Jan Majkutewicz
commited on
Model save
Browse files- README.md +3 -16
- adapter_model.safetensors +1 -1
- all_results.json +3 -16
- train_results.json +3 -3
- trainer_state.json +959 -959
README.md
CHANGED
|
@@ -1,11 +1,8 @@
|
|
| 1 |
---
|
| 2 |
base_model: alignment-handbook/zephyr-7b-sft-full
|
| 3 |
-
datasets:
|
| 4 |
-
- SeniorKabanocci/oasst1-binarized
|
| 5 |
library_name: peft
|
| 6 |
license: apache-2.0
|
| 7 |
tags:
|
| 8 |
-
- alignment-handbook
|
| 9 |
- trl
|
| 10 |
- dpo
|
| 11 |
- generated_from_trainer
|
|
@@ -17,20 +14,10 @@ model-index:
|
|
| 17 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 18 |
should probably proofread and complete it, then remove this comment. -->
|
| 19 |
|
| 20 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/jan-majkutewicz/huggingface/runs/
|
| 21 |
# zephyr-7b-oasst1
|
| 22 |
|
| 23 |
-
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on
|
| 24 |
-
It achieves the following results on the evaluation set:
|
| 25 |
-
- Loss: 0.6059
|
| 26 |
-
- Rewards/chosen: -0.9430
|
| 27 |
-
- Rewards/rejected: -1.3459
|
| 28 |
-
- Rewards/accuracies: 0.6348
|
| 29 |
-
- Rewards/margins: 0.4030
|
| 30 |
-
- Logps/rejected: -239.4801
|
| 31 |
-
- Logps/chosen: -256.8972
|
| 32 |
-
- Logits/rejected: -2.8973
|
| 33 |
-
- Logits/chosen: -2.9273
|
| 34 |
|
| 35 |
## Model description
|
| 36 |
|
|
@@ -49,7 +36,7 @@ More information needed
|
|
| 49 |
### Training hyperparameters
|
| 50 |
|
| 51 |
The following hyperparameters were used during training:
|
| 52 |
-
- learning_rate:
|
| 53 |
- train_batch_size: 8
|
| 54 |
- eval_batch_size: 8
|
| 55 |
- seed: 42
|
|
|
|
| 1 |
---
|
| 2 |
base_model: alignment-handbook/zephyr-7b-sft-full
|
|
|
|
|
|
|
| 3 |
library_name: peft
|
| 4 |
license: apache-2.0
|
| 5 |
tags:
|
|
|
|
| 6 |
- trl
|
| 7 |
- dpo
|
| 8 |
- generated_from_trainer
|
|
|
|
| 14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
|
| 17 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/jan-majkutewicz/huggingface/runs/vp4emkhb)
|
| 18 |
# zephyr-7b-oasst1
|
| 19 |
|
| 20 |
+
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on an unknown dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 36 |
### Training hyperparameters
|
| 37 |
|
| 38 |
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 1e-06
|
| 40 |
- train_batch_size: 8
|
| 41 |
- eval_batch_size: 8
|
| 42 |
- seed: 42
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 335605144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73157727aa31d029e720b3348bbf61b47cd9d6ed4e229cd5281fd4749b57e865
|
| 3 |
size 335605144
|
all_results.json
CHANGED
|
@@ -1,22 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9994353472614342,
|
| 3 |
-
"eval_logits/chosen": -2.9273033142089844,
|
| 4 |
-
"eval_logits/rejected": -2.8973140716552734,
|
| 5 |
-
"eval_logps/chosen": -256.8972473144531,
|
| 6 |
-
"eval_logps/rejected": -239.4800567626953,
|
| 7 |
-
"eval_loss": 0.6058722734451294,
|
| 8 |
-
"eval_rewards/accuracies": 0.6348314881324768,
|
| 9 |
-
"eval_rewards/chosen": -0.9429696798324585,
|
| 10 |
-
"eval_rewards/margins": 0.4029598534107208,
|
| 11 |
-
"eval_rewards/rejected": -1.3459293842315674,
|
| 12 |
-
"eval_runtime": 230.8014,
|
| 13 |
-
"eval_samples": 712,
|
| 14 |
-
"eval_samples_per_second": 3.085,
|
| 15 |
-
"eval_steps_per_second": 0.386,
|
| 16 |
"total_flos": 0.0,
|
| 17 |
-
"train_loss": 0.
|
| 18 |
-
"train_runtime":
|
| 19 |
"train_samples": 14167,
|
| 20 |
-
"train_samples_per_second": 1.
|
| 21 |
"train_steps_per_second": 0.089
|
| 22 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9994353472614342,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.6718437324135991,
|
| 5 |
+
"train_runtime": 9969.8232,
|
| 6 |
"train_samples": 14167,
|
| 7 |
+
"train_samples_per_second": 1.421,
|
| 8 |
"train_steps_per_second": 0.089
|
| 9 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9994353472614342,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 14167,
|
| 7 |
-
"train_samples_per_second": 1.
|
| 8 |
"train_steps_per_second": 0.089
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9994353472614342,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.6718437324135991,
|
| 5 |
+
"train_runtime": 9969.8232,
|
| 6 |
"train_samples": 14167,
|
| 7 |
+
"train_samples_per_second": 1.421,
|
| 8 |
"train_steps_per_second": 0.089
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -11,7 +11,7 @@
|
|
| 11 |
{
|
| 12 |
"epoch": 0.001129305477131564,
|
| 13 |
"grad_norm": 10.5625,
|
| 14 |
-
"learning_rate":
|
| 15 |
"logits/chosen": -2.8912107944488525,
|
| 16 |
"logits/rejected": -2.9116690158843994,
|
| 17 |
"logps/chosen": -205.43255615234375,
|
|
@@ -25,1331 +25,1331 @@
|
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.01129305477131564,
|
| 28 |
-
"grad_norm": 11.
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"logits/chosen": -2.
|
| 31 |
-
"logits/rejected": -2.
|
| 32 |
-
"logps/chosen": -286.
|
| 33 |
-
"logps/rejected": -234.
|
| 34 |
-
"loss": 0.
|
| 35 |
-
"rewards/accuracies": 0.
|
| 36 |
-
"rewards/chosen": 0.
|
| 37 |
-
"rewards/margins": 0.
|
| 38 |
-
"rewards/rejected": 0.
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.02258610954263128,
|
| 43 |
-
"grad_norm": 8.
|
| 44 |
-
"learning_rate":
|
| 45 |
-
"logits/chosen": -2.
|
| 46 |
-
"logits/rejected": -2.
|
| 47 |
-
"logps/chosen": -242.
|
| 48 |
-
"logps/rejected": -230.
|
| 49 |
-
"loss": 0.
|
| 50 |
-
"rewards/accuracies": 0.
|
| 51 |
-
"rewards/chosen": -0.
|
| 52 |
-
"rewards/margins": -0.
|
| 53 |
-
"rewards/rejected": 0.
|
| 54 |
"step": 20
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.03387916431394692,
|
| 58 |
"grad_norm": 10.4375,
|
| 59 |
-
"learning_rate":
|
| 60 |
-
"logits/chosen": -2.
|
| 61 |
-
"logits/rejected": -2.
|
| 62 |
-
"logps/chosen": -
|
| 63 |
-
"logps/rejected": -206.
|
| 64 |
-
"loss": 0.
|
| 65 |
-
"rewards/accuracies": 0.
|
| 66 |
-
"rewards/chosen":
|
| 67 |
-
"rewards/margins": 0.
|
| 68 |
-
"rewards/rejected": -0.
|
| 69 |
"step": 30
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.04517221908526256,
|
| 73 |
-
"grad_norm": 10.
|
| 74 |
-
"learning_rate":
|
| 75 |
-
"logits/chosen": -2.
|
| 76 |
-
"logits/rejected": -2.
|
| 77 |
-
"logps/chosen": -239.
|
| 78 |
-
"logps/rejected": -223.
|
| 79 |
-
"loss": 0.
|
| 80 |
-
"rewards/accuracies": 0.
|
| 81 |
-
"rewards/chosen": -0.
|
| 82 |
-
"rewards/margins": 0.
|
| 83 |
-
"rewards/rejected":
|
| 84 |
"step": 40
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"epoch": 0.05646527385657821,
|
| 88 |
"grad_norm": 11.25,
|
| 89 |
-
"learning_rate":
|
| 90 |
-
"logits/chosen": -2.
|
| 91 |
-
"logits/rejected": -2.
|
| 92 |
-
"logps/chosen": -252.
|
| 93 |
-
"logps/rejected": -
|
| 94 |
-
"loss": 0.
|
| 95 |
-
"rewards/accuracies": 0.
|
| 96 |
-
"rewards/chosen": -0.
|
| 97 |
-
"rewards/margins": 0.
|
| 98 |
-
"rewards/rejected": -0.
|
| 99 |
"step": 50
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.06775832862789384,
|
| 103 |
-
"grad_norm": 9.
|
| 104 |
-
"learning_rate":
|
| 105 |
-
"logits/chosen": -2.
|
| 106 |
-
"logits/rejected": -2.
|
| 107 |
-
"logps/chosen": -
|
| 108 |
-
"logps/rejected": -
|
| 109 |
-
"loss": 0.
|
| 110 |
-
"rewards/accuracies": 0.
|
| 111 |
-
"rewards/chosen": -0.
|
| 112 |
-
"rewards/margins": 0.
|
| 113 |
-
"rewards/rejected": -0.
|
| 114 |
"step": 60
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.07905138339920949,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"logits/chosen": -2.
|
| 121 |
-
"logits/rejected": -2.
|
| 122 |
-
"logps/chosen": -249.
|
| 123 |
-
"logps/rejected": -
|
| 124 |
-
"loss": 0.
|
| 125 |
-
"rewards/accuracies": 0.
|
| 126 |
-
"rewards/chosen": -0.
|
| 127 |
-
"rewards/margins": 0.
|
| 128 |
-
"rewards/rejected": -0.
|
| 129 |
"step": 70
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.09034443817052512,
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"logits/chosen": -2.
|
| 136 |
-
"logits/rejected": -2.
|
| 137 |
-
"logps/chosen": -244.
|
| 138 |
-
"logps/rejected": -
|
| 139 |
-
"loss": 0.
|
| 140 |
-
"rewards/accuracies": 0.
|
| 141 |
-
"rewards/chosen": -0.
|
| 142 |
-
"rewards/margins": 0.
|
| 143 |
-
"rewards/rejected": -0.
|
| 144 |
"step": 80
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.10163749294184077,
|
| 148 |
-
"grad_norm": 6.
|
| 149 |
-
"learning_rate":
|
| 150 |
-
"logits/chosen": -2.
|
| 151 |
-
"logits/rejected": -2.
|
| 152 |
-
"logps/chosen": -
|
| 153 |
-
"logps/rejected": -
|
| 154 |
-
"loss": 0.
|
| 155 |
-
"rewards/accuracies": 0.
|
| 156 |
-
"rewards/chosen": -0.
|
| 157 |
-
"rewards/margins": 0.
|
| 158 |
-
"rewards/rejected": -0.
|
| 159 |
"step": 90
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.11293054771315642,
|
| 163 |
-
"grad_norm": 8.
|
| 164 |
-
"learning_rate":
|
| 165 |
-
"logits/chosen": -2.
|
| 166 |
-
"logits/rejected": -2.
|
| 167 |
-
"logps/chosen": -
|
| 168 |
-
"logps/rejected": -
|
| 169 |
-
"loss": 0.
|
| 170 |
-
"rewards/accuracies": 0.
|
| 171 |
-
"rewards/chosen": -0.
|
| 172 |
-
"rewards/margins": 0.
|
| 173 |
-
"rewards/rejected": -0.
|
| 174 |
"step": 100
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.12422360248447205,
|
| 178 |
-
"grad_norm":
|
| 179 |
-
"learning_rate":
|
| 180 |
-
"logits/chosen": -2.
|
| 181 |
-
"logits/rejected": -2.
|
| 182 |
-
"logps/chosen": -
|
| 183 |
-
"logps/rejected": -
|
| 184 |
-
"loss": 0.
|
| 185 |
-
"rewards/accuracies": 0.
|
| 186 |
-
"rewards/chosen": -0.
|
| 187 |
-
"rewards/margins": 0.
|
| 188 |
-
"rewards/rejected": -0.
|
| 189 |
"step": 110
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.13551665725578768,
|
| 193 |
-
"grad_norm": 8.
|
| 194 |
-
"learning_rate":
|
| 195 |
-
"logits/chosen": -2.
|
| 196 |
-
"logits/rejected": -2.
|
| 197 |
-
"logps/chosen": -
|
| 198 |
-
"logps/rejected": -
|
| 199 |
-
"loss": 0.
|
| 200 |
-
"rewards/accuracies": 0.
|
| 201 |
-
"rewards/chosen": -0.
|
| 202 |
-
"rewards/margins": 0.
|
| 203 |
-
"rewards/rejected": -0.
|
| 204 |
"step": 120
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 0.14680971202710333,
|
| 208 |
-
"grad_norm":
|
| 209 |
-
"learning_rate":
|
| 210 |
-
"logits/chosen": -2.
|
| 211 |
-
"logits/rejected": -2.
|
| 212 |
-
"logps/chosen": -
|
| 213 |
-
"logps/rejected": -
|
| 214 |
-
"loss": 0.
|
| 215 |
-
"rewards/accuracies": 0.
|
| 216 |
-
"rewards/chosen": -0.
|
| 217 |
-
"rewards/margins": 0.
|
| 218 |
-
"rewards/rejected": -0.
|
| 219 |
"step": 130
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.15810276679841898,
|
| 223 |
-
"grad_norm": 7.
