kl": 0.00019311904907226562,
|
| 273 |
"learning_rate": 9.866330768241983e-07,
|
| 274 |
+
"loss": 0.0971,
|
| 275 |
+
"reward": 0.4652777761220932,
|
| 276 |
+
"reward_std": 0.4706720970571041,
|
| 277 |
+
"rewards/accuracy_reward": 0.12500000186264515,
|
| 278 |
"rewards/difficulty_following_reward": 0.0,
|
| 279 |
+
"rewards/format_reward": 0.2152777798473835,
|
| 280 |
"step": 17,
|
| 281 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"clip_ratio": 0.0,
|
| 285 |
+
"completion_length": 2977.625,
|
| 286 |
"epoch": 0.030848329048843187,
|
| 287 |
+
"grad_norm": 0.11323793232440948,
|
| 288 |
+
"kl": 0.00023287534713745117,
|
| 289 |
"learning_rate": 9.825677631722435e-07,
|
| 290 |
+
"loss": 0.0141,
|
| 291 |
+
"reward": 0.35416666977107525,
|
| 292 |
+
"reward_std": 0.40337391197681427,
|
| 293 |
+
"rewards/accuracy_reward": 0.11111111473292112,
|
| 294 |
"rewards/difficulty_following_reward": 0.0,
|
| 295 |
+
"rewards/format_reward": 0.1319444477558136,
|
| 296 |
"step": 18,
|
| 297 |
+
"vanishing_advantage_ratio": 0.5000000074505806
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"clip_ratio": 0.0,
|
| 301 |
+
"completion_length": 2997.7222290039062,
|
| 302 |
"epoch": 0.032562125107112254,
|
| 303 |
+
"grad_norm": 0.12171383947134018,
|
| 304 |
+
"kl": 0.00020396709442138672,
|
| 305 |
"learning_rate": 9.779754323328192e-07,
|
| 306 |
+
"loss": 0.0399,
|
| 307 |
+
"reward": 0.3472222313284874,
|
| 308 |
+
"reward_std": 0.39615704864263535,
|
| 309 |
+
"rewards/accuracy_reward": 0.11111111473292112,
|
| 310 |
"rewards/difficulty_following_reward": 0.0,
|
| 311 |
+
"rewards/format_reward": 0.1250000037252903,
|
| 312 |
"step": 19,
|
| 313 |
+
"vanishing_advantage_ratio": 0.4444444514811039
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"clip_ratio": 0.0,
|
| 317 |
+
"completion_length": 3004.875,
|
| 318 |
"epoch": 0.03427592116538132,
|
| 319 |
+
"grad_norm": 0.15534202754497528,
|
| 320 |
+
"kl": 0.0001380443572998047,
|
| 321 |
"learning_rate": 9.728616793536587e-07,
|
| 322 |
+
"loss": 0.1373,
|
| 323 |
+
"reward": 0.3541666651144624,
|
| 324 |
+
"reward_std": 0.5004552379250526,
|
| 325 |
+
"rewards/accuracy_reward": 0.11111111287027597,
|
| 326 |
"rewards/difficulty_following_reward": 0.0,
|
| 327 |
+
"rewards/format_reward": 0.13194444682449102,
|
| 328 |
"step": 20,
|
| 329 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 330 |
},
|
| 331 |
{
|
| 332 |
"clip_ratio": 0.0,
|
| 333 |
+
"completion_length": 2767.9166870117188,
|
| 334 |
"epoch": 0.03598971722365039,
|
| 335 |
+
"grad_norm": 0.12289870530366898,
|
| 336 |
+
"kl": 0.0005383491516113281,
|
| 337 |
"learning_rate": 9.672327345550543e-07,
|
| 338 |
+
"loss": 0.0652,
|
| 339 |
+
"reward": 0.4722222313284874,
|
| 340 |
+
"reward_std": 0.4577459469437599,
|
| 341 |
+
"rewards/accuracy_reward": 0.1250000037252903,
|
| 342 |
"rewards/difficulty_following_reward": 0.0,
|
| 343 |
+
"rewards/format_reward": 0.2222222276031971,
|
| 344 |
"step": 21,
|
| 345 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"clip_ratio": 0.0,
|
| 349 |
+
"completion_length": 2990.611083984375,
|
| 350 |
"epoch": 0.037703513281919454,
|
| 351 |
+
"grad_norm": 0.1411183923482895,
|
| 352 |
+
"kl": 0.00030428171157836914,
|
| 353 |
"learning_rate": 9.610954559391704e-07,
|
| 354 |
+
"loss": 0.0744,
|
| 355 |
+
"reward": 0.2152777803130448,
|
| 356 |
+
"reward_std": 0.3392846863716841,
|
| 357 |
+
"rewards/accuracy_reward": 0.0555555559694767,
|
| 358 |
"rewards/difficulty_following_reward": 0.0,
|
| 359 |
+
"rewards/format_reward": 0.10416666883975267,
|
| 360 |
"step": 22,
|
| 361 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"clip_ratio": 0.0,
|
| 365 |
+
"completion_length": 3004.9583129882812,
|
| 366 |
"epoch": 0.03941730934018852,
|
| 367 |
+
"grad_norm": 0.10731787979602814,
|
| 368 |
+
"kl": 0.0005457401275634766,
|
| 369 |
"learning_rate": 9.54457320834625e-07,
|
| 370 |
+
"loss": 0.0293,
|
| 371 |
+
"reward": 0.20833333674818277,
|
| 372 |
+
"reward_std": 0.2911061681807041,
|
| 373 |
+
"rewards/accuracy_reward": 0.041666666977107525,
|
| 374 |
"rewards/difficulty_following_reward": 0.0,
|
| 375 |
+
"rewards/format_reward": 0.12500000139698386,
|
| 376 |
"step": 23,
|
| 377 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"clip_ratio": 0.0,
|
| 381 |
+
"completion_length": 2584.2916870117188,
|
| 382 |
"epoch": 0.04113110539845758,
|
| 383 |
+
"grad_norm": 0.15161892771720886,
|
| 384 |
+
"kl": 0.0003027915954589844,
|
| 385 |
"learning_rate": 9.473264167865171e-07,
|
| 386 |
+
"loss": 0.06,
|
| 387 |
+
"reward": 0.4861111044883728,
|
| 388 |
+
"reward_std": 0.47357870638370514,
|
| 389 |
+
"rewards/accuracy_reward": 0.13888889364898205,
|
| 390 |
"rewards/difficulty_following_reward": 0.0,
|
| 391 |
+
"rewards/format_reward": 0.2083333358168602,
|
| 392 |
"step": 24,
|
| 393 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"clip_ratio": 0.0,
|
| 397 |
+
"completion_length": 2536.3195190429688,
|
| 398 |
"epoch": 0.04284490145672665,
|
| 399 |
+
"grad_norm": 0.14601115882396698,
|
| 400 |
+
"kl": 0.0003859996795654297,
|
| 401 |
"learning_rate": 9.397114317029974e-07,
|
| 402 |
+
"loss": 0.0375,
|
| 403 |
+
"reward": 0.5277777910232544,
|
| 404 |
+
"reward_std": 0.5513055957853794,
|
| 405 |
+
"rewards/accuracy_reward": 0.13888889085501432,
|
| 406 |
"rewards/difficulty_following_reward": 0.0,
|
| 407 |
+
"rewards/format_reward": 0.25000000931322575,
|
| 408 |
"step": 25,
|
| 409 |
+
"vanishing_advantage_ratio": 0.2083333432674408
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"clip_ratio": 0.0,
|
| 413 |
+
"completion_length": 2963.8333129882812,
|
| 414 |
"epoch": 0.044558697514995714,
|
| 415 |
+
"grad_norm": 0.14391477406024933,
|
| 416 |
+
"kl": 0.00029397010803222656,
|
| 417 |
"learning_rate": 9.316216432703916e-07,
|
| 418 |
+
"loss": 0.0872,
|
| 419 |
+
"reward": 0.31944445334374905,
|
| 420 |
+
"reward_std": 0.39608718268573284,
|
| 421 |
+
"rewards/accuracy_reward": 0.08333333395421505,
|
| 422 |
"rewards/difficulty_following_reward": 0.0,
|
| 423 |
+
"rewards/format_reward": 0.15277778217568994,
|
| 424 |
"step": 26,
|
| 425 |
+
"vanishing_advantage_ratio": 0.34722223319113255
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"clip_ratio": 0.0,
|
| 429 |
+
"completion_length": 2815.4861450195312,
|
| 430 |
"epoch": 0.04627249357326478,
|
| 431 |
+
"grad_norm": 0.14442721009254456,
|
| 432 |
+
"kl": 0.00036406517028808594,
|
| 433 |
"learning_rate": 9.230669076497687e-07,
|
| 434 |
+
"loss": 0.062,
|
| 435 |
+
"reward": 0.3333333358168602,
|
| 436 |
+
"reward_std": 0.37698132544755936,
|
| 437 |
+
"rewards/accuracy_reward": 0.09722222294658422,
|
| 438 |
"rewards/difficulty_following_reward": 0.0,
|
| 439 |
+
"rewards/format_reward": 0.13888889364898205,
|
| 440 |
"step": 27,
|
| 441 |
+
