Model save
Browse files- README.md +1 -1
- all_results.json +5 -5
- model.safetensors +1 -1
- train_results.json +5 -5
- trainer_state.json +14 -703
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -27,7 +27,7 @@ print(output["generated_text"])
|
|
| 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/riddickzhou/huggingface/runs/
|
| 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).
|
|
|
|
| 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/riddickzhou/huggingface/runs/25d8vd1w)
|
| 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":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": -7.761021455128988e-10,
|
| 4 |
+
"train_runtime": 74.9061,
|
| 5 |
+
"train_samples": 10,
|
| 6 |
+
"train_samples_per_second": 0.401,
|
| 7 |
+
"train_steps_per_second": 0.04
|
| 8 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3554214752
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:16bea0a35dba47a078ac1750e001512a02837da37c70c3c806244efbe9efefc8
|
| 3 |
size 3554214752
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"total_flos": 0.0,
|
| 3 |
+
"train_loss": -7.761021455128988e-10,
|
| 4 |
+
"train_runtime": 74.9061,
|
| 5 |
+
"train_samples": 10,
|
| 6 |
+
"train_samples_per_second": 0.401,
|
| 7 |
+
"train_steps_per_second": 0.04
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,720 +1,31 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"completion_length":
|
| 13 |
-
"epoch": 0
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
| 18 |
-
"reward": 0.1682291713077575,
|
| 19 |
-
"reward_std": 0.14682963252998887,
|
| 20 |
-
"rewards/accuracy_reward": 0.1682291713077575,
|
| 21 |
"rewards/format_reward": 0.0,
|
| 22 |
-
"step":
|
| 23 |
-
},
|
| 24 |
-
{
|
| 25 |
-
"completion_length": 841.7924652099609,
|
| 26 |
-
"epoch": 0.11263639563533967,
|
| 27 |
-
"grad_norm": 0.004771426951179321,
|
| 28 |
-
"kl": 0.002738553285598755,
|
| 29 |
-
"learning_rate": 7.4074074074074075e-06,
|
| 30 |
-
"loss": 0.0001,
|
| 31 |
-
"reward": 0.30807292461395264,
|
| 32 |
-
"reward_std": 0.14928703159093856,
|
| 33 |
-
"rewards/accuracy_reward": 0.30807292461395264,
|
| 34 |
-
"rewards/format_reward": 0.0,
|
| 35 |
-
"step": 20
|
| 36 |
-
},
|
| 37 |
-
{
|
| 38 |
-
"completion_length": 762.6487173080444,
|
| 39 |
-
"epoch": 0.1689545934530095,
|
| 40 |
-
"grad_norm": 0.007118948305002948,
|
| 41 |
-
"kl": 0.009932708740234376,
|
| 42 |
-
"learning_rate": 1.1111111111111113e-05,
|
| 43 |
-
"loss": 0.0004,
|
| 44 |
-
"reward": 0.3781250110128894,
|
| 45 |
-
"reward_std": 0.20807096506468953,
|
| 46 |
-
"rewards/accuracy_reward": 0.3781250110128894,
|
| 47 |
-
"rewards/format_reward": 0.0,
|
| 48 |
-
"step": 30
|
| 49 |
-
},
|
| 50 |
-
{
|
| 51 |
-
"completion_length": 798.5023654937744,
|
| 52 |
-
"epoch": 0.22527279127067934,
|
| 53 |
-
"grad_norm": 0.007503736791945914,
|
| 54 |
-
"kl": 0.015509796142578126,
|
| 55 |
-
"learning_rate": 1.4814814814814815e-05,
|
| 56 |
-
"loss": 0.0006,
|
| 57 |
-
"reward": 0.38020834540948273,
|
| 58 |
-
"reward_std": 0.2114361953921616,
|
| 59 |
-
"rewards/accuracy_reward": 0.38020834540948273,
|
| 60 |
-
"rewards/format_reward": 0.0,
|
| 61 |
-
"step": 40
|
| 62 |
-
},
|
| 63 |
-
{
|
| 64 |
-
"completion_length": 719.8515825271606,
|
| 65 |
-
"epoch": 0.2815909890883492,
|
| 66 |
-
"grad_norm": 0.006071145947288392,
|
| 67 |
-
"kl": 0.034136962890625,
|
| 68 |
-
"learning_rate": 1.851851851851852e-05,
|
| 69 |
-
"loss": 0.0014,
|
| 70 |
-
"reward": 0.45156251119915397,
|
| 71 |
-
"reward_std": 0.22436260846443473,
|
| 72 |
-
"rewards/accuracy_reward": 0.45156251119915397,
|
| 73 |
-
"rewards/format_reward": 0.0,
|
| 74 |
-
"step": 50
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"completion_length": 788.8078327178955,