|
| 224 |
-
"learning_rate":
|
| 225 |
-
"logits/chosen": -2.
|
| 226 |
-
"logits/rejected": -2.
|
| 227 |
-
"logps/chosen": -
|
| 228 |
-
"logps/rejected": -
|
| 229 |
-
"loss": 0.
|
| 230 |
-
"rewards/accuracies": 0.
|
| 231 |
-
"rewards/chosen": -0.
|
| 232 |
-
"rewards/margins": 0.
|
| 233 |
-
"rewards/rejected": -0.
|
| 234 |
"step": 140
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.16939582156973462,
|
| 238 |
-
"grad_norm": 10.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"logits/chosen": -2.
|
| 241 |
-
"logits/rejected": -2.
|
| 242 |
-
"logps/chosen": -
|
| 243 |
-
"logps/rejected": -
|
| 244 |
-
"loss": 0.
|
| 245 |
-
"rewards/accuracies": 0.
|
| 246 |
-
"rewards/chosen": -0.
|
| 247 |
-
"rewards/margins": 0.
|
| 248 |
-
"rewards/rejected": -0.
|
| 249 |
"step": 150
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.18068887634105024,
|
| 253 |
-
"grad_norm":
|
| 254 |
-
"learning_rate":
|
| 255 |
-
"logits/chosen": -2.
|
| 256 |
-
"logits/rejected": -2.
|
| 257 |
-
"logps/chosen": -
|
| 258 |
-
"logps/rejected": -
|
| 259 |
-
"loss": 0.
|
| 260 |
-
"rewards/accuracies": 0.
|
| 261 |
-
"rewards/chosen": -0.
|
| 262 |
-
"rewards/margins": 0.
|
| 263 |
-
"rewards/rejected": -0.
|
| 264 |
"step": 160
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 0.1919819311123659,
|
| 268 |
-
"grad_norm":
|
| 269 |
-
"learning_rate":
|
| 270 |
-
"logits/chosen": -2.
|
| 271 |
-
"logits/rejected": -2.
|
| 272 |
-
"logps/chosen": -
|
| 273 |
-
"logps/rejected": -
|
| 274 |
-
"loss": 0.
|
| 275 |
"rewards/accuracies": 0.581250011920929,
|
| 276 |
-
"rewards/chosen": -0.
|
| 277 |
-
"rewards/margins": 0.
|
| 278 |
-
"rewards/rejected": -0.
|
| 279 |
"step": 170
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 0.20327498588368154,
|
| 283 |
-
"grad_norm":
|
| 284 |
-
"learning_rate":
|
| 285 |
-
"logits/chosen": -2.
|
| 286 |
-
"logits/rejected": -2.
|
| 287 |
-
"logps/chosen": -
|
| 288 |
-
"logps/rejected": -
|
| 289 |
-
"loss": 0.
|
| 290 |
-
"rewards/accuracies": 0.
|
| 291 |
-
"rewards/chosen": -0.
|
| 292 |
-
"rewards/margins": 0.
|
| 293 |
-
"rewards/rejected": -0.
|
| 294 |
"step": 180
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 0.21456804065499718,
|
| 298 |
-
"grad_norm":
|
| 299 |
-
"learning_rate":
|
| 300 |
-
"logits/chosen": -2.
|
| 301 |
-
"logits/rejected": -2.
|
| 302 |
-
"logps/chosen": -
|
| 303 |
-
"logps/rejected": -
|
| 304 |
-
"loss": 0.
|
| 305 |
-
"rewards/accuracies": 0.
|
| 306 |
-
"rewards/chosen": -0.
|
| 307 |
-
"rewards/margins": 0.
|
| 308 |
-
"rewards/rejected": -0.
|
| 309 |
"step": 190
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 0.22586109542631283,
|
| 313 |
-
"grad_norm":
|
| 314 |
-
"learning_rate":
|
| 315 |
-
"logits/chosen": -2.
|
| 316 |
-
"logits/rejected": -2.
|
| 317 |
-
"logps/chosen": -
|
| 318 |
-
"logps/rejected": -
|
| 319 |
-
"loss": 0.
|
| 320 |
-
"rewards/accuracies": 0.
|
| 321 |
-
"rewards/chosen": -0.
|
| 322 |
-
"rewards/margins": 0.
|
| 323 |
-
"rewards/rejected": -0.
|
| 324 |
"step": 200
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.23715415019762845,
|
| 328 |
-
"grad_norm":
|
| 329 |
-
"learning_rate":
|
| 330 |
-
"logits/chosen": -2.
|
| 331 |
-
"logits/rejected": -2.
|
| 332 |
-
"logps/chosen": -
|
| 333 |
-
"logps/rejected": -
|
| 334 |
-
"loss": 0.
|
| 335 |
-
"rewards/accuracies": 0.
|
| 336 |
-
"rewards/chosen": -0.
|
| 337 |
-
"rewards/margins": 0.
|
| 338 |
-
"rewards/rejected": -0.
|
| 339 |
"step": 210
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.2484472049689441,
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"logits/chosen": -2.
|
| 346 |
-
"logits/rejected": -2.
|
| 347 |
-
"logps/chosen": -
|
| 348 |
-
"logps/rejected": -
|
| 349 |
-
"loss": 0.
|
| 350 |
-
"rewards/accuracies": 0.
|
| 351 |
-
"rewards/chosen": -0.
|
| 352 |
-
"rewards/margins": 0.
|
| 353 |
-
"rewards/rejected": -0.
|
| 354 |
"step": 220
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.2597402597402597,
|
| 358 |
-
"grad_norm": 10.
|
| 359 |
-
"learning_rate":
|
| 360 |
-
"logits/chosen": -2.
|
| 361 |
-
"logits/rejected": -2.
|
| 362 |
-
"logps/chosen": -
|
| 363 |
-
"logps/rejected": -
|
| 364 |
-
"loss": 0.
|
| 365 |
-
"rewards/accuracies": 0.
|
| 366 |
-
"rewards/chosen": -0.
|
| 367 |
-
"rewards/margins": 0.
|
| 368 |
-
"rewards/rejected": -0.
|
| 369 |
"step": 230
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 0.27103331451157536,
|
| 373 |
-
"grad_norm":
|
| 374 |
-
"learning_rate":
|
| 375 |
-
"logits/chosen": -2.
|
| 376 |
-
"logits/rejected": -2.
|
| 377 |
-
"logps/chosen": -
|
| 378 |
-
"logps/rejected": -
|
| 379 |
-
"loss": 0.
|
| 380 |
-
"rewards/accuracies": 0.
|
| 381 |
-
"rewards/chosen": -0.
|
| 382 |
-
"rewards/margins": 0.
|
| 383 |
-
"rewards/rejected": -0.
|
| 384 |
"step": 240
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 0.282326369282891,
|
| 388 |
-
"grad_norm": 8.
|
| 389 |
-
"learning_rate":
|
| 390 |
-
"logits/chosen": -2.
|
| 391 |
-
"logits/rejected": -2.
|
| 392 |
-
"logps/chosen": -
|
| 393 |
-
"logps/rejected": -
|
| 394 |
-
"loss": 0.
|
| 395 |
-
"rewards/accuracies": 0.
|
| 396 |
-
"rewards/chosen": -0.
|
| 397 |
-
"rewards/margins": 0.
|
| 398 |
-
"rewards/rejected": -
|
| 399 |
"step": 250
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 0.29361942405420666,
|
| 403 |
-
"grad_norm":
|
| 404 |
-
"learning_rate":
|
| 405 |
-
"logits/chosen": -2.
|
| 406 |
-
"logits/rejected": -2.
|
| 407 |
-
"logps/chosen": -
|
| 408 |
-
"logps/rejected": -
|
| 409 |
-
"loss": 0.
|
| 410 |
"rewards/accuracies": 0.625,
|
| 411 |
-
"rewards/chosen": -0.
|
| 412 |
-
"rewards/margins": 0.
|
| 413 |
-
"rewards/rejected": -0.
|
| 414 |
"step": 260
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 0.3049124788255223,
|
| 418 |
-
"grad_norm": 9.
|
| 419 |
-
"learning_rate":
|
| 420 |
-
"logits/chosen": -2.
|
| 421 |
-
"logits/rejected": -2.
|
| 422 |
-
"logps/chosen": -
|
| 423 |
-
"logps/rejected": -
|
| 424 |
-
"loss": 0.
|
| 425 |
-
"rewards/accuracies": 0.
|
| 426 |
-
"rewards/chosen": -0.
|
| 427 |
-
"rewards/margins": 0.
|
| 428 |
-
"rewards/rejected": -0.
|
| 429 |
"step": 270
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.31620553359683795,
|
| 433 |
-
"grad_norm":
|
| 434 |
-
"learning_rate":
|
| 435 |
-
"logits/chosen": -2.
|
| 436 |
-
"logits/rejected": -2.
|
| 437 |
-
"logps/chosen": -
|
| 438 |
-
"logps/rejected": -
|
| 439 |
-
"loss": 0.
|
| 440 |
-
"rewards/accuracies": 0.
|
| 441 |
-
"rewards/chosen": -0.
|
| 442 |
-
"rewards/margins": 0.
|
| 443 |
-
"rewards/rejected": -0.
|
| 444 |
"step": 280
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.3274985883681536,
|
| 448 |
-
"grad_norm":
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"logits/chosen": -2.
|
| 451 |
-
"logits/rejected": -2.
|
| 452 |
-
"logps/chosen": -
|
| 453 |
-
"logps/rejected": -
|
| 454 |
-
"loss": 0.
|
| 455 |
-
"rewards/accuracies": 0.
|
| 456 |
-
"rewards/chosen": -0.
|
| 457 |
-
"rewards/margins": 0.
|
| 458 |
-
"rewards/rejected": -0.
|
| 459 |
"step": 290
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.33879164313946925,
|
| 463 |
-
"grad_norm": 11.
|
| 464 |
-
"learning_rate":
|
| 465 |
-
"logits/chosen": -2.
|
| 466 |
-
"logits/rejected": -2.
|
| 467 |
-
"logps/chosen": -
|
| 468 |
-
"logps/rejected": -
|
| 469 |
-
"loss": 0.
|
| 470 |
-
"rewards/accuracies": 0.
|
| 471 |
-
"rewards/chosen": -0.
|
| 472 |
-
"rewards/margins": 0.
|
| 473 |
-
"rewards/rejected": -0.
|
| 474 |
"step": 300
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.3500846979107849,
|
| 478 |
-
"grad_norm":
|
| 479 |
-
"learning_rate":
|
| 480 |
-
"logits/chosen": -2.
|
| 481 |
-
"logits/rejected": -2.
|
| 482 |
-
"logps/chosen": -
|
| 483 |
-
"logps/rejected": -
|
| 484 |
-
"loss": 0.
|
| 485 |
-
"rewards/accuracies": 0.
|
| 486 |
-
"rewards/chosen": -0.
|
| 487 |
-
"rewards/margins": 0.
|
| 488 |
-
"rewards/rejected": -0.
|
| 489 |
"step": 310
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.3613777526821005,
|
| 493 |
-
"grad_norm": 9.
|
| 494 |
-
"learning_rate":
|
| 495 |
-
"logits/chosen": -2.
|
| 496 |
-
"logits/rejected": -2.
|
| 497 |
-
"logps/chosen": -
|
| 498 |
-
"logps/rejected": -
|
| 499 |
-
"loss": 0.
|
| 500 |
-
"rewards/accuracies": 0.
|
| 501 |
-
"rewards/chosen": -0.
|
| 502 |
-
"rewards/margins": 0.
|
| 503 |
-
"rewards/rejected": -0.
|
| 504 |
"step": 320
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 0.37267080745341613,
|
| 508 |
-
"grad_norm": 9.
|
| 509 |
-
"learning_rate":
|
| 510 |
-
"logits/chosen": -2.
|
| 511 |
-
"logits/rejected": -2.
|
| 512 |
-
"logps/chosen": -
|
| 513 |
-
"logps/rejected": -
|
| 514 |
-
"loss": 0.
|
| 515 |
-
"rewards/accuracies": 0.
|
| 516 |
-
"rewards/chosen": -0.
|
| 517 |
-
"rewards/margins": 0.
|
| 518 |
-
"rewards/rejected": -0.
|
| 519 |
"step": 330
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.3839638622247318,
|
| 523 |
-
"grad_norm":
|
| 524 |
-
"learning_rate":
|
| 525 |
-
"logits/chosen": -2.
|
| 526 |
-
"logits/rejected": -2.
|
| 527 |
-
"logps/chosen": -
|
| 528 |
-
"logps/rejected": -
|
| 529 |
-
"loss": 0.
|
| 530 |
-
"rewards/accuracies": 0.
|
| 531 |
-
"rewards/chosen": -0.
|
| 532 |
-
"rewards/margins": 0.
|
| 533 |
-
"rewards/rejected": -
|
| 534 |
"step": 340
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3952569169960474,
|
| 538 |
-
"grad_norm":
|
| 539 |
-
"learning_rate":
|
| 540 |
-
"logits/chosen": -2.
|
| 541 |
-
"logits/rejected": -2.
|
| 542 |
-
"logps/chosen": -
|
| 543 |
-
"logps/rejected": -
|
| 544 |
-
"loss": 0.
|
| 545 |
-
"rewards/accuracies": 0.
|
| 546 |
-
"rewards/chosen": -0.
|
| 547 |
-
"rewards/margins": 0.
|
| 548 |
-
"rewards/rejected": -
|
| 549 |
"step": 350
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.40654997176736307,
|
| 553 |
-
"grad_norm":
|
| 554 |
-
"learning_rate":
|
| 555 |
-
"logits/chosen": -2.