"vanishing_advantage_ratio": 0.5000000149011612
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"clip_ratio": 0.0,
|
| 445 |
+
"completion_length": 2111.8333740234375,
|
| 446 |
"epoch": 0.04798628963153385,
|
| 447 |
+
"grad_norm": 0.20688049495220184,
|
| 448 |
+
"kl": 0.0007753372192382812,
|
| 449 |
"learning_rate": 9.140576474687263e-07,
|
| 450 |
+
"loss": 0.0729,
|
| 451 |
+
"reward": 0.5416666716337204,
|
| 452 |
+
"reward_std": 0.5932498127222061,
|
| 453 |
+
"rewards/accuracy_reward": 0.15277778171002865,
|
| 454 |
"rewards/difficulty_following_reward": 0.0,
|
| 455 |
+
"rewards/format_reward": 0.23611111007630825,
|
| 456 |
"step": 28,
|
| 457 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"clip_ratio": 0.0,
|
| 461 |
+
"completion_length": 2817.1111450195312,
|
| 462 |
"epoch": 0.049700085689802914,
|
| 463 |
+
"grad_norm": 0.1729794591665268,
|
| 464 |
+
"kl": 0.000820159912109375,
|
| 465 |
"learning_rate": 9.046048391230247e-07,
|
| 466 |
+
"loss": 0.0789,
|
| 467 |
+
"reward": 0.3958333288319409,
|
| 468 |
+
"reward_std": 0.45582315884530544,
|
| 469 |
+
"rewards/accuracy_reward": 0.15277778171002865,
|
| 470 |
"rewards/difficulty_following_reward": 0.0,
|
| 471 |
+
"rewards/format_reward": 0.09027778031304479,
|
| 472 |
"step": 29,
|
| 473 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"clip_ratio": 0.0,
|
| 477 |
+
"completion_length": 2341.2222595214844,
|
| 478 |
"epoch": 0.05141388174807198,
|
| 479 |
+
"grad_norm": 0.16700980067253113,
|
| 480 |
+
"kl": 0.000972747802734375,
|
| 481 |
"learning_rate": 8.9471999940354e-07,
|
| 482 |
+
"loss": 0.0022,
|
| 483 |
+
"reward": 0.3958333358168602,
|
| 484 |
+
"reward_std": 0.4832002595067024,
|
| 485 |
+
"rewards/accuracy_reward": 0.08333333395421505,
|
| 486 |
"rewards/difficulty_following_reward": 0.0,
|
| 487 |
+
"rewards/format_reward": 0.2291666679084301,
|
| 488 |
"step": 30,
|
| 489 |
+
"vanishing_advantage_ratio": 0.20833334140479565
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"clip_ratio": 0.0,
|
| 493 |
+
"completion_length": 2607.0416259765625,
|
| 494 |
"epoch": 0.05312767780634105,
|
| 495 |
+
"grad_norm": 0.13790923357009888,
|
| 496 |
+
"kl": 0.0006513595581054688,
|
| 497 |
"learning_rate": 8.844151714648274e-07,
|
| 498 |
+
"loss": 0.0044,
|
| 499 |
+
"reward": 0.3263888880610466,
|
| 500 |
+
"reward_std": 0.33597812056541443,
|
| 501 |
+
"rewards/accuracy_reward": 0.06944444496184587,
|
| 502 |
"rewards/difficulty_following_reward": 0.0,
|
| 503 |
+
"rewards/format_reward": 0.18750000558793545,
|
| 504 |
"step": 31,
|
| 505 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"clip_ratio": 0.0,
|
| 509 |
+
"completion_length": 2935.6806030273438,
|
| 510 |
"epoch": 0.054841473864610114,
|
| 511 |
+
"grad_norm": 0.13338877260684967,
|
| 512 |
+
"kl": 0.000720977783203125,
|
| 513 |
"learning_rate": 8.737029101523929e-07,
|
| 514 |
+
"loss": 0.0753,
|
| 515 |
+
"reward": 0.5416666716337204,
|
| 516 |
+
"reward_std": 0.6741950437426567,
|
| 517 |
+
"rewards/accuracy_reward": 0.16666667070239782,
|
| 518 |
"rewards/difficulty_following_reward": 0.0,
|
| 519 |
+
"rewards/format_reward": 0.2083333358168602,
|
| 520 |
"step": 32,
|
| 521 |
+
"vanishing_advantage_ratio": 0.1111111156642437
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"clip_ratio": 0.0,
|
| 525 |
+
"completion_length": 2178.9861755371094,
|
| 526 |
"epoch": 0.056555269922879174,
|
| 527 |
+
"grad_norm": 0.20104320347309113,
|
| 528 |
+
"kl": 0.0009927749633789062,
|
| 529 |
"learning_rate": 8.625962667065487e-07,
|
| 530 |
+
"loss": 0.0859,
|
| 531 |
+
"reward": 0.6458333283662796,
|
| 532 |
+
"reward_std": 0.5696279220283031,
|
| 533 |
+
"rewards/accuracy_reward": 0.19444444589316845,
|
| 534 |
"rewards/difficulty_following_reward": 0.0,
|
| 535 |
+
"rewards/format_reward": 0.2569444440305233,
|
| 536 |
"step": 33,
|
| 537 |
+
"vanishing_advantage_ratio": 0.1944444514811039
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"clip_ratio": 0.0,
|
| 541 |
+
"completion_length": 2810.7222290039062,
|
| 542 |
"epoch": 0.05826906598114824,
|
| 543 |
+
"grad_norm": 0.11122416704893112,
|
| 544 |
+
"kl": 0.0006494522094726562,
|
| 545 |
"learning_rate": 8.511087728614862e-07,
|
| 546 |
+
"loss": 0.0723,
|
| 547 |
+
"reward": 0.3611111119389534,
|
| 548 |
+
"reward_std": 0.2897772639989853,
|
| 549 |
+
"rewards/accuracy_reward": 0.09722222574055195,
|
| 550 |
"rewards/difficulty_following_reward": 0.0,
|
| 551 |
+
"rewards/format_reward": 0.16666666977107525,
|
| 552 |
"step": 34,
|
| 553 |
+
"vanishing_advantage_ratio": 0.5416666809469461
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"clip_ratio": 0.0,
|
| 557 |
+
"completion_length": 3160.1249389648438,
|
| 558 |
"epoch": 0.05998286203941731,
|
| 559 |
+
"grad_norm": 0.08742042630910873,
|
| 560 |
+
"kl": 0.0006799697875976562,
|
| 561 |
"learning_rate": 8.392544243589427e-07,
|
| 562 |
+
"loss": 0.011,
|
| 563 |
+
"reward": 0.180555556435138,
|
| 564 |
+
"reward_std": 0.20691198110580444,
|
| 565 |
+
"rewards/accuracy_reward": 0.041666666977107525,
|
| 566 |
"rewards/difficulty_following_reward": 0.0,
|
| 567 |
+
"rewards/format_reward": 0.09722222527489066,
|
| 568 |
"step": 35,
|
| 569 |
+
"vanishing_advantage_ratio": 0.5833333432674408
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"clip_ratio": 0.0,
|
| 573 |
+
"completion_length": 2915.0556030273438,
|
| 574 |
"epoch": 0.061696658097686374,
|
| 575 |
+
"grad_norm": 0.1445658802986145,
|
| 576 |
+
"kl": 0.0008459091186523438,
|
| 577 |
"learning_rate": 8.270476638965461e-07,
|
| 578 |
+
"loss": 0.0597,
|
| 579 |
+
"reward": 0.19444444589316845,
|
| 580 |
+
"reward_std": 0.22136816568672657,
|
| 581 |
+
"rewards/accuracy_reward": 0.041666666977107525,
|
| 582 |
"rewards/difficulty_following_reward": 0.0,
|
| 583 |
+
"rewards/format_reward": 0.11111111240461469,
|
| 584 |
"step": 36,
|
| 585 |
"vanishing_advantage_ratio": 0.5000000149011612
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"clip_ratio": 0.0,
|
| 589 |
+
"completion_length": 3040.625,
|
| 590 |
"epoch": 0.06341045415595545,
|
| 591 |
+
"grad_norm": 0.10913708806037903,
|
| 592 |
+
"kl": 0.0008916854858398438,
|
| 593 |
"learning_rate": 8.145033635316128e-07,
|
| 594 |
+
"loss": 0.0071,
|
| 595 |
+
"reward": 0.1597222238779068,
|
| 596 |
+
"reward_std": 0.17662217281758785,
|
| 597 |
+
"rewards/accuracy_reward": 0.041666666977107525,
|
| 598 |
"rewards/difficulty_following_reward": 0.0,
|
| 599 |
+
"rewards/format_reward": 0.07638889225199819,
|
| 600 |
"step": 37,
|
| 601 |
+
"vanishing_advantage_ratio": 0.7500000149011612
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"clip_ratio": 0.0,
|
| 605 |
+
"completion_length": 3170.638916015625,
|
| 606 |
"epoch": 0.06512425021422451,
|
| 607 |
+
"grad_norm": 0.11745505779981613,
|
| 608 |
+
"kl": 0.0007038116455078125,
|
| 609 |