|
| 78 |
-
"epoch": 0.337909186906019,
|
| 79 |
-
"grad_norm": 0.02698349636255445,
|
| 80 |
-
"kl": 0.13095016479492189,
|
| 81 |
-
"learning_rate": 1.999219309353572e-05,
|
| 82 |
-
"loss": 0.0052,
|
| 83 |
-
"reward": 0.38776042775716635,
|
| 84 |
-
"reward_std": 0.22218369678594171,
|
| 85 |
-
"rewards/accuracy_reward": 0.38776042775716635,
|
| 86 |
-
"rewards/format_reward": 0.0,
|
| 87 |
-
"step": 60
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"completion_length": 809.6302291870118,
|
| 91 |
-
"epoch": 0.39422738472368885,
|
| 92 |
-
"grad_norm": 0.004711601163642177,
|
| 93 |
-
"kl": 0.1918182373046875,
|
| 94 |
-
"learning_rate": 1.99445283568046e-05,
|
| 95 |
-
"loss": 0.0077,
|
| 96 |
-
"reward": 0.37812500933650883,
|
| 97 |
-
"reward_std": 0.22222802694886923,
|
| 98 |
-
"rewards/accuracy_reward": 0.37812500933650883,
|
| 99 |
-
"rewards/format_reward": 0.0,
|
| 100 |
-
"step": 70
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"completion_length": 802.3885620117187,
|
| 104 |
-
"epoch": 0.4505455825413587,
|
| 105 |
-
"grad_norm": 0.004723358467824132,
|
| 106 |
-
"kl": 0.15126266479492187,
|
| 107 |
-
"learning_rate": 1.985374248499546e-05,
|
| 108 |
-
"loss": 0.0061,
|
| 109 |
-
"reward": 0.3697916771983728,
|
| 110 |
-
"reward_std": 0.20327610801905394,
|
| 111 |
-
"rewards/accuracy_reward": 0.3697916771983728,
|
| 112 |
-
"rewards/format_reward": 0.0,
|
| 113 |
-
"step": 80
|
| 114 |
-
},
|
| 115 |
-
{
|
| 116 |
-
"completion_length": 802.2953327178955,
|
| 117 |
-
"epoch": 0.5068637803590285,
|
| 118 |
-
"grad_norm": 0.009074543078939649,
|
| 119 |
-
"kl": 0.262213134765625,
|
| 120 |
-
"learning_rate": 1.972022914080411e-05,
|
| 121 |
-
"loss": 0.0105,
|
| 122 |
-
"reward": 0.37083334419876335,
|
| 123 |
-
"reward_std": 0.22432688772678375,
|
| 124 |
-
"rewards/accuracy_reward": 0.37083334419876335,
|
| 125 |
-
"rewards/format_reward": 0.0,
|
| 126 |
-
"step": 90
|
| 127 |
-
},
|
| 128 |
-
{
|
| 129 |
-
"completion_length": 787.9052295684814,
|
| 130 |
-
"epoch": 0.5631819781766984,
|
| 131 |
-
"grad_norm": 0.006486195069627925,
|
| 132 |
-
"kl": 0.28298492431640626,
|
| 133 |
-
"learning_rate": 1.9544567260379455e-05,
|
| 134 |
-
"loss": 0.0113,
|
| 135 |
-
"reward": 0.39479167736135423,
|
| 136 |
-
"reward_std": 0.23836338161490858,
|
| 137 |
-
"rewards/accuracy_reward": 0.39479167736135423,
|
| 138 |
-
"rewards/format_reward": 0.0,
|
| 139 |
-
"step": 100
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"completion_length": 815.8523633956909,
|
| 143 |
-
"epoch": 0.6195001759943681,
|
| 144 |
-
"grad_norm": 0.005538478051469189,
|
| 145 |
-
"kl": 0.23237152099609376,
|
| 146 |
-
"learning_rate": 1.9327518542959717e-05,
|
| 147 |
-
"loss": 0.0093,
|
| 148 |
-
"reward": 0.37942709189374,
|
| 149 |
-
"reward_std": 0.2303738916758448,
|
| 150 |
-
"rewards/accuracy_reward": 0.37942709189374,
|
| 151 |
-
"rewards/format_reward": 0.0,
|
| 152 |
-
"step": 110
|
| 153 |
-
},
|
| 154 |
-
{
|
| 155 |
-
"completion_length": 811.012520980835,
|
| 156 |
-
"epoch": 0.675818373812038,
|
| 157 |
-
"grad_norm": 0.004328950832672593,
|
| 158 |
-
"kl": 0.16385421752929688,
|
| 159 |
-
"learning_rate": 1.9070024148017375e-05,
|
| 160 |
-
"loss": 0.0066,
|
| 161 |
-
"reward": 0.38645834478083996,
|
| 162 |
-
"reward_std": 0.2094512910116464,
|
| 163 |
-
"rewards/accuracy_reward": 0.38645834478083996,
|
| 164 |
-
"rewards/format_reward": 0.0,
|
| 165 |
-
"step": 120
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"completion_length": 812.9247604370117,
|
| 169 |
-
"epoch": 0.7321365716297078,
|
| 170 |
-
"grad_norm": 0.006602899810792365,
|
| 171 |
-
"kl": 0.26991424560546873,
|
| 172 |
-