|
| 556 |
-
"logits/rejected": -2.
|
| 557 |
-
"logps/chosen": -
|
| 558 |
-
"logps/rejected": -
|
| 559 |
-
"loss": 0.
|
| 560 |
-
"rewards/accuracies": 0.
|
| 561 |
-
"rewards/chosen": -0.
|
| 562 |
-
"rewards/margins": 0.
|
| 563 |
-
"rewards/rejected": -
|
| 564 |
"step": 360
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 0.4178430265386787,
|
| 568 |
-
"grad_norm": 10.
|
| 569 |
-
"learning_rate":
|
| 570 |
-
"logits/chosen": -2.
|
| 571 |
-
"logits/rejected": -2.
|
| 572 |
-
"logps/chosen": -
|
| 573 |
-
"logps/rejected": -
|
| 574 |
-
"loss": 0.
|
| 575 |
-
"rewards/accuracies": 0.
|
| 576 |
-
"rewards/chosen": -
|
| 577 |
-
"rewards/margins": 0.
|
| 578 |
-
"rewards/rejected": -
|
| 579 |
"step": 370
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.42913608130999437,
|
| 583 |
-
"grad_norm":
|
| 584 |
-
"learning_rate":
|
| 585 |
-
"logits/chosen": -2.
|
| 586 |
-
"logits/rejected": -2.
|
| 587 |
-
"logps/chosen": -
|
| 588 |
-
"logps/rejected": -
|
| 589 |
-
"loss": 0.
|
| 590 |
"rewards/accuracies": 0.59375,
|
| 591 |
-
"rewards/chosen": -0.
|
| 592 |
-
"rewards/margins": 0.
|
| 593 |
-
"rewards/rejected": -
|
| 594 |
"step": 380
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 0.44042913608131,
|
| 598 |
-
"grad_norm":
|
| 599 |
-
"learning_rate":
|
| 600 |
-
"logits/chosen": -2.
|
| 601 |
-
"logits/rejected": -2.
|
| 602 |
-
"logps/chosen": -
|
| 603 |
-
"logps/rejected": -
|
| 604 |
-
"loss": 0.
|
| 605 |
-
"rewards/accuracies": 0.
|
| 606 |
-
"rewards/chosen": -0.
|
| 607 |
-
"rewards/margins": 0.
|
| 608 |
-
"rewards/rejected": -
|
| 609 |
"step": 390
|
| 610 |
},
|
| 611 |
{
|
| 612 |
"epoch": 0.45172219085262566,
|
| 613 |
-
"grad_norm":
|
| 614 |
-
"learning_rate":
|
| 615 |
-
"logits/chosen": -2.
|
| 616 |
-
"logits/rejected": -2.
|
| 617 |
-
"logps/chosen": -
|
| 618 |
-
"logps/rejected": -
|
| 619 |
-
"loss": 0.
|
| 620 |
-
"rewards/accuracies": 0.
|
| 621 |
-
"rewards/chosen": -0.
|
| 622 |
-
"rewards/margins": 0.
|
| 623 |
-
"rewards/rejected": -
|
| 624 |
"step": 400
|
| 625 |
},
|
| 626 |
{
|
| 627 |
"epoch": 0.46301524562394125,
|
| 628 |
-
"grad_norm":
|
| 629 |
-
"learning_rate":
|
| 630 |
-
"logits/chosen": -2.
|
| 631 |
-
"logits/rejected": -2.
|
| 632 |
-
"logps/chosen": -
|
| 633 |
-
"logps/rejected": -
|
| 634 |
-
"loss": 0.
|
| 635 |
-
"rewards/accuracies": 0.
|
| 636 |
-
"rewards/chosen": -
|
| 637 |
-
"rewards/margins": 0.
|
| 638 |
-
"rewards/rejected": -
|
| 639 |
"step": 410
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.4743083003952569,
|
| 643 |
-
"grad_norm":
|
| 644 |
-
"learning_rate":
|
| 645 |
-
"logits/chosen": -2.
|
| 646 |
-
"logits/rejected": -2.
|
| 647 |
-
"logps/chosen": -
|
| 648 |
-
"logps/rejected": -
|
| 649 |
-
"loss": 0.
|
| 650 |
-
"rewards/accuracies": 0.
|
| 651 |
-
"rewards/chosen": -0.
|
| 652 |
-
"rewards/margins": 0.
|
| 653 |
-
"rewards/rejected": -
|
| 654 |
"step": 420
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.48560135516657255,
|
| 658 |
-
"grad_norm":
|
| 659 |
-
"learning_rate":
|
| 660 |
-
"logits/chosen": -2.
|
| 661 |
-
"logits/rejected": -2.
|
| 662 |
-
"logps/chosen": -
|
| 663 |
-
"logps/rejected": -
|
| 664 |
-
"loss": 0.
|
| 665 |
-
"rewards/accuracies": 0.
|
| 666 |
-
"rewards/chosen": -0.
|
| 667 |
-
"rewards/margins": 0.
|
| 668 |
-
"rewards/rejected": -
|
| 669 |
"step": 430
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 0.4968944099378882,
|
| 673 |
-
"grad_norm": 8.
|
| 674 |
-
"learning_rate":
|
| 675 |
-
"logits/chosen": -2.
|
| 676 |
-
"logits/rejected": -2.
|
| 677 |
-
"logps/chosen": -
|
| 678 |
-
"logps/rejected": -
|
| 679 |
-
"loss": 0.
|
| 680 |
-
"rewards/accuracies": 0.
|
| 681 |
-
"rewards/chosen": -0.
|
| 682 |
-
"rewards/margins": 0.
|
| 683 |
-
"rewards/rejected": -
|
| 684 |
"step": 440
|
| 685 |
},
|
| 686 |
{
|
| 687 |
"epoch": 0.5081874647092038,
|
| 688 |
-
"grad_norm":
|
| 689 |
-
"learning_rate":
|
| 690 |
-
"logits/chosen": -2.
|
| 691 |
-
"logits/rejected": -2.
|
| 692 |
-
"logps/chosen": -
|
| 693 |
-
"logps/rejected": -
|
| 694 |
-
"loss": 0.
|
| 695 |
-
"rewards/accuracies": 0.
|
| 696 |
-
"rewards/chosen": -0.
|
| 697 |
-
"rewards/margins": 0.
|
| 698 |
-
"rewards/rejected": -
|
| 699 |
"step": 450
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.5194805194805194,
|
| 703 |
-
"grad_norm": 9.
|
| 704 |
-
"learning_rate":
|
| 705 |
-
"logits/chosen": -2.
|
| 706 |
-
"logits/rejected": -2.
|
| 707 |
-
"logps/chosen": -
|
| 708 |
-
"logps/rejected": -
|
| 709 |
-
"loss": 0.
|
| 710 |
-
"rewards/accuracies": 0.
|
| 711 |
-
"rewards/chosen": -0.
|
| 712 |
-
"rewards/margins": 0.
|
| 713 |
-
"rewards/rejected": -
|
| 714 |
"step": 460
|
| 715 |
},
|
| 716 |
{
|
| 717 |
"epoch": 0.5307735742518351,
|
| 718 |
-
"grad_norm": 9.
|
| 719 |
-
"learning_rate":
|
| 720 |
-
"logits/chosen": -2.
|
| 721 |
-
"logits/rejected": -2.
|
| 722 |
-
"logps/chosen": -
|
| 723 |
-
"logps/rejected": -
|
| 724 |
-
"loss": 0.
|
| 725 |
-
"rewards/accuracies": 0.
|
| 726 |
-
"rewards/chosen": -0.
|
| 727 |
-
"rewards/margins": 0.
|
| 728 |
-
"rewards/rejected": -0.
|
| 729 |
"step": 470
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.5420666290231507,
|
| 733 |
-
"grad_norm": 8.
|
| 734 |
-
"learning_rate":
|
| 735 |
-
"logits/chosen": -2.
|
| 736 |
-
"logits/rejected": -2.
|
| 737 |
-
"logps/chosen": -
|
| 738 |
-
"logps/rejected": -
|
| 739 |
-
"loss": 0.
|
| 740 |
-
"rewards/accuracies": 0.
|
| 741 |
-
"rewards/chosen": -0.
|
| 742 |
-
"rewards/margins": 0.
|
| 743 |
-
"rewards/rejected": -
|
| 744 |
"step": 480
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.5533596837944664,
|
| 748 |
-
"grad_norm":
|
| 749 |
-
"learning_rate":
|
| 750 |
-
"logits/chosen": -2.
|
| 751 |
-
"logits/rejected": -2.
|
| 752 |
-
"logps/chosen": -
|
| 753 |
-
"logps/rejected": -
|
| 754 |
-
"loss": 0.
|
| 755 |
-
"rewards/accuracies": 0.
|
| 756 |
-
"rewards/chosen": -0.
|
| 757 |
-
"rewards/margins": 0.
|
| 758 |
-
"rewards/rejected": -
|
| 759 |
"step": 490
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.564652738565782,
|
| 763 |
-
"grad_norm": 8.
|
| 764 |
-
"learning_rate":
|
| 765 |
-
"logits/chosen": -2.
|
| 766 |
-
"logits/rejected": -2.
|
| 767 |
-
"logps/chosen": -
|
| 768 |
-
"logps/rejected": -
|
| 769 |
-
"loss": 0.
|
| 770 |
-
"rewards/accuracies": 0.
|
| 771 |
-
"rewards/chosen": -0.
|
| 772 |
-
"rewards/margins": 0.
|
| 773 |
-
"rewards/rejected": -
|
| 774 |
"step": 500
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 0.5759457933370977,
|
| 778 |
-
"grad_norm":
|
| 779 |
-
"learning_rate":
|
| 780 |
-
"logits/chosen": -2.
|
| 781 |
-
"logits/rejected": -2.
|
| 782 |
-
"logps/chosen": -
|
| 783 |
-
"logps/rejected": -
|
| 784 |
-
"loss": 0.
|
| 785 |
-
"rewards/accuracies": 0.
|
| 786 |
-
"rewards/chosen": -0.
|
| 787 |
-
"rewards/margins": 0.
|
| 788 |
-
"rewards/rejected": -0.
|
| 789 |
"step": 510
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 0.5872388481084133,
|
| 793 |
-
"grad_norm":
|
| 794 |
-
"learning_rate":
|
| 795 |
-
"logits/chosen": -2.
|
| 796 |
-
"logits/rejected": -2.
|
| 797 |
-
"logps/chosen": -
|
| 798 |
-
"logps/rejected": -
|
| 799 |
-
"loss": 0.
|
| 800 |
-
"rewards/accuracies": 0.
|
| 801 |
-
"rewards/chosen": -0.
|
| 802 |
-
"rewards/margins": 0.
|
| 803 |
-
"rewards/rejected": -
|
| 804 |
"step": 520
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 0.598531902879729,
|
| 808 |
"grad_norm": 10.0,
|
| 809 |
-
"learning_rate":
|
| 810 |
-
"logits/chosen": -2.
|
| 811 |
-
"logits/rejected": -2.
|
| 812 |
-
"logps/chosen": -
|
| 813 |
-
"logps/rejected": -
|
| 814 |
-
"loss": 0.
|
| 815 |
"rewards/accuracies": 0.518750011920929,
|
| 816 |
-
"rewards/chosen": -0.
|
| 817 |
-
"rewards/margins": 0.
|
| 818 |
-
"rewards/rejected": -
|
| 819 |
"step": 530
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.6098249576510446,
|
| 823 |
-
"grad_norm":
|
| 824 |
-
"learning_rate":
|
| 825 |
-
"logits/chosen": -2.
|
| 826 |
-
"logits/rejected": -2.
|
| 827 |
-
"logps/chosen": -
|
| 828 |
-
"logps/rejected": -
|
| 829 |
-
"loss": 0.
|
| 830 |
-
"rewards/accuracies": 0.
|
| 831 |
-
"rewards/chosen": -0.
|
| 832 |
-
"rewards/margins": 0.
|
| 833 |
-
"rewards/rejected": -
|
| 834 |
"step": 540
|
| 835 |
},
|
| 836 |
{
|
| 837 |
"epoch": 0.6211180124223602,
|
| 838 |
-
"grad_norm":
|
| 839 |
-
"learning_rate":
|
| 840 |
-
"logits/chosen": -2.
|
| 841 |
-
"logits/rejected": -2.
|
| 842 |
-
"logps/chosen": -
|
| 843 |
-
"logps/rejected": -
|
| 844 |
-
"loss": 0.
|
| 845 |
-
"rewards/accuracies": 0.
|
| 846 |
-
"rewards/chosen": -0.
|
| 847 |
-
"rewards/margins": 0.
|
| 848 |
-
"rewards/rejected": -
|
| 849 |
"step": 550
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.6324110671936759,
|
| 853 |
-
"grad_norm":
|
| 854 |
-
"learning_rate":
|
| 855 |
-
"logits/chosen": -2.
|
| 856 |
-
"logits/rejected": -2.
|
| 857 |
-
"logps/chosen": -
|
| 858 |
-
"logps/rejected": -
|
| 859 |
-
"loss": 0.
|
| 860 |
-
"rewards/accuracies": 0.
|
| 861 |
-
"rewards/chosen": -0.
|
| 862 |
-
"rewards/margins": 0.
|
| 863 |
-
"rewards/rejected": -
|
| 864 |
"step": 560
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.6437041219649915,
|
| 868 |
-
"grad_norm": 9.
|
| 869 |
-
"learning_rate":
|
| 870 |
-
"logits/chosen": -2.
|
| 871 |
-
"logits/rejected": -2.