"learning_rate": 8.01636806561836e-07,
|
| 610 |
+
"loss": 0.0956,
|
| 611 |
+
"reward": 0.19444444216787815,
|
| 612 |
+
"reward_std": 0.2374982163310051,
|
| 613 |
+
"rewards/accuracy_reward": 0.0555555559694767,
|
| 614 |
"rewards/difficulty_following_reward": 0.0,
|
| 615 |
+
"rewards/format_reward": 0.0833333358168602,
|
| 616 |
"step": 38,
|
| 617 |
+
"vanishing_advantage_ratio": 0.5000000074505806
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"clip_ratio": 0.0,
|
| 621 |
+
"completion_length": 2920.8472290039062,
|
| 622 |
"epoch": 0.06683804627249357,
|
| 623 |
+
"grad_norm": 0.12855492532253265,
|
| 624 |
+
"kl": 0.0009975433349609375,
|
| 625 |
"learning_rate": 7.884636689049422e-07,
|
| 626 |
+
"loss": 0.0913,
|
| 627 |
+
"reward": 0.4791666832752526,
|
| 628 |
+
"reward_std": 0.48048618994653225,
|
| 629 |
+
"rewards/accuracy_reward": 0.15277777891606092,
|
| 630 |
"rewards/difficulty_following_reward": 0.0,
|
| 631 |
+
"rewards/format_reward": 0.17361111054196954,
|
| 632 |
"step": 39,
|
| 633 |
"vanishing_advantage_ratio": 0.2500000074505806
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"clip_ratio": 0.0,
|
| 637 |
+
"completion_length": 2656.6666870117188,
|
| 638 |
"epoch": 0.06855184233076264,
|
| 639 |
+
"grad_norm": 0.17364995181560516,
|
| 640 |
+
"kl": 0.0011653900146484375,
|
| 641 |
"learning_rate": 7.75e-07,
|
| 642 |
+
"loss": 0.0587,
|
| 643 |
+
"reward": 0.4444444552063942,
|
| 644 |
+
"reward_std": 0.5394208431243896,
|
| 645 |
+
"rewards/accuracy_reward": 0.13888889271765947,
|
| 646 |
"rewards/difficulty_following_reward": 0.0,
|
| 647 |
+
"rewards/format_reward": 0.1666666679084301,
|
| 648 |
"step": 40,
|
| 649 |
"vanishing_advantage_ratio": 0.1944444552063942
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"clip_ratio": 0.0,
|
| 653 |
+
"completion_length": 2850.90283203125,
|
| 654 |
"epoch": 0.0702656383890317,
|
| 655 |
+
"grad_norm": 0.15265439450740814,
|
| 656 |
+
"kl": 0.0010967254638671875,
|
| 657 |
"learning_rate": 7.612622032536507e-07,
|
| 658 |
+
"loss": 0.0992,
|
| 659 |
+
"reward": 0.4305555606260896,
|
| 660 |
+
"reward_std": 0.4596329629421234,
|
| 661 |
+
"rewards/accuracy_reward": 0.12500000093132257,
|
| 662 |
"rewards/difficulty_following_reward": 0.0,
|
| 663 |
"rewards/format_reward": 0.18055556062608957,
|
| 664 |
"step": 41,
|
| 665 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"clip_ratio": 0.0,
|
| 669 |
+
"completion_length": 2313.3612060546875,
|
| 670 |
"epoch": 0.07197943444730077,
|
| 671 |
+
"grad_norm": 0.23461805284023285,
|
| 672 |
+
"kl": 0.001087188720703125,
|
| 673 |
"learning_rate": 7.472670160550848e-07,
|
| 674 |
+
"loss": 0.1506,
|
| 675 |
+
"reward": 0.4305555634200573,
|
| 676 |
+
"reward_std": 0.33141833916306496,
|
| 677 |
"rewards/accuracy_reward": 0.09722222294658422,
|
| 678 |
"rewards/difficulty_following_reward": 0.0,
|
| 679 |
+
"rewards/format_reward": 0.2361111119389534,
|
| 680 |
"step": 42,
|
| 681 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"clip_ratio": 0.0,
|
| 685 |
+
"completion_length": 2540.5139770507812,
|
| 686 |
"epoch": 0.07369323050556983,
|
| 687 |
+
"grad_norm": 0.1409270167350769,
|
| 688 |
+
"kl": 0.001255035400390625,
|
| 689 |
"learning_rate": 7.330314893841101e-07,
|
| 690 |
+
"loss": 0.0387,
|
| 691 |
+
"reward": 0.645833345130086,
|
| 692 |
+
"reward_std": 0.4597787447273731,
|
| 693 |
+
"rewards/accuracy_reward": 0.22222222480922937,
|
| 694 |
"rewards/difficulty_following_reward": 0.0,
|
| 695 |
+
"rewards/format_reward": 0.20138889085501432,
|
| 696 |
"step": 43,
|
| 697 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"clip_ratio": 0.0,
|
| 701 |
+
"completion_length": 2670.291748046875,
|
| 702 |
"epoch": 0.07540702656383891,
|
| 703 |
+
"grad_norm": 0.15075387060642242,
|
| 704 |
+
"kl": 0.001132965087890625,
|
| 705 |
"learning_rate": 7.185729670371604e-07,
|
| 706 |
+
"loss": 0.0947,
|
| 707 |
+
"reward": 0.2500000074505806,
|
| 708 |
+
"reward_std": 0.2931553889065981,
|
| 709 |
+
"rewards/accuracy_reward": 0.02777777798473835,
|
| 710 |
"rewards/difficulty_following_reward": 0.0,
|
| 711 |
+
"rewards/format_reward": 0.1944444477558136,
|
| 712 |
"step": 44,
|
| 713 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"clip_ratio": 0.0,
|
| 717 |
+
"completion_length": 2781.5416870117188,
|
| 718 |
"epoch": 0.07712082262210797,
|
| 719 |
+
"grad_norm": 0.13281068205833435,
|
| 720 |
+
"kl": 0.0013933181762695312,
|
| 721 |
"learning_rate": 7.039090644965509e-07,
|
| 722 |
+
"loss": 0.0486,
|
| 723 |
+
"reward": 0.3819444486871362,
|
| 724 |
+
"reward_std": 0.3621416501700878,
|
| 725 |
+
"rewards/accuracy_reward": 0.1111111156642437,
|
| 726 |
"rewards/difficulty_following_reward": 0.0,
|
| 727 |
+
"rewards/format_reward": 0.15972222480922937,
|
| 728 |
"step": 45,
|
| 729 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"clip_ratio": 0.0,
|
| 733 |
+
"completion_length": 2386.4722290039062,
|
| 734 |
"epoch": 0.07883461868037704,
|
| 735 |
+
"grad_norm": 0.125954732298851,
|
| 736 |
+
"kl": 0.00160980224609375,
|
| 737 |
"learning_rate": 6.890576474687263e-07,
|
| 738 |
+
"loss": 0.0362,
|
| 739 |
+
"reward": 0.555555559694767,
|
| 740 |
+
"reward_std": 0.49604532122612,
|
| 741 |
+
"rewards/accuracy_reward": 0.15277778264135122,
|
| 742 |
"rewards/difficulty_following_reward": 0.0,
|
| 743 |
+
"rewards/format_reward": 0.25000000186264515,
|
| 744 |
"step": 46,
|
| 745 |
+
"vanishing_advantage_ratio": 0.43055556900799274
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"clip_ratio": 0.0,
|
| 749 |
+
"completion_length": 2793.875,
|
| 750 |
"epoch": 0.0805484147386461,
|
| 751 |
+
"grad_norm": 0.15522564947605133,
|
| 752 |
+
"kl": 0.001346588134765625,
|
| 753 |
"learning_rate": 6.740368101176495e-07,
|
| 754 |
+
"loss": 0.1015,
|
| 755 |
+
"reward": 0.29166666977107525,
|
| 756 |
+
"reward_std": 0.34645168110728264,
|
| 757 |
+
"rewards/accuracy_reward": 0.0555555559694767,
|
| 758 |
"rewards/difficulty_following_reward": 0.0,
|
| 759 |
+
"rewards/format_reward": 0.18055555783212185,
|
| 760 |
"step": 47,
|
| 761 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"clip_ratio": 0.0,
|
| 765 |
+
"completion_length": 2424.4306030273438,
|
| 766 |
"epoch": 0.08226221079691516,
|
| 767 |
+
"grad_norm": 0.162226602435112,
|
| 768 |
+
"kl": 0.002109527587890625,
|
| 769 |
"learning_rate": 6.588648530198504e-07,
|
| 770 |
+
"loss": 0.0622,
|
| 771 |
+
"reward": 0.486111119389534,
|
| 772 |
+
"reward_std": 0.49431629478931427,
|
| 773 |
+
"rewards/accuracy_reward": 0.12500000186264515,
|
| 774 |
"rewards/difficulty_following_reward": 0.0,
|
| 775 |
+
"rewards/format_reward": 0.2361111119389534,
|
| 776 |
"step": 48,
|
| 777 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"clip_ratio": 0.0,
|
| 781 |
+
"completion_length": 2259.90283203125,
|
| 782 |