"learning_rate": 1.8773200614234587e-05,
|
| 173 |
-
"loss": 0.0108,
|
| 174 |
-
"reward": 0.36484376071020963,
|
| 175 |
-
"reward_std": 0.22123059891164304,
|
| 176 |
-
"rewards/accuracy_reward": 0.36484376071020963,
|
| 177 |
-
"rewards/format_reward": 0.0,
|
| 178 |
-
"step": 130
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"completion_length": 817.9567909240723,
|
| 182 |
-
"epoch": 0.7884547694473777,
|
| 183 |
-
"grad_norm": 0.0053164002751646695,
|
| 184 |
-
"kl": 0.302496337890625,
|
| 185 |
-
"learning_rate": 1.8438335018005052e-05,
|
| 186 |
-
"loss": 0.0121,
|
| 187 |
-
"reward": 0.3606770923361182,
|
| 188 |
-
"reward_std": 0.2574935433920473,
|
| 189 |
-
"rewards/accuracy_reward": 0.3606770923361182,
|
| 190 |
-
"rewards/format_reward": 0.0,
|
| 191 |
-
"step": 140
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"completion_length": 827.4060081481933,
|
| 195 |
-
"epoch": 0.8447729672650475,
|
| 196 |
-
"grad_norm": 0.006126912969513216,
|
| 197 |
-
"kl": 0.20762939453125,
|
| 198 |
-
"learning_rate": 1.8066879392455932e-05,
|
| 199 |
-
"loss": 0.0083,
|
| 200 |
-
"reward": 0.31015625975560396,
|
| 201 |
-
"reward_std": 0.18832489829510451,
|
| 202 |
-
"rewards/accuracy_reward": 0.31015625975560396,
|
| 203 |
-
"rewards/format_reward": 0.0,
|
| 204 |
-
"step": 150
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"completion_length": 821.6466341018677,
|
| 208 |
-
"epoch": 0.9010911650827174,
|
| 209 |
-
"grad_norm": 0.004946811830026607,
|
| 210 |
-
"kl": 0.3302276611328125,
|
| 211 |
-
"learning_rate": 1.766044443118978e-05,
|
| 212 |
-
"loss": 0.0132,
|
| 213 |
-
"reward": 0.2981770907761529,
|
| 214 |
-
"reward_std": 0.2436295541934669,
|
| 215 |
-
"rewards/accuracy_reward": 0.2981770907761529,
|
| 216 |
-
"rewards/format_reward": 0.0,
|
| 217 |
-
"step": 160
|
| 218 |
-
},
|
| 219 |
-
{
|
| 220 |
-
"completion_length": 755.7635591506958,
|
| 221 |
-
"epoch": 0.9574093629003871,
|
| 222 |
-
"grad_norm": 0.005692446906633277,
|
| 223 |
-
"kl": 0.19274749755859374,
|
| 224 |
-
"learning_rate": 1.7220792504048227e-05,
|
| 225 |
-
"loss": 0.0077,
|
| 226 |
-
"reward": 0.41692709631752223,
|
| 227 |
-
"reward_std": 0.2269477566704154,
|
| 228 |
-
"rewards/accuracy_reward": 0.41692709631752223,
|
| 229 |
-
"rewards/format_reward": 0.0,
|
| 230 |
-
"step": 170
|
| 231 |
-
},
|
| 232 |
-
{
|
| 233 |
-
"completion_length": 770.9676110760059,
|
| 234 |
-
"epoch": 1.011263639563534,
|
| 235 |
-
"grad_norm": 0.0062556198464566865,
|
| 236 |
-
"kl": 0.243722554125817,
|
| 237 |
-
"learning_rate": 1.6749830015182106e-05,
|
| 238 |
-
"loss": 0.0093,
|
| 239 |
-
"reward": 0.33496732995206235,
|
| 240 |
-
"reward_std": 0.25141142401235556,
|
| 241 |
-
"rewards/accuracy_reward": 0.33496732995206235,
|
| 242 |
-
"rewards/format_reward": 0.0,
|
| 243 |
-
"step": 180
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"completion_length": 733.9075698852539,
|
| 247 |
-
"epoch": 1.0675818373812038,
|
| 248 |
-
"grad_norm": 0.013478379314182602,
|
| 249 |
-
"kl": 0.1884002685546875,
|
| 250 |
-
"learning_rate": 1.6249599136564837e-05,
|
| 251 |
-
"loss": 0.0075,
|
| 252 |
-
"reward": 0.4039062622701749,
|
| 253 |
-
"reward_std": 0.2399413853418082,
|
| 254 |
-
"rewards/accuracy_reward": 0.4039062622701749,
|
| 255 |
-
"rewards/format_reward": 0.0,
|
| 256 |
-
"step": 190
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"completion_length": 792.2987182617187,
|
| 260 |
-
"epoch": 1.1239000351988737,
|
| 261 |
-
"grad_norm": 0.011208080356335833,
|
| 262 |
-
"kl": 0.34430999755859376,
|
| 263 |
-
"learning_rate": 1.5722268952793806e-05,
|
| 264 |
-
"loss": 0.0138,
|
| 265 |
-
"reward": 0.33385417598765343,
|
| 266 |
-
"reward_std": 0.2561748518142849,
