Model save
Browse files- README.md +3 -5
- all_results.json +4 -4
- generation_config.json +5 -10
- train_results.json +4 -4
- trainer_state.json +444 -1144
README.md
CHANGED
|
@@ -1,11 +1,9 @@
|
|
| 1 |
---
|
| 2 |
-
base_model:
|
| 3 |
-
datasets: chansung/verifiable-coding-problems-python
|
| 4 |
library_name: transformers
|
| 5 |
model_name: Qwen2.5-1.5B-CCRL-1
|
| 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 Qwen2.5-1.5B-CCRL-1
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [
|
| 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/chansung18/huggingface/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: coolcui/DeepSeek-R1-Distill-Qwen-1.5B-GRPO
|
|
|
|
| 3 |
library_name: transformers
|
| 4 |
model_name: Qwen2.5-1.5B-CCRL-1
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
| 7 |
- trl
|
| 8 |
- grpo
|
| 9 |
licence: license
|
|
|
|
| 11 |
|
| 12 |
# Model Card for Qwen2.5-1.5B-CCRL-1
|
| 13 |
|
| 14 |
+
This model is a fine-tuned version of [coolcui/DeepSeek-R1-Distill-Qwen-1.5B-GRPO](https://huggingface.co/coolcui/DeepSeek-R1-Distill-Qwen-1.5B-GRPO).
|
| 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/chansung18/huggingface/runs/opz13onj)
|
| 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": 949,
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.0,
|
| 4 |
+
"train_runtime": 13359.4688,
|
| 5 |
"train_samples": 949,
|
| 6 |
+
"train_samples_per_second": 0.479,
|
| 7 |
+
"train_steps_per_second": 0.004
|
| 8 |
}
|
generation_config.json
CHANGED
|
@@ -1,14 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
|
|
|
| 3 |
"do_sample": true,
|
| 4 |
-
"eos_token_id":
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
],
|
| 8 |
-
"pad_token_id": 151643,
|
| 9 |
-
"repetition_penalty": 1.1,
|
| 10 |
-
"temperature": 0.7,
|
| 11 |
-
"top_k": 20,
|
| 12 |
-
"top_p": 0.8,
|
| 13 |
"transformers_version": "4.50.0"
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151646,
|
| 4 |
"do_sample": true,
|
| 5 |
+
"eos_token_id": 151643,
|
| 6 |
+
"temperature": 0.6,
|
| 7 |
+
"top_p": 0.95,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
"transformers_version": "4.50.0"
|
| 9 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 949,
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.0,
|
| 4 |
+
"train_runtime": 13359.4688,
|
| 5 |
"train_samples": 949,
|
| 6 |
+
"train_samples_per_second": 0.479,
|
| 7 |
+
"train_steps_per_second": 0.004
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -2,1427 +2,727 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"clip_ratio": 0.0,
|
| 14 |
-
"completion_length":
|
| 15 |
"epoch": 0.01680672268907563,
|
| 16 |
-
"grad_norm": 0.
|
| 17 |
"kl": 0.0,
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss":
|
| 20 |
-
"num_tokens":
|
| 21 |
-
"reward": 0.
|
| 22 |
-
"reward_std": 0.
|
| 23 |
-
"rewards/
|
| 24 |
"step": 1
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"clip_ratio": 0.0,
|
| 28 |
-
"completion_length":
|
| 29 |
"epoch": 0.03361344537815126,
|
| 30 |
-
"grad_norm": 0.
|
| 31 |
"kl": 0.0,
|
| 32 |
-
"learning_rate":
|
| 33 |
-
"loss": 0.
|
| 34 |
-
"num_tokens":
|
| 35 |
-
"reward": 0.
|
| 36 |
-
"reward_std": 0.
|
| 37 |
-
"rewards/
|
| 38 |
"step": 2
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"clip_ratio": 0.0,
|
| 42 |
-
"completion_length":
|
| 43 |
"epoch": 0.05042016806722689,
|
| 44 |
-
"grad_norm": 0.
|
| 45 |
-
"kl": 0.
|
| 46 |
-
"learning_rate":
|
| 47 |
-
"loss": 0.
|
| 48 |
-
"num_tokens":
|
| 49 |
-
"reward": 0.
|
| 50 |
-
"reward_std": 0.
|
| 51 |
-
"rewards/
|
| 52 |
"step": 3
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"clip_ratio": 0.0,
|
| 56 |
-
"completion_length":
|
| 57 |
"epoch": 0.06722689075630252,
|
| 58 |
-
"grad_norm": 0.
|
| 59 |
-
"kl": 0.
|
| 60 |
-
"learning_rate": 9.
|
| 61 |
"loss": 0.0,
|
| 62 |
-
"num_tokens":
|
| 63 |
"reward": 0.0,
|
| 64 |
"reward_std": 0.0,
|
| 65 |
-
"rewards/
|
| 66 |
"step": 4
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"clip_ratio": 0.0,
|
| 70 |
-
"completion_length":
|
| 71 |
"epoch": 0.08403361344537816,
|
| 72 |
-
"grad_norm": 0.
|
| 73 |
-
"kl": 0.
|
| 74 |
-
"learning_rate": 9.
|
| 75 |
-
"loss": 0.
|
| 76 |
-
"num_tokens":
|
| 77 |
-
"reward": 0.
|
| 78 |
-
"reward_std": 0.
|
| 79 |
-
"rewards/
|
| 80 |
"step": 5
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"clip_ratio": 0.0,
|
| 84 |
-
"completion_length":
|
| 85 |
"epoch": 0.10084033613445378,
|
| 86 |
-
"grad_norm": 0.
|
| 87 |
-
"kl": 0.
|
| 88 |
-
"learning_rate": 9.
|
| 89 |
"loss": 0.0,
|
| 90 |
-
"num_tokens":
|
| 91 |
"reward": 0.0,
|
| 92 |
"reward_std": 0.0,
|
| 93 |
-
"rewards/
|
| 94 |
"step": 6
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"clip_ratio": 0.0,
|
| 98 |
-
"completion_length":
|
| 99 |
"epoch": 0.11764705882352941,
|
| 100 |
-
"grad_norm": 0.
|
| 101 |
-
"kl": 0.
|
| 102 |
-
"learning_rate": 9.
|
| 103 |
-
"loss": 0.
|
| 104 |
-
"num_tokens":
|
| 105 |
-
"reward": 0.
|
| 106 |
-
"reward_std": 0.
|
| 107 |
-
"rewards/
|
| 108 |
"step": 7
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"clip_ratio": 0.0,
|
| 112 |
-
"completion_length":
|
| 113 |
"epoch": 0.13445378151260504,
|
| 114 |
-
"grad_norm": 0.
|
| 115 |
-
"kl": 0.
|
| 116 |
-
"learning_rate": 9.
|
| 117 |
"loss": 0.0,
|
| 118 |
-
"num_tokens":
|
| 119 |
"reward": 0.0,
|
| 120 |
"reward_std": 0.0,
|
| 121 |
-
"rewards/
|
| 122 |
"step": 8
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"clip_ratio": 0.0,
|
| 126 |
-
"completion_length":
|
| 127 |
"epoch": 0.15126050420168066,
|
| 128 |
-
"grad_norm": 0.
|
| 129 |
-
"kl": 0.
|
| 130 |
-
"learning_rate": 9.
|
| 131 |
-
"loss": 0.
|
| 132 |
-
"num_tokens":
|
| 133 |
-
"reward": 0.
|
| 134 |
-
"reward_std": 0.
|
| 135 |
-
"rewards/
|
| 136 |
"step": 9
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"clip_ratio": 0.0,
|
| 140 |
-
"completion_length":
|
| 141 |
"epoch": 0.16806722689075632,
|
| 142 |
-
"grad_norm": 0.
|
| 143 |
-
"kl": 0.
|
| 144 |
-
"learning_rate": 9.
|
| 145 |
-
"loss":
|
| 146 |
-
"num_tokens":
|
| 147 |
-
"reward": 0.
|
| 148 |
-
"reward_std": 0.
|
| 149 |
-
"rewards/
|
| 150 |
"step": 10
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"clip_ratio": 0.0,
|
| 154 |
-
"completion_length":
|
| 155 |
"epoch": 0.18487394957983194,
|
| 156 |
-
"grad_norm": 0.
|
| 157 |
-
"kl": 0.
|
| 158 |
-
"learning_rate": 9.
|
| 159 |
-
"loss":
|
| 160 |
-
"num_tokens":
|
| 161 |
-
"reward": 0.
|
| 162 |
-
"reward_std": 0.
|
| 163 |
-
"rewards/
|
| 164 |
"step": 11
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"clip_ratio": 0.0,
|
| 168 |
-
"completion_length":
|
| 169 |
"epoch": 0.20168067226890757,
|
| 170 |
-
"grad_norm": 0.
|
| 171 |
-
"kl": 0.
|
| 172 |
-
"learning_rate": 9.
|
| 173 |
-
"loss": 0.
|
| 174 |
-
"num_tokens":
|
| 175 |
-
"reward": 0.
|
| 176 |
-
"reward_std": 0.
|
| 177 |
-
"rewards/
|
| 178 |
"step": 12
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"clip_ratio": 0.0,
|
| 182 |
-
"completion_length":
|
| 183 |
"epoch": 0.2184873949579832,
|
| 184 |
-
"grad_norm": 0.
|
| 185 |
-
"kl": 0.
|
| 186 |
-
"learning_rate":
|
| 187 |
-
"loss": 0.
|
| 188 |
-
"num_tokens":
|
| 189 |
-
"reward": 0.
|
| 190 |
-
"reward_std": 0.
|
| 191 |
-
"rewards/
|
| 192 |
"step": 13
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"clip_ratio": 0.0,
|
| 196 |
-
"completion_length":
|
| 197 |
"epoch": 0.23529411764705882,
|
| 198 |
-
"grad_norm": 0.
|
| 199 |
-
"kl": 0.
|
| 200 |
-
"learning_rate":
|
| 201 |
-
"loss": 0.
|
| 202 |
-
"num_tokens":
|
| 203 |
-
"reward": 0.
|
| 204 |
-
"reward_std": 0.
|
| 205 |
-
"rewards/
|
| 206 |
"step": 14
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"clip_ratio": 0.0,
|
| 210 |
-
"completion_length":
|
| 211 |
"epoch": 0.25210084033613445,
|
| 212 |
-
"grad_norm": 0.
|
| 213 |
-
"kl": 0.
|
| 214 |
-
"learning_rate":
|
| 215 |
-
"loss": 0.
|
| 216 |
-
"num_tokens":
|
| 217 |
-
"reward": 0.
|
| 218 |
-
"reward_std": 0.
|
| 219 |
-
"rewards/
|
| 220 |
"step": 15
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"clip_ratio": 0.0,
|
| 224 |
-
"completion_length":
|
| 225 |
"epoch": 0.2689075630252101,
|
| 226 |
-
"grad_norm": 0.
|
| 227 |
-
"kl": 0.
|
| 228 |
-
"learning_rate":
|
| 229 |
-
"loss":
|
| 230 |
-
"num_tokens":
|
| 231 |
-
"reward": 0.
|
| 232 |
-
"reward_std": 0.
|
| 233 |
-
"rewards/
|
| 234 |
"step": 16
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"clip_ratio": 0.0,
|
| 238 |
-
"completion_length":
|
| 239 |
"epoch": 0.2857142857142857,
|
| 240 |
-
"grad_norm": 0.
|
| 241 |
-
"kl": 0.
|
| 242 |
-
"learning_rate":
|
| 243 |
-
"loss": 0.
|
| 244 |
-
"num_tokens":
|
| 245 |
-
"reward": 0.
|
| 246 |
-
"reward_std": 0.
|
| 247 |
-
"rewards/
|
| 248 |
"step": 17
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"clip_ratio": 0.0,
|
| 252 |
-
"completion_length":
|
| 253 |
"epoch": 0.3025210084033613,
|
| 254 |
-
"grad_norm": 0.
|
| 255 |
-
"kl": 0.
|
| 256 |
-
"learning_rate":
|
| 257 |
-
"loss": 0.
|
| 258 |
-
"num_tokens":
|
| 259 |
-
"reward": 0.
|
| 260 |
-
"reward_std": 0.