"epoch": 0.08397600685518423,
|
| 783 |
+
"grad_norm": 0.19547973573207855,
|
| 784 |
+
"kl": 0.00144195556640625,
|
| 785 |
"learning_rate": 6.435602608679916e-07,
|
| 786 |
+
"loss": 0.0995,
|
| 787 |
+
"reward": 0.6180555671453476,
|
| 788 |
+
"reward_std": 0.4886632487177849,
|
| 789 |
+
"rewards/accuracy_reward": 0.16666667256504297,
|
| 790 |
"rewards/difficulty_following_reward": 0.0,
|
| 791 |
+
"rewards/format_reward": 0.2847222276031971,
|
| 792 |
"step": 49,
|
| 793 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"clip_ratio": 0.0,
|
| 797 |
+
"completion_length": 2402.8333435058594,
|
| 798 |
"epoch": 0.0856898029134533,
|
| 799 |
+
"grad_norm": 0.18848426640033722,
|
| 800 |
+
"kl": 0.002063751220703125,
|
| 801 |
"learning_rate": 6.281416799501187e-07,
|
| 802 |
+
"loss": 0.0181,
|
| 803 |
+
"reward": 0.3472222201526165,
|
| 804 |
+
"reward_std": 0.407356571406126,
|
| 805 |
+
"rewards/accuracy_reward": 0.0555555559694767,
|
| 806 |
"rewards/difficulty_following_reward": 0.0,
|
| 807 |
+
"rewards/format_reward": 0.2361111156642437,
|
| 808 |
"step": 50,
|
| 809 |
+
"vanishing_advantage_ratio": 0.0833333358168602
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"clip_ratio": 0.0,
|
| 813 |
+
"completion_length": 2527.8333740234375,
|
| 814 |
"epoch": 0.08740359897172237,
|
| 815 |
+
"grad_norm": 0.16279014945030212,
|
| 816 |
+
"kl": 0.00167083740234375,
|
| 817 |
"learning_rate": 6.126278954320294e-07,
|
| 818 |
+
"loss": 0.0806,
|
| 819 |
+
"reward": 0.6458333432674408,
|
| 820 |
+
"reward_std": 0.522053524851799,
|
| 821 |
+
"rewards/accuracy_reward": 0.2083333358168602,
|
| 822 |
"rewards/difficulty_following_reward": 0.0,
|
| 823 |
+
"rewards/format_reward": 0.22916667349636555,
|
| 824 |
"step": 51,
|
| 825 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"clip_ratio": 0.0,
|
| 829 |
+
"completion_length": 2555.5555725097656,
|
| 830 |
"epoch": 0.08911739502999143,
|
| 831 |
+
"grad_norm": 0.19262254238128662,
|
| 832 |
+
"kl": 0.001934051513671875,
|
| 833 |
"learning_rate": 5.97037808470444e-07,
|
| 834 |
+
"loss": 0.1354,
|
| 835 |
+
"reward": 0.5763888992369175,
|
| 836 |
+
"reward_std": 0.6699204966425896,
|
| 837 |
+
"rewards/accuracy_reward": 0.18055556062608957,
|
| 838 |
"rewards/difficulty_following_reward": 0.0,
|
| 839 |
+
"rewards/format_reward": 0.21527778171002865,
|
| 840 |
"step": 52,
|
| 841 |
+
"vanishing_advantage_ratio": 0.1111111156642437
|
| 842 |
},
|
| 843 |
{
|
| 844 |
"clip_ratio": 0.0,
|
| 845 |
+
"completion_length": 2766.388916015625,
|
| 846 |
"epoch": 0.0908311910882605,
|
| 847 |
+
"grad_norm": 0.185680091381073,
|
| 848 |
+
"kl": 0.001705169677734375,
|
| 849 |
"learning_rate": 5.813904131848564e-07,
|
| 850 |
+
"loss": 0.1234,
|
| 851 |
+
"reward": 0.28472222574055195,
|
| 852 |
+
"reward_std": 0.2922345921397209,
|
| 853 |
+
"rewards/accuracy_reward": 0.055555556900799274,
|
| 854 |
"rewards/difficulty_following_reward": 0.0,
|
| 855 |
+
"rewards/format_reward": 0.1736111119389534,
|
| 856 |
"step": 53,
|
| 857 |
+
"vanishing_advantage_ratio": 0.2777777910232544
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"clip_ratio": 0.0,
|
| 861 |
+
"completion_length": 2823.0,
|
| 862 |
"epoch": 0.09254498714652956,
|
| 863 |
+
"grad_norm": 0.10592877119779587,
|
| 864 |
+
"kl": 0.001628875732421875,
|
| 865 |
"learning_rate": 5.657047735161255e-07,
|
| 866 |
+
"loss": 0.002,
|
| 867 |
+
"reward": 0.31249999441206455,
|
| 868 |
+
"reward_std": 0.2363711018115282,
|
| 869 |
+
"rewards/accuracy_reward": 0.08333333395421505,
|
| 870 |
"rewards/difficulty_following_reward": 0.0,
|
| 871 |
+
"rewards/format_reward": 0.14583333767950535,
|
| 872 |
"step": 54,
|
| 873 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"clip_ratio": 0.0,
|
| 877 |
+
"completion_length": 3061.5834350585938,
|
| 878 |
"epoch": 0.09425878320479864,
|
| 879 |
+
"grad_norm": 0.13768742978572845,
|
| 880 |
+
"kl": 0.00177764892578125,
|
| 881 |
"learning_rate": 5.5e-07,
|
| 882 |
+
"loss": 0.1519,
|
| 883 |
+
"reward": 0.29166667722165585,
|
| 884 |
+
"reward_std": 0.4167081881314516,
|
| 885 |
+
"rewards/accuracy_reward": 0.06944444589316845,
|
| 886 |
"rewards/difficulty_following_reward": 0.0,
|
| 887 |
+
"rewards/format_reward": 0.1527777798473835,
|
| 888 |
"step": 55,
|
| 889 |
+
"vanishing_advantage_ratio": 0.26388889737427235
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"clip_ratio": 0.0,
|
| 893 |
+
"completion_length": 2675.4584045410156,
|
| 894 |
"epoch": 0.0959725792630677,
|
| 895 |
+
"grad_norm": 0.15516793727874756,
|
| 896 |
+
"kl": 0.0018901824951171875,
|
| 897 |
"learning_rate": 5.342952264838747e-07,
|
| 898 |
+
"loss": 0.0732,
|
| 899 |
+
"reward": 0.3888888880610466,
|
| 900 |
+
"reward_std": 0.3425787817686796,
|
| 901 |
+
"rewards/accuracy_reward": 0.11111111380159855,
|
| 902 |
"rewards/difficulty_following_reward": 0.0,
|
| 903 |
+
"rewards/format_reward": 0.1666666679084301,
|
| 904 |
"step": 56,
|
| 905 |
+
"vanishing_advantage_ratio": 0.4166666716337204
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"clip_ratio": 0.0,
|
| 909 |
+
"completion_length": 2887.638916015625,
|
| 910 |
"epoch": 0.09768637532133675,
|
| 911 |
+
"grad_norm": 0.13576894998550415,
|
| 912 |
+
"kl": 0.0018768310546875,
|
| 913 |
"learning_rate": 5.186095868151436e-07,
|
| 914 |
+
"loss": 0.0397,
|
| 915 |
+
"reward": 0.5138888955116272,
|
| 916 |
+
"reward_std": 0.4866417311131954,
|
| 917 |
+
"rewards/accuracy_reward": 0.16666667349636555,
|
| 918 |
"rewards/difficulty_following_reward": 0.0,
|
| 919 |
+
"rewards/format_reward": 0.18055556155741215,
|
| 920 |
"step": 57,
|
| 921 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"clip_ratio": 0.0,
|
| 925 |
+
"completion_length": 2637.7083129882812,
|
| 926 |
"epoch": 0.09940017137960583,
|
| 927 |
+
"grad_norm": 0.17882922291755676,
|
| 928 |
+
"kl": 0.00240325927734375,
|
| 929 |
"learning_rate": 5.02962191529556e-07,
|
| 930 |
+
"loss": 0.1296,
|
| 931 |
+
"reward": 0.34722222946584225,
|
| 932 |
+
"reward_std": 0.39691098034381866,
|
| 933 |
+
"rewards/accuracy_reward": 0.06944444496184587,
|
| 934 |
"rewards/difficulty_following_reward": 0.0,
|
| 935 |
+
"rewards/format_reward": 0.20833333767950535,
|
| 936 |
"step": 58,
|
| 937 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"clip_ratio": 0.0,
|
| 941 |
+
"completion_length": 2386.9444274902344,
|
| 942 |
"epoch": 0.10111396743787489,
|
| 943 |
+
"grad_norm": 0.15917612612247467,
|
| 944 |
+
"kl": 0.0022869110107421875,
|
| 945 |
"learning_rate": 4.873721045679706e-07,
|
| 946 |
+
"loss": 0.0046,
|
| 947 |
+
"reward": 0.8402777761220932,
|
| 948 |
+
"reward_std": 0.6729260385036469,
|
| 949 |
+
"rewards/accuracy_reward": 0.2916666716337204,
|
| 950 |
"rewards/difficulty_following_reward": 0.0,
|
| 951 |
+
"rewards/format_reward": 0.2569444477558136,
|
| 952 |