|
| 872 |
-
"logps/chosen": -
|
| 873 |
-
"logps/rejected": -
|
| 874 |
-
"loss": 0.
|
| 875 |
-
"rewards/accuracies": 0.
|
| 876 |
-
"rewards/chosen": -0.
|
| 877 |
-
"rewards/margins": 0.
|
| 878 |
-
"rewards/rejected": -
|
| 879 |
"step": 570
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 0.6549971767363072,
|
| 883 |
-
"grad_norm": 11.
|
| 884 |
-
"learning_rate":
|
| 885 |
-
"logits/chosen": -2.
|
| 886 |
-
"logits/rejected": -2.
|
| 887 |
-
"logps/chosen": -
|
| 888 |
-
"logps/rejected": -
|
| 889 |
-
"loss": 0.
|
| 890 |
-
"rewards/accuracies": 0.
|
| 891 |
-
"rewards/chosen": -
|
| 892 |
-
"rewards/margins": 0.
|
| 893 |
-
"rewards/rejected": -
|
| 894 |
"step": 580
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 0.6662902315076228,
|
| 898 |
-
"grad_norm":
|
| 899 |
-
"learning_rate":
|
| 900 |
-
"logits/chosen": -2.
|
| 901 |
-
"logits/rejected": -2.
|
| 902 |
-
"logps/chosen": -
|
| 903 |
-
"logps/rejected": -
|
| 904 |
-
"loss": 0.
|
| 905 |
-
"rewards/accuracies": 0.
|
| 906 |
-
"rewards/chosen": -0.
|
| 907 |
-
"rewards/margins": 0.
|
| 908 |
-
"rewards/rejected": -
|
| 909 |
"step": 590
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 0.6775832862789385,
|
| 913 |
-
"grad_norm": 11.
|
| 914 |
-
"learning_rate":
|
| 915 |
-
"logits/chosen": -2.
|
| 916 |
-
"logits/rejected": -2.
|
| 917 |
-
"logps/chosen": -
|
| 918 |
-
"logps/rejected": -
|
| 919 |
-
"loss": 0.
|
| 920 |
-
"rewards/accuracies": 0.
|
| 921 |
-
"rewards/chosen": -
|
| 922 |
-
"rewards/margins": 0.
|
| 923 |
-
"rewards/rejected": -
|
| 924 |
"step": 600
|
| 925 |
},
|
| 926 |
{
|
| 927 |
"epoch": 0.6888763410502541,
|
| 928 |
-
"grad_norm": 7.
|
| 929 |
-
"learning_rate":
|
| 930 |
-
"logits/chosen": -2.
|
| 931 |
-
"logits/rejected": -2.
|
| 932 |
-
"logps/chosen": -
|
| 933 |
-
"logps/rejected": -
|
| 934 |
-
"loss": 0.
|
| 935 |
-
"rewards/accuracies": 0.
|
| 936 |
-
"rewards/chosen": -0.
|
| 937 |
-
"rewards/margins": 0.
|
| 938 |
-
"rewards/rejected": -
|
| 939 |
"step": 610
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 0.7001693958215698,
|
| 943 |
-
"grad_norm":
|
| 944 |
-
"learning_rate":
|
| 945 |
-
"logits/chosen": -2.
|
| 946 |
-
"logits/rejected": -2.
|
| 947 |
-
"logps/chosen": -
|
| 948 |
-
"logps/rejected": -
|
| 949 |
-
"loss": 0.
|
| 950 |
-
"rewards/accuracies": 0.
|
| 951 |
-
"rewards/chosen": -
|
| 952 |
-
"rewards/margins": 0.
|
| 953 |
-
"rewards/rejected": -
|
| 954 |
"step": 620
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.7114624505928854,
|
| 958 |
-
"grad_norm": 11.
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"logits/chosen": -2.
|
| 961 |
-
"logits/rejected": -2.
|
| 962 |
-
"logps/chosen": -
|
| 963 |
-
"logps/rejected": -
|
| 964 |
-
"loss": 0.
|
| 965 |
-
"rewards/accuracies": 0.
|
| 966 |
-
"rewards/chosen": -0.
|
| 967 |
-
"rewards/margins": 0.
|
| 968 |
-
"rewards/rejected": -
|
| 969 |
"step": 630
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.722755505364201,
|
| 973 |
-
"grad_norm":
|
| 974 |
-
"learning_rate":
|
| 975 |
-
"logits/chosen": -2.
|
| 976 |
-
"logits/rejected": -2.
|
| 977 |
-
"logps/chosen": -
|
| 978 |
-
"logps/rejected": -
|
| 979 |
-
"loss": 0.
|
| 980 |
"rewards/accuracies": 0.5687500238418579,
|
| 981 |
-
"rewards/chosen": -
|
| 982 |
-
"rewards/margins": 0.
|
| 983 |
-
"rewards/rejected": -
|
| 984 |
"step": 640
|
| 985 |
},
|
| 986 |
{
|
| 987 |
"epoch": 0.7340485601355167,
|
| 988 |
-
"grad_norm":
|
| 989 |
-
"learning_rate":
|
| 990 |
-
"logits/chosen": -2.
|
| 991 |
-
"logits/rejected": -2.
|
| 992 |
-
"logps/chosen": -
|
| 993 |
-
"logps/rejected": -
|
| 994 |
-
"loss": 0.
|
| 995 |
"rewards/accuracies": 0.5562499761581421,
|
| 996 |
-
"rewards/chosen": -
|
| 997 |
-
"rewards/margins": 0.
|
| 998 |
-
"rewards/rejected": -
|
| 999 |
"step": 650
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 0.7453416149068323,
|
| 1003 |
-
"grad_norm":
|
| 1004 |
-
"learning_rate":
|
| 1005 |
-
"logits/chosen": -2.
|
| 1006 |
-
"logits/rejected": -2.
|
| 1007 |
-
"logps/chosen": -
|
| 1008 |
-
"logps/rejected": -
|
| 1009 |
-
"loss": 0.
|
| 1010 |
-
"rewards/accuracies": 0.
|
| 1011 |
-
"rewards/chosen": -
|
| 1012 |
-
"rewards/margins": 0.
|
| 1013 |
-
"rewards/rejected": -
|
| 1014 |
"step": 660
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
"epoch": 0.756634669678148,
|
| 1018 |
-
"grad_norm":
|
| 1019 |
-
"learning_rate":
|
| 1020 |
-
"logits/chosen": -2.
|
| 1021 |
-
"logits/rejected": -2.
|
| 1022 |
-
"logps/chosen": -
|
| 1023 |
-
"logps/rejected": -
|
| 1024 |
-
"loss": 0.
|
| 1025 |
-
"rewards/accuracies": 0.
|
| 1026 |
-
"rewards/chosen": -0.
|
| 1027 |
-
"rewards/margins": 0.
|
| 1028 |
-
"rewards/rejected": -
|
| 1029 |
"step": 670
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
"epoch": 0.7679277244494636,
|
| 1033 |
-
"grad_norm": 11.
|
| 1034 |
-
"learning_rate":
|
| 1035 |
-
"logits/chosen": -2.
|
| 1036 |
-
"logits/rejected": -2.
|
| 1037 |
-
"logps/chosen": -
|
| 1038 |
-
"logps/rejected": -
|
| 1039 |
-
"loss": 0.
|
| 1040 |
-
"rewards/accuracies": 0.
|
| 1041 |
-
"rewards/chosen": -
|
| 1042 |
-
"rewards/margins": 0.
|
| 1043 |
-
"rewards/rejected": -
|
| 1044 |
"step": 680
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.7792207792207793,
|
| 1048 |
-
"grad_norm": 10.
|
| 1049 |
-
"learning_rate":
|
| 1050 |
-
"logits/chosen": -2.
|
| 1051 |
-
"logits/rejected": -2.
|
| 1052 |
-
"logps/chosen": -
|
| 1053 |
-
"logps/rejected": -
|
| 1054 |
-
"loss": 0.
|
| 1055 |
-
"rewards/accuracies": 0.
|
| 1056 |
-
"rewards/chosen": -
|
| 1057 |
-
"rewards/margins": 0.
|
| 1058 |
-
"rewards/rejected": -
|
| 1059 |
"step": 690
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.7905138339920948,
|
| 1063 |
-
"grad_norm": 9.
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"logits/chosen": -2.
|
| 1066 |
-
"logits/rejected": -2.
|
| 1067 |
-
"logps/chosen": -
|
| 1068 |
-
"logps/rejected": -
|
| 1069 |
-
"loss": 0.
|
| 1070 |
-
"rewards/accuracies": 0.
|
| 1071 |
-
"rewards/chosen": -
|
| 1072 |
-
"rewards/margins": 0.
|
| 1073 |
-
"rewards/rejected": -
|
| 1074 |
"step": 700
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 0.8018068887634106,
|
| 1078 |
"grad_norm": 9.5625,
|
| 1079 |
-
"learning_rate":
|
| 1080 |
-
"logits/chosen": -2.
|
| 1081 |
-
"logits/rejected": -2.
|
| 1082 |
-
"logps/chosen": -
|
| 1083 |
-
"logps/rejected": -
|
| 1084 |
-
"loss": 0.
|
| 1085 |
-
"rewards/accuracies": 0.
|
| 1086 |
-
"rewards/chosen": -
|
| 1087 |
-
"rewards/margins": 0.
|
| 1088 |
-
"rewards/rejected": -
|
| 1089 |
"step": 710
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 0.8130999435347261,
|
| 1093 |
-
"grad_norm": 8.
|
| 1094 |
-
"learning_rate":
|
| 1095 |
-
"logits/chosen": -2.
|
| 1096 |
-
"logits/rejected": -2.
|
| 1097 |
-
"logps/chosen": -
|
| 1098 |
-
"logps/rejected": -
|
| 1099 |
-
"loss": 0.
|
| 1100 |
"rewards/accuracies": 0.581250011920929,
|
| 1101 |
-
"rewards/chosen": -
|
| 1102 |
-
"rewards/margins": 0.
|
| 1103 |
-
"rewards/rejected": -
|
| 1104 |
"step": 720
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 0.8243929983060417,
|
| 1108 |
-
"grad_norm":
|
| 1109 |
-
"learning_rate":
|
| 1110 |
-
"logits/chosen": -2.
|
| 1111 |
-
"logits/rejected": -2.
|
| 1112 |
-
"logps/chosen": -
|
| 1113 |
-
"logps/rejected": -
|
| 1114 |
-
"loss": 0.
|
| 1115 |
-
"rewards/accuracies": 0.
|
| 1116 |
-
"rewards/chosen": -
|
| 1117 |
-
"rewards/margins": 0.
|
| 1118 |
-
"rewards/rejected": -
|
| 1119 |
"step": 730
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
"epoch": 0.8356860530773574,
|
| 1123 |
-
"grad_norm":
|
| 1124 |
-
"learning_rate":
|
| 1125 |
-
"logits/chosen": -2.
|
| 1126 |
-
"logits/rejected": -2.
|
| 1127 |
-
"logps/chosen": -
|
| 1128 |
-
"logps/rejected": -
|
| 1129 |
-
"loss": 0.
|
| 1130 |
-
"rewards/accuracies": 0.
|
| 1131 |
-
"rewards/chosen": -
|
| 1132 |
-
"rewards/margins": 0.
|
| 1133 |
-
"rewards/rejected": -
|
| 1134 |
"step": 740
|
| 1135 |
},
|
| 1136 |
{
|
| 1137 |
"epoch": 0.846979107848673,
|
| 1138 |
-
"grad_norm": 7.
|
| 1139 |
-
"learning_rate":
|
| 1140 |
-
"logits/chosen": -2.
|
| 1141 |
-
"logits/rejected": -2.
|
| 1142 |
-
"logps/chosen": -
|
| 1143 |
-
"logps/rejected": -
|
| 1144 |
-
"loss": 0.
|
| 1145 |
-
"rewards/accuracies": 0.
|
| 1146 |
-
"rewards/chosen": -
|
| 1147 |
-
"rewards/margins": 0.
|
| 1148 |
-
"rewards/rejected": -
|
| 1149 |
"step": 750
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
"epoch": 0.8582721626199887,
|
| 1153 |
-
"grad_norm":
|
| 1154 |
-
"learning_rate":
|
| 1155 |
-
"logits/chosen": -2.
|
| 1156 |
-
"logits/rejected": -2.
|
| 1157 |
-
"logps/chosen": -
|
| 1158 |
-
"logps/rejected": -
|
| 1159 |
-
"loss": 0.
|
| 1160 |
-
"rewards/accuracies": 0.
|
| 1161 |
-
"rewards/chosen": -
|
| 1162 |
-
"rewards/margins": 0.
|
| 1163 |
-
"rewards/rejected": -
|
| 1164 |
"step": 760
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.8695652173913043,
|
| 1168 |
"grad_norm": 9.6875,
|
| 1169 |
-
"learning_rate":
|
| 1170 |
-
"logits/chosen": -2.
|
| 1171 |
-
"logits/rejected": -2.
|
| 1172 |
-
"logps/chosen": -
|
| 1173 |
-
"logps/rejected": -
|
| 1174 |
-
"loss": 0.
|
| 1175 |
-
"rewards/accuracies": 0.
|
| 1176 |
-
"rewards/chosen": -
|
| 1177 |
-
"rewards/margins": 0.
|
| 1178 |
-
"rewards/rejected": -
|
| 1179 |
"step": 770
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.88085827216262,
|
| 1183 |
-
"grad_norm":
|
| 1184 |
-
"learning_rate":
|
| 1185 |
-
"logits/chosen": -2.
|
| 1186 |
-
"logits/rejected": -2.