|
| 261 |
-
"rewards/
|
| 262 |
"step": 18
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"clip_ratio": 0.0,
|
| 266 |
-
"completion_length":
|
| 267 |
"epoch": 0.31932773109243695,
|
| 268 |
-
"grad_norm": 0.
|
| 269 |
-
"kl": 0.
|
| 270 |
-
"learning_rate":
|
| 271 |
-
"loss":
|
| 272 |
-
"num_tokens":
|
| 273 |
-
"reward": 0.
|
| 274 |
-
"reward_std": 0.
|
| 275 |
-
"rewards/
|
| 276 |
"step": 19
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"clip_ratio": 0.0,
|
| 280 |
-
"completion_length":
|
| 281 |
"epoch": 0.33613445378151263,
|
| 282 |
-
"grad_norm": 0.
|
| 283 |
-
"kl": 0.
|
| 284 |
-
"learning_rate":
|
| 285 |
-
"loss": 0.
|
| 286 |
-
"num_tokens":
|
| 287 |
-
"reward": 0.
|
| 288 |
-
"reward_std": 0.
|
| 289 |
-
"rewards/
|
| 290 |
"step": 20
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"clip_ratio": 0.0,
|
| 294 |
-
"completion_length":
|
| 295 |
"epoch": 0.35294117647058826,
|
| 296 |
-
"grad_norm": 0.
|
| 297 |
-
"kl": 0.
|
| 298 |
-
"learning_rate":
|
| 299 |
-
"loss": 0.
|
| 300 |
-
"num_tokens":
|
| 301 |
-
"reward": 0.
|
| 302 |
-
"reward_std": 0.
|
| 303 |
-
"rewards/
|
| 304 |
"step": 21
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"clip_ratio": 0.0,
|
| 308 |
-
"completion_length":
|
| 309 |
"epoch": 0.3697478991596639,
|
| 310 |
-
"grad_norm": 0.
|
| 311 |
-
"kl": 0.
|
| 312 |
-
"learning_rate":
|
| 313 |
-
"loss":
|
| 314 |
-
"num_tokens":
|
| 315 |
-
"reward": 0.
|
| 316 |
-
"reward_std": 0.
|
| 317 |
-
"rewards/
|
| 318 |
"step": 22
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"clip_ratio": 0.0,
|
| 322 |
-
"completion_length":
|
| 323 |
"epoch": 0.3865546218487395,
|
| 324 |
-
"grad_norm": 0.
|
| 325 |
-
"kl": 0.
|
| 326 |
-
"learning_rate":
|
| 327 |
-
"loss": 0.
|
| 328 |
-
"num_tokens":
|
| 329 |
-
"reward": 0.
|
| 330 |
-
"reward_std": 0.
|
| 331 |
-
"rewards/
|
| 332 |
"step": 23
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"clip_ratio": 0.0,
|
| 336 |
-
"completion_length":
|
| 337 |
"epoch": 0.40336134453781514,
|
| 338 |
-
"grad_norm": 0.
|
| 339 |
-
"kl": 0.
|
| 340 |
-
"learning_rate":
|
| 341 |
-
"loss":
|
| 342 |
-
"num_tokens":
|
| 343 |
-
"reward": 0.
|
| 344 |
-
"reward_std": 0.
|
| 345 |
-
"rewards/
|
| 346 |
"step": 24
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"clip_ratio": 0.0,
|
| 350 |
-
"completion_length":
|
| 351 |
"epoch": 0.42016806722689076,
|
| 352 |
-
"grad_norm": 0.
|
| 353 |
-
"kl": 0.
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss": 0.
|
| 356 |
-
"num_tokens":
|
| 357 |
-
"reward": 0.
|
| 358 |
-
"reward_std": 0.
|
| 359 |
-
"rewards/
|
| 360 |
"step": 25
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"clip_ratio": 0.0,
|
| 364 |
-
"completion_length":
|
| 365 |
"epoch": 0.4369747899159664,
|
| 366 |
-
"grad_norm": 0.
|
| 367 |
-
"kl": 0.
|
| 368 |
-
"learning_rate":
|
| 369 |
-
"loss": 0.
|
| 370 |
-
"num_tokens":
|
| 371 |
-
"reward": 0.
|
| 372 |
-
"reward_std": 0.
|
| 373 |
-
"rewards/
|
| 374 |
"step": 26
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"clip_ratio": 0.0,
|
| 378 |
-
"completion_length":
|
| 379 |
"epoch": 0.453781512605042,
|
| 380 |
-
"grad_norm": 0.
|
| 381 |
-
"kl": 0.
|
| 382 |
-
"learning_rate":
|
| 383 |
-
"loss":
|
| 384 |
-
"num_tokens":
|
| 385 |
-
"reward": 0.
|
| 386 |
-
"reward_std": 0.
|
| 387 |
-
"rewards/
|
| 388 |
"step": 27
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"clip_ratio": 0.0,
|
| 392 |
-
"completion_length":
|
| 393 |
"epoch": 0.47058823529411764,
|
| 394 |
-
"grad_norm": 0.
|
| 395 |
-
"kl": 0.
|
| 396 |
-
"learning_rate":
|
| 397 |
-
"loss": 0.
|
| 398 |
-
"num_tokens":
|
| 399 |
-
"reward": 0.
|
| 400 |
-
"reward_std": 0.
|
| 401 |
-
"rewards/
|
| 402 |
"step": 28
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"clip_ratio": 0.0,
|
| 406 |
-
"completion_length":
|
| 407 |
"epoch": 0.48739495798319327,
|
| 408 |
-
"grad_norm": 0.
|
| 409 |
-
"kl": 0.
|
| 410 |
-
"learning_rate":
|
| 411 |
-
"loss": 0.
|
| 412 |
-
"num_tokens":
|
| 413 |
-
"reward": 0.
|
| 414 |
-
"reward_std": 0.
|
| 415 |
-
"rewards/
|
| 416 |
"step": 29
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"clip_ratio": 0.0,
|
| 420 |
-
"completion_length":
|
| 421 |
"epoch": 0.5042016806722689,
|
| 422 |
-
"grad_norm": 0.
|
| 423 |
-
"kl": 0.
|
| 424 |
-
"learning_rate":
|
| 425 |
-
"loss":
|
| 426 |
-
"num_tokens":
|
| 427 |
-
"reward": 0.
|
| 428 |
-
"reward_std": 0.
|
| 429 |
-
"rewards/
|
| 430 |
"step": 30
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"clip_ratio": 0.0,
|
| 434 |
-
"completion_length":
|
| 435 |
"epoch": 0.5210084033613446,
|
| 436 |
-
"grad_norm": 0.
|
| 437 |
-
"kl": 0.
|
| 438 |
-
"learning_rate":
|
| 439 |
-
"loss": 0.
|
| 440 |
-
"num_tokens":
|
| 441 |
-
"reward": 0.
|
| 442 |
-
"reward_std": 0.
|
| 443 |
-
"rewards/
|
| 444 |
"step": 31
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"clip_ratio": 0.0,
|
| 448 |
-
"completion_length":
|
| 449 |
"epoch": 0.5378151260504201,
|
| 450 |
-
"grad_norm": 0.
|
| 451 |
-
"kl": 0.
|
| 452 |
-
"learning_rate":
|
| 453 |
-
"loss":
|
| 454 |
-
"num_tokens":
|
| 455 |
-
"reward": 0.
|
| 456 |
-
"reward_std": 0.
|
| 457 |
-
"rewards/
|
| 458 |
"step": 32
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"clip_ratio": 0.0,
|
| 462 |
-
"completion_length":
|
| 463 |
"epoch": 0.5546218487394958,
|
| 464 |
-
"grad_norm": 0.
|
| 465 |
-
"kl": 0.
|
| 466 |
-
"learning_rate":
|
| 467 |
-
"loss": 0.
|
| 468 |
-
"num_tokens":
|
| 469 |
-
"reward": 0.
|
| 470 |
-
"reward_std": 0.
|
| 471 |
-
"rewards/
|
| 472 |
"step": 33
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"clip_ratio": 0.0,
|
| 476 |
-
"completion_length":
|
| 477 |
"epoch": 0.5714285714285714,
|
| 478 |
-
"grad_norm": 0.
|
| 479 |
-
"kl": 0.
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 0.
|
| 482 |
-
"num_tokens":
|
| 483 |
-
"reward": 0.
|
| 484 |
-
"reward_std": 0.
|
| 485 |
-
"rewards/
|
| 486 |
"step": 34
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"clip_ratio": 0.0,
|
| 490 |
-
"completion_length":
|
| 491 |
"epoch": 0.5882352941176471,
|
| 492 |
-
"grad_norm": 0.
|
| 493 |
-
"kl": 0.
|
| 494 |
-
"learning_rate":
|
| 495 |
-
"loss": 0.
|
| 496 |
-
"num_tokens":
|
| 497 |
-
"reward": 0.
|
| 498 |
-
"reward_std": 0.
|
| 499 |
-
"rewards/
|
| 500 |
"step": 35
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"clip_ratio": 0.0,
|
| 504 |
-
"completion_length":
|
| 505 |
"epoch": 0.6050420168067226,
|
| 506 |
-
"grad_norm": 0.
|
| 507 |
-
"kl": 0.
|
| 508 |
-
"learning_rate":
|
| 509 |
-
"loss":
|
| 510 |
-
"num_tokens":
|
| 511 |
-
"reward": 0.
|
| 512 |
-
"reward_std": 0.
|
| 513 |
-
"rewards/
|
| 514 |
"step": 36
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"clip_ratio": 0.0,
|
| 518 |
-
"completion_length":
|
| 519 |
"epoch": 0.6218487394957983,
|
| 520 |
-
"grad_norm": 0.
|
| 521 |
-
"kl": 0.
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss":
|
| 524 |
-
"num_tokens":
|
| 525 |
-
"reward": 0.
|
| 526 |
-
"reward_std": 0.
|
| 527 |
-
"rewards/
|
| 528 |
"step": 37
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"clip_ratio": 0.0,
|
| 532 |
-
"completion_length":
|
| 533 |
"epoch": 0.6386554621848739,
|
| 534 |
-
"grad_norm": 0.
|
| 535 |
-
"kl": 0.
|
| 536 |
-
"learning_rate":
|
| 537 |
-
"loss": 0.
|
| 538 |
-
"num_tokens":
|
| 539 |
-
"reward": 0.
|
| 540 |
"reward_std": 0.0,
|
| 541 |
-
"rewards/
|
| 542 |
"step": 38
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"clip_ratio": 0.0,
|
| 546 |
-
"completion_length":
|
| 547 |
"epoch": 0.6554621848739496,
|
| 548 |
-
"grad_norm": 0.
|
| 549 |
-
"kl": 0.
|
| 550 |
-
"learning_rate":
|
| 551 |
-
"loss": 0.
|
| 552 |
-
"num_tokens":
|
| 553 |
-
"reward": 0.
|
| 554 |
"reward_std": 0.0,
|
| 555 |
-
"rewards/
|
| 556 |
"step": 39
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"clip_ratio": 0.0,
|
| 560 |
-
"completion_length":
|
| 561 |
"epoch": 0.6722689075630253,
|
| 562 |
-
"grad_norm": 0.
|
| 563 |
-
"kl": 0.
|
| 564 |
-
"learning_rate":
|
| 565 |
-
"loss":
|
| 566 |
-
"num_tokens":
|
| 567 |
-
"reward": 0.
|
| 568 |
-
"reward_std": 0.
|
| 569 |
-
"rewards/
|
| 570 |
"step": 40
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"clip_ratio": 0.0,
|
| 574 |
-
"completion_length":
|
| 575 |
"epoch": 0.6890756302521008,
|
| 576 |
-
"grad_norm": 0.
|
| 577 |
-
"kl": 0.
|
| 578 |
-
"learning_rate":
|
| 579 |
-
"loss":
|
| 580 |
-
"num_tokens":
|
| 581 |
-
"reward": 0.
|
| 582 |
-
"reward_std": 0.
|
| 583 |
-
"rewards/
|
| 584 |
"step": 41
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"clip_ratio": 0.0,
|
| 588 |
-
"completion_length":
|
| 589 |
"epoch": 0.7058823529411765,
|
| 590 |
-
"grad_norm": 0.
|
| 591 |
-
"kl": 0.
|
| 592 |
-
"learning_rate":
|
| 593 |
-
"loss": 0.