"step": 59,
|
| 953 |
"vanishing_advantage_ratio": 0.1666666716337204
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"clip_ratio": 0.0,
|
| 957 |
+
"completion_length": 3012.5000610351562,
|
| 958 |
"epoch": 0.10282776349614396,
|
| 959 |
+
"grad_norm": 0.13718292117118835,
|
| 960 |
+
"kl": 0.0018711090087890625,
|
| 961 |
"learning_rate": 4.7185832004988133e-07,
|
| 962 |
+
"loss": 0.0949,
|
| 963 |
+
"reward": 0.5069444496184587,
|
| 964 |
+
"reward_std": 0.5694549642503262,
|
| 965 |
+
"rewards/accuracy_reward": 0.180555559694767,
|
| 966 |
"rewards/difficulty_following_reward": 0.0,
|
| 967 |
+
"rewards/format_reward": 0.14583333767950535,
|
| 968 |
"step": 60,
|
| 969 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"clip_ratio": 0.0,
|
| 973 |
+
"completion_length": 2862.763916015625,
|
| 974 |
"epoch": 0.10454155955441302,
|
| 975 |
+
"grad_norm": 0.15042024850845337,
|
| 976 |
+
"kl": 0.002170562744140625,
|
| 977 |
"learning_rate": 4.5643973913200837e-07,
|
| 978 |
+
"loss": 0.1462,
|
| 979 |
+
"reward": 0.29861111007630825,
|
| 980 |
+
"reward_std": 0.30184047669172287,
|
| 981 |
+
"rewards/accuracy_reward": 0.0555555559694767,
|
| 982 |
"rewards/difficulty_following_reward": 0.0,
|
| 983 |
+
"rewards/format_reward": 0.18750000558793545,
|
| 984 |
"step": 61,
|
| 985 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"clip_ratio": 0.0,
|
| 989 |
+
"completion_length": 2483.3195190429688,
|
| 990 |
"epoch": 0.1062553556126821,
|
| 991 |
+
"grad_norm": 0.1718958169221878,
|
| 992 |
+
"kl": 0.0028839111328125,
|
| 993 |
"learning_rate": 4.4113514698014953e-07,
|
| 994 |
+
"loss": 0.0138,
|
| 995 |
+
"reward": 0.4583333358168602,
|
| 996 |
+
"reward_std": 0.45289405435323715,
|
| 997 |
+
"rewards/accuracy_reward": 0.1388888917863369,
|
| 998 |
"rewards/difficulty_following_reward": 0.0,
|
| 999 |
+
"rewards/format_reward": 0.180555559694767,
|
| 1000 |
"step": 62,
|
| 1001 |
+
"vanishing_advantage_ratio": 0.26388889737427235
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"clip_ratio": 0.0,
|
| 1005 |
+
"completion_length": 3120.638916015625,
|
| 1006 |
"epoch": 0.10796915167095116,
|
| 1007 |
+
"grad_norm": 0.15143321454524994,
|
| 1008 |
+
"kl": 0.002471923828125,
|
| 1009 |
"learning_rate": 4.2596318988235037e-07,
|
| 1010 |
+
"loss": 0.0997,
|
| 1011 |
+
"reward": 0.4444444519467652,
|
| 1012 |
+
"reward_std": 0.5317340567708015,
|
| 1013 |
+
"rewards/accuracy_reward": 0.16666667349636555,
|
| 1014 |
"rewards/difficulty_following_reward": 0.0,
|
| 1015 |
+
"rewards/format_reward": 0.11111111147329211,
|
| 1016 |
"step": 63,
|
| 1017 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"clip_ratio": 0.0,
|
| 1021 |
+
"completion_length": 2806.263885498047,
|
| 1022 |
"epoch": 0.10968294772922023,
|
| 1023 |
+
"grad_norm": 0.18540050089359283,
|
| 1024 |
+
"kl": 0.00319671630859375,
|
| 1025 |
"learning_rate": 4.1094235253127374e-07,
|
| 1026 |
+
"loss": 0.0537,
|
| 1027 |
+
"reward": 0.3055555569007993,
|
| 1028 |
+
"reward_std": 0.346462732180953,
|
| 1029 |
+
"rewards/accuracy_reward": 0.055555556900799274,
|
| 1030 |
"rewards/difficulty_following_reward": 0.0,
|
| 1031 |
+
"rewards/format_reward": 0.19444445054978132,
|
| 1032 |
"step": 64,
|
| 1033 |
+
"vanishing_advantage_ratio": 0.1944444514811039
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"clip_ratio": 0.0,
|
| 1037 |
+
"completion_length": 2921.3333740234375,
|
| 1038 |
"epoch": 0.11139674378748929,
|
| 1039 |
+
"grad_norm": 0.1683017462491989,
|
| 1040 |
+
"kl": 0.00262451171875,
|
| 1041 |
"learning_rate": 3.9609093550344907e-07,
|
| 1042 |
+
"loss": 0.1098,
|
| 1043 |
+
"reward": 0.3055555638857186,
|
| 1044 |
+
"reward_std": 0.3952417355030775,
|
| 1045 |
+
"rewards/accuracy_reward": 0.06944444496184587,
|
| 1046 |
"rewards/difficulty_following_reward": 0.0,
|
| 1047 |
+
"rewards/format_reward": 0.1666666683740914,
|
| 1048 |
"step": 65,
|
| 1049 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
"clip_ratio": 0.0,
|
| 1053 |
+
"completion_length": 2583.1666870117188,
|
| 1054 |
"epoch": 0.11311053984575835,
|
| 1055 |
+
"grad_norm": 0.21984979510307312,
|
| 1056 |
+
"kl": 0.002838134765625,
|
| 1057 |
"learning_rate": 3.8142703296283953e-07,
|
| 1058 |
+
"loss": 0.1025,
|
| 1059 |
+
"reward": 0.6250000149011612,
|
| 1060 |
+
"reward_std": 0.46652648597955704,
|
| 1061 |
+
"rewards/accuracy_reward": 0.236111119389534,
|
| 1062 |
"rewards/difficulty_following_reward": 0.0,
|
| 1063 |
+
"rewards/format_reward": 0.15277778171002865,
|
| 1064 |
"step": 66,
|
| 1065 |
+
"vanishing_advantage_ratio": 0.5000000149011612
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"clip_ratio": 0.0,
|
| 1069 |
+
"completion_length": 2952.7083740234375,
|
| 1070 |
"epoch": 0.11482433590402742,
|
| 1071 |
+
"grad_norm": 0.1412336826324463,
|
| 1072 |
+
"kl": 0.003108978271484375,
|
| 1073 |
"learning_rate": 3.6696851061588994e-07,
|
| 1074 |
+
"loss": 0.0266,
|
| 1075 |
+
"reward": 0.2569444477558136,
|
| 1076 |
+
"reward_std": 0.35272519290447235,
|
| 1077 |
+
"rewards/accuracy_reward": 0.0555555559694767,
|
| 1078 |
"rewards/difficulty_following_reward": 0.0,
|
| 1079 |
+
"rewards/format_reward": 0.1458333395421505,
|
| 1080 |
"step": 67,
|
| 1081 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"clip_ratio": 0.0,
|
| 1085 |
+
"completion_length": 2923.541748046875,
|
| 1086 |
"epoch": 0.11653813196229648,
|
| 1087 |
+
"grad_norm": 0.13442060351371765,
|
| 1088 |
+
"kl": 0.003047943115234375,
|
| 1089 |
"learning_rate": 3.5273298394491515e-07,
|
| 1090 |
+
"loss": -0.002,
|
| 1091 |
+
"reward": 0.4027777723968029,
|
| 1092 |
+
"reward_std": 0.4289739280939102,
|
| 1093 |
+
"rewards/accuracy_reward": 0.12500000279396772,
|
| 1094 |
"rewards/difficulty_following_reward": 0.0,
|
| 1095 |
+
"rewards/format_reward": 0.15277778171002865,
|
| 1096 |
"step": 68,
|
| 1097 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"clip_ratio": 0.0,
|
| 1101 |
+
"completion_length": 2733.9444580078125,
|
| 1102 |
"epoch": 0.11825192802056556,
|
| 1103 |
+
"grad_norm": 0.14998486638069153,
|
| 1104 |
+
"kl": 0.0030517578125,
|
| 1105 |
"learning_rate": 3.387377967463493e-07,
|
| 1106 |
+
"loss": 0.0939,
|
| 1107 |
+
"reward": 0.5763889029622078,
|
| 1108 |
+
"reward_std": 0.5283073335886002,
|
| 1109 |
+
"rewards/accuracy_reward": 0.18055555876344442,
|
| 1110 |
"rewards/difficulty_following_reward": 0.0,
|
| 1111 |
+
"rewards/format_reward": 0.21527778171002865,
|
| 1112 |
"step": 69,
|
| 1113 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"clip_ratio": 0.0,
|
| 1117 |
+
"completion_length": 2523.2361450195312,
|
| 1118 |
"epoch": 0.11996572407883462,
|
| 1119 |
+
"grad_norm": 0.1801416426897049,
|
| 1120 |
+
"kl": 0.0034027099609375,
|
| 1121 |
"learning_rate": 3.250000000000001e-07,
|
| 1122 |
+
"loss": 0.0797,
|
| 1123 |
+