|
| 1187 |
-
"logps/chosen": -
|
| 1188 |
-
"logps/rejected": -
|
| 1189 |
-
"loss": 0.
|
| 1190 |
-
"rewards/accuracies": 0.
|
| 1191 |
-
"rewards/chosen": -
|
| 1192 |
-
"rewards/margins": 0.
|
| 1193 |
-
"rewards/rejected": -
|
| 1194 |
"step": 780
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
"epoch": 0.8921513269339356,
|
| 1198 |
-
"grad_norm":
|
| 1199 |
-
"learning_rate":
|
| 1200 |
-
"logits/chosen": -2.
|
| 1201 |
-
"logits/rejected": -2.
|
| 1202 |
-
"logps/chosen": -
|
| 1203 |
-
"logps/rejected": -
|
| 1204 |
-
"loss": 0.
|
| 1205 |
-
"rewards/accuracies": 0.
|
| 1206 |
-
"rewards/chosen": -0.
|
| 1207 |
-
"rewards/margins": 0.
|
| 1208 |
-
"rewards/rejected": -
|
| 1209 |
"step": 790
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 0.9034443817052513,
|
| 1213 |
-
"grad_norm": 9.
|
| 1214 |
-
"learning_rate":
|
| 1215 |
-
"logits/chosen": -2.
|
| 1216 |
-
"logits/rejected": -2.
|
| 1217 |
-
"logps/chosen": -
|
| 1218 |
-
"logps/rejected": -
|
| 1219 |
-
"loss": 0.
|
| 1220 |
-
"rewards/accuracies": 0.
|
| 1221 |
-
"rewards/chosen": -
|
| 1222 |
-
"rewards/margins": 0.
|
| 1223 |
-
"rewards/rejected": -
|
| 1224 |
"step": 800
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 0.9147374364765669,
|
| 1228 |
-
"grad_norm": 8.
|
| 1229 |
-
"learning_rate":
|
| 1230 |
-
"logits/chosen": -2.
|
| 1231 |
-
"logits/rejected": -2.
|
| 1232 |
-
"logps/chosen": -
|
| 1233 |
-
"logps/rejected": -
|
| 1234 |
-
"loss": 0.
|
| 1235 |
-
"rewards/accuracies": 0.
|
| 1236 |
-
"rewards/chosen": -0.
|
| 1237 |
-
"rewards/margins": 0.
|
| 1238 |
-
"rewards/rejected": -
|
| 1239 |
"step": 810
|
| 1240 |
},
|
| 1241 |
{
|
| 1242 |
"epoch": 0.9260304912478825,
|
| 1243 |
-
"grad_norm": 10.
|
| 1244 |
-
"learning_rate":
|
| 1245 |
-
"logits/chosen": -2.
|
| 1246 |
-
"logits/rejected": -2.
|
| 1247 |
-
"logps/chosen": -
|
| 1248 |
-
"logps/rejected": -
|
| 1249 |
-
"loss": 0.
|
| 1250 |
"rewards/accuracies": 0.5375000238418579,
|
| 1251 |
-
"rewards/chosen": -
|
| 1252 |
-
"rewards/margins": 0.
|
| 1253 |
-
"rewards/rejected": -
|
| 1254 |
"step": 820
|
| 1255 |
},
|
| 1256 |
{
|
| 1257 |
"epoch": 0.9373235460191982,
|
| 1258 |
-
"grad_norm":
|
| 1259 |
-
"learning_rate":
|
| 1260 |
-
"logits/chosen": -2.
|
| 1261 |
-
"logits/rejected": -2.
|
| 1262 |
-
"logps/chosen": -
|
| 1263 |
-
"logps/rejected": -
|
| 1264 |
-
"loss": 0.
|
| 1265 |
-
"rewards/accuracies": 0.
|
| 1266 |
-
"rewards/chosen": -
|
| 1267 |
-
"rewards/margins": 0.
|
| 1268 |
-
"rewards/rejected": -
|
| 1269 |
"step": 830
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.9486166007905138,
|
| 1273 |
-
"grad_norm":
|
| 1274 |
-
"learning_rate":
|
| 1275 |
-
"logits/chosen": -2.
|
| 1276 |
-
"logits/rejected": -2.
|
| 1277 |
-
"logps/chosen": -
|
| 1278 |
-
"logps/rejected": -
|
| 1279 |
-
"loss": 0.
|
| 1280 |
-
"rewards/accuracies": 0.
|
| 1281 |
-
"rewards/chosen": -
|
| 1282 |
-
"rewards/margins": 0.
|
| 1283 |
-
"rewards/rejected": -
|
| 1284 |
"step": 840
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.9599096555618295,
|
| 1288 |
-
"grad_norm":
|
| 1289 |
-
"learning_rate":
|
| 1290 |
-
"logits/chosen": -2.
|
| 1291 |
-
"logits/rejected": -2.
|
| 1292 |
-
"logps/chosen": -
|
| 1293 |
-
"logps/rejected": -
|
| 1294 |
-
"loss": 0.
|
| 1295 |
-
"rewards/accuracies": 0.
|
| 1296 |
-
"rewards/chosen": -
|
| 1297 |
-
"rewards/margins": 0.
|
| 1298 |
-
"rewards/rejected": -
|
| 1299 |
"step": 850
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
"epoch": 0.9712027103331451,
|
| 1303 |
-
"grad_norm":
|
| 1304 |
-
"learning_rate":
|
| 1305 |
-
"logits/chosen": -2.
|
| 1306 |
-
"logits/rejected": -2.
|
| 1307 |
-
"logps/chosen": -
|
| 1308 |
-
"logps/rejected": -
|
| 1309 |
-
"loss": 0.
|
| 1310 |
-
"rewards/accuracies": 0.
|
| 1311 |
-
"rewards/chosen": -
|
| 1312 |
-
"rewards/margins": 0.
|
| 1313 |
-
"rewards/rejected": -
|
| 1314 |
"step": 860
|
| 1315 |
},
|
| 1316 |
{
|
| 1317 |
"epoch": 0.9824957651044608,
|
| 1318 |
-
"grad_norm":
|
| 1319 |
-
"learning_rate":
|
| 1320 |
-
"logits/chosen": -2.
|
| 1321 |
-
"logits/rejected": -2.
|
| 1322 |
-
"logps/chosen": -
|
| 1323 |
-
"logps/rejected": -
|
| 1324 |
-
"loss": 0.
|
| 1325 |
-
"rewards/accuracies": 0.
|
| 1326 |
-
"rewards/chosen": -0.
|
| 1327 |
-
"rewards/margins": 0.
|
| 1328 |
-
"rewards/rejected": -
|
| 1329 |
"step": 870
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 0.9937888198757764,
|
| 1333 |
-
"grad_norm": 9.
|
| 1334 |
-
"learning_rate":
|
| 1335 |
-
"logits/chosen": -2.
|
| 1336 |
-
"logits/rejected": -2.
|
| 1337 |
-
"logps/chosen": -
|
| 1338 |
-
"logps/rejected": -
|
| 1339 |
-
"loss": 0.
|
| 1340 |
-
"rewards/accuracies": 0.
|
| 1341 |
-
"rewards/chosen": -
|
| 1342 |
-
"rewards/margins": 0.
|
| 1343 |
-
"rewards/rejected": -
|
| 1344 |
"step": 880
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
"epoch": 0.9994353472614342,
|
| 1348 |
"step": 885,
|
| 1349 |
"total_flos": 0.0,
|
| 1350 |
-
"train_loss": 0.
|
| 1351 |
-
"train_runtime":
|
| 1352 |
-
"train_samples_per_second": 1.
|
| 1353 |
"train_steps_per_second": 0.089
|
| 1354 |
}
|
| 1355 |
],
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.001129305477131564,
|
| 13 |
"grad_norm": 10.5625,
|
| 14 |
+
"learning_rate": 1.1235955056179774e-08,
|
| 15 |
"logits/chosen": -2.8912107944488525,
|
| 16 |
"logits/rejected": -2.9116690158843994,
|
| 17 |
"logps/chosen": -205.43255615234375,
|
|
|
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.01129305477131564,
|
| 28 |
+
"grad_norm": 11.5,
|
| 29 |
+
"learning_rate": 1.1235955056179774e-07,
|
| 30 |
+
"logits/chosen": -2.9240243434906006,
|
| 31 |
+
"logits/rejected": -2.895505666732788,
|
| 32 |
+
"logps/chosen": -286.3099365234375,
|
| 33 |
+
"logps/rejected": -234.07386779785156,
|
| 34 |
+
"loss": 0.6929,
|
| 35 |
+
"rewards/accuracies": 0.3888888955116272,
|
| 36 |
+
"rewards/chosen": 0.00452011963352561,
|
| 37 |
+
"rewards/margins": 0.0007920749485492706,
|
| 38 |
+
"rewards/rejected": 0.0037280453834682703,
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.02258610954263128,
|
| 43 |
+
"grad_norm": 8.125,
|
| 44 |
+
"learning_rate": 2.2471910112359549e-07,
|
| 45 |
+
"logits/chosen": -2.9115793704986572,
|
| 46 |
+
"logits/rejected": -2.8940234184265137,
|
| 47 |
+
"logps/chosen": -242.3488311767578,
|
| 48 |
+
"logps/rejected": -230.69912719726562,
|
| 49 |
+
"loss": 0.6965,
|
| 50 |
+
"rewards/accuracies": 0.40625,
|
| 51 |
+
"rewards/chosen": -0.006699255667626858,
|
| 52 |
+
"rewards/margins": -0.006314368452876806,
|
| 53 |
+
"rewards/rejected": -0.0003848885535262525,
|
| 54 |
"step": 20
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.03387916431394692,
|
| 58 |
"grad_norm": 10.4375,
|
| 59 |
+
"learning_rate": 3.3707865168539325e-07,
|
| 60 |
+
"logits/chosen": -2.933945894241333,
|
| 61 |
+
"logits/rejected": -2.9031498432159424,
|
| 62 |
+
"logps/chosen": -227.9883270263672,
|
| 63 |
+
"logps/rejected": -206.83169555664062,
|
| 64 |
+
"loss": 0.6914,
|
| 65 |
+
"rewards/accuracies": 0.5,
|
| 66 |
+
"rewards/chosen": 0.0010180685203522444,
|
| 67 |
+
"rewards/margins": 0.003906534984707832,
|
| 68 |
+
"rewards/rejected": -0.0028884666971862316,
|
| 69 |
"step": 30
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.04517221908526256,
|
| 73 |
+
"grad_norm": 10.5,
|
| 74 |
+
"learning_rate": 4.4943820224719097e-07,
|
| 75 |
+
"logits/chosen": -2.940782308578491,
|
| 76 |
+
"logits/rejected": -2.903311252593994,
|
| 77 |
+
"logps/chosen": -239.7572021484375,
|
| 78 |
+
"logps/rejected": -223.0744171142578,
|
| 79 |
+
"loss": 0.6943,
|
| 80 |
+
"rewards/accuracies": 0.46875,
|
| 81 |
+
"rewards/chosen": -0.001310513005591929,
|
| 82 |
+
"rewards/margins": -0.0019447545055299997,
|
| 83 |
+
"rewards/rejected": 0.0006342411506921053,
|
| 84 |
"step": 40
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"epoch": 0.05646527385657821,
|
| 88 |
"grad_norm": 11.25,
|
| 89 |
+
"learning_rate": 5.617977528089887e-07,
|
| 90 |
+
"logits/chosen": -2.926548719406128,
|
| 91 |
+
"logits/rejected": -2.891540050506592,
|
| 92 |
+
"logps/chosen": -252.4969024658203,
|
| 93 |
+
"logps/rejected": -208.90573120117188,
|
| 94 |
+
"loss": 0.6926,
|
| 95 |
+
"rewards/accuracies": 0.48124998807907104,
|
| 96 |
+
"rewards/chosen": -0.002666835905984044,
|
| 97 |
+
"rewards/margins": 0.0014006331330165267,
|
| 98 |
+
"rewards/rejected": -0.004067468456923962,
|
| 99 |
"step": 50
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.06775832862789384,
|
| 103 |
+
"grad_norm": 9.0625,
|
| 104 |
+
"learning_rate": 6.741573033707865e-07,
|
| 105 |
+
"logits/chosen": -2.951221227645874,
|
| 106 |
+
"logits/rejected": -2.9214468002319336,
|
| 107 |
+
"logps/chosen": -230.9364013671875,
|
| 108 |
+
"logps/rejected": -222.9247283935547,
|
| 109 |
+
"loss": 0.696,
|
| 110 |
+
"rewards/accuracies": 0.42500001192092896,
|
| 111 |
+
"rewards/chosen": -0.008724420331418514,
|
| 112 |
+
"rewards/margins": -0.0053501129150390625,
|
| 113 |
+
"rewards/rejected": -0.003374306485056877,
|
| 114 |
"step": 60
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.07905138339920949,
|
| 118 |
+
"grad_norm": 15.75,
|
| 119 |
+
"learning_rate": 7.865168539325843e-07,
|
| 120 |
+
"logits/chosen": -2.915444850921631,
|
| 121 |
+
"logits/rejected": -2.8806443214416504,
|
| 122 |
+
"logps/chosen": -249.3389434814453,
|
| 123 |
+
"logps/rejected": -235.85330200195312,
|
| 124 |
+
"loss": 0.6919,
|
| 125 |
+
"rewards/accuracies": 0.512499988079071,
|
| 126 |
+
"rewards/chosen": -0.006868250668048859,
|
| 127 |
+
"rewards/margins": 0.003010864369571209,
|
| 128 |
+
"rewards/rejected": -0.009879115037620068,
|
| 129 |
"step": 70
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.09034443817052512,
|
| 133 |
+
"grad_norm": 13.375,
|
| 134 |
+
"learning_rate": 8.988764044943819e-07,
|
| 135 |
+
"logits/chosen": -2.893014430999756,
|
| 136 |
+
"logits/rejected": -2.8584141731262207,
|
| 137 |
+
"logps/chosen": -244.0336456298828,
|
| 138 |
+
"logps/rejected": -241.87136840820312,
|
| 139 |
+
"loss": 0.6888,
|
| 140 |
+
"rewards/accuracies": 0.543749988079071,
|
| 141 |
+
"rewards/chosen": -0.01003711111843586,
|
| 142 |
+
"rewards/margins": 0.00937909446656704,
|
| 143 |
+
"rewards/rejected": -0.0194162018597126,
|
| 144 |
"step": 80
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.10163749294184077,
|
| 148 |
+