|
| 594 |
-
"num_tokens":
|
| 595 |
-
"reward": 0.
|
| 596 |
-
"reward_std": 0.
|
| 597 |
-
"rewards/
|
| 598 |
"step": 42
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"clip_ratio": 0.0,
|
| 602 |
-
"completion_length":
|
| 603 |
"epoch": 0.7226890756302521,
|
| 604 |
-
"grad_norm": 0.
|
| 605 |
-
"kl": 0.
|
| 606 |
-
"learning_rate":
|
| 607 |
-
"loss":
|
| 608 |
-
"num_tokens":
|
| 609 |
-
"reward": 0.
|
| 610 |
-
"reward_std": 0.
|
| 611 |
-
"rewards/
|
| 612 |
"step": 43
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"clip_ratio": 0.0,
|
| 616 |
-
"completion_length":
|
| 617 |
"epoch": 0.7394957983193278,
|
| 618 |
-
"grad_norm": 0.
|
| 619 |
-
"kl": 0.
|
| 620 |
-
"learning_rate":
|
| 621 |
-
"loss": 0.
|
| 622 |
-
"num_tokens":
|
| 623 |
-
"reward": 0.
|
| 624 |
"reward_std": 0.0,
|
| 625 |
-
"rewards/
|
| 626 |
"step": 44
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"clip_ratio": 0.0,
|
| 630 |
-
"completion_length":
|
| 631 |
"epoch": 0.7563025210084033,
|
| 632 |
-
"grad_norm": 0.
|
| 633 |
-
"kl": 0.
|
| 634 |
-
"learning_rate":
|
| 635 |
-
"loss": 0.
|
| 636 |
-
"num_tokens":
|
| 637 |
-
"reward": 0.
|
| 638 |
-
"reward_std": 0.
|
| 639 |
-
"rewards/
|
| 640 |
"step": 45
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"clip_ratio": 0.0,
|
| 644 |
-
"completion_length":
|
| 645 |
"epoch": 0.773109243697479,
|
| 646 |
-
"grad_norm": 0.
|
| 647 |
-
"kl": 0.
|
| 648 |
-
"learning_rate":
|
| 649 |
-
"loss": 0.
|
| 650 |
-
"num_tokens":
|
| 651 |
-
"reward": 0.
|
| 652 |
"reward_std": 0.0,
|
| 653 |
-
"rewards/
|
| 654 |
"step": 46
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"clip_ratio": 0.0,
|
| 658 |
-
"completion_length":
|
| 659 |
"epoch": 0.7899159663865546,
|
| 660 |
-
"grad_norm": 0.
|
| 661 |
-
"kl": 0.
|
| 662 |
-
"learning_rate":
|
| 663 |
-
"loss": 0.0023,
|
| 664 |
-
"num_tokens": 6159890.0,
|
| 665 |
-
"reward": 0.359375,
|
| 666 |
-
"reward_std": 0.04419417306780815,
|
| 667 |
-
"rewards/code_format_reward": 0.359375,
|
| 668 |
-
"step": 47
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"clip_ratio": 0.0,
|
| 672 |
-
"completion_length": 558.078125,
|
| 673 |
-
"epoch": 0.8067226890756303,
|
| 674 |
-
"grad_norm": 0.3406285345554352,
|
| 675 |
-
"kl": 0.007928848266601562,
|
| 676 |
-
"learning_rate": 6.00901221787878e-07,
|
| 677 |
-
"loss": -0.024,
|
| 678 |
-
"num_tokens": 6292572.0,
|
| 679 |
-
"reward": 0.421875,
|
| 680 |
-
"reward_std": 0.04419417306780815,
|
| 681 |
-
"rewards/code_format_reward": 0.421875,
|
| 682 |
-
"step": 48
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"clip_ratio": 0.0,
|
| 686 |
-
"completion_length": 543.0625,
|
| 687 |
-
"epoch": 0.8235294117647058,
|
| 688 |
-
"grad_norm": 0.2923266887664795,
|
| 689 |
-
"kl": 0.009540557861328125,
|
| 690 |
-
"learning_rate": 5.86396198162632e-07,
|
| 691 |
-
"loss": -0.011,
|
| 692 |
-
"num_tokens": 6423868.0,
|
| 693 |
-
"reward": 0.421875,
|
| 694 |
-
"reward_std": 0.04419417306780815,
|
| 695 |
-
"rewards/code_format_reward": 0.421875,
|
| 696 |
-
"step": 49
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"clip_ratio": 0.0,
|
| 700 |
-
"completion_length": 571.78125,
|
| 701 |
-
"epoch": 0.8403361344537815,
|
| 702 |
-
"grad_norm": 0.3251660466194153,
|
| 703 |
-
"kl": 0.00586700439453125,
|
| 704 |
-
"learning_rate": 5.718529999515017e-07,
|
| 705 |
-
"loss": 0.0111,
|
| 706 |
-
"num_tokens": 6556552.0,
|
| 707 |
-
"reward": 0.3046875,
|
| 708 |
-
"reward_std": 0.022097086533904076,
|
| 709 |
-
"rewards/code_format_reward": 0.3046875,
|
| 710 |
-
"step": 50
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"clip_ratio": 0.0,
|
| 714 |
-
"completion_length": 581.1640625,
|
| 715 |
-
"epoch": 0.8571428571428571,
|
| 716 |
-
"grad_norm": 0.17117629945278168,
|
| 717 |
-
"kl": 0.007695198059082031,
|
| 718 |
-
"learning_rate": 5.572868809611257e-07,
|
| 719 |
-
"loss": -0.019,
|
| 720 |
-
"num_tokens": 6694565.0,
|
| 721 |
-
"reward": 0.296875,
|
| 722 |
-
"reward_std": 0.0289318785071373,
|
| 723 |
-
"rewards/code_format_reward": 0.296875,
|
| 724 |
-
"step": 51
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"clip_ratio": 0.0,
|
| 728 |
-
"completion_length": 507.4453125,
|
| 729 |
-
"epoch": 0.8739495798319328,
|
| 730 |
-
"grad_norm": 0.2713668644428253,
|
| 731 |
-
"kl": 0.00688934326171875,
|
| 732 |
-
"learning_rate": 5.427131190388743e-07,
|
| 733 |
-
"loss": 0.0161,
|
| 734 |
-
"num_tokens": 6817558.0,
|
| 735 |
-
"reward": 0.4921875,
|
| 736 |
-
"reward_std": 0.0657544732093811,
|
| 737 |
-
"rewards/code_format_reward": 0.4921875,
|
| 738 |
-
"step": 52
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"clip_ratio": 0.0,
|
| 742 |
-
"completion_length": 524.703125,
|
| 743 |
-
"epoch": 0.8907563025210085,
|
| 744 |
-
"grad_norm": 0.2056041955947876,
|
| 745 |
-
"kl": 0.012969970703125,
|
| 746 |
-
"learning_rate": 5.281470000484985e-07,
|
| 747 |
-
"loss": 0.0066,
|
| 748 |
-
"num_tokens": 6943264.0,
|
| 749 |
-
"reward": 0.4296875,
|
| 750 |
-
"reward_std": 0.022097086533904076,
|
| 751 |
-
"rewards/code_format_reward": 0.4296875,
|
| 752 |
-
"step": 53
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"clip_ratio": 0.0,
|
| 756 |
-
"completion_length": 549.6875,
|
| 757 |
-
"epoch": 0.907563025210084,
|
| 758 |
-
"grad_norm": 0.23319439589977264,
|
| 759 |
-
"kl": 0.007537841796875,
|
| 760 |
-
"learning_rate": 5.136038018373682e-07,
|
| 761 |
-
"loss": 0.0105,
|
| 762 |
-
"num_tokens": 7073792.0,
|
| 763 |
-
"reward": 0.4296875,
|
| 764 |
-
"reward_std": 0.022097086533904076,
|
| 765 |
-
"rewards/code_format_reward": 0.4296875,
|
| 766 |
-
"step": 54
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"clip_ratio": 0.0,
|
| 770 |
-
"completion_length": 529.8125,
|
| 771 |
-
"epoch": 0.9243697478991597,
|
| 772 |
-
"grad_norm": 0.19098737835884094,
|
| 773 |
-
"kl": 0.00901031494140625,
|
| 774 |
-
"learning_rate": 4.990987782121221e-07,
|
| 775 |
-
"loss": -0.0105,
|
| 776 |
-
"num_tokens": 7201304.0,
|
| 777 |
-
"reward": 0.3671875,
|
| 778 |
-
"reward_std": 0.022097086533904076,
|
| 779 |
-
"rewards/code_format_reward": 0.3671875,
|
| 780 |
-
"step": 55
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"clip_ratio": 0.0,
|
| 784 |
-
"completion_length": 526.3515625,
|
| 785 |
-
"epoch": 0.9411764705882353,
|
| 786 |
-
"grad_norm": 0.019878050312399864,
|
| 787 |
-
"kl": 0.00701141357421875,
|
| 788 |
-
"learning_rate": 4.846471429395295e-07,
|
| 789 |
-
"loss": 0.0001,
|
| 790 |
-
"num_tokens": 7331461.0,
|
| 791 |
-
"reward": 0.375,
|
| 792 |
-
"reward_std": 0.0,
|
| 793 |
-
"rewards/code_format_reward": 0.375,
|
| 794 |
-
"step": 56
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"clip_ratio": 0.0,
|
| 798 |
-
"completion_length": 486.59375,
|
| 799 |
-
"epoch": 0.957983193277311,
|
| 800 |
-
"grad_norm": 0.010634235106408596,
|
| 801 |
-
"kl": 0.0092620849609375,
|
| 802 |
-
"learning_rate": 4.7026405378934975e-07,
|
| 803 |
-
"loss": 0.0001,
|
| 804 |
-
"num_tokens": 7452657.0,
|
| 805 |
-
"reward": 0.5625,
|
| 806 |
-
"reward_std": 0.0,
|
| 807 |
-
"rewards/code_format_reward": 0.5625,
|
| 808 |
-
"step": 57
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"clip_ratio": 0.0,
|
| 812 |
-
"completion_length": 413.234375,
|
| 813 |
-
"epoch": 0.9747899159663865,
|
| 814 |
-
"grad_norm": 0.210678830742836,
|
| 815 |
-
"kl": 0.011688232421875,
|
| 816 |
-
"learning_rate": 4.5596459663592625e-07,
|
| 817 |
-
"loss": 0.0237,
|
| 818 |
-
"num_tokens": 7560991.0,
|
| 819 |
-
"reward": 0.6171875,
|
| 820 |
-
"reward_std": 0.022097086533904076,
|
| 821 |
-
"rewards/code_format_reward": 0.6171875,
|
| 822 |
-
"step": 58
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"clip_ratio": 0.0,
|
| 826 |
-
"completion_length": 552.8452529907227,
|
| 827 |
-
"epoch": 0.9915966386554622,
|
| 828 |
-
"grad_norm": 0.3087288439273834,
|
| 829 |
-
"kl": 0.00782012939453125,
|
| 830 |
-
"learning_rate": 4.41763769635186e-07,
|
| 831 |
-
"loss": -0.0165,
|
| 832 |
-
"num_tokens": 7690230.0,
|
| 833 |
-
"reward": 0.4140625,
|
| 834 |
-
"reward_std": 0.06629125960171223,
|
| 835 |
-
"rewards/code_format_reward": 0.4140625,
|
| 836 |
-
"step": 59
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"clip_ratio": 0.0,
|
| 840 |
-
"completion_length": 571.84375,
|
| 841 |
-
"epoch": 1.0168067226890756,
|
| 842 |
-
"grad_norm": 0.2766449451446533,
|
| 843 |
-
"kl": 0.009830474853515625,
|
| 844 |
-
"learning_rate": 4.2767646749364574e-07,
|
| 845 |
-
"loss": 0.0012,
|
| 846 |
-
"num_tokens": 7825994.0,
|
| 847 |
-
"reward": 0.359375,