"reward": 0.36111112125217915,
|
| 1124 |
+
"reward_std": 0.3477918654680252,
|
| 1125 |
+
"rewards/accuracy_reward": 0.06944444496184587,
|
| 1126 |
"rewards/difficulty_following_reward": 0.0,
|
| 1127 |
+
"rewards/format_reward": 0.22222222574055195,
|
| 1128 |
"step": 70,
|
| 1129 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"clip_ratio": 0.0,
|
| 1133 |
+
"completion_length": 2869.7361450195312,
|
| 1134 |
"epoch": 0.12167952013710369,
|
| 1135 |
+
"grad_norm": 0.13626869022846222,
|
| 1136 |
+
"kl": 0.002971649169921875,
|
| 1137 |
"learning_rate": 3.115363310950578e-07,
|
| 1138 |
+
"loss": -0.0103,
|
| 1139 |
+
"reward": 0.618055559694767,
|
| 1140 |
+
"reward_std": 0.49544404074549675,
|
| 1141 |
+
"rewards/accuracy_reward": 0.20833333767950535,
|
| 1142 |
"rewards/difficulty_following_reward": 0.0,
|
| 1143 |
+
"rewards/format_reward": 0.2013888917863369,
|
| 1144 |
"step": 71,
|
| 1145 |
+
"vanishing_advantage_ratio": 0.34722223319113255
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"clip_ratio": 0.0,
|
| 1149 |
+
"completion_length": 2333.15283203125,
|
| 1150 |
"epoch": 0.12339331619537275,
|
| 1151 |
+
"grad_norm": 0.22007954120635986,
|
| 1152 |
+
"kl": 0.003650665283203125,
|
| 1153 |
"learning_rate": 2.9836319343816397e-07,
|
| 1154 |
+
"loss": 0.0953,
|
| 1155 |
+
"reward": 0.5486111231148243,
|
| 1156 |
+
"reward_std": 0.38520985655486584,
|
| 1157 |
+
"rewards/accuracy_reward": 0.16666666883975267,
|
| 1158 |
"rewards/difficulty_following_reward": 0.0,
|
| 1159 |
+
"rewards/format_reward": 0.2152777798473835,
|
| 1160 |
"step": 72,
|
| 1161 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"clip_ratio": 0.0,
|
| 1165 |
+
"completion_length": 2609.166748046875,
|
| 1166 |
"epoch": 0.12510711225364182,
|
| 1167 |
+
"grad_norm": 0.14889562129974365,
|
| 1168 |
+
"kl": 0.00350189208984375,
|
| 1169 |
"learning_rate": 2.854966364683872e-07,
|
| 1170 |
+
"loss": 0.0133,
|
| 1171 |
+
"reward": 0.5972222238779068,
|
| 1172 |
+
"reward_std": 0.49283862113952637,
|
| 1173 |
+
"rewards/accuracy_reward": 0.18055556248873472,
|
| 1174 |
"rewards/difficulty_following_reward": 0.0,
|
| 1175 |
+
"rewards/format_reward": 0.23611111287027597,
|
| 1176 |
"step": 73,
|
| 1177 |
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"clip_ratio": 0.0,
|
| 1181 |
+
"completion_length": 1885.9722290039062,
|
| 1182 |
"epoch": 0.1268209083119109,
|
| 1183 |
+
"grad_norm": 0.18624606728553772,
|
| 1184 |
+
"kl": 0.003803253173828125,
|
| 1185 |
"learning_rate": 2.729523361034538e-07,
|
| 1186 |
+
"loss": 0.0368,
|
| 1187 |
+
"reward": 0.4791666641831398,
|
| 1188 |
+
"reward_std": 0.41904643177986145,
|
| 1189 |
+
"rewards/accuracy_reward": 0.11111111473292112,
|
| 1190 |
"rewards/difficulty_following_reward": 0.0,
|
| 1191 |
+
"rewards/format_reward": 0.2569444514811039,
|
| 1192 |
"step": 74,
|
| 1193 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"clip_ratio": 0.0,
|
| 1197 |
+
"completion_length": 3035.8333740234375,
|
| 1198 |
"epoch": 0.12853470437017994,
|
| 1199 |
+
"grad_norm": 0.13360603153705597,
|
| 1200 |
+
"kl": 0.003421783447265625,
|
| 1201 |
"learning_rate": 2.6074557564105724e-07,
|
| 1202 |
+
"loss": 0.081,
|
| 1203 |
+
"reward": 0.4583333432674408,
|
| 1204 |
+
"reward_std": 0.5805492922663689,
|
| 1205 |
+
"rewards/accuracy_reward": 0.13888889085501432,
|
| 1206 |
"rewards/difficulty_following_reward": 0.0,
|
| 1207 |
+
"rewards/format_reward": 0.180555559694767,
|
| 1208 |
"step": 75,
|
| 1209 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"clip_ratio": 0.0,
|
| 1213 |
+
"completion_length": 2006.5000305175781,
|
| 1214 |
"epoch": 0.13024850042844902,
|
| 1215 |
+
"grad_norm": 0.186117023229599,
|
| 1216 |
+
"kl": 0.003589630126953125,
|
| 1217 |
"learning_rate": 2.488912271385139e-07,
|
| 1218 |
+
"loss": 0.0093,
|
| 1219 |
+
"reward": 0.5555555522441864,
|
| 1220 |
+
"reward_std": 0.5216063931584358,
|
| 1221 |
+
"rewards/accuracy_reward": 0.13888888992369175,
|
| 1222 |
"rewards/difficulty_following_reward": 0.0,
|
| 1223 |
+
"rewards/format_reward": 0.2777777835726738,
|
| 1224 |
"step": 76,
|
| 1225 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"clip_ratio": 0.0,
|
| 1229 |
+
"completion_length": 2852.1111450195312,
|
| 1230 |
"epoch": 0.1319622964867181,
|
| 1231 |
+
"grad_norm": 0.14473779499530792,
|
| 1232 |
+
"kl": 0.003143310546875,
|
| 1233 |
"learning_rate": 2.374037332934512e-07,
|
| 1234 |
+
"loss": 0.094,
|
| 1235 |
+
"reward": 0.4930555522441864,
|
| 1236 |
+
"reward_std": 0.5302782356739044,
|
| 1237 |
+
"rewards/accuracy_reward": 0.16666667070239782,
|
| 1238 |
"rewards/difficulty_following_reward": 0.0,
|
| 1239 |
+
"rewards/format_reward": 0.15972222900018096,
|
| 1240 |
"step": 77,
|
| 1241 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"clip_ratio": 0.0,
|
| 1245 |
+
"completion_length": 3228.5277709960938,
|
| 1246 |
"epoch": 0.13367609254498714,
|
| 1247 |
+
"grad_norm": 0.12343769520521164,
|
| 1248 |
+
"kl": 0.003009796142578125,
|
| 1249 |
"learning_rate": 2.2629708984760706e-07,
|
| 1250 |
+
"loss": -0.016,
|
| 1251 |
+
"reward": 0.2638888843357563,
|
| 1252 |
+
"reward_std": 0.3488573133945465,
|
| 1253 |
+
"rewards/accuracy_reward": 0.08333333488553762,
|
| 1254 |
"rewards/difficulty_following_reward": 0.0,
|
| 1255 |
+
"rewards/format_reward": 0.09722222248092294,
|
| 1256 |
"step": 78,
|
| 1257 |
+
"vanishing_advantage_ratio": 0.5000000074505806
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"clip_ratio": 0.0,
|
| 1261 |
+
"completion_length": 2360.986083984375,
|
| 1262 |
"epoch": 0.1353898886032562,
|
| 1263 |
+
"grad_norm": 0.18473461270332336,
|
| 1264 |
+
"kl": 0.0040740966796875,
|
| 1265 |
"learning_rate": 2.1558482853517253e-07,
|
| 1266 |
+
"loss": 0.0521,
|
| 1267 |
+
"reward": 0.4027777761220932,
|
| 1268 |
+
"reward_std": 0.3384508527815342,
|
| 1269 |
+
"rewards/accuracy_reward": 0.06944444496184587,
|
| 1270 |
"rewards/difficulty_following_reward": 0.0,
|
| 1271 |
+
"rewards/format_reward": 0.2638888955116272,
|
| 1272 |
"step": 79,
|
| 1273 |
+
"vanishing_advantage_ratio": 0.1805555671453476
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"clip_ratio": 0.0,
|
| 1277 |
+
"completion_length": 3112.1111450195312,
|
| 1278 |
"epoch": 0.13710368466152528,
|
| 1279 |
+
"grad_norm": 0.12150842696428299,
|
| 1280 |
+
"kl": 0.003692626953125,
|
| 1281 |
"learning_rate": 2.0528000059645995e-07,
|
| 1282 |
+
"loss": 0.072,
|
| 1283 |
+
"reward": 0.3125000009313226,
|
| 1284 |
+
"reward_std": 0.3745912276208401,
|
| 1285 |
+
"rewards/accuracy_reward": 0.08333333395421505,
|
| 1286 |
"rewards/difficulty_following_reward": 0.0,
|
| 1287 |
+
"rewards/format_reward": 0.14583333674818277,
|
| 1288 |
"step": 80,
|
| 1289 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
"clip_ratio": 0.0,
|
| 1293 |
+
"completion_length": 2774.513916015625,