"grad_norm": 6.90625,
|
| 149 |
+
"learning_rate": 9.99996105846605e-07,
|
| 150 |
+
"logits/chosen": -2.912137031555176,
|
| 151 |
+
"logits/rejected": -2.8917088508605957,
|
| 152 |
+
"logps/chosen": -230.53872680664062,
|
| 153 |
+
"logps/rejected": -230.29702758789062,
|
| 154 |
+
"loss": 0.6905,
|
| 155 |
+
"rewards/accuracies": 0.5,
|
| 156 |
+
"rewards/chosen": -0.01865602657198906,
|
| 157 |
+
"rewards/margins": 0.006008770316839218,
|
| 158 |
+
"rewards/rejected": -0.024664796888828278,
|
| 159 |
"step": 90
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.11293054771315642,
|
| 163 |
+
"grad_norm": 8.0625,
|
| 164 |
+
"learning_rate": 9.99528880830604e-07,
|
| 165 |
+
"logits/chosen": -2.954880952835083,
|
| 166 |
+
"logits/rejected": -2.9029383659362793,
|
| 167 |
+
"logps/chosen": -230.21240234375,
|
| 168 |
+
"logps/rejected": -242.08786010742188,
|
| 169 |
+
"loss": 0.6888,
|
| 170 |
+
"rewards/accuracies": 0.5,
|
| 171 |
+
"rewards/chosen": -0.03368181362748146,
|
| 172 |
+
"rewards/margins": 0.010173236019909382,
|
| 173 |
+
"rewards/rejected": -0.04385504871606827,
|
| 174 |
"step": 100
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.12422360248447205,
|
| 178 |
+
"grad_norm": 10.4375,
|
| 179 |
+
"learning_rate": 9.982836589657042e-07,
|
| 180 |
+
"logits/chosen": -2.893187999725342,
|
| 181 |
+
"logits/rejected": -2.855039596557617,
|
| 182 |
+
"logps/chosen": -253.92001342773438,
|
| 183 |
+
"logps/rejected": -249.1540069580078,
|
| 184 |
+
"loss": 0.6881,
|
| 185 |
+
"rewards/accuracies": 0.53125,
|
| 186 |
+
"rewards/chosen": -0.04104708880186081,
|
| 187 |
+
"rewards/margins": 0.011361517943441868,
|
| 188 |
+
"rewards/rejected": -0.0524086132645607,
|
| 189 |
"step": 110
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.13551665725578768,
|
| 193 |
+
"grad_norm": 8.1875,
|
| 194 |
+
"learning_rate": 9.962623796366428e-07,
|
| 195 |
+
"logits/chosen": -2.9364559650421143,
|
| 196 |
+
"logits/rejected": -2.921806812286377,
|
| 197 |
+
"logps/chosen": -233.13034057617188,
|
| 198 |
+
"logps/rejected": -215.0299072265625,
|
| 199 |
+
"loss": 0.6917,
|
| 200 |
+
"rewards/accuracies": 0.5062500238418579,
|
| 201 |
+
"rewards/chosen": -0.0493025965988636,
|
| 202 |
+
"rewards/margins": 0.004110801964998245,
|
| 203 |
+
"rewards/rejected": -0.05341339856386185,
|
| 204 |
"step": 120
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 0.14680971202710333,
|
| 208 |
+
"grad_norm": 10.875,
|
| 209 |
+
"learning_rate": 9.934681909075434e-07,
|
| 210 |
+
"logits/chosen": -2.9207844734191895,
|
| 211 |
+
"logits/rejected": -2.894355297088623,
|
| 212 |
+
"logps/chosen": -246.37173461914062,
|
| 213 |
+
"logps/rejected": -252.4733428955078,
|
| 214 |
+
"loss": 0.6846,
|
| 215 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 216 |
+
"rewards/chosen": -0.056733161211013794,
|
| 217 |
+
"rewards/margins": 0.019008487462997437,
|
| 218 |
+
"rewards/rejected": -0.07574164122343063,
|
| 219 |
"step": 130
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.15810276679841898,
|
| 223 |
+
"grad_norm": 7.75,
|
| 224 |
+
"learning_rate": 9.899054446189302e-07,
|
| 225 |
+
"logits/chosen": -2.9487192630767822,
|
| 226 |
+
"logits/rejected": -2.9129204750061035,
|
| 227 |
+
"logps/chosen": -248.5870361328125,
|
| 228 |
+
"logps/rejected": -237.9082489013672,
|
| 229 |
+
"loss": 0.6859,
|
| 230 |
+
"rewards/accuracies": 0.4937500059604645,
|
| 231 |
+
"rewards/chosen": -0.06868256628513336,
|
| 232 |
+
"rewards/margins": 0.016593391075730324,
|
| 233 |
+
"rewards/rejected": -0.08527596294879913,
|
| 234 |
"step": 140
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.16939582156973462,
|
| 238 |
+
"grad_norm": 10.75,
|
| 239 |
+
"learning_rate": 9.855796896099044e-07,
|
| 240 |
+
"logits/chosen": -2.915994167327881,
|
| 241 |
+
"logits/rejected": -2.8766446113586426,
|
| 242 |
+
"logps/chosen": -257.65740966796875,
|
| 243 |
+
"logps/rejected": -226.9933624267578,
|
| 244 |
+
"loss": 0.6854,
|
| 245 |
+
"rewards/accuracies": 0.550000011920929,
|
| 246 |
+
"rewards/chosen": -0.07911563664674759,
|
| 247 |
+
"rewards/margins": 0.018480664119124413,
|
| 248 |
+
"rewards/rejected": -0.09759628772735596,
|
| 249 |
"step": 150
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.18068887634105024,
|
| 253 |
+
"grad_norm": 9.375,
|
| 254 |
+
"learning_rate": 9.804976630760418e-07,
|
| 255 |
+
"logits/chosen": -2.918802261352539,
|
| 256 |
+
"logits/rejected": -2.87815260887146,
|
| 257 |
+
"logps/chosen": -219.70858764648438,
|
| 258 |
+
"logps/rejected": -212.9764404296875,
|
| 259 |
+
"loss": 0.6801,
|
| 260 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 261 |
+
"rewards/chosen": -0.07974254339933395,
|
| 262 |
+
"rewards/margins": 0.029045332223176956,
|
| 263 |
+
"rewards/rejected": -0.10878787934780121,
|
| 264 |
"step": 160
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 0.1919819311123659,
|
| 268 |
+
"grad_norm": 58.5,
|
| 269 |
+
"learning_rate": 9.746672800764734e-07,
|
| 270 |
+
"logits/chosen": -2.9482781887054443,
|
| 271 |
+
"logits/rejected": -2.911285400390625,
|
| 272 |
+
"logps/chosen": -248.09774780273438,
|
| 273 |
+
"logps/rejected": -235.35397338867188,
|
| 274 |
+
"loss": 0.6792,
|
| 275 |
"rewards/accuracies": 0.581250011920929,
|
| 276 |
+
"rewards/chosen": -0.09282518178224564,
|
| 277 |
+
"rewards/margins": 0.03139440342783928,
|
| 278 |
+
"rewards/rejected": -0.12421958148479462,
|
| 279 |
"step": 170
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 0.20327498588368154,
|
| 283 |
+
"grad_norm": 9.8125,
|
| 284 |
+
"learning_rate": 9.680976212064874e-07,
|
| 285 |
+
"logits/chosen": -2.9101614952087402,
|
| 286 |
+
"logits/rejected": -2.865682363510132,
|
| 287 |
+
"logps/chosen": -244.28964233398438,
|
| 288 |
+
"logps/rejected": -202.06594848632812,
|
| 289 |
+
"loss": 0.6829,
|
| 290 |
+
"rewards/accuracies": 0.53125,
|
| 291 |
+
"rewards/chosen": -0.10991182178258896,
|
| 292 |
+
"rewards/margins": 0.024404151365160942,
|
| 293 |
+
"rewards/rejected": -0.13431595265865326,
|
| 294 |
"step": 180
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 0.21456804065499718,
|
| 298 |
+
"grad_norm": 9.5,
|
| 299 |
+
"learning_rate": 9.607989184548542e-07,
|
| 300 |
+
"logits/chosen": -2.928318738937378,
|
| 301 |
+
"logits/rejected": -2.9046478271484375,
|
| 302 |
+
"logps/chosen": -249.5603790283203,
|
| 303 |
+
"logps/rejected": -234.435791015625,
|
| 304 |
+
"loss": 0.6893,
|
| 305 |
+
"rewards/accuracies": 0.53125,
|
| 306 |
+
"rewards/chosen": -0.11975125968456268,
|
| 307 |
+
"rewards/margins": 0.012782303616404533,
|
| 308 |
+
"rewards/rejected": -0.13253356516361237,
|
| 309 |
"step": 190
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 0.22586109542631283,
|
| 313 |
+
"grad_norm": 10.0625,
|
| 314 |
+
"learning_rate": 9.52782539267901e-07,
|
| 315 |
+
"logits/chosen": -2.9089949131011963,
|
| 316 |
+
"logits/rejected": -2.8771417140960693,
|
| 317 |
+
"logps/chosen": -238.6053009033203,
|
| 318 |
+
"logps/rejected": -244.843994140625,
|
| 319 |
+
"loss": 0.6775,
|
| 320 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 321 |
+
"rewards/chosen": -0.12132523953914642,
|
| 322 |
+
"rewards/margins": 0.03615022823214531,
|
| 323 |
+
"rewards/rejected": -0.15747547149658203,
|
| 324 |
"step": 200
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.23715415019762845,
|
| 328 |
+
"grad_norm": 10.75,
|
| 329 |
+
"learning_rate": 9.44060968845156e-07,
|
| 330 |
+
"logits/chosen": -2.957009792327881,
|
| 331 |
+
"logits/rejected": -2.917191982269287,
|
| 332 |
+
"logps/chosen": -254.15109252929688,
|
| 333 |
+
"logps/rejected": -213.61257934570312,
|
| 334 |
+
"loss": 0.6785,
|
| 335 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 336 |
+
"rewards/chosen": -0.12494392693042755,
|
| 337 |
+
"rewards/margins": 0.034127604216337204,
|
| 338 |
+
"rewards/rejected": -0.15907153487205505,
|
| 339 |
"step": 210
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.2484472049689441,
|
| 343 |
+
"grad_norm": 11.6875,
|
| 344 |
+
"learning_rate": 9.346477906941331e-07,
|
| 345 |
+
"logits/chosen": -2.9415574073791504,
|
| 346 |
+
"logits/rejected": -2.908458709716797,
|
| 347 |
+
"logps/chosen": -273.68597412109375,
|
| 348 |
+
"logps/rejected": -242.5727996826172,
|
| 349 |
+
"loss": 0.6812,
|
| 350 |
+
"rewards/accuracies": 0.550000011920929,
|
| 351 |
+
"rewards/chosen": -0.14751294255256653,
|
| 352 |
+
"rewards/margins": 0.03075326606631279,
|
| 353 |
+
"rewards/rejected": -0.17826619744300842,
|
| 354 |
"step": 220
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.2597402597402597,
|
| 358 |
+
"grad_norm": 10.75,
|
| 359 |
+
"learning_rate": 9.245576654745471e-07,
|
| 360 |
+
"logits/chosen": -2.961373805999756,
|
| 361 |
+
"logits/rejected": -2.9270126819610596,
|
| 362 |
+
"logps/chosen": -271.6230163574219,
|
| 363 |
+
"logps/rejected": -242.1273956298828,
|
| 364 |
+
"loss": 0.6785,
|
| 365 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 366 |
+
"rewards/chosen": -0.15283231437206268,
|
| 367 |
+
"rewards/margins": 0.035538919270038605,
|
| 368 |
+
"rewards/rejected": -0.18837125599384308,
|
| 369 |
"step": 230
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 0.27103331451157536,
|
| 373 |
+
"grad_norm": 11.4375,
|
| 374 |
+
"learning_rate": 9.13806308164905e-07,
|
| 375 |
+
"logits/chosen": -2.9348835945129395,
|
| 376 |
+
"logits/rejected": -2.8914520740509033,
|
| 377 |
+
"logps/chosen": -231.8206024169922,
|
| 378 |
+
"logps/rejected": -218.68106079101562,
|
| 379 |
+
"loss": 0.6784,
|
| 380 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 381 |
+
"rewards/chosen": -0.1429007351398468,
|
| 382 |
+
"rewards/margins": 0.037423450499773026,
|
| 383 |
+
"rewards/rejected": -0.18032419681549072,
|
| 384 |
"step": 240
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 0.282326369282891,
|
| 388 |
+
"grad_norm": 8.9375,
|
| 389 |
+
"learning_rate": 9.024104635870367e-07,
|
| 390 |
+
"logits/chosen": -2.9596941471099854,
|
| 391 |
+
"logits/rejected": -2.9176220893859863,
|
| 392 |
+
"logps/chosen": -234.1117706298828,
|
| 393 |
+
"logps/rejected": -221.8756866455078,
|
| 394 |
+
"loss": 0.6625,
|
| 395 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 396 |
+
"rewards/chosen": -0.16116580367088318,
|
| 397 |
+
"rewards/margins": 0.07054189592599869,
|
| 398 |
+
"rewards/rejected": -0.23170769214630127,