|
| 848 |
-
"reward_std": 0.04419417306780815,
|
| 849 |
-
"rewards/code_format_reward": 0.359375,
|
| 850 |
-
"step": 60
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"clip_ratio": 0.0,
|
| 854 |
-
"completion_length": 549.734375,
|
| 855 |
-
"epoch": 1.0336134453781514,
|
| 856 |
-
"grad_norm": 0.008657719939947128,
|
| 857 |
-
"kl": 0.00926971435546875,
|
| 858 |
-
"learning_rate": 4.1371746584592227e-07,
|
| 859 |
-
"loss": 0.0001,
|
| 860 |
-
"num_tokens": 7954624.0,
|
| 861 |
-
"reward": 0.4375,
|
| 862 |
-
"reward_std": 0.0,
|
| 863 |
-
"rewards/code_format_reward": 0.4375,
|
| 864 |
-
"step": 61
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"clip_ratio": 0.0,
|
| 868 |
-
"completion_length": 607.4296875,
|
| 869 |
-
"epoch": 1.050420168067227,
|
| 870 |
-
"grad_norm": 0.00941454991698265,
|
| 871 |
-
"kl": 0.00757598876953125,
|
| 872 |
-
"learning_rate": 3.999014057571308e-07,
|
| 873 |
-
"loss": 0.0001,
|
| 874 |
-
"num_tokens": 8095783.0,
|
| 875 |
-
"reward": 0.3125,
|
| 876 |
-
"reward_std": 0.0,
|
| 877 |
-
"rewards/code_format_reward": 0.3125,
|
| 878 |
-
"step": 62
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"clip_ratio": 0.0,
|
| 882 |
-
"completion_length": 466.0625,
|
| 883 |
-
"epoch": 1.0672268907563025,
|
| 884 |
-
"grad_norm": 0.2536774277687073,
|
| 885 |
-
"kl": 0.0111541748046875,
|
| 886 |
-
"learning_rate": 3.862427783664306e-07,
|
| 887 |
-
"loss": -0.0146,
|
| 888 |
-
"num_tokens": 8209831.0,
|
| 889 |
-
"reward": 0.609375,
|
| 890 |
-
"reward_std": 0.04419417306780815,
|
| 891 |
-
"rewards/code_format_reward": 0.609375,
|
| 892 |
-
"step": 63
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"clip_ratio": 0.0,
|
| 896 |
-
"completion_length": 585.3203125,
|
| 897 |
-
"epoch": 1.084033613445378,
|
| 898 |
-
"grad_norm": 0.20105506479740143,
|
| 899 |
-
"kl": 0.004669189453125,
|
| 900 |
-
"learning_rate": 3.7275590968782087e-07,
|
| 901 |
-
"loss": 0.0095,
|
| 902 |
-
"num_tokens": 8347160.0,
|
| 903 |
-
"reward": 0.2421875,
|
| 904 |
-
"reward_std": 0.022097086533904076,
|
| 905 |
-
"rewards/code_format_reward": 0.2421875,
|
| 906 |
-
"step": 64
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"clip_ratio": 0.0,
|
| 910 |
-
"completion_length": 493.7578125,
|
| 911 |
-
"epoch": 1.1008403361344539,
|
| 912 |
-
"grad_norm": 0.010381446219980717,
|
| 913 |
-
"kl": 0.00858306884765625,
|
| 914 |
-
"learning_rate": 3.594549455841296e-07,
|
| 915 |
-
"loss": 0.0001,
|
| 916 |
-
"num_tokens": 8471065.0,
|
| 917 |
-
"reward": 0.5,
|
| 918 |
-
"reward_std": 0.0,
|
| 919 |
-
"rewards/code_format_reward": 0.5,
|
| 920 |
-
"step": 65
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"clip_ratio": 0.0,
|
| 924 |
-
"completion_length": 487.8125,
|
| 925 |
-
"epoch": 1.1176470588235294,
|
| 926 |
-
"grad_norm": 0.21298491954803467,
|
| 927 |
-
"kl": 0.007595062255859375,
|
| 928 |
-
"learning_rate": 3.4635383692995755e-07,
|
| 929 |
-
"loss": 0.011,
|
| 930 |
-
"num_tokens": 8591161.0,
|
| 931 |
-
"reward": 0.4296875,
|
| 932 |
-
"reward_std": 0.022097086533904076,
|
| 933 |
-
"rewards/code_format_reward": 0.4296875,
|
| 934 |
-
"step": 66
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"clip_ratio": 0.0,
|
| 938 |
-
"completion_length": 570.90625,
|
| 939 |
-
"epoch": 1.134453781512605,
|
| 940 |
-
"grad_norm": 0.005000706762075424,
|
| 941 |
-
"kl": 0.005382537841796875,
|
| 942 |
-
"learning_rate": 3.3346632497913773e-07,
|
| 943 |
-
"loss": 0.0001,
|
| 944 |
-
"num_tokens": 8723157.0,
|
| 945 |
-
"reward": 0.25,
|
| 946 |
-
"reward_std": 0.0,
|
| 947 |
-
"rewards/code_format_reward": 0.25,
|
| 948 |
-
"step": 67
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"clip_ratio": 0.0,
|
| 952 |
-
"completion_length": 406.5625,
|
| 953 |
-
"epoch": 1.1512605042016806,
|
| 954 |
-
"grad_norm": 0.2167414426803589,
|
| 955 |
-
"kl": 0.011280059814453125,
|
| 956 |
-
"learning_rate": 3.208059269520568e-07,
|
| 957 |
-
"loss": 0.0024,
|
| 958 |
-
"num_tokens": 8834893.0,
|
| 959 |
-
"reward": 0.5546875,
|
| 960 |
-
"reward_std": 0.022097086533904076,
|
| 961 |
-
"rewards/code_format_reward": 0.5546875,
|
| 962 |
-
"step": 68
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"clip_ratio": 0.0,
|
| 966 |
-
"completion_length": 519.734375,
|
| 967 |
-
"epoch": 1.1680672268907564,
|
| 968 |
-
"grad_norm": 0.22041496634483337,
|
| 969 |
-
"kl": 0.011386871337890625,
|
| 970 |
-
"learning_rate": 3.083859218579573e-07,
|
| 971 |
-
"loss": 0.0064,
|
| 972 |
-
"num_tokens": 8962931.0,
|
| 973 |
-
"reward": 0.4765625,
|
| 974 |
-
"reward_std": 0.03234682232141495,
|
| 975 |
-
"rewards/code_format_reward": 0.4765625,
|
| 976 |
-
"step": 69
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"clip_ratio": 0.0,
|
| 980 |
-
"completion_length": 552.875,
|
| 981 |
-
"epoch": 1.184873949579832,
|
| 982 |
-
"grad_norm": 0.008582229726016521,
|
| 983 |
-
"kl": 0.0049285888671875,
|
| 984 |
-
"learning_rate": 2.9621933656709207e-07,
|
| 985 |
"loss": 0.0,
|
| 986 |
-
"num_tokens":
|
| 987 |
-
"reward": 0.
|
| 988 |
"reward_std": 0.0,
|
| 989 |
-
"rewards/
|
| 990 |
-
"step":
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"clip_ratio": 0.0,
|
| 994 |
-
"completion_length": 505.3203125,
|
| 995 |
-
"epoch": 1.2016806722689075,
|
| 996 |
-
"grad_norm": 0.22665004432201385,
|
| 997 |
-
"kl": 0.00914764404296875,
|
| 998 |
-
"learning_rate": 2.843189321473349e-07,
|
| 999 |
-
"loss": -0.0036,
|
| 1000 |
-
"num_tokens": 9220108.0,
|
| 1001 |
-
"reward": 0.4921875,
|
| 1002 |
-
"reward_std": 0.022097086533904076,
|
| 1003 |
-
"rewards/code_format_reward": 0.4921875,
|
| 1004 |
-
"step": 71
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"clip_ratio": 0.0,
|
| 1008 |
-
"completion_length":
|
| 1009 |
-
"epoch":
|
| 1010 |
-
"grad_norm": 0.
|
| 1011 |
-
"kl": 0.
|
| 1012 |
-
"learning_rate":
|
| 1013 |
"loss": 0.0,
|
| 1014 |
-
"num_tokens":
|
| 1015 |
-
"reward": 0.
|
| 1016 |
-
"reward_std": 0.0,
|
| 1017 |
-
"rewards/code_format_reward": 0.25,
|
| 1018 |
-
"step": 72
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"clip_ratio": 0.0,
|
| 1022 |
-
"completion_length": 546.234375,
|
| 1023 |
-
"epoch": 1.2352941176470589,
|
| 1024 |
-
"grad_norm": 0.007375677116215229,
|
| 1025 |
-
"kl": 0.005222320556640625,
|
| 1026 |
-
"learning_rate": 2.613663011659871e-07,
|
| 1027 |
-
"loss": 0.0001,
|
| 1028 |
-
"num_tokens": 9490949.0,
|
| 1029 |
-
"reward": 0.3125,
|
| 1030 |
-
"reward_std": 0.0,
|
| 1031 |
-
"rewards/code_format_reward": 0.3125,
|
| 1032 |
-
"step": 73
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"clip_ratio": 0.0,
|
| 1036 |
-
"completion_length": 489.0859375,
|
| 1037 |
-
"epoch": 1.2521008403361344,
|
| 1038 |
-
"grad_norm": 0.24114979803562164,
|
| 1039 |
-
"kl": 0.0102996826171875,
|
| 1040 |
-
"learning_rate": 2.5033814874474356e-07,
|
| 1041 |
-
"loss": 0.0336,
|
| 1042 |
-
"num_tokens": 9614784.0,
|
| 1043 |
-
"reward": 0.5546875,
|
| 1044 |
-
"reward_std": 0.022097086533904076,
|
| 1045 |
-
"rewards/code_format_reward": 0.5546875,
|
| 1046 |
-
"step": 74
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"clip_ratio": 0.0,
|
| 1050 |
-
"completion_length": 477.3828125,
|
| 1051 |
-
"epoch": 1.26890756302521,
|
| 1052 |
-
"grad_norm": 0.013432592153549194,
|
| 1053 |
-
"kl": 0.01093292236328125,
|
| 1054 |
-
"learning_rate": 2.3962430022485305e-07,
|
| 1055 |
-
"loss": 0.0001,
|
| 1056 |
-
"num_tokens": 9735377.0,
|
| 1057 |
-
"reward": 0.625,
|
| 1058 |
-
"reward_std": 0.0,
|
| 1059 |
-
"rewards/code_format_reward": 0.625,
|
| 1060 |
-
"step": 75
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"clip_ratio": 0.0,
|
| 1064 |
-
"completion_length": 555.890625,
|
| 1065 |
-
"epoch": 1.2857142857142856,
|
| 1066 |
-
"grad_norm": 0.013460615649819374,
|
| 1067 |
-
"kl": 0.0073699951171875,
|
| 1068 |
-
"learning_rate": 2.2923599295392694e-07,
|
| 1069 |
-
"loss": 0.0001,
|
| 1070 |
-
"num_tokens": 9867139.0,
|
| 1071 |
-
"reward": 0.4375,
|
| 1072 |
-
"reward_std": 0.0,
|
| 1073 |
-
"rewards/code_format_reward": 0.4375,
|
| 1074 |
-
"step": 76
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"clip_ratio": 0.0,
|
| 1078 |
-
"completion_length": 409.7734375,
|
| 1079 |
-
"epoch": 1.3025210084033614,
|
| 1080 |
-
"grad_norm": 0.3048118054866791,
|
| 1081 |
-
"kl": 0.0102081298828125,
|
| 1082 |
-
"learning_rate": 2.1918412283175994e-07,
|
| 1083 |
-
"loss": -0.003,
|
| 1084 |
-
"num_tokens": 9980222.0,
|
| 1085 |
-
"reward": 0.671875,
|
| 1086 |
-
"reward_std": 0.04419417306780815,
|
| 1087 |
-
"rewards/code_format_reward": 0.671875,