|
| 1294 |
"epoch": 0.13881748071979436,
|
| 1295 |
+
"grad_norm": 0.1500415951013565,
|
| 1296 |
+
"kl": 0.003559112548828125,
|
| 1297 |
"learning_rate": 1.9539516087697517e-07,
|
| 1298 |
+
"loss": 0.0482,
|
| 1299 |
+
"reward": 0.41666668001562357,
|
| 1300 |
+
"reward_std": 0.4251419398933649,
|
| 1301 |
+
"rewards/accuracy_reward": 0.1388888955116272,
|
| 1302 |
"rewards/difficulty_following_reward": 0.0,
|
| 1303 |
+
"rewards/format_reward": 0.13888889271765947,
|
| 1304 |
"step": 81,
|
| 1305 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"clip_ratio": 0.0,
|
| 1309 |
+
"completion_length": 2676.375,
|
| 1310 |
"epoch": 0.1405312767780634,
|
| 1311 |
+
"grad_norm": 0.1790565550327301,
|
| 1312 |
+
"kl": 0.003726959228515625,
|
| 1313 |
"learning_rate": 1.8594235253127372e-07,
|
| 1314 |
+
"loss": 0.0556,
|
| 1315 |
+
"reward": 0.506944440305233,
|
| 1316 |
+
"reward_std": 0.4319311436265707,
|
| 1317 |
+
"rewards/accuracy_reward": 0.13888889271765947,
|
| 1318 |
"rewards/difficulty_following_reward": 0.0,
|
| 1319 |
+
"rewards/format_reward": 0.2291666716337204,
|
| 1320 |
"step": 82,
|
| 1321 |
+
"vanishing_advantage_ratio": 0.2777777872979641
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"clip_ratio": 0.0,
|
| 1325 |
+
"completion_length": 2491.4861755371094,
|
| 1326 |
"epoch": 0.14224507283633248,
|
| 1327 |
+
"grad_norm": 0.16082800924777985,
|
| 1328 |
+
"kl": 0.0035247802734375,
|
| 1329 |
"learning_rate": 1.7693309235023127e-07,
|
| 1330 |
+
"loss": 0.0628,
|
| 1331 |
+
"reward": 0.4444444477558136,
|
| 1332 |
+
"reward_std": 0.41663530841469765,
|
| 1333 |
+
"rewards/accuracy_reward": 0.11111111287027597,
|
| 1334 |
"rewards/difficulty_following_reward": 0.0,
|
| 1335 |
+
"rewards/format_reward": 0.2222222276031971,
|
| 1336 |
"step": 83,
|
| 1337 |
+
"vanishing_advantage_ratio": 0.34722223319113255
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"clip_ratio": 0.0,
|
| 1341 |
+
"completion_length": 2634.1805725097656,
|
| 1342 |
"epoch": 0.14395886889460155,
|
| 1343 |
+
"grad_norm": 0.15972262620925903,
|
| 1344 |
+
"kl": 0.003818511962890625,
|
| 1345 |
"learning_rate": 1.6837835672960831e-07,
|
| 1346 |
+
"loss": 0.0673,
|
| 1347 |
+
"reward": 0.4375000074505806,
|
| 1348 |
+
"reward_std": 0.48190969228744507,
|
| 1349 |
+
"rewards/accuracy_reward": 0.09722222294658422,
|
| 1350 |
"rewards/difficulty_following_reward": 0.0,
|
| 1351 |
+
"rewards/format_reward": 0.24305556062608957,
|
| 1352 |
"step": 84,
|
| 1353 |
+
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"clip_ratio": 0.0,
|
| 1357 |
+
"completion_length": 2900.65283203125,
|
| 1358 |
"epoch": 0.1456726649528706,
|
| 1359 |
+
"grad_norm": 0.15328432619571686,
|
| 1360 |
+
"kl": 0.003406524658203125,
|
| 1361 |
"learning_rate": 1.6028856829700258e-07,
|
| 1362 |
+
"loss": 0.0828,
|
| 1363 |
+
"reward": 0.4861111044883728,
|
| 1364 |
+
"reward_std": 0.4899628609418869,
|
| 1365 |
+
"rewards/accuracy_reward": 0.1666666679084301,
|
| 1366 |
"rewards/difficulty_following_reward": 0.0,
|
| 1367 |
+
"rewards/format_reward": 0.15277778171002865,
|
| 1368 |
"step": 85,
|
| 1369 |
"vanishing_advantage_ratio": 0.2500000074505806
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"clip_ratio": 0.0,
|
| 1373 |
+
"completion_length": 2665.902801513672,
|
| 1374 |
"epoch": 0.14738646101113967,
|
| 1375 |
+
"grad_norm": 0.16509687900543213,
|
| 1376 |
+
"kl": 0.0034637451171875,
|
| 1377 |
"learning_rate": 1.5267358321348285e-07,
|
| 1378 |
+
"loss": 0.0968,
|
| 1379 |
+
"reward": 0.24999999813735485,
|
| 1380 |
+
"reward_std": 0.30046750232577324,
|
| 1381 |
+
"rewards/accuracy_reward": 0.02777777798473835,
|
| 1382 |
"rewards/difficulty_following_reward": 0.0,
|
| 1383 |
+
"rewards/format_reward": 0.19444444589316845,
|
| 1384 |
"step": 86,
|
| 1385 |
+
"vanishing_advantage_ratio": 0.26388889737427235
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"clip_ratio": 0.0,
|
| 1389 |
+
"completion_length": 3268.4306030273438,
|
| 1390 |
"epoch": 0.14910025706940874,
|
| 1391 |
+
"grad_norm": 0.10825953632593155,
|
| 1392 |
+
"kl": 0.003009796142578125,
|
| 1393 |
"learning_rate": 1.4554267916537495e-07,
|
| 1394 |
+
"loss": 0.0666,
|
| 1395 |
+
"reward": 0.4027777696028352,
|
| 1396 |
+
"reward_std": 0.3645976707339287,
|
| 1397 |
+
"rewards/accuracy_reward": 0.15277778171002865,
|
| 1398 |
"rewards/difficulty_following_reward": 0.0,
|
| 1399 |
+
"rewards/format_reward": 0.09722222574055195,
|
| 1400 |
"step": 87,
|
| 1401 |
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1402 |
},
|
| 1403 |
{
|
| 1404 |
"clip_ratio": 0.0,
|
| 1405 |
+
"completion_length": 3164.0277709960938,
|
| 1406 |
"epoch": 0.15081405312767782,
|
| 1407 |
+
"grad_norm": 0.1394396275281906,
|
| 1408 |
+
"kl": 0.003810882568359375,
|
| 1409 |
"learning_rate": 1.3890454406082956e-07,
|
| 1410 |
+
"loss": 0.0838,
|
| 1411 |
+
"reward": 0.31250000558793545,
|
| 1412 |
+
"reward_std": 0.385826725512743,
|
| 1413 |
+
"rewards/accuracy_reward": 0.09722222574055195,
|
| 1414 |
"rewards/difficulty_following_reward": 0.0,
|
| 1415 |
+
"rewards/format_reward": 0.11805555876344442,
|
| 1416 |
"step": 88,
|
| 1417 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"clip_ratio": 0.0,
|
| 1421 |
+
"completion_length": 2483.1527709960938,
|
| 1422 |
"epoch": 0.15252784918594686,
|
| 1423 |
+
"grad_norm": 0.15645141899585724,
|
| 1424 |
+
"kl": 0.00453948974609375,
|
| 1425 |
"learning_rate": 1.3276726544494571e-07,
|
| 1426 |
+
"loss": 0.0977,
|
| 1427 |
+
"reward": 0.34722223225980997,
|
| 1428 |
+
"reward_std": 0.3350673224776983,
|
| 1429 |
+
"rewards/accuracy_reward": 0.055555556900799274,
|
| 1430 |
"rewards/difficulty_following_reward": 0.0,
|
| 1431 |
+
"rewards/format_reward": 0.23611111659556627,
|
| 1432 |
"step": 89,
|
| 1433 |
+
"vanishing_advantage_ratio": 0.4166666716337204
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
"clip_ratio": 0.0,
|
| 1437 |
+
"completion_length": 2722.499969482422,
|
| 1438 |
"epoch": 0.15424164524421594,
|
| 1439 |
+
"grad_norm": 0.150289386510849,
|
| 1440 |
+
"kl": 0.003353118896484375,
|
| 1441 |
"learning_rate": 1.2713832064634125e-07,
|
| 1442 |
+
"loss": 0.0805,
|
| 1443 |
+
"reward": 0.27777778171002865,
|
| 1444 |
+
"reward_std": 0.3167807497084141,
|
| 1445 |
"rewards/accuracy_reward": 0.055555556900799274,
|
| 1446 |
"rewards/difficulty_following_reward": 0.0,
|
| 1447 |
+
"rewards/format_reward": 0.16666666697710752,
|
| 1448 |
"step": 90,
|
| 1449 |
+
"vanishing_advantage_ratio": 0.5138888955116272
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"clip_ratio": 0.0,
|
| 1453 |
+
"completion_length": 2716.0833129882812,
|
| 1454 |
"epoch": 0.155955441302485,
|
| 1455 |
+
"grad_norm": 0.17992275953292847,
|
| 1456 |
+
"kl": 0.004726409912109375,
|
| 1457 |
"learning_rate": 1.220245676671809e-07,