|
| 399 |
"step": 250
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 0.29361942405420666,
|
| 403 |
+
"grad_norm": 10.875,
|
| 404 |
+
"learning_rate": 8.90387880326684e-07,
|
| 405 |
+
"logits/chosen": -2.9382996559143066,
|
| 406 |
+
"logits/rejected": -2.9059150218963623,
|
| 407 |
+
"logps/chosen": -236.49020385742188,
|
| 408 |
+
"logps/rejected": -239.8851318359375,
|
| 409 |
+
"loss": 0.6765,
|
| 410 |
"rewards/accuracies": 0.625,
|
| 411 |
+
"rewards/chosen": -0.15588422119617462,
|
| 412 |
+
"rewards/margins": 0.04047652706503868,
|
| 413 |
+
"rewards/rejected": -0.1963607370853424,
|
| 414 |
"step": 260
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 0.3049124788255223,
|
| 418 |
+
"grad_norm": 9.0,
|
| 419 |
+
"learning_rate": 8.777572830907684e-07,
|
| 420 |
+
"logits/chosen": -2.950742721557617,
|
| 421 |
+
"logits/rejected": -2.935871124267578,
|
| 422 |
+
"logps/chosen": -235.8411407470703,
|
| 423 |
+
"logps/rejected": -237.037109375,
|
| 424 |
+
"loss": 0.6797,
|
| 425 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 426 |
+
"rewards/chosen": -0.18154451251029968,
|
| 427 |
+
"rewards/margins": 0.03585205227136612,
|
| 428 |
+
"rewards/rejected": -0.2173965722322464,
|
| 429 |
"step": 270
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.31620553359683795,
|
| 433 |
+
"grad_norm": 9.5,
|
| 434 |
+
"learning_rate": 8.645383435443851e-07,
|
| 435 |
+
"logits/chosen": -2.944791793823242,
|
| 436 |
+
"logits/rejected": -2.9193873405456543,
|
| 437 |
+
"logps/chosen": -258.33331298828125,
|
| 438 |
+
"logps/rejected": -272.0276184082031,
|
| 439 |
+
"loss": 0.6716,
|
| 440 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 441 |
+
"rewards/chosen": -0.1910783350467682,
|
| 442 |
+
"rewards/margins": 0.051886141300201416,
|
| 443 |
+
"rewards/rejected": -0.242964506149292,
|
| 444 |
"step": 280
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.3274985883681536,
|
| 448 |
+
"grad_norm": 11.1875,
|
| 449 |
+
"learning_rate": 8.507516496729493e-07,
|
| 450 |
+
"logits/chosen": -2.9242568016052246,
|
| 451 |
+
"logits/rejected": -2.9168529510498047,
|
| 452 |
+
"logps/chosen": -251.5631103515625,
|
| 453 |
+
"logps/rejected": -245.46444702148438,
|
| 454 |
+
"loss": 0.703,
|
| 455 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 456 |
+
"rewards/chosen": -0.23275713622570038,
|
| 457 |
+
"rewards/margins": -0.012155926786363125,
|
| 458 |
+
"rewards/rejected": -0.22060124576091766,
|
| 459 |
"step": 290
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.33879164313946925,
|
| 463 |
+
"grad_norm": 11.6875,
|
| 464 |
+
"learning_rate": 8.364186737172068e-07,
|
| 465 |
+
"logits/chosen": -2.9377264976501465,
|
| 466 |
+
"logits/rejected": -2.9192919731140137,
|
| 467 |
+
"logps/chosen": -272.1338806152344,
|
| 468 |
+
"logps/rejected": -252.2882537841797,
|
| 469 |
+
"loss": 0.6809,
|
| 470 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 471 |
+
"rewards/chosen": -0.19352756440639496,
|
| 472 |
+
"rewards/margins": 0.03381720930337906,
|
| 473 |
+
"rewards/rejected": -0.2273447960615158,
|
| 474 |
"step": 300
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.3500846979107849,
|
| 478 |
+
"grad_norm": 9.4375,
|
| 479 |
+
"learning_rate": 8.215617387310522e-07,
|
| 480 |
+
"logits/chosen": -2.961489200592041,
|
| 481 |
+
"logits/rejected": -2.9303905963897705,
|
| 482 |
+
"logps/chosen": -255.98681640625,
|
| 483 |
+
"logps/rejected": -215.4890594482422,
|
| 484 |
+
"loss": 0.6801,
|
| 485 |
+
"rewards/accuracies": 0.518750011920929,
|
| 486 |
+
"rewards/chosen": -0.19549933075904846,
|
| 487 |
+
"rewards/margins": 0.035571686923503876,
|
| 488 |
+
"rewards/rejected": -0.23107102513313293,
|
| 489 |
"step": 310
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.3613777526821005,
|
| 493 |
+
"grad_norm": 9.6875,
|
| 494 |
+
"learning_rate": 8.062039838142401e-07,
|
| 495 |
+
"logits/chosen": -2.9097650051116943,
|
| 496 |
+
"logits/rejected": -2.879265546798706,
|
| 497 |
+
"logps/chosen": -233.59078979492188,
|
| 498 |
+
"logps/rejected": -255.03677368164062,
|
| 499 |
+
"loss": 0.6642,
|
| 500 |
+
"rewards/accuracies": 0.59375,
|
| 501 |
+
"rewards/chosen": -0.1973254680633545,
|
| 502 |
+
"rewards/margins": 0.06654740124940872,
|
| 503 |
+
"rewards/rejected": -0.2638728618621826,
|
| 504 |
"step": 320
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 0.37267080745341613,
|
| 508 |
+
"grad_norm": 9.5,
|
| 509 |
+
"learning_rate": 7.903693280741331e-07,
|
| 510 |
+
"logits/chosen": -2.932699203491211,
|
| 511 |
+
"logits/rejected": -2.8970255851745605,
|
| 512 |
+
"logps/chosen": -245.6687469482422,
|
| 513 |
+
"logps/rejected": -264.0479736328125,
|
| 514 |
+
"loss": 0.6635,
|
| 515 |
+
"rewards/accuracies": 0.59375,
|
| 516 |
+
"rewards/chosen": -0.1542760580778122,
|
| 517 |
+
"rewards/margins": 0.07145790755748749,
|
| 518 |
+
"rewards/rejected": -0.22573396563529968,
|
| 519 |
"step": 330
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.3839638622247318,
|
| 523 |
+
"grad_norm": 9.5625,
|
| 524 |
+
"learning_rate": 7.740824333726213e-07,
|
| 525 |
+
"logits/chosen": -2.974498748779297,
|
| 526 |
+
"logits/rejected": -2.9419915676116943,
|
| 527 |
+
"logps/chosen": -254.0312957763672,
|
| 528 |
+
"logps/rejected": -241.7189178466797,
|
| 529 |
+
"loss": 0.6755,
|
| 530 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 531 |
+
"rewards/chosen": -0.19228370487689972,
|
| 532 |
+
"rewards/margins": 0.04604003205895424,
|
| 533 |
+
"rewards/rejected": -0.23832373321056366,
|
| 534 |
"step": 340
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3952569169960474,
|
| 538 |
+
"grad_norm": 12.375,
|
| 539 |
+
"learning_rate": 7.573686659162293e-07,
|
| 540 |
+
"logits/chosen": -2.959407329559326,
|
| 541 |
+
"logits/rejected": -2.937516212463379,
|
| 542 |
+
"logps/chosen": -246.8042449951172,
|
| 543 |
+
"logps/rejected": -242.3812713623047,
|
| 544 |
+
"loss": 0.6706,
|
| 545 |
+
"rewards/accuracies": 0.550000011920929,
|
| 546 |
+
"rewards/chosen": -0.20731408894062042,
|
| 547 |
+
"rewards/margins": 0.061324410140514374,
|
| 548 |
+
"rewards/rejected": -0.2686384916305542,
|
| 549 |
"step": 350
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.40654997176736307,
|
| 553 |
+
"grad_norm": 9.5625,
|
| 554 |
+
"learning_rate": 7.402540567492336e-07,
|
| 555 |
+
"logits/chosen": -2.9905340671539307,
|
| 556 |
+
"logits/rejected": -2.95076060295105,
|
| 557 |
+
"logps/chosen": -243.1343536376953,
|
| 558 |
+
"logps/rejected": -226.71694946289062,
|
| 559 |
+
"loss": 0.6703,
|
| 560 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 561 |
+
"rewards/chosen": -0.22629483044147491,
|
| 562 |
+
"rewards/margins": 0.06099146604537964,
|
| 563 |
+
"rewards/rejected": -0.28728628158569336,
|
| 564 |
"step": 360
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 0.4178430265386787,
|
| 568 |
+
"grad_norm": 10.125,
|
| 569 |
+
"learning_rate": 7.227652612113213e-07,
|
| 570 |
+
"logits/chosen": -2.8893442153930664,
|
| 571 |
+
"logits/rejected": -2.852905035018921,
|
| 572 |
+
"logps/chosen": -266.652099609375,
|
| 573 |
+
"logps/rejected": -236.76708984375,
|
| 574 |
+
"loss": 0.6789,
|
| 575 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 576 |
+
"rewards/chosen": -0.2726292014122009,
|
| 577 |
+
"rewards/margins": 0.046290718019008636,
|
| 578 |
+
"rewards/rejected": -0.31891992688179016,
|
| 579 |
"step": 370
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.42913608130999437,
|
| 583 |
+
"grad_norm": 10.4375,
|
| 584 |
+
"learning_rate": 7.049295174229328e-07,
|
| 585 |
+
"logits/chosen": -2.9600396156311035,
|
| 586 |
+
"logits/rejected": -2.916116237640381,
|
| 587 |
+
"logps/chosen": -279.30035400390625,
|
| 588 |
+
"logps/rejected": -240.49032592773438,
|
| 589 |
+
"loss": 0.6625,
|
| 590 |
"rewards/accuracies": 0.59375,
|
| 591 |
+
"rewards/chosen": -0.2531939148902893,
|
| 592 |
+
"rewards/margins": 0.0769796222448349,
|
| 593 |
+
"rewards/rejected": -0.3301735520362854,
|
| 594 |
"step": 380
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 0.44042913608131,
|
| 598 |
+
"grad_norm": 11.8125,
|
| 599 |
+
"learning_rate": 6.867746038629462e-07,
|
| 600 |
+
"logits/chosen": -2.9024429321289062,
|
| 601 |
+
"logits/rejected": -2.8742191791534424,
|
| 602 |
+
"logps/chosen": -242.5234375,
|
| 603 |
+
"logps/rejected": -207.02584838867188,
|
| 604 |
+
"loss": 0.6728,
|
| 605 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 606 |
+
"rewards/chosen": -0.26158207654953003,
|
| 607 |
+
"rewards/margins": 0.05569303780794144,
|
| 608 |
+
"rewards/rejected": -0.31727513670921326,
|
| 609 |
"step": 390
|
| 610 |
},
|
| 611 |
{
|
| 612 |
"epoch": 0.45172219085262566,
|
| 613 |
+
"grad_norm": 9.0625,
|
| 614 |
+
"learning_rate": 6.683287961047741e-07,
|
| 615 |
+
"logits/chosen": -2.9038491249084473,
|
| 616 |
+
"logits/rejected": -2.8683886528015137,
|
| 617 |
+
"logps/chosen": -287.3917541503906,
|
| 618 |
+
"logps/rejected": -234.09072875976562,
|
| 619 |
+
"loss": 0.6692,
|
| 620 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 621 |
+
"rewards/chosen": -0.23289862275123596,
|
| 622 |
+
"rewards/margins": 0.06903888285160065,
|
| 623 |
+
"rewards/rejected": -0.3019375205039978,
|
| 624 |
"step": 400
|
| 625 |
},
|
| 626 |
{
|
| 627 |
"epoch": 0.46301524562394125,
|
| 628 |
+
"grad_norm": 13.125,
|
| 629 |
+
"learning_rate": 6.496208227782556e-07,
|
| 630 |
+
"logits/chosen": -2.8886725902557373,
|
| 631 |
+
"logits/rejected": -2.8595938682556152,
|
| 632 |
+
"logps/chosen": -253.18283081054688,
|
| 633 |
+
"logps/rejected": -246.322509765625,
|
| 634 |
+
"loss": 0.6765,
|
| 635 |
+
"rewards/accuracies": 0.53125,
|
| 636 |
+
"rewards/chosen": -0.27256402373313904,
|
| 637 |
+
"rewards/margins": 0.05145202949643135,
|
| 638 |
+
"rewards/rejected": -0.3240160644054413,
|
| 639 |
"step": 410
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.4743083003952569,
|
| 643 |
+
"grad_norm": 10.375,
|
| 644 |
+
"learning_rate": 6.306798208259297e-07,
|
| 645 |
+
"logits/chosen": -2.9606752395629883,
|
| 646 |
+
"logits/rejected": -2.9306626319885254,
|
| 647 |
+
"logps/chosen": -253.44491577148438,
|
| 648 |
+
"logps/rejected": -240.0795135498047,
|
| 649 |
+
"loss": 0.6778,
|
| 650 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 651 |
+
"rewards/chosen": -0.2641372084617615,