|
| 267 |
-
"rewards/accuracy_reward": 0.33385417598765343,
|
| 268 |
-
"rewards/format_reward": 0.0,
|
| 269 |
-
"step": 200
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"completion_length": 791.714602470398,
|
| 273 |
-
"epoch": 1.1802182330165434,
|
| 274 |
-
"grad_norm": 0.07302433070784739,
|
| 275 |
-
"kl": 0.28140411376953123,
|
| 276 |
-
"learning_rate": 1.517012605557746e-05,
|
| 277 |
-
"loss": 0.0113,
|
| 278 |
-
"reward": 0.38307292729150505,
|
| 279 |
-
"reward_std": 0.2577817846089602,
|
| 280 |
-
"rewards/accuracy_reward": 0.38307292729150505,
|
| 281 |
-
"rewards/format_reward": 0.0,
|
| 282 |
-
"step": 210
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"completion_length": 772.5690313339234,
|
| 286 |
-
"epoch": 1.2365364308342133,
|
| 287 |
-
"grad_norm": 0.0053274694242660915,
|
| 288 |
-
"kl": 0.15906524658203125,
|
| 289 |
-
"learning_rate": 1.4595564628691944e-05,
|
| 290 |
-
"loss": 0.0064,
|
| 291 |
-
"reward": 0.41510418029502033,
|
| 292 |
-
"reward_std": 0.2146047185175121,
|
| 293 |
-
"rewards/accuracy_reward": 0.41510418029502033,
|
| 294 |
-
"rewards/format_reward": 0.0,
|
| 295 |
-
"step": 220
|
| 296 |
-
},
|
| 297 |
-
{
|
| 298 |
-
"completion_length": 781.6310104370117,
|
| 299 |
-
"epoch": 1.2928546286518832,
|
| 300 |
-
"grad_norm": 0.004489343094465525,
|
| 301 |
-
"kl": 0.2008056640625,
|
| 302 |
-
"learning_rate": 1.400107606640056e-05,
|
| 303 |
-
"loss": 0.008,
|
| 304 |
-
"reward": 0.421354177640751,
|
| 305 |
-
"reward_std": 0.21437433324754238,
|
| 306 |
-
"rewards/accuracy_reward": 0.421354177640751,
|
| 307 |
-
"rewards/format_reward": 0.0,
|
| 308 |
-
"step": 230
|
| 309 |
-
},
|
| 310 |
-
{
|
| 311 |
-
"completion_length": 789.0828311920166,
|
| 312 |
-
"epoch": 1.349172826469553,
|
| 313 |
-
"grad_norm": 0.005360028572566901,
|
| 314 |
-
"kl": 0.24453582763671874,
|
| 315 |
-
"learning_rate": 1.3389238170352318e-05,
|
| 316 |
-
"loss": 0.0098,
|
| 317 |
-
"reward": 0.3609375107800588,
|
| 318 |
-
"reward_std": 0.23370700590312482,
|
| 319 |
-
"rewards/accuracy_reward": 0.3609375107800588,
|
| 320 |
-
"rewards/format_reward": 0.0,
|
| 321 |
-
"step": 240
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"completion_length": 772.9205938339234,
|
| 325 |
-
"epoch": 1.405491024287223,
|
| 326 |
-
"grad_norm": 0.006362180220231643,
|
| 327 |
-
"kl": 0.1741455078125,
|
| 328 |
-
"learning_rate": 1.2762703971803684e-05,
|
| 329 |
-
"loss": 0.007,
|
| 330 |
-
"reward": 0.40182292759418486,
|
| 331 |
-
"reward_std": 0.23207554304972292,
|
| 332 |
-
"rewards/accuracy_reward": 0.40182292759418486,
|
| 333 |
-
"rewards/format_reward": 0.0,
|
| 334 |
-
"step": 250
|
| 335 |
-
},
|
| 336 |
-
{
|
| 337 |
-
"completion_length": 781.4427295684815,
|
| 338 |
-
"epoch": 1.4618092221048926,
|
| 339 |
-
"grad_norm": 0.00462894748969255,
|
| 340 |
-
"kl": 0.22792816162109375,
|
| 341 |
-
"learning_rate": 1.2124190227632138e-05,
|
| 342 |
-
"loss": 0.0091,
|
| 343 |
-
"reward": 0.3945312611758709,
|
| 344 |
-
"reward_std": 0.2250285217538476,
|
| 345 |
-
"rewards/accuracy_reward": 0.3945312611758709,
|
| 346 |
-
"rewards/format_reward": 0.0,
|
| 347 |
-
"step": 260
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"completion_length": 771.0130414962769,
|
| 351 |
-
"epoch": 1.5181274199225625,
|
| 352 |
-
"grad_norm": 0.004564509677450662,
|
| 353 |
-
"kl": 0.18439483642578125,
|
| 354 |
-
"learning_rate": 1.1476465640024814e-05,
|
| 355 |
-
"loss": 0.0074,
|
| 356 |
-
"reward": 0.4356770958285779,
|
| 357 |
-
"reward_std": 0.24487592168152333,
|
| 358 |
-
"rewards/accuracy_reward": 0.4356770958285779,
|
| 359 |
-
"rewards/format_reward": 0.0,
|
| 360 |
-
"step": 270