|
| 1458 |
+
"loss": 0.0844,
|
| 1459 |
+
"reward": 0.604166679084301,
|
| 1460 |
+
"reward_std": 0.6643020883202553,
|
| 1461 |
+
"rewards/accuracy_reward": 0.18055555690079927,
|
| 1462 |
"rewards/difficulty_following_reward": 0.0,
|
| 1463 |
+
"rewards/format_reward": 0.2430555559694767,
|
| 1464 |
"step": 91,
|
| 1465 |
+
"vanishing_advantage_ratio": 0.0833333358168602
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"clip_ratio": 0.0,
|
| 1469 |
+
"completion_length": 2535.277801513672,
|
| 1470 |
"epoch": 0.15766923736075408,
|
| 1471 |
+
"grad_norm": 0.15826836228370667,
|
| 1472 |
+
"kl": 0.00363922119140625,
|
| 1473 |
"learning_rate": 1.1743223682775649e-07,
|
| 1474 |
+
"loss": 0.087,
|
| 1475 |
+
"reward": 0.3472222238779068,
|
| 1476 |
+
"reward_std": 0.333256833255291,
|
| 1477 |
+
"rewards/accuracy_reward": 0.08333333395421505,
|
| 1478 |
"rewards/difficulty_following_reward": 0.0,
|
| 1479 |
+
"rewards/format_reward": 0.180555559694767,
|
| 1480 |
"step": 92,
|
| 1481 |
+
"vanishing_advantage_ratio": 0.4166666716337204
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"clip_ratio": 0.0,
|
| 1485 |
+
"completion_length": 2722.6806030273438,
|
| 1486 |
"epoch": 0.15938303341902313,
|
| 1487 |
+
"grad_norm": 0.15519429743289948,
|
| 1488 |
+
"kl": 0.003940582275390625,
|
| 1489 |
"learning_rate": 1.1336692317580158e-07,
|
| 1490 |
+
"loss": 0.0829,
|
| 1491 |
+
"reward": 0.5763888955116272,
|
| 1492 |
+
"reward_std": 0.48380210250616074,
|
| 1493 |
+
"rewards/accuracy_reward": 0.20833334047347307,
|
| 1494 |
"rewards/difficulty_following_reward": 0.0,
|
| 1495 |
+
"rewards/format_reward": 0.1597222243435681,
|
| 1496 |
"step": 93,
|
| 1497 |
+
"vanishing_advantage_ratio": 0.1666666716337204
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"clip_ratio": 0.0,
|
| 1501 |
+
"completion_length": 2634.277801513672,
|
| 1502 |
"epoch": 0.1610968294772922,
|
| 1503 |
+
"grad_norm": 0.1622547060251236,
|
| 1504 |
+
"kl": 0.004276275634765625,
|
| 1505 |
"learning_rate": 1.0983357966978745e-07,
|
| 1506 |
+
"loss": 0.0776,
|
| 1507 |
+
"reward": 0.3888888955116272,
|
| 1508 |
+
"reward_std": 0.31960033625364304,
|
| 1509 |
+
"rewards/accuracy_reward": 0.1111111119389534,
|
| 1510 |
"rewards/difficulty_following_reward": 0.0,
|
| 1511 |
+
"rewards/format_reward": 0.1666666679084301,
|
| 1512 |
"step": 94,
|
| 1513 |
"vanishing_advantage_ratio": 0.4166666716337204
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
"clip_ratio": 0.0,
|
| 1517 |
+
"completion_length": 2685.6666564941406,
|
| 1518 |
"epoch": 0.16281062553556128,
|
| 1519 |
+
"grad_norm": 0.19985106587409973,
|
| 1520 |
+
"kl": 0.00397491455078125,
|
| 1521 |
"learning_rate": 1.068365111445064e-07,
|
| 1522 |
+
"loss": 0.0964,
|
| 1523 |
+
"reward": 0.4166666641831398,
|
| 1524 |
+
"reward_std": 0.5311803258955479,
|
| 1525 |
+
"rewards/accuracy_reward": 0.11111111473292112,
|
| 1526 |
"rewards/difficulty_following_reward": 0.0,
|
| 1527 |
+
"rewards/format_reward": 0.19444444589316845,
|
| 1528 |
"step": 95,
|
| 1529 |
"vanishing_advantage_ratio": 0.0833333358168602
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"clip_ratio": 0.0,
|
| 1533 |
+
"completion_length": 2594.0972290039062,
|
| 1534 |
"epoch": 0.16452442159383032,
|
| 1535 |
+
"grad_norm": 0.16063077747821808,
|
| 1536 |
+
"kl": 0.0034027099609375,
|
| 1537 |
"learning_rate": 1.0437936906629334e-07,
|
| 1538 |
+
"loss": 0.0895,
|
| 1539 |
+
"reward": 0.5902777835726738,
|
| 1540 |
+
"reward_std": 0.5667251460254192,
|
| 1541 |
+
"rewards/accuracy_reward": 0.2083333320915699,
|
| 1542 |
"rewards/difficulty_following_reward": 0.0,
|
| 1543 |
+
"rewards/format_reward": 0.17361111380159855,
|
| 1544 |
"step": 96,
|
| 1545 |
+
"vanishing_advantage_ratio": 0.2500000111758709
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"clip_ratio": 0.0,
|
| 1549 |
+
"completion_length": 3230.77783203125,
|
| 1550 |
"epoch": 0.1662382176520994,
|
| 1551 |
+
"grad_norm": 0.19083616137504578,
|
| 1552 |
+
"kl": 0.00445556640625,
|
| 1553 |
"learning_rate": 1.0246514708427701e-07,
|
| 1554 |
+
"loss": 0.0987,
|
| 1555 |
+
"reward": 0.29861112125217915,
|
| 1556 |
+
"reward_std": 0.4562392868101597,
|
| 1557 |
+
"rewards/accuracy_reward": 0.08333333488553762,
|
| 1558 |
"rewards/difficulty_following_reward": 0.0,
|
| 1559 |
+
"rewards/format_reward": 0.13194444589316845,
|
| 1560 |
"step": 97,
|
| 1561 |
+
"vanishing_advantage_ratio": 0.0
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"clip_ratio": 0.0,
|
| 1565 |
+
"completion_length": 2813.527801513672,
|
| 1566 |
"epoch": 0.16795201371036847,
|
| 1567 |
+
"grad_norm": 0.1406376212835312,
|
| 1568 |
+
"kl": 0.00418853759765625,
|
| 1569 |
"learning_rate": 1.0109617738307911e-07,
|
| 1570 |
+
"loss": 0.0603,
|
| 1571 |
+
"reward": 0.4027777872979641,
|
| 1572 |
+
"reward_std": 0.40550027787685394,
|
| 1573 |
+
"rewards/accuracy_reward": 0.12500000186264515,
|
| 1574 |
"rewards/difficulty_following_reward": 0.0,
|
| 1575 |
+
"rewards/format_reward": 0.15277777891606092,
|
| 1576 |
"step": 98,
|
| 1577 |
+
"vanishing_advantage_ratio": 0.3333333432674408
|
| 1578 |
},
|
| 1579 |
{
|
| 1580 |
"clip_ratio": 0.0,
|
| 1581 |
+
"completion_length": 2933.4166870117188,
|
| 1582 |
"epoch": 0.16966580976863754,
|
| 1583 |
+
"grad_norm": 0.1805686354637146,
|
| 1584 |
+
"kl": 0.0044708251953125,
|
| 1585 |
"learning_rate": 1.002741278414069e-07,
|
| 1586 |
+
"loss": 0.1209,
|
| 1587 |
+
"reward": 0.4236111082136631,
|
| 1588 |
+
"reward_std": 0.5723845213651657,
|
| 1589 |
+
"rewards/accuracy_reward": 0.1250000037252903,
|
| 1590 |
"rewards/difficulty_following_reward": 0.0,
|
| 1591 |
+
"rewards/format_reward": 0.1736111156642437,
|
| 1592 |
"step": 99,
|
| 1593 |
+
"vanishing_advantage_ratio": 0.0833333358168602
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"clip_ratio": 0.0,
|
| 1597 |
+
"completion_length": 2795.6805419921875,
|
| 1598 |
"epoch": 0.1713796058269066,
|
| 1599 |
+
"grad_norm": 0.16270823776721954,
|
| 1600 |
+
"kl": 0.004749298095703125,
|
| 1601 |
"learning_rate": 1e-07,
|
| 1602 |
+
"loss": 0.1155,
|
| 1603 |
+
"reward": 0.24999999720603228,
|
| 1604 |
+
"reward_std": 0.32094707898795605,
|
| 1605 |
+
"rewards/accuracy_reward": 0.02777777798473835,
|
| 1606 |
"rewards/difficulty_following_reward": 0.0,
|
| 1607 |
+
"rewards/format_reward": 0.19444444309920073,
|
| 1608 |
"step": 100,
|
| 1609 |
+
"vanishing_advantage_ratio": 0.416666679084301
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
"epoch": 0.1713796058269066,
|
| 1613 |
"step": 100,
|
| 1614 |
"total_flos": 0.0,
|
| 1615 |
+
"train_loss": 0.0719883194193244,
|
| 1616 |
+
"train_runtime": 18385.1788,
|
| 1617 |
+
"train_samples_per_second": 0.392,
|
| 1618 |
+
"train_steps_per_second": 0.005
|
| 1619 |
}
|
| 1620 |
],
|
| 1621 |
"logging_steps": 1,
|