|
| 1088 |
-
"step": 77
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"clip_ratio": 0.0,
|
| 1092 |
-
"completion_length": 542.71875,
|
| 1093 |
-
"epoch": 1.319327731092437,
|
| 1094 |
-
"grad_norm": 0.2550560235977173,
|
| 1095 |
-
"kl": 0.007965087890625,
|
| 1096 |
-
"learning_rate": 2.0947923288203713e-07,
|
| 1097 |
-
"loss": 0.0955,
|
| 1098 |
-
"num_tokens": 10107666.0,
|
| 1099 |
-
"reward": 0.484375,
|
| 1100 |
-
"reward_std": 0.04419417306780815,
|
| 1101 |
-
"rewards/code_format_reward": 0.484375,
|
| 1102 |
-
"step": 78
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"clip_ratio": 0.0,
|
| 1106 |
-
"completion_length": 517.8125,
|
| 1107 |
-
"epoch": 1.3361344537815127,
|
| 1108 |
-
"grad_norm": 0.2195437252521515,
|
| 1109 |
-
"kl": 0.00948333740234375,
|
| 1110 |
-
"learning_rate": 2.0013150219415793e-07,
|
| 1111 |
-
"loss": -0.0028,
|
| 1112 |
-
"num_tokens": 10232114.0,
|
| 1113 |
-
"reward": 0.46875,
|
| 1114 |
-
"reward_std": 0.033407654613256454,
|
| 1115 |
-
"rewards/code_format_reward": 0.46875,
|
| 1116 |
-
"step": 79
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"clip_ratio": 0.0,
|
| 1120 |
-
"completion_length": 521.390625,
|
| 1121 |
-
"epoch": 1.3529411764705883,
|
| 1122 |
-
"grad_norm": 0.008532202802598476,
|
| 1123 |
-
"kl": 0.006832122802734375,
|
| 1124 |
-
"learning_rate": 1.9115073524677572e-07,
|
| 1125 |
-
"loss": 0.0001,
|
| 1126 |
-
"num_tokens": 10357436.0,
|
| 1127 |
-
"reward": 0.4375,
|
| 1128 |
-
"reward_std": 0.0,
|
| 1129 |
-
"rewards/code_format_reward": 0.4375,
|
| 1130 |
-
"step": 80
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"clip_ratio": 0.0,
|
| 1134 |
-
"completion_length": 485.96875,
|
| 1135 |
-
"epoch": 1.3697478991596639,
|
| 1136 |
-
"grad_norm": 0.011241392232477665,
|
| 1137 |
-
"kl": 0.00920867919921875,
|
| 1138 |
-
"learning_rate": 1.8254635162425503e-07,
|
| 1139 |
-
"loss": 0.0001,
|
| 1140 |
-
"num_tokens": 10481120.0,
|
| 1141 |
-
"reward": 0.625,
|
| 1142 |
-
"reward_std": 0.0,
|
| 1143 |
-
"rewards/code_format_reward": 0.625,
|
| 1144 |
-
"step": 81
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"clip_ratio": 0.0,
|
| 1148 |
-
"completion_length": 485.390625,
|
| 1149 |
-
"epoch": 1.3865546218487395,
|
| 1150 |
-
"grad_norm": 0.2583635151386261,
|
| 1151 |
-
"kl": 0.01049041748046875,
|
| 1152 |
-
"learning_rate": 1.7432737613682807e-07,
|
| 1153 |
-
"loss": 0.007,
|
| 1154 |
-
"num_tokens": 10603258.0,
|
| 1155 |
-
"reward": 0.5546875,
|
| 1156 |
-
"reward_std": 0.022097086533904076,
|
| 1157 |
-
"rewards/code_format_reward": 0.5546875,
|
| 1158 |
-
"step": 82
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"clip_ratio": 0.0,
|
| 1162 |
-
"completion_length": 583.484375,
|
| 1163 |
-
"epoch": 1.403361344537815,
|
| 1164 |
-
"grad_norm": 0.007445089519023895,
|
| 1165 |
-
"kl": 0.00675201416015625,
|
| 1166 |
-
"learning_rate": 1.6650242935481388e-07,
|
| 1167 |
-
"loss": 0.0001,
|
| 1168 |
-
"num_tokens": 10736448.0,
|
| 1169 |
-
"reward": 0.375,
|
| 1170 |
-
"reward_std": 0.0,
|
| 1171 |
-
"rewards/code_format_reward": 0.375,
|
| 1172 |
-
"step": 83
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"clip_ratio": 0.0,
|
| 1176 |
-
"completion_length": 565.8515625,
|
| 1177 |
-
"epoch": 1.4201680672268908,
|
| 1178 |
-
"grad_norm": 0.012575228698551655,
|
| 1179 |
-
"kl": 0.006989479064941406,
|
| 1180 |
-
"learning_rate": 1.59079718566832e-07,
|
| 1181 |
-
"loss": 0.0001,
|
| 1182 |
-
"num_tokens": 10867749.0,
|
| 1183 |
-
"reward": 0.375,
|
| 1184 |
-
"reward_std": 0.0,
|
| 1185 |
-
"rewards/code_format_reward": 0.375,
|
| 1186 |
-
"step": 84
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"clip_ratio": 0.0,
|
| 1190 |
-
"completion_length": 466.5,
|
| 1191 |
-
"epoch": 1.4369747899159664,
|
| 1192 |
-
"grad_norm": 0.22415108978748322,
|
| 1193 |
-
"kl": 0.0146026611328125,
|
| 1194 |
-
"learning_rate": 1.5206702917148945e-07,
|
| 1195 |
-
"loss": 0.0055,
|
| 1196 |
-
"num_tokens": 10988109.0,
|
| 1197 |
-
"reward": 0.6171875,
|
| 1198 |
-
"reward_std": 0.022097086533904076,
|
| 1199 |
-
"rewards/code_format_reward": 0.6171875,
|
| 1200 |
-
"step": 85
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"clip_ratio": 0.0,
|
| 1204 |
-
"completion_length": 425.671875,
|
| 1205 |
-
"epoch": 1.453781512605042,
|
| 1206 |
-
"grad_norm": 0.012631588615477085,
|
| 1207 |
-
"kl": 0.00989532470703125,
|
| 1208 |
-
"learning_rate": 1.4547171651157214e-07,
|
| 1209 |
-
"loss": 0.0001,
|
| 1210 |
-
"num_tokens": 11100171.0,
|
| 1211 |
-
"reward": 0.5625,
|
| 1212 |
-
"reward_std": 0.0,
|
| 1213 |
-
"rewards/code_format_reward": 0.5625,
|
| 1214 |
-
"step": 86
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"clip_ratio": 0.0,
|
| 1218 |
-
"completion_length": 484.8125,
|
| 1219 |
-
"epoch": 1.4705882352941178,
|
| 1220 |
-
"grad_norm": 0.19023600220680237,
|
| 1221 |
-
"kl": 0.01287078857421875,
|
| 1222 |
-
"learning_rate": 1.3930069815930697e-07,
|
| 1223 |
-
"loss": -0.0073,
|
| 1224 |
-
"num_tokens": 11218611.0,
|
| 1225 |
-
"reward": 0.5546875,
|
| 1226 |
-
"reward_std": 0.022097086533904076,
|
| 1227 |
-
"rewards/code_format_reward": 0.5546875,
|
| 1228 |
-
"step": 87
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"clip_ratio": 0.0,
|
| 1232 |
-
"completion_length": 542.421875,
|
| 1233 |
-
"epoch": 1.4873949579831933,
|
| 1234 |
-
"grad_norm": 0.18787921965122223,
|
| 1235 |
-
"kl": 0.005619049072265625,
|
| 1236 |
-
"learning_rate": 1.3356044666078315e-07,
|
| 1237 |
-
"loss": -0.0035,
|
| 1238 |
-
"num_tokens": 11349593.0,
|
| 1239 |
-
"reward": 0.2421875,
|
| 1240 |
-
"reward_std": 0.022097086533904076,
|
| 1241 |
-
"rewards/code_format_reward": 0.2421875,
|
| 1242 |
-
"step": 88
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"clip_ratio": 0.0,
|
| 1246 |
-
"completion_length": 414.28125,
|
| 1247 |
-
"epoch": 1.504201680672269,
|
| 1248 |
-
"grad_norm": 0.2461971491575241,
|
| 1249 |
-
"kl": 0.01172637939453125,
|
| 1250 |
-
"learning_rate": 1.2825698274714542e-07,
|
| 1251 |
-
"loss": 0.0123,
|
| 1252 |
-
"num_tokens": 11459157.0,
|
| 1253 |
-
"reward": 0.7421875,
|
| 1254 |
-
"reward_std": 0.022097086533904076,
|
| 1255 |
-
"rewards/code_format_reward": 0.7421875,
|
| 1256 |
-
"step": 89
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"clip_ratio": 0.0,
|
| 1260 |
-
"completion_length": 535.8671875,
|
| 1261 |
-
"epoch": 1.5210084033613445,
|
| 1262 |
-
"grad_norm": 0.009749805554747581,
|
| 1263 |
-
"kl": 0.0081024169921875,
|
| 1264 |
-
"learning_rate": 1.233958690196783e-07,
|
| 1265 |
-
"loss": 0.0001,
|
| 1266 |
-
"num_tokens": 11587884.0,
|
| 1267 |
-
"reward": 0.4375,
|
| 1268 |
-
"reward_std": 0.0,
|
| 1269 |
-
"rewards/code_format_reward": 0.4375,
|
| 1270 |
-
"step": 90
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"clip_ratio": 0.0,
|
| 1274 |
-
"completion_length": 599.015625,
|
| 1275 |
-
"epoch": 1.53781512605042,
|
| 1276 |
-
"grad_norm": 0.0069138724356889725,
|
| 1277 |
-
"kl": 0.00567626953125,
|
| 1278 |
-
"learning_rate": 1.1898220411540583e-07,
|
| 1279 |
-
"loss": 0.0001,
|
| 1280 |
-
"num_tokens": 11723398.0,
|
| 1281 |
-
"reward": 0.375,
|
| 1282 |
-
"reward_std": 0.0,
|
| 1283 |
-
"rewards/code_format_reward": 0.375,
|
| 1284 |
-
"step": 91
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"clip_ratio": 0.0,
|
| 1288 |
-
"completion_length": 516.984375,
|
| 1289 |
-
"epoch": 1.5546218487394958,
|
| 1290 |
-
"grad_norm": 0.26793172955513,
|
| 1291 |
-
"kl": 0.00763702392578125,
|
| 1292 |
-
"learning_rate": 1.1502061735932499e-07,
|
| 1293 |
-
"loss": -0.0202,
|
| 1294 |
-
"num_tokens": 11850164.0,
|
| 1295 |
-
"reward": 0.359375,
|
| 1296 |
-
"reward_std": 0.04419417306780815,
|
| 1297 |
-
"rewards/code_format_reward": 0.359375,
|
| 1298 |
-
"step": 92
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"clip_ratio": 0.0,
|
| 1302 |
-
"completion_length": 506.890625,
|
| 1303 |
-
"epoch": 1.5714285714285714,
|
| 1304 |
-
"grad_norm": 0.010911713354289532,
|
| 1305 |
-
"kl": 0.00727081298828125,
|
| 1306 |
-
"learning_rate": 1.115152639088833e-07,
|
| 1307 |
-
"loss": 0.0001,
|
| 1308 |
-
"num_tokens": 11975358.0,
|
| 1309 |
-
"reward": 0.5,
|
| 1310 |
"reward_std": 0.0,
|
| 1311 |
-
"rewards/
|
| 1312 |
-
"step":
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"clip_ratio": 0.0,
|
| 1316 |
-
"completion_length":
|
| 1317 |
-
"epoch":
|
| 1318 |
-
"grad_norm": 0.
|
| 1319 |
-
"kl": 0.
|
| 1320 |
-
"learning_rate": 1.
|
| 1321 |
"loss": 0.0,
|
| 1322 |
-
"num_tokens":
|
| 1323 |
-
"reward": 0.