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"completion_length": 760.0971549987793,
|
| 364 |
-
"epoch": 1.5744456177402322,
|
| 365 |
-
"grad_norm": 0.004478890033122169,
|
| 366 |
-
"kl": 0.15024261474609374,
|
| 367 |
-
"learning_rate": 1.0822338850923644e-05,
|
| 368 |
-
"loss": 0.006,
|
| 369 |
-
"reward": 0.44088542780373247,
|
| 370 |
-
"reward_std": 0.2205974128562957,
|
| 371 |
-
"rewards/accuracy_reward": 0.44088542780373247,
|
| 372 |
-
"rewards/format_reward": 0.0,
|
| 373 |
-
"step": 280
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"completion_length": 726.0036645889282,
|
| 377 |
-
"epoch": 1.630763815557902,
|
| 378 |
-
"grad_norm": 0.004052676061378306,
|
| 379 |
-
"kl": 0.14538421630859374,
|
| 380 |
-
"learning_rate": 1.0164646263284993e-05,
|
| 381 |
-
"loss": 0.0058,
|
| 382 |
-
"reward": 0.4242187628289685,
|
| 383 |
-
"reward_std": 0.21298313974402844,
|
| 384 |
-
"rewards/accuracy_reward": 0.4242187628289685,
|
| 385 |
-
"rewards/format_reward": 0.0,
|
| 386 |
-
"step": 290
|
| 387 |
-
},
|
| 388 |
-
{
|
| 389 |
-
"completion_length": 740.6461137771606,
|
| 390 |
-
"epoch": 1.687082013375572,
|
| 391 |
-
"grad_norm": 0.004367463252999239,
|
| 392 |
-
"kl": 0.150811767578125,
|
| 393 |
-
"learning_rate": 9.506239741962971e-06,
|
| 394 |
-
"loss": 0.006,
|
| 395 |
-
"reward": 0.43411459552589804,
|
| 396 |
-
"reward_std": 0.22852448388002813,
|
| 397 |
-
"rewards/accuracy_reward": 0.43411459552589804,
|
| 398 |
-
"rewards/format_reward": 0.0,
|
| 399 |
-
"step": 300
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"completion_length": 761.9489791870117,
|
| 403 |
-
"epoch": 1.7434002111932418,
|
| 404 |
-
"grad_norm": 0.005632246817382834,
|
| 405 |
-
"kl": 0.16129150390625,
|
| 406 |
-
"learning_rate": 8.849974247547307e-06,
|
| 407 |
-
"loss": 0.0065,
|
| 408 |
-
"reward": 0.41744792722165586,
|
| 409 |
-
"reward_std": 0.24217631085775793,
|
| 410 |
-
"rewards/accuracy_reward": 0.41744792722165586,
|
| 411 |
-
"rewards/format_reward": 0.0,
|
| 412 |
-
"step": 310
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"completion_length": 756.2075763702393,
|
| 416 |
-
"epoch": 1.7997184090109117,
|
| 417 |
-
"grad_norm": 0.004607559124840718,
|
| 418 |
-
"kl": 0.178369140625,
|
| 419 |
-
"learning_rate": 8.198695456777653e-06,
|
| 420 |
-
"loss": 0.0071,
|
| 421 |
-
"reward": 0.43723959578201177,
|
| 422 |
-
"reward_std": 0.25495934183709323,
|
| 423 |
-
"rewards/accuracy_reward": 0.43723959578201177,
|
| 424 |
-
"rewards/format_reward": 0.0,
|
| 425 |
-
"step": 320
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"completion_length": 762.5203329086304,
|
| 429 |
-
"epoch": 1.8560366068285816,
|
| 430 |
-
"grad_norm": 0.004599726892888303,
|
| 431 |
-
"kl": 0.20076904296875,
|
| 432 |
-
"learning_rate": 7.555227423214174e-06,
|
| 433 |
-
"loss": 0.008,
|
| 434 |
-
"reward": 0.40234376094304025,
|
| 435 |
-
"reward_std": 0.23350222799926995,
|
| 436 |
-
"rewards/accuracy_reward": 0.40234376094304025,
|
| 437 |
-
"rewards/format_reward": 0.0,
|
| 438 |
-
"step": 330
|
| 439 |
-
},
|
| 440 |
-
{
|
| 441 |
-
"completion_length": 769.9359592437744,
|
| 442 |
-
"epoch": 1.9123548046462513,
|
| 443 |
-
"grad_norm": 0.020436084881252883,
|
| 444 |
-
"kl": 0.20345611572265626,
|
| 445 |
-
"learning_rate": 6.92236033166988e-06,
|
| 446 |
-
"loss": 0.0081,
|
| 447 |
-
"reward": 0.4013020948739722,
|
| 448 |
-
"reward_std": 0.23616144466213881,
|
| 449 |
-
"rewards/accuracy_reward": 0.4013020948739722,
|
| 450 |
-
"rewards/format_reward": 0.0,
|
| 451 |
-
"step": 340
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"completion_length": 752.683092880249,
|
| 455 |
-
"epoch": 1.9686730024639212,
|
| 456 |
-
"grad_norm": 0.005642833494237804,
|
| 457 |
-
"kl": 0.17805633544921876,
|
| 458 |
-
"learning_rate": 6.302838399503477e-06,
|
| 459 |
-
"loss": 0.0071,
|
| 460 |
-
"reward": 0.41770834459457545,
|
| 461 |
-
"reward_std": 0.2381706119515002,
|
| 462 |
-
"rewards/accuracy_reward": 0.41770834459457545,
|
| 463 |
-
"rewards/format_reward": 0.0,
|
| 464 |
-
"step": 350
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"completion_length": 733.1225681678922,
|
| 468 |
-
"epoch": 2.022527279127068,
|
| 469 |
-
"grad_norm": 0.004337322419551108,
|
| 470 |
-
"kl": 0.15886693218954248,
|
| 471 |
-
"learning_rate": 5.699347977234799e-06,
|
| 472 |
-
"loss": 0.0061,
|
| 473 |
-
"reward": 0.45098040661975447,
|
| 474 |
-
"reward_std": 0.22561242217137142,
|
| 475 |
-
"rewards/accuracy_reward": 0.45098040661975447,
|
| 476 |
-
"rewards/format_reward": 0.0,
|
| 477 |
-
"step": 360
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"completion_length": 758.8823127746582,
|
| 481 |
-
"epoch": 2.0788454769447378,
|
| 482 |
-
"grad_norm": 0.007541831613575845,
|
| 483 |
-
"kl": 0.21207427978515625,
|
| 484 |
-
"learning_rate": 5.114505900080473e-06,
|
| 485 |
-
"loss": 0.0085,
|
| 486 |
-
"reward": 0.4036458440357819,
|
| 487 |
-
"reward_std": 0.23944172961637378,
|
| 488 |
-
"rewards/accuracy_reward": 0.4036458440357819,
|
| 489 |
-
"rewards/format_reward": 0.0,
|
| 490 |
-
"step": 370
|
| 491 |
-
},
|
| 492 |
-
{
|
| 493 |
-
"completion_length": 753.5510608673096,
|
| 494 |
-
"epoch": 2.1351636747624076,
|
| 495 |
-
"grad_norm": 0.006621490780654525,
|
| 496 |
-
"kl": 0.23095550537109374,
|
| 497 |
-
"learning_rate": 4.550848140919606e-06,
|
| 498 |
-
"loss": 0.0092,
|
| 499 |
-
"reward": 0.3807291773846373,
|
| 500 |
-
"reward_std": 0.2663448811043054,
|
| 501 |
-
"rewards/accuracy_reward": 0.3807291773846373,
|
| 502 |
-
"rewards/format_reward": 0.0,
|
| 503 |
-
"step": 380
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"completion_length": 758.0966333389282,
|
| 507 |
-
"epoch": 2.1914818725800775,
|
| 508 |
-
"grad_norm": 0.008553486900879023,
|
| 509 |
-
"kl": 0.2570648193359375,
|
| 510 |
-
"learning_rate": 4.010818813892e-06,
|
| 511 |
-
"loss": 0.0103,
|
| 512 |
-
"reward": 0.3971354293404147,
|
| 513 |
-
"reward_std": 0.2736111101694405,
|
| 514 |
-
"rewards/accuracy_reward": 0.3971354293404147,
|
| 515 |
-
"rewards/format_reward": 0.0,
|
| 516 |
-
"step": 390
|
| 517 |
-
},
|
| 518 |
-
{
|
| 519 |
-
"completion_length": 744.2885627746582,
|
| 520 |
-
"epoch": 2.2478000703977474,
|
| 521 |
-
"grad_norm": 0.00846416546487227,
|
| 522 |
-
"kl": 0.23549957275390626,
|
| 523 |
-
"learning_rate": 3.496759576311235e-06,
|
| 524 |
-
"loss": 0.0094,
|
| 525 |
-
"reward": 0.4242187612457201,
|
| 526 |
-
"reward_std": 0.25463975183665755,
|
| 527 |
-
"rewards/accuracy_reward": 0.4242187612457201,
|
| 528 |
-
"rewards/format_reward": 0.0,
|
| 529 |
-
"step": 400
|
| 530 |
-
},
|
| 531 |
-
{
|
| 532 |
-
"completion_length": 732.2635631561279,
|
| 533 |
-
"epoch": 2.304118268215417,
|
| 534 |
-
"grad_norm": 0.007949535937296182,
|
| 535 |
-
"kl": 0.21009674072265624,
|
| 536 |
-
"learning_rate": 3.010899474847655e-06,
|
| 537 |
-
"loss": 0.0084,
|
| 538 |
-
"reward": 0.46302084617782385,
|
| 539 |
-
"reward_std": 0.2583510285709053,
|
| 540 |
-
"rewards/accuracy_reward": 0.46302084617782385,
|
| 541 |
-
"rewards/format_reward": 0.0,
|
| 542 |
-
"step": 410
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"completion_length": 753.3565294265748,
|
| 546 |
-
"epoch": 2.3604364660330868,
|
| 547 |
-
"grad_norm": 0.0061687048404628015,
|
| 548 |
-
"kl": 0.20720977783203126,
|
| 549 |
-
"learning_rate": 2.555345280009872e-06,
|
| 550 |
-
"loss": 0.0083,
|
| 551 |
-
"reward": 0.4119791783392429,