|
| 1324 |
-
"reward_std": 0.0,
|
| 1325 |
-
"rewards/code_format_reward": 0.3125,
|
| 1326 |
-
"step": 94
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"clip_ratio": 0.0,
|
| 1330 |
-
"completion_length": 391.6796875,
|
| 1331 |
-
"epoch": 1.6050420168067228,
|
| 1332 |
-
"grad_norm": 0.014589727856218815,
|
| 1333 |
-
"kl": 0.0121002197265625,
|
| 1334 |
-
"learning_rate": 1.0588748106974918e-07,
|
| 1335 |
-
"loss": 0.0001,
|
| 1336 |
-
"num_tokens": 12219404.0,
|
| 1337 |
-
"reward": 0.625,
|
| 1338 |
-
"reward_std": 0.0,
|
| 1339 |
-
"rewards/code_format_reward": 0.625,
|
| 1340 |
-
"step": 95
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"clip_ratio": 0.0,
|
| 1344 |
-
"completion_length": 583.4453125,
|
| 1345 |
-
"epoch": 1.6218487394957983,
|
| 1346 |
-
"grad_norm": 0.010287419892847538,
|
| 1347 |
-
"kl": 0.0075836181640625,
|
| 1348 |
-
"learning_rate": 1.0377095444810871e-07,
|
| 1349 |
-
"loss": 0.0001,
|
| 1350 |
-
"num_tokens": 12354245.0,
|
| 1351 |
-
"reward": 0.4375,
|
| 1352 |
-
"reward_std": 0.0,
|
| 1353 |
-
"rewards/code_format_reward": 0.4375,
|
| 1354 |
-
"step": 96
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"clip_ratio": 0.0,
|
| 1358 |
-
"completion_length": 551.484375,
|
| 1359 |
-
"epoch": 1.638655462184874,
|
| 1360 |
-
"grad_norm": 0.21819841861724854,
|
| 1361 |
-
"kl": 0.00997161865234375,
|
| 1362 |
-
"learning_rate": 1.0212246047502372e-07,
|
| 1363 |
-
"loss": -0.0046,
|
| 1364 |
-
"num_tokens": 12482835.0,
|
| 1365 |
-
"reward": 0.4296875,
|
| 1366 |
-
"reward_std": 0.022097086533904076,
|
| 1367 |
-
"rewards/code_format_reward": 0.4296875,
|
| 1368 |
-
"step": 97
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"clip_ratio": 0.0,
|
| 1372 |
-
"completion_length": 486.953125,
|
| 1373 |
-
"epoch": 1.6554621848739495,
|
| 1374 |
-
"grad_norm": 0.011982027441263199,
|
| 1375 |
-
"kl": 0.0090484619140625,
|
| 1376 |
-
"learning_rate": 1.0094372819302977e-07,
|
| 1377 |
-
"loss": 0.0001,
|
| 1378 |
-
"num_tokens": 12605909.0,
|
| 1379 |
-
"reward": 0.5625,
|
| 1380 |
-
"reward_std": 0.0,
|
| 1381 |
-
"rewards/code_format_reward": 0.5625,
|
| 1382 |
-
"step": 98
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"clip_ratio": 0.0,
|
| 1386 |
-
"completion_length": 516.625,
|
| 1387 |
-
"epoch": 1.6722689075630253,
|
| 1388 |
-
"grad_norm": 0.008852253668010235,
|
| 1389 |
-
"kl": 0.00608062744140625,
|
| 1390 |
-
"learning_rate": 1.0023599392951829e-07,
|
| 1391 |
-
"loss": 0.0001,
|
| 1392 |
-
"num_tokens": 12732141.0,
|
| 1393 |
-
"reward": 0.3125,
|
| 1394 |
"reward_std": 0.0,
|
| 1395 |
-
"rewards/
|
| 1396 |
-
"step":
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"clip_ratio": 0.0,
|
| 1400 |
-
"completion_length":
|
| 1401 |
-
"epoch":
|
| 1402 |
-
"grad_norm": 0.
|
| 1403 |
-
"kl": 0.
|
| 1404 |
"learning_rate": 1e-07,
|
| 1405 |
-
"loss": 0.
|
| 1406 |
-
"num_tokens":
|
| 1407 |
-
"reward": 0.
|
| 1408 |
"reward_std": 0.0,
|
| 1409 |
-
"rewards/
|
| 1410 |
-
"step":
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
-
"epoch":
|
| 1414 |
-
"step":
|
| 1415 |
"total_flos": 0.0,
|
| 1416 |
-
"train_loss": 0.
|
| 1417 |
-
"train_runtime":
|
| 1418 |
-
"train_samples_per_second": 0.
|
| 1419 |
-
"train_steps_per_second": 0.
|
| 1420 |
}
|
| 1421 |
],
|
| 1422 |
"logging_steps": 1,
|
| 1423 |
-
"max_steps":
|
| 1424 |
"num_input_tokens_seen": 0,
|
| 1425 |
-
"num_train_epochs":
|
| 1426 |
"save_steps": 50,
|
| 1427 |
"stateful_callbacks": {
|
| 1428 |
"TrainerControl": {
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8403361344537815,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 50,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"clip_ratio": 0.0,
|
| 14 |
+
"completion_length": 1599.671875,
|
| 15 |
"epoch": 0.01680672268907563,
|
| 16 |
+
"grad_norm": 0.0,
|
| 17 |
"kl": 0.0,
|
| 18 |
+
"learning_rate": 5e-07,
|
| 19 |
+
"loss": 0.0,
|
| 20 |
+
"num_tokens": 266094.0,
|
| 21 |
+
"reward": 0.0,
|
| 22 |
+
"reward_std": 0.0,
|
| 23 |
+
"rewards/format_reward": 0.0,
|
| 24 |
"step": 1
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"clip_ratio": 0.0,
|
| 28 |
+
"completion_length": 1544.015625,
|
| 29 |
"epoch": 0.03361344537815126,
|
| 30 |
+
"grad_norm": 0.0,
|
| 31 |
"kl": 0.0,
|
| 32 |
+
"learning_rate": 1e-06,
|
| 33 |
+
"loss": 0.0,
|
| 34 |
+
"num_tokens": 523136.0,
|
| 35 |
+
"reward": 0.0,
|
| 36 |
+
"reward_std": 0.0,
|
| 37 |
+
"rewards/format_reward": 0.0,
|
| 38 |
"step": 2
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"clip_ratio": 0.0,
|
| 42 |
+
"completion_length": 1510.03125,
|
| 43 |
"epoch": 0.05042016806722689,
|
| 44 |
+
"grad_norm": 0.0,
|
| 45 |
+
"kl": 0.0,
|
| 46 |
+
"learning_rate": 9.990365154573716e-07,
|
| 47 |
+
"loss": 0.0,
|
| 48 |
+
"num_tokens": 778148.0,
|
| 49 |
+
"reward": 0.0,
|
| 50 |
+
"reward_std": 0.0,
|
| 51 |
+
"rewards/format_reward": 0.0,
|
| 52 |
"step": 3
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"clip_ratio": 0.0,
|
| 56 |
+
"completion_length": 1525.9296875,
|
| 57 |
"epoch": 0.06722689075630252,
|
| 58 |
+
"grad_norm": 0.0,
|
| 59 |
+
"kl": 0.0,
|
| 60 |
+
"learning_rate": 9.961501876182148e-07,
|
| 61 |
"loss": 0.0,
|
| 62 |
+
"num_tokens": 1032547.0,
|
| 63 |
"reward": 0.0,
|
| 64 |
"reward_std": 0.0,
|
| 65 |
+
"rewards/format_reward": 0.0,
|
| 66 |
"step": 4
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"clip_ratio": 0.0,
|
| 70 |
+
"completion_length": 1469.1953125,
|
| 71 |
"epoch": 0.08403361344537816,
|
| 72 |
+
"grad_norm": 0.0,
|
| 73 |
+
"kl": 0.0,
|
| 74 |
+
"learning_rate": 9.913533761814537e-07,
|
| 75 |
+
"loss": 0.0,
|
| 76 |
+
"num_tokens": 1278956.0,
|
| 77 |
+
"reward": 0.0,
|
| 78 |
+
"reward_std": 0.0,
|
| 79 |
+
"rewards/format_reward": 0.0,
|
| 80 |
"step": 5
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"clip_ratio": 0.0,
|
| 84 |
+
"completion_length": 1516.4375,
|
| 85 |
"epoch": 0.10084033613445378,
|
| 86 |
+
"grad_norm": 0.0,
|
| 87 |
+
"kl": 0.0,
|
| 88 |
+
"learning_rate": 9.846666218300807e-07,
|
| 89 |
"loss": 0.0,
|
| 90 |
+
"num_tokens": 1532156.0,
|
| 91 |
"reward": 0.0,
|
| 92 |
"reward_std": 0.0,
|
| 93 |
+
"rewards/format_reward": 0.0,
|
| 94 |
"step": 6
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"clip_ratio": 0.0,
|
| 98 |
+
"completion_length": 1353.75,
|
| 99 |
"epoch": 0.11764705882352941,
|
| 100 |
+
"grad_norm": 0.0,
|
| 101 |
+
"kl": 0.0,
|
| 102 |
+
"learning_rate": 9.761185582727976e-07,
|
| 103 |
+
"loss": 0.0,
|
| 104 |
+
"num_tokens": 1760652.0,
|
| 105 |
+
"reward": 0.0,
|
| 106 |
+
"reward_std": 0.0,
|
| 107 |
+
"rewards/format_reward": 0.0,
|
| 108 |
"step": 7
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"clip_ratio": 0.0,
|
| 112 |
+
"completion_length": 1524.59375,
|
| 113 |
"epoch": 0.13445378151260504,
|
| 114 |
+
"grad_norm": 0.0,
|
| 115 |
+
"kl": 0.0,
|
| 116 |
+
"learning_rate": 9.657457896300791e-07,
|
| 117 |
"loss": 0.0,
|
| 118 |
+
"num_tokens": 2017904.0,
|
| 119 |
"reward": 0.0,
|
| 120 |
"reward_std": 0.0,
|
| 121 |
+
"rewards/format_reward": 0.0,
|
| 122 |
"step": 8
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"clip_ratio": 0.0,
|
| 126 |
+
"completion_length": 1266.0703125,
|
| 127 |
"epoch": 0.15126050420168066,
|
| 128 |
+
"grad_norm": 0.0,
|
| 129 |
+
"kl": 0.0,
|
| 130 |
+
"learning_rate": 9.535927336897097e-07,
|
| 131 |
+
"loss": 0.0,
|
| 132 |
+
"num_tokens": 2233457.0,
|
| 133 |
+
"reward": 0.0,
|
| 134 |
+
"reward_std": 0.0,
|
| 135 |
+
"rewards/format_reward": 0.0,
|
| 136 |
"step": 9
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"clip_ratio": 0.0,
|
| 140 |
+
"completion_length": 1721.0390625,
|
| 141 |
"epoch": 0.16806722689075632,
|
| 142 |
+
"grad_norm": 0.0,
|
| 143 |
+
"kl": 0.0,
|
| 144 |
+
"learning_rate": 9.397114317029974e-07,
|
| 145 |
+
"loss": 0.0,
|
| 146 |
+
"num_tokens": 2516198.0,
|
| 147 |
+
"reward": 0.0,
|
| 148 |
+
"reward_std": 0.0,
|
| 149 |
+
"rewards/format_reward": 0.0,
|
| 150 |
"step": 10
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"clip_ratio": 0.0,
|
| 154 |
+
"completion_length": 1270.0546875,
|
| 155 |
"epoch": 0.18487394957983194,
|
| 156 |
+
"grad_norm": 0.0,
|
| 157 |
+
"kl": 0.0,
|
| 158 |
+
"learning_rate": 9.241613255361454e-07,
|
| 159 |
+
"loss": 0.0,
|
| 160 |
+
"num_tokens": 2734181.0,
|
| 161 |
+
"reward": 0.0,
|
| 162 |
+
"reward_std": 0.0,
|
| 163 |
+
"rewards/format_reward": 0.0,
|
| 164 |
"step": 11
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"clip_ratio": 0.0,
|
| 168 |
+
"completion_length": 1756.8515625,
|
| 169 |
"epoch": 0.20168067226890757,
|
| 170 |
+
"grad_norm": 0.0,
|
| 171 |
+
"kl": 0.0,
|
| 172 |
+
"learning_rate": 9.070090031310558e-07,
|
| 173 |
+
"loss": 0.0,
|
| 174 |
+
"num_tokens": 3022002.0,
|
| 175 |
+
"reward": 0.0,
|
| 176 |
+
"reward_std": 0.0,
|
| 177 |
+