|
| 552 |
-
"reward_std": 0.23550991672091187,
|
| 553 |
-
"rewards/accuracy_reward": 0.4119791783392429,
|
| 554 |
-
"rewards/format_reward": 0.0,
|
| 555 |
-
"step": 420
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"completion_length": 753.3632999420166,
|
| 559 |
-
"epoch": 2.4167546638507567,
|
| 560 |
-
"grad_norm": 0.005803641731146478,
|
| 561 |
-
"kl": 0.22764892578125,
|
| 562 |
-
"learning_rate": 2.132072350836164e-06,
|
| 563 |
-
"loss": 0.0091,
|
| 564 |
-
"reward": 0.4026041769655421,
|
| 565 |
-
"reward_std": 0.24279093905352056,
|
| 566 |
-
"rewards/accuracy_reward": 0.4026041769655421,
|
| 567 |
-
"rewards/format_reward": 0.0,
|
| 568 |
-
"step": 430
|
| 569 |
-
},
|
| 570 |
-
{
|
| 571 |
-
"completion_length": 767.0419466018677,
|
| 572 |
-
"epoch": 2.4730728616684265,
|
| 573 |
-
"grad_norm": 0.007478292393973801,
|
| 574 |
-
"kl": 0.2535888671875,
|
| 575 |
-
"learning_rate": 1.7429160694078983e-06,
|
| 576 |
-
"loss": 0.0101,
|
| 577 |
-
"reward": 0.3908854263834655,
|
| 578 |
-
"reward_std": 0.2595780567731708,
|
| 579 |
-
"rewards/accuracy_reward": 0.3908854263834655,
|
| 580 |
-
"rewards/format_reward": 0.0,
|
| 581 |
-
"step": 440
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"completion_length": 785.3596576690674,
|
| 585 |
-
"epoch": 2.5293910594860964,
|
| 586 |
-
"grad_norm": 0.0048942895216092,
|
| 587 |
-
"kl": 0.24837188720703124,
|
| 588 |
-
"learning_rate": 1.3895638823264447e-06,
|
| 589 |
-
"loss": 0.0099,
|
| 590 |
-
"reward": 0.36744792812969534,
|
| 591 |
-
"reward_std": 0.268617476709187,
|
| 592 |
-
"rewards/accuracy_reward": 0.36744792812969534,
|
| 593 |
-
"rewards/format_reward": 0.0,
|
| 594 |
-
"step": 450
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"completion_length": 775.6502796173096,
|
| 598 |
-
"epoch": 2.5857092573037663,
|
| 599 |
-
"grad_norm": 0.005204813465921492,
|
| 600 |
-
"kl": 0.232977294921875,
|
| 601 |
-
"learning_rate": 1.0735479836630136e-06,
|
| 602 |
-
"loss": 0.0093,
|
| 603 |
-
"reward": 0.38828126061707735,
|
| 604 |
-
"reward_std": 0.26048703310079874,
|
| 605 |
-
"rewards/accuracy_reward": 0.38828126061707735,
|
| 606 |
-
"rewards/format_reward": 0.0,
|
| 607 |
-
"step": 460
|
| 608 |
-
},
|
| 609 |
-
{
|
| 610 |
-
"completion_length": 757.7630428314209,
|
| 611 |
-
"epoch": 2.642027455121436,
|
| 612 |
-
"grad_norm": 0.004898162790105863,
|
| 613 |
-
"kl": 0.2490570068359375,
|
| 614 |
-
"learning_rate": 7.96238671109374e-07,
|
| 615 |
-
"loss": 0.01,
|
| 616 |
-
"reward": 0.4135416790144518,
|
| 617 |
-
"reward_std": 0.27739356658421455,
|
| 618 |
-
"rewards/accuracy_reward": 0.4135416790144518,
|
| 619 |
-
"rewards/format_reward": 0.0,
|
| 620 |
-
"step": 470
|
| 621 |
-
},
|
| 622 |
-
{
|
| 623 |
-
"completion_length": 747.8677282333374,
|
| 624 |
-
"epoch": 2.698345652939106,
|
| 625 |
-
"grad_norm": 0.004969150074366622,
|
| 626 |
-
"kl": 0.24514312744140626,
|
| 627 |
-
"learning_rate": 5.588384041383089e-07,
|
| 628 |
-
"loss": 0.0098,
|
| 629 |
-
"reward": 0.4010416786186397,
|
| 630 |
-
"reward_std": 0.25757726905867456,
|
| 631 |
-
"rewards/accuracy_reward": 0.4010416786186397,
|
| 632 |
-
"rewards/format_reward": 0.0,
|
| 633 |
-
"step": 480
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"completion_length": 762.9106948852539,
|
| 637 |
-
"epoch": 2.754663850756776,
|
| 638 |
-
"grad_norm": 0.006926745354414065,
|
| 639 |
-
"kl": 0.2514556884765625,
|
| 640 |
-
"learning_rate": 3.6237658993861114e-07,
|
| 641 |
-
"loss": 0.0101,
|
| 642 |
-
"reward": 0.3640625093365088,
|
| 643 |
-
"reward_std": 0.24269089489243925,
|
| 644 |
-
"rewards/accuracy_reward": 0.3640625093365088,
|
| 645 |
-
"rewards/format_reward": 0.0,
|
| 646 |
-
"step": 490
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"completion_length": 757.0679887771606,
|
| 650 |
-
"epoch": 2.810982048574446,
|
| 651 |
-
"grad_norm": 0.0057848948123537984,
|
| 652 |
-
"kl": 0.238916015625,
|
| 653 |
-
"learning_rate": 2.0770511973365436e-07,
|
| 654 |
-
"loss": 0.0096,
|
| 655 |
-
"reward": 0.37890625961590557,
|
| 656 |
-
"reward_std": 0.25779678178951143,
|
| 657 |
-
"rewards/accuracy_reward": 0.37890625961590557,
|
| 658 |
-
"rewards/format_reward": 0.0,
|
| 659 |
-
"step": 500
|
| 660 |
-
},
|
| 661 |
-
{
|
| 662 |
-
"completion_length": 759.939081954956,
|
| 663 |
-
"epoch": 2.8673002463921153,
|
| 664 |
-
"grad_norm": 0.005638420102813596,
|
| 665 |
-
"kl": 0.22593994140625,
|
| 666 |
-
"learning_rate": 9.549467483884412e-08,
|
| 667 |
-
"loss": 0.009,
|
| 668 |
-
"reward": 0.39687501240987333,
|
| 669 |
-
"reward_std": 0.2561142744962126,
|
| 670 |
-
"rewards/accuracy_reward": 0.39687501240987333,
|
| 671 |
-
"rewards/format_reward": 0.0,
|
| 672 |
-
"step": 510
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"completion_length": 753.216947555542,
|
| 676 |
-
"epoch": 2.9236184442097852,
|
| 677 |
-
"grad_norm": 0.005654304233128449,
|
| 678 |
-
"kl": 0.22020263671875,
|
| 679 |
-
"learning_rate": 2.6231818475468407e-08,
|
| 680 |
-
"loss": 0.0088,
|
| 681 |
-
"reward": 0.4106770950369537,
|
| 682 |
-
"reward_std": 0.26481629475019874,
|
| 683 |
-
"rewards/accuracy_reward": 0.4106770950369537,
|
| 684 |
-
"rewards/format_reward": 0.0,
|
| 685 |
-
"step": 520
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"completion_length": 753.779185295105,
|
| 689 |
-
"epoch": 2.979936642027455,
|
| 690 |
-
"grad_norm": 0.005106567352866245,
|
| 691 |
-
"kl": 0.2215423583984375,
|
| 692 |
-
"learning_rate": 2.1688595134516932e-10,
|
| 693 |
-
"loss": 0.0089,
|
| 694 |
-
"reward": 0.3789062601979822,
|
| 695 |
-
"reward_std": 0.23878278099000455,
|
| 696 |
-
"rewards/accuracy_reward": 0.3789062601979822,
|
| 697 |
-
"rewards/format_reward": 0.0,
|
| 698 |
-
"step": 530
|
| 699 |
-
},
|
| 700 |
-
{
|
| 701 |
-
"completion_length": 707.6640872955322,
|
| 702 |
-
"epoch": 2.985568461809222,
|
| 703 |
-
"kl": 0.210693359375,
|
| 704 |
-
"reward": 0.5104166846722364,
|
| 705 |
-
"reward_std": 0.3139621028676629,
|
| 706 |
-
"rewards/accuracy_reward": 0.5104166846722364,
|
| 707 |
-
"rewards/format_reward": 0.0,
|
| 708 |
-
"step": 531,
|
| 709 |
"total_flos": 0.0,
|
| 710 |
-
"train_loss":
|
| 711 |
-
"train_runtime":
|
| 712 |
-
"train_samples_per_second": 0.
|
| 713 |
-
"train_steps_per_second": 0.
|
| 714 |
}
|
| 715 |
],
|
| 716 |
"logging_steps": 10,
|
| 717 |
-
"max_steps":
|
| 718 |
"num_input_tokens_seen": 0,
|
| 719 |
"num_train_epochs": 3,
|
| 720 |
"save_steps": 500,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 3,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"completion_length": 888.5694444444445,
|
| 13 |
+
"epoch": 3.0,
|
| 14 |
+
"kl": 0.0,
|
| 15 |
+
"reward": 0.34375,
|
| 16 |
+
"reward_std": 0.36623653107219273,
|
| 17 |
+
"rewards/accuracy_reward": 0.34375,
|
|
|
|
|
|
|
|
|
|
| 18 |
"rewards/format_reward": 0.0,
|
| 19 |
+
"step": 3,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
"total_flos": 0.0,
|
| 21 |
+
"train_loss": -7.761021455128988e-10,
|
| 22 |
+
"train_runtime": 74.9061,
|
| 23 |
+
"train_samples_per_second": 0.401,
|
| 24 |
+
"train_steps_per_second": 0.04
|
| 25 |
}
|
| 26 |
],
|
| 27 |
"logging_steps": 10,
|
| 28 |
+
"max_steps": 3,
|
| 29 |
"num_input_tokens_seen": 0,
|
| 30 |
"num_train_epochs": 3,
|
| 31 |
"save_steps": 500,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95a7bdf9bb202c53a98818044b703a04b0b5ab1785dfb042611f590d9ac11794
|
| 3 |
size 7032
|