"rewards/format_reward": 0.0,
|
| 178 |
"step": 12
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"clip_ratio": 0.0,
|
| 182 |
+
"completion_length": 1439.25,
|
| 183 |
"epoch": 0.2184873949579832,
|
| 184 |
+
"grad_norm": 0.0,
|
| 185 |
+
"kl": 0.0,
|
| 186 |
+
"learning_rate": 8.883279133655398e-07,
|
| 187 |
+
"loss": 0.0,
|
| 188 |
+
"num_tokens": 3265898.0,
|
| 189 |
+
"reward": 0.0,
|
| 190 |
+
"reward_std": 0.0,
|
| 191 |
+
"rewards/format_reward": 0.0,
|
| 192 |
"step": 13
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"clip_ratio": 0.0,
|
| 196 |
+
"completion_length": 1287.7109375,
|
| 197 |
"epoch": 0.23529411764705882,
|
| 198 |
+
"grad_norm": 0.0,
|
| 199 |
+
"kl": 0.0,
|
| 200 |
+
"learning_rate": 8.681980515339463e-07,
|
| 201 |
+
"loss": 0.0,
|
| 202 |
+
"num_tokens": 3485349.0,
|
| 203 |
+
"reward": 0.0,
|
| 204 |
+
"reward_std": 0.0,
|
| 205 |
+
"rewards/format_reward": 0.0,
|
| 206 |
"step": 14
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"clip_ratio": 0.0,
|
| 210 |
+
"completion_length": 1291.4765625,
|
| 211 |
"epoch": 0.25210084033613445,
|
| 212 |
+
"grad_norm": 0.0,
|
| 213 |
+
"kl": 0.0,
|
| 214 |
+
"learning_rate": 8.46705616795031e-07,
|
| 215 |
+
"loss": 0.0,
|
| 216 |
+
"num_tokens": 3709706.0,
|
| 217 |
+
"reward": 0.0,
|
| 218 |
+
"reward_std": 0.0,
|
| 219 |
+
"rewards/format_reward": 0.0,
|
| 220 |
"step": 15
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"clip_ratio": 0.0,
|
| 224 |
+
"completion_length": 1431.7890625,
|
| 225 |
"epoch": 0.2689075630252101,
|
| 226 |
+
"grad_norm": 0.0,
|
| 227 |
+
"kl": 0.0,
|
| 228 |
+
"learning_rate": 8.239426430539243e-07,
|
| 229 |
+
"loss": 0.0,
|
| 230 |
+
"num_tokens": 3949471.0,
|
| 231 |
+
"reward": 0.0,
|
| 232 |
+
"reward_std": 0.0,
|
| 233 |
+
"rewards/format_reward": 0.0,
|
| 234 |
"step": 16
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"clip_ratio": 0.0,
|
| 238 |
+
"completion_length": 1147.71875,
|
| 239 |
"epoch": 0.2857142857142857,
|
| 240 |
+
"grad_norm": 0.0,
|
| 241 |
+
"kl": 0.0,
|
| 242 |
+
"learning_rate": 8.00006604858821e-07,
|
| 243 |
+
"loss": 0.0,
|
| 244 |
+
"num_tokens": 4152771.0,
|
| 245 |
+
"reward": 0.0,
|
| 246 |
+
"reward_std": 0.0,
|
| 247 |
+
"rewards/format_reward": 0.0,
|
| 248 |
"step": 17
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"clip_ratio": 0.0,
|
| 252 |
+
"completion_length": 1743.6484375,
|
| 253 |
"epoch": 0.3025210084033613,
|
| 254 |
+
"grad_norm": 0.0,
|
| 255 |
+
"kl": 0.0,
|
| 256 |
+
"learning_rate": 7.75e-07,
|
| 257 |
+
"loss": 0.0,
|
| 258 |
+
"num_tokens": 4439918.0,
|
| 259 |
+
"reward": 0.0,
|
| 260 |
+
"reward_std": 0.0,
|
| 261 |
+
"rewards/format_reward": 0.0,
|
| 262 |
"step": 18
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"clip_ratio": 0.0,
|
| 266 |
+
"completion_length": 1624.828125,
|
| 267 |
"epoch": 0.31932773109243695,
|
| 268 |
+
"grad_norm": 0.0,
|
| 269 |
+
"kl": 0.0,
|
| 270 |
+
"learning_rate": 7.490299105985506e-07,
|
| 271 |
+
"loss": 0.0,
|
| 272 |
+
"num_tokens": 4708000.0,
|
| 273 |
+
"reward": 0.0,
|
| 274 |
+
"reward_std": 0.0,
|
| 275 |
+
"rewards/format_reward": 0.0,
|
| 276 |
"step": 19
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"clip_ratio": 0.0,
|
| 280 |
+
"completion_length": 1281.0859375,
|
| 281 |
"epoch": 0.33613445378151263,
|
| 282 |
+
"grad_norm": 0.0,
|
| 283 |
+
"kl": 0.0,
|
| 284 |
+
"learning_rate": 7.222075445642904e-07,
|
| 285 |
+
"loss": 0.0,
|
| 286 |
+
"num_tokens": 4930731.0,
|
| 287 |
+
"reward": 0.0,
|
| 288 |
+
"reward_std": 0.0,
|
| 289 |
+
"rewards/format_reward": 0.0,
|
| 290 |
"step": 20
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"clip_ratio": 0.0,
|
| 294 |
+
"completion_length": 1734.984375,
|
| 295 |
"epoch": 0.35294117647058826,
|
| 296 |
+
"grad_norm": 0.0,
|
| 297 |
+
"kl": 0.0,
|
| 298 |
+
"learning_rate": 6.946477593864227e-07,
|
| 299 |
+
"loss": 0.0,
|
| 300 |
+
"num_tokens": 5215121.0,
|
| 301 |
+
"reward": 0.0,
|
| 302 |
+
"reward_std": 0.0,
|
| 303 |
+
"rewards/format_reward": 0.0,
|
| 304 |
"step": 21
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"clip_ratio": 0.0,
|
| 308 |
+
"completion_length": 1469.5625,
|
| 309 |
"epoch": 0.3697478991596639,
|
| 310 |
+
"grad_norm": 0.0,
|
| 311 |
+
"kl": 0.0,
|
| 312 |
+
"learning_rate": 6.664685702961344e-07,
|
| 313 |
+
"loss": 0.0,
|
| 314 |
+
"num_tokens": 5459289.0,
|
| 315 |
+
"reward": 0.0,
|
| 316 |
+
"reward_std": 0.0,
|
| 317 |
+
"rewards/format_reward": 0.0,
|
| 318 |
"step": 22
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"clip_ratio": 0.0,
|
| 322 |
+
"completion_length": 1616.2734375,
|
| 323 |
"epoch": 0.3865546218487395,
|
| 324 |
+
"grad_norm": 0.0,
|
| 325 |
+
"kl": 0.0,
|
| 326 |
+
"learning_rate": 6.377906449072577e-07,
|
| 327 |
+
"loss": 0.0,
|
| 328 |
+
"num_tokens": 5727860.0,
|
| 329 |
+
"reward": 0.0,
|
| 330 |
+
"reward_std": 0.0,
|
| 331 |
+
"rewards/format_reward": 0.0,
|
| 332 |
"step": 23
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"clip_ratio": 0.0,
|
| 336 |
+
"completion_length": 1414.109375,
|
| 337 |
"epoch": 0.40336134453781514,
|
| 338 |
+
"grad_norm": 0.0,
|
| 339 |
+
"kl": 0.0,
|
| 340 |
+
"learning_rate": 6.087367864990232e-07,
|
| 341 |
+
"loss": 0.0,
|
| 342 |
+
"num_tokens": 5971898.0,
|
| 343 |
+
"reward": 0.0,
|
| 344 |
+
"reward_std": 0.0,
|
| 345 |
+
"rewards/format_reward": 0.0,
|
| 346 |
"step": 24
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"clip_ratio": 0.0,
|
| 350 |
+
"completion_length": 1353.9296875,
|
| 351 |
"epoch": 0.42016806722689076,
|
| 352 |
+
"grad_norm": 0.0,
|
| 353 |
+
"kl": 0.0,
|
| 354 |
+
"learning_rate": 5.794314081535644e-07,
|
| 355 |
+
"loss": 0.0,
|
| 356 |
+
"num_tokens": 6203001.0,
|
| 357 |
+
"reward": 0.0,
|
| 358 |
+
"reward_std": 0.0,
|
| 359 |
+
"rewards/format_reward": 0.0,
|
| 360 |
"step": 25
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"clip_ratio": 0.0,
|
| 364 |
+
"completion_length": 1247.484375,
|
| 365 |
"epoch": 0.4369747899159664,
|
| 366 |
+
"grad_norm": 0.0,
|
| 367 |
+
"kl": 0.0,
|
| 368 |
+
"learning_rate": 5.5e-07,
|
| 369 |
+
"loss": 0.0,
|
| 370 |
+
"num_tokens": 6419583.0,
|
| 371 |
+
"reward": 0.0,
|
| 372 |
+
"reward_std": 0.0,
|
| 373 |
+
"rewards/format_reward": 0.0,
|
| 374 |
"step": 26
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"clip_ratio": 0.0,
|
| 378 |
+
"completion_length": 1523.859375,
|
| 379 |
"epoch": 0.453781512605042,
|
| 380 |
+
"grad_norm": 0.0,
|
| 381 |
+
"kl": 0.0,
|
| 382 |
+
"learning_rate": 5.205685918464355e-07,
|
| 383 |
+
"loss": 0.0,
|
| 384 |
+
"num_tokens": 6674853.0,
|
| 385 |
+
"reward": 0.0,
|
| 386 |
+
"reward_std": 0.0,
|
| 387 |
+
"rewards/format_reward": 0.0,
|
| 388 |
"step": 27
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"clip_ratio": 0.0,
|
| 392 |
+
"completion_length": 1352.765625,
|
| 393 |
"epoch": 0.47058823529411764,
|
| 394 |
+
"grad_norm": 0.0,
|
| 395 |
+
"kl": 0.0,
|
| 396 |
+
"learning_rate": 4.912632135009769e-07,
|
| 397 |
+
"loss": 0.0,
|
| 398 |
+
"num_tokens": 6904127.0,
|
| 399 |
+
"reward": 0.0,
|
| 400 |
+
"reward_std": 0.0,
|
| 401 |
+
"rewards/format_reward": 0.0,
|
| 402 |
"step": 28
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"clip_ratio": 0.0,
|
| 406 |
+
"completion_length": 1583.96875,
|
| 407 |
"epoch": 0.48739495798319327,
|
| 408 |
+
"grad_norm": 0.0,
|
| 409 |
+
"kl": 0.0,
|
| 410 |
+
"learning_rate": 4.6220935509274227e-07,
|
| 411 |
+
"loss": 0.0,
|
| 412 |
+
"num_tokens": 7167227.0,
|
| 413 |
+
"reward": 0.0,
|
| 414 |
+
"reward_std": 0.0,
|
| 415 |
+
"rewards/format_reward": 0.0,
|
| 416 |
"step": 29
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"clip_ratio": 0.0,
|
| 420 |
+
"completion_length": 1578.2421875,
|
| 421 |
"epoch": 0.5042016806722689,
|
| 422 |
+
"grad_norm": 0.0,
|
| 423 |
+
"kl": 0.0,
|
| 424 |
+
"learning_rate": 4.3353142970386557e-07,
|
| 425 |
+
"loss": 0.0,
|
| 426 |
+
"num_tokens": 7432202.0,
|
| 427 |
+
"reward": 0.0,
|
| 428 |
+
"reward_std": 0.0,
|
| 429 |
+
"rewards/format_reward": 0.0,
|
| 430 |
"step": 30
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"clip_ratio": 0.0,
|
| 434 |
+
"completion_length": 1786.71875,
|
| 435 |
"epoch": 0.5210084033613446,
|
| 436 |
+
"grad_norm": 0.0,
|
| 437 |
+
"kl": 0.0,
|
| 438 |
+
"learning_rate": 4.053522406135774e-07,
|
| 439 |
+
"loss": 0.0,
|
| 440 |
+
"num_tokens": 7721798.0,
|
| 441 |
+
"reward": 0.0,
|
| 442 |
+
"reward_std": 0.0,
|
| 443 |
+
"rewards/format_reward": 0.0,
|
| 444 |
"step": 31
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"clip_ratio": 0.0,
|
| 448 |
+
"completion_length": 1568.609375,
|
| 449 |
"epoch": 0.5378151260504201,
|
| 450 |
+
"grad_norm": 0.0,
|
| 451 |
+
"kl": 0.0,
|
| 452 |
+
"learning_rate": 3.777924554357096e-07,
|
| 453 |
+
"loss": 0.0,
|
| 454 |
+
"num_tokens": 7986012.0,
|
| 455 |
+
"reward": 0.0,
|
| 456 |
+
"reward_std": 0.0,
|
| 457 |
+
"rewards/format_reward": 0.0,
|
| 458 |
"step": 32
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"clip_ratio": 0.0,
|
| 462 |
+
"completion_length": 1542.75,
|
| 463 |
"epoch": 0.5546218487394958,
|
| 464 |
+
"grad_norm": 0.0,
|
| 465 |
+
"kl": 0.0,
|
| 466 |
+
"learning_rate": 3.5097008940144955e-07,
|
| 467 |
+
"loss": 0.0,
|
| 468 |
+
"num_tokens": 8244964.0,
|
| 469 |
+
"reward": 0.0,
|
| 470 |
+
"reward_std": 0.0,
|
| 471 |
+
"rewards/format_reward": 0.0,
|
| 472 |
"step": 33
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"clip_ratio": 0.0,
|
| 476 |
+
"completion_length": 1326.6875,
|
| 477 |
"epoch": 0.5714285714285714,
|
| 478 |
+
"grad_norm": 0.0,
|
| 479 |
+
"kl": 0.0,
|
| 480 |
+
"learning_rate": 3.250000000000001e-07,
|
| 481 |
+
"loss": 0.0,
|
| 482 |
+
"num_tokens": 8476132.0,
|
| 483 |
+
"reward": 0.0,
|
| 484 |
+
"reward_std": 0.0,
|
| 485 |
+
"rewards/format_reward": 0.0,
|
| 486 |
"step": 34
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"clip_ratio": 0.0,
|
| 490 |
+
"completion_length": 1582.921875,
|
| 491 |
"epoch": 0.5882352941176471,
|
| 492 |
+
"grad_norm": 0.0,
|
| 493 |
+
"kl": 0.0,
|
| 494 |
+
"learning_rate": 2.999933951411791e-07,
|
| 495 |
+
"loss": 0.0,
|
| 496 |
+
"num_tokens": 8739482.0,
|
| 497 |
+
"reward": 0.0,
|
| 498 |
+
"reward_std": 0.0,
|
| 499 |
+
"rewards/format_reward": 0.0,
|
| 500 |
"step": 35
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"clip_ratio": 0.0,
|
| 504 |
+
"completion_length": 1426.2890625,
|
| 505 |
"epoch": 0.6050420168067226,
|
| 506 |
+
"grad_norm": 0.0,
|
| 507 |
+
"kl": 0.0,
|
| 508 |
+
"learning_rate": 2.760573569460757e-07,
|
| 509 |
+
"loss": 0.0,
|
| 510 |
+
"num_tokens": 8981671.0,
|
| 511 |
+
"reward": 0.0,
|
| 512 |
+
"reward_std": 0.0,
|
| 513 |
+
"rewards/format_reward": 0.0,
|
| 514 |
"step": 36
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"clip_ratio": 0.0,
|
| 518 |
+
"completion_length": 1474.1796875,
|
| 519 |
"epoch": 0.6218487394957983,
|
| 520 |
+
"grad_norm": 0.0,
|
| 521 |
+
"kl": 0.0,
|
| 522 |
+
"learning_rate": 2.53294383204969e-07,
|
| 523 |
+
"loss": 0.0,
|
| 524 |
+
"num_tokens": 9227398.0,
|
| 525 |
+
"reward": 0.0,
|
| 526 |
+
"reward_std": 0.0,
|
| 527 |
+
"rewards/format_reward": 0.0,
|
| 528 |
"step": 37
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"clip_ratio": 0.0,
|
| 532 |
+
"completion_length": 1203.25,
|
| 533 |
"epoch": 0.6386554621848739,
|
| 534 |
+
"grad_norm": 0.0,
|
| 535 |
+
"kl": 0.0,
|
| 536 |
+
"learning_rate": 2.3180194846605364e-07,
|
| 537 |
+
"loss": 0.0,
|
| 538 |
+
"num_tokens": 9436654.0,
|
| 539 |
+
"reward": 0.0,
|
| 540 |
"reward_std": 0.0,
|
| 541 |
+
"rewards/format_reward": 0.0,
|
| 542 |
"step": 38
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"clip_ratio": 0.0,
|
| 546 |
+
"completion_length": 1524.3671875,
|
| 547 |
"epoch": 0.6554621848739496,
|
| 548 |
+
"grad_norm": 0.0,
|
| 549 |
+
"kl": 0.0,
|
| 550 |
+
"learning_rate": 2.1167208663446022e-07,
|
| 551 |
+
"loss": 0.0,
|
| 552 |
+
"num_tokens": 9690789.0,
|
| 553 |
+
"reward": 0.0,
|
| 554 |
"reward_std": 0.0,
|
| 555 |
+
"rewards/format_reward": 0.0,
|
| 556 |
"step": 39
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"clip_ratio": 0.0,
|
| 560 |
+
"completion_length": 1663.359375,
|
| 561 |
"epoch": 0.6722689075630253,
|
| 562 |
+
"grad_norm": 0.0,
|
| 563 |
+
"kl": 0.0,
|
| 564 |
+
"learning_rate": 1.929909968689442e-07,
|
| 565 |
+
"loss": 0.0,
|
| 566 |
+
"num_tokens": 9964307.0,
|
| 567 |
+
"reward": 0.0,
|
| 568 |
+
"reward_std": 0.0,
|
| 569 |
+
"rewards/format_reward": 0.0,
|
| 570 |
"step": 40
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"clip_ratio": 0.0,
|
| 574 |
+
"completion_length": 1360.765625,
|
| 575 |
"epoch": 0.6890756302521008,
|
| 576 |
+
"grad_norm": 0.0,
|
| 577 |
+
"kl": 0.0,
|
| 578 |
+
"learning_rate": 1.758386744638546e-07,
|
| 579 |
+
"loss": 0.0,
|
| 580 |
+
"num_tokens": 10195365.0,
|
| 581 |
+
"reward": 0.0,
|
| 582 |
+
"reward_std": 0.0,
|
| 583 |
+
"rewards/format_reward": 0.0,
|
| 584 |
"step": 41
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"clip_ratio": 0.0,
|
| 588 |
+
"completion_length": 1217.7421875,
|
| 589 |
"epoch": 0.7058823529411765,
|
| 590 |
+
"grad_norm": 0.0,
|
| 591 |
+
"kl": 0.0,
|
| 592 |
+
"learning_rate": 1.6028856829700258e-07,
|
| 593 |
+
"loss": 0.0,
|
| 594 |
+
"num_tokens": 10409540.0,
|
| 595 |
+
"reward": 0.0,
|
| 596 |
+
"reward_std": 0.0,
|
| 597 |
+
"rewards/format_reward": 0.0,
|
| 598 |
"step": 42
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"clip_ratio": 0.0,
|
| 602 |
+
"completion_length": 1373.4375,
|
| 603 |
"epoch": 0.7226890756302521,
|
| 604 |
+
"grad_norm": 0.0,
|
| 605 |
+
"kl": 0.0,
|
| 606 |
+
"learning_rate": 1.464072663102903e-07,
|
| 607 |
+
"loss": 0.0,
|
| 608 |
+
"num_tokens": 10643324.0,
|
| 609 |
+
"reward": 0.0,
|
| 610 |
+
"reward_std": 0.0,
|
| 611 |
+
"rewards/format_reward": 0.0,
|
| 612 |
"step": 43
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"clip_ratio": 0.0,
|
| 616 |
+
"completion_length": 1329.703125,
|
| 617 |
"epoch": 0.7394957983193278,
|
| 618 |
+
"grad_norm": 0.0,
|
| 619 |
+
"kl": 0.0,
|
| 620 |
+
"learning_rate": 1.3425421036992097e-07,
|
| 621 |
+
"loss": 0.0,
|
| 622 |
+
"num_tokens": 10876006.0,
|
| 623 |
+
"reward": 0.0,
|
| 624 |
"reward_std": 0.0,
|
| 625 |
+
"rewards/format_reward": 0.0,
|
| 626 |
"step": 44
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"clip_ratio": 0.0,
|
| 630 |
+
"completion_length": 1481.7890625,
|
| 631 |
"epoch": 0.7563025210084033,
|
| 632 |
+
"grad_norm": 0.0,
|
| 633 |
+
"kl": 0.0,
|
| 634 |
+
"learning_rate": 1.238814417272025e-07,
|
| 635 |
+
"loss": 0.0,
|
| 636 |
+
"num_tokens": 11124203.0,
|
| 637 |
+
"reward": 0.0,
|
| 638 |
+
"reward_std": 0.0,
|
| 639 |
+
"rewards/format_reward": 0.0,
|
| 640 |
"step": 45
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"clip_ratio": 0.0,
|
| 644 |
+
"completion_length": 1337.3515625,
|
| 645 |
"epoch": 0.773109243697479,
|
| 646 |
+
"grad_norm": 0.0,
|
| 647 |
+
"kl": 0.0,
|
| 648 |
+
"learning_rate": 1.1533337816991931e-07,
|
| 649 |
+
"loss": 0.0,
|
| 650 |
+
"num_tokens": 11350872.0,
|
| 651 |
+
"reward": 0.0,
|
| 652 |
"reward_std": 0.0,
|
| 653 |
+
"rewards/format_reward": 0.0,
|
| 654 |
"step": 46
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"clip_ratio": 0.0,
|
| 658 |
+
"completion_length": 1556.4453125,
|
| 659 |
"epoch": 0.7899159663865546,
|
| 660 |
+
"grad_norm": 0.0,
|
| 661 |
+
"kl": 0.0,
|
| 662 |
+
"learning_rate": 1.086466238185463e-07,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 663 |
"loss": 0.0,
|
| 664 |
+
"num_tokens": 11611593.0,
|
| 665 |
+
"reward": 0.0,
|
| 666 |
"reward_std": 0.0,
|
| 667 |
+
"rewards/format_reward": 0.0,
|
| 668 |
+
"step": 47
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"clip_ratio": 0.0,
|
| 672 |
+
"completion_length": 1598.1484375,
|
| 673 |
+
"epoch": 0.8067226890756303,
|
| 674 |
+
"grad_norm": 0.0,
|
| 675 |
+
"kl": 0.0,
|
| 676 |
+
"learning_rate": 1.0384981238178533e-07,
|
| 677 |
"loss": 0.0,
|
| 678 |
+
"num_tokens": 11876876.0,
|
| 679 |
+
"reward": 0.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 680 |
"reward_std": 0.0,
|
| 681 |
+
"rewards/format_reward": 0.0,
|
| 682 |
+
"step": 48
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"clip_ratio": 0.0,
|
| 686 |
+
"completion_length": 1476.125,
|
| 687 |
+
"epoch": 0.8235294117647058,
|
| 688 |
+
"grad_norm": 0.0,
|
| 689 |
+
"kl": 0.0,
|
| 690 |
+
"learning_rate": 1.0096348454262843e-07,
|
| 691 |
"loss": 0.0,
|
| 692 |
+
"num_tokens": 12127164.0,
|
| 693 |
+
"reward": 0.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 694 |
"reward_std": 0.0,
|
| 695 |
+
"rewards/format_reward": 0.0,
|
| 696 |
+
"step": 49
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"clip_ratio": 0.0,
|
| 700 |
+
"completion_length": 1574.3984375,
|
| 701 |
+
"epoch": 0.8403361344537815,
|
| 702 |
+
"grad_norm": 0.0,
|
| 703 |
+
"kl": 0.0,
|
| 704 |
"learning_rate": 1e-07,
|
| 705 |
+
"loss": 0.0,
|
| 706 |
+
"num_tokens": 12387831.0,
|
| 707 |
+
"reward": 0.0,
|
| 708 |
"reward_std": 0.0,
|
| 709 |
+
"rewards/format_reward": 0.0,
|
| 710 |
+
"step": 50
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 0.8403361344537815,
|
| 714 |
+
"step": 50,
|
| 715 |
"total_flos": 0.0,
|
| 716 |
+
"train_loss": 0.0,
|
| 717 |
+
"train_runtime": 13359.4688,
|
| 718 |
+
"train_samples_per_second": 0.479,
|
| 719 |
+
"train_steps_per_second": 0.004
|
| 720 |
}
|
| 721 |
],
|
| 722 |
"logging_steps": 1,
|
| 723 |
+
"max_steps": 50,
|
| 724 |
"num_input_tokens_seen": 0,
|
| 725 |
+
"num_train_epochs": 1,
|
| 726 |
"save_steps": 50,
|
| 727 |
"stateful_callbacks": {
|
| 728 |
"TrainerControl": {
|