|
| 652 |
+
"rewards/margins": 0.0465705581009388,
|
| 653 |
+
"rewards/rejected": -0.31070777773857117,
|
| 654 |
"step": 420
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.48560135516657255,
|
| 658 |
+
"grad_norm": 10.8125,
|
| 659 |
+
"learning_rate": 6.115352901233778e-07,
|
| 660 |
+
"logits/chosen": -2.9563944339752197,
|
| 661 |
+
"logits/rejected": -2.9226722717285156,
|
| 662 |
+
"logps/chosen": -275.2793273925781,
|
| 663 |
+
"logps/rejected": -250.719482421875,
|
| 664 |
+
"loss": 0.6735,
|
| 665 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 666 |
+
"rewards/chosen": -0.27404916286468506,
|
| 667 |
+
"rewards/margins": 0.056207090616226196,
|
| 668 |
+
"rewards/rejected": -0.33025628328323364,
|
| 669 |
"step": 430
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 0.4968944099378882,
|
| 673 |
+
"grad_norm": 8.9375,
|
| 674 |
+
"learning_rate": 5.922170475343124e-07,
|
| 675 |
+
"logits/chosen": -2.9221410751342773,
|
| 676 |
+
"logits/rejected": -2.891664505004883,
|
| 677 |
+
"logps/chosen": -252.1197509765625,
|
| 678 |
+
"logps/rejected": -239.79946899414062,
|
| 679 |
+
"loss": 0.6643,
|
| 680 |
+
"rewards/accuracies": 0.612500011920929,
|
| 681 |
+
"rewards/chosen": -0.2773807644844055,
|
| 682 |
+
"rewards/margins": 0.08373435586690903,
|
| 683 |
+
"rewards/rejected": -0.36111515760421753,
|
| 684 |
"step": 440
|
| 685 |
},
|
| 686 |
{
|
| 687 |
"epoch": 0.5081874647092038,
|
| 688 |
+
"grad_norm": 8.0,
|
| 689 |
+
"learning_rate": 5.727551804719693e-07,
|
| 690 |
+
"logits/chosen": -2.9440019130706787,
|
| 691 |
+
"logits/rejected": -2.9178504943847656,
|
| 692 |
+
"logps/chosen": -240.7253875732422,
|
| 693 |
+
"logps/rejected": -245.9892120361328,
|
| 694 |
+
"loss": 0.6604,
|
| 695 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 696 |
+
"rewards/chosen": -0.2647743225097656,
|
| 697 |
+
"rewards/margins": 0.08178414404392242,
|
| 698 |
+
"rewards/rejected": -0.34655848145484924,
|
| 699 |
"step": 450
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.5194805194805194,
|
| 703 |
+
"grad_norm": 9.9375,
|
| 704 |
+
"learning_rate": 5.531800000391275e-07,
|
| 705 |
+
"logits/chosen": -2.9381301403045654,
|
| 706 |
+
"logits/rejected": -2.899775743484497,
|
| 707 |
+
"logps/chosen": -253.4866485595703,
|
| 708 |
+
"logps/rejected": -227.26168823242188,
|
| 709 |
+
"loss": 0.6578,
|
| 710 |
+
"rewards/accuracies": 0.612500011920929,
|
| 711 |
+
"rewards/chosen": -0.2723321318626404,
|
| 712 |
+
"rewards/margins": 0.09097979962825775,
|
| 713 |
+
"rewards/rejected": -0.3633119463920593,
|
| 714 |
"step": 460
|
| 715 |
},
|
| 716 |
{
|
| 717 |
"epoch": 0.5307735742518351,
|
| 718 |
+
"grad_norm": 9.625,
|
| 719 |
+
"learning_rate": 5.335219938197445e-07,
|
| 720 |
+
"logits/chosen": -2.9399991035461426,
|
| 721 |
+
"logits/rejected": -2.903505802154541,
|
| 722 |
+
"logps/chosen": -246.1412353515625,
|
| 723 |
+
"logps/rejected": -221.0483856201172,
|
| 724 |
+
"loss": 0.6679,
|
| 725 |
+
"rewards/accuracies": 0.53125,
|
| 726 |
+
"rewards/chosen": -0.24423566460609436,
|
| 727 |
+
"rewards/margins": 0.06885796785354614,
|
| 728 |
+
"rewards/rejected": -0.3130936622619629,
|
| 729 |
"step": 470
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.5420666290231507,
|
| 733 |
+
"grad_norm": 8.8125,
|
| 734 |
+
"learning_rate": 5.138117783957261e-07,
|
| 735 |
+
"logits/chosen": -2.9549431800842285,
|
| 736 |
+
"logits/rejected": -2.9279115200042725,
|
| 737 |
+
"logps/chosen": -258.9583435058594,
|
| 738 |
+
"logps/rejected": -230.93106079101562,
|
| 739 |
+
"loss": 0.6651,
|
| 740 |
+
"rewards/accuracies": 0.5625,
|
| 741 |
+
"rewards/chosen": -0.28696444630622864,
|
| 742 |
+
"rewards/margins": 0.07692141085863113,
|
| 743 |
+
"rewards/rejected": -0.3638857901096344,
|
| 744 |
"step": 480
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.5533596837944664,
|
| 748 |
+
"grad_norm": 11.25,
|
| 749 |
+
"learning_rate": 4.940800516627885e-07,
|
| 750 |
+
"logits/chosen": -2.9042859077453613,
|
| 751 |
+
"logits/rejected": -2.8782927989959717,
|
| 752 |
+
"logps/chosen": -259.2146911621094,
|
| 753 |
+
"logps/rejected": -238.78005981445312,
|
| 754 |
+
"loss": 0.6506,
|
| 755 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 756 |
+
"rewards/chosen": -0.25122493505477905,
|
| 757 |
+
"rewards/margins": 0.10116108506917953,
|
| 758 |
+
"rewards/rejected": -0.3523860573768616,
|
| 759 |
"step": 490
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.564652738565782,
|
| 763 |
+
"grad_norm": 8.125,
|
| 764 |
+
"learning_rate": 4.743575450196772e-07,
|
| 765 |
+
"logits/chosen": -2.9411609172821045,
|
| 766 |
+
"logits/rejected": -2.9064881801605225,
|
| 767 |
+
"logps/chosen": -245.3742218017578,
|
| 768 |
+
"logps/rejected": -242.74740600585938,
|
| 769 |
+
"loss": 0.6671,
|
| 770 |
+
"rewards/accuracies": 0.53125,
|
| 771 |
+
"rewards/chosen": -0.29283052682876587,
|
| 772 |
+
"rewards/margins": 0.07133854180574417,
|
| 773 |
+
"rewards/rejected": -0.36416909098625183,
|
| 774 |
"step": 500
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 0.5759457933370977,
|
| 778 |
+
"grad_norm": 9.0,
|
| 779 |
+
"learning_rate": 4.5467497550520505e-07,
|
| 780 |
+
"logits/chosen": -2.894622325897217,
|
| 781 |
+
"logits/rejected": -2.8733580112457275,
|
| 782 |
+
"logps/chosen": -222.1239471435547,
|
| 783 |
+
"logps/rejected": -213.3123779296875,
|
| 784 |
+
"loss": 0.6705,
|
| 785 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 786 |
+
"rewards/chosen": -0.2835238575935364,
|
| 787 |
+
"rewards/margins": 0.060517072677612305,
|
| 788 |
+
"rewards/rejected": -0.3440409302711487,
|
| 789 |
"step": 510
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 0.5872388481084133,
|
| 793 |
+
"grad_norm": 9.4375,
|
| 794 |
+
"learning_rate": 4.350629979576569e-07,
|
| 795 |
+
"logits/chosen": -2.9461731910705566,
|
| 796 |
+
"logits/rejected": -2.8937017917633057,
|
| 797 |
+
"logps/chosen": -232.8549346923828,
|
| 798 |
+
"logps/rejected": -213.99911499023438,
|
| 799 |
+
"loss": 0.6405,
|
| 800 |
+
"rewards/accuracies": 0.675000011920929,
|
| 801 |
+
"rewards/chosen": -0.27412381768226624,
|
| 802 |
+
"rewards/margins": 0.13067342340946198,
|
| 803 |
+
"rewards/rejected": -0.404797226190567,
|
| 804 |
"step": 520
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 0.598531902879729,
|
| 808 |
"grad_norm": 10.0,
|
| 809 |
+
"learning_rate": 4.155521572710684e-07,
|
| 810 |
+
"logits/chosen": -2.9578232765197754,
|
| 811 |
+
"logits/rejected": -2.920867919921875,
|
| 812 |
+
"logps/chosen": -230.7373809814453,
|
| 813 |
+
"logps/rejected": -210.01296997070312,
|
| 814 |
+
"loss": 0.6765,
|
| 815 |
"rewards/accuracies": 0.518750011920929,
|
| 816 |
+
"rewards/chosen": -0.27560561895370483,
|
| 817 |
+
"rewards/margins": 0.05005854368209839,
|
| 818 |
+
"rewards/rejected": -0.325664222240448,
|
| 819 |
"step": 530
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.6098249576510446,
|
| 823 |
+
"grad_norm": 10.375,
|
| 824 |
+
"learning_rate": 3.9617284082273836e-07,
|
| 825 |
+
"logits/chosen": -2.9541661739349365,
|
| 826 |
+
"logits/rejected": -2.905620574951172,
|
| 827 |
+
"logps/chosen": -233.76467895507812,
|
| 828 |
+
"logps/rejected": -225.522705078125,
|
| 829 |
+
"loss": 0.6498,
|
| 830 |
+
"rewards/accuracies": 0.625,
|
| 831 |
+
"rewards/chosen": -0.2639801800251007,
|
| 832 |
+
"rewards/margins": 0.10957106202840805,
|
| 833 |
+
"rewards/rejected": -0.37355121970176697,
|
| 834 |
"step": 540
|
| 835 |
},
|
| 836 |
{
|
| 837 |
"epoch": 0.6211180124223602,
|
| 838 |
+
"grad_norm": 10.6875,
|
| 839 |
+
"learning_rate": 3.7695523114606836e-07,
|
| 840 |
+
"logits/chosen": -2.936779499053955,
|
| 841 |
+
"logits/rejected": -2.9056625366210938,
|
| 842 |
+
"logps/chosen": -244.0745086669922,
|
| 843 |
+
"logps/rejected": -241.7302703857422,
|
| 844 |
+
"loss": 0.6858,
|
| 845 |
+
"rewards/accuracies": 0.5,
|
| 846 |
+
"rewards/chosen": -0.3084772527217865,
|
| 847 |
+
"rewards/margins": 0.04285974055528641,
|
| 848 |
+
"rewards/rejected": -0.3513370156288147,
|
| 849 |
"step": 550
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.6324110671936759,
|
| 853 |
+
"grad_norm": 10.75,
|
| 854 |
+
"learning_rate": 3.579292589224374e-07,
|
| 855 |
+
"logits/chosen": -2.9537434577941895,
|
| 856 |
+
"logits/rejected": -2.9143412113189697,
|
| 857 |
+
"logps/chosen": -263.7715759277344,
|
| 858 |
+
"logps/rejected": -249.03793334960938,
|
| 859 |
+
"loss": 0.6506,
|
| 860 |
+
"rewards/accuracies": 0.612500011920929,
|
| 861 |
+
"rewards/chosen": -0.28075623512268066,
|
| 862 |
+
"rewards/margins": 0.10373449325561523,
|
| 863 |
+
"rewards/rejected": -0.3844907283782959,
|
| 864 |
"step": 560
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.6437041219649915,
|
| 868 |
+
"grad_norm": 9.4375,
|
| 869 |
+
"learning_rate": 3.391245563653276e-07,
|
| 870 |
+
"logits/chosen": -2.957152843475342,
|
| 871 |
+
"logits/rejected": -2.9082839488983154,
|
| 872 |
+
"logps/chosen": -252.3236846923828,
|
| 873 |
+
"logps/rejected": -221.5681610107422,
|
| 874 |
+
"loss": 0.6473,
|
| 875 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 876 |
+
"rewards/chosen": -0.2503497004508972,
|
| 877 |
+
"rewards/margins": 0.11741576343774796,
|
| 878 |
+
"rewards/rejected": -0.36776548624038696,
|
| 879 |
"step": 570
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 0.6549971767363072,
|
| 883 |
+
"grad_norm": 11.25,
|
| 884 |
+
"learning_rate": 3.20570411069301e-07,
|
| 885 |
+
"logits/chosen": -2.947129487991333,
|
| 886 |
+
"logits/rejected": -2.910438060760498,
|
| 887 |
+
"logps/chosen": -247.65579223632812,
|
| 888 |
+
"logps/rejected": -219.4207763671875,
|
| 889 |
+
"loss": 0.6602,
|
| 890 |
+
"rewards/accuracies": 0.574999988079071,
|
| 891 |
+
"rewards/chosen": -0.34559428691864014,
|
| 892 |
+
"rewards/margins": 0.08174686133861542,
|
| 893 |
+
"rewards/rejected": -0.4273412227630615,
|
| 894 |
"step": 580
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 0.6662902315076228,
|
| 898 |
+
"grad_norm": 9.9375,
|
| 899 |
+
"learning_rate": 3.0229572039570826e-07,
|
| 900 |
+
"logits/chosen": -2.9612464904785156,
|
| 901 |
+
"logits/rejected": -2.912680149078369,
|
| 902 |
+
"logps/chosen": -237.4712677001953,
|
| 903 |
+
"logps/rejected": -232.04373168945312,
|
| 904 |
+
"loss": 0.653,
|
| 905 |
+
"rewards/accuracies": 0.606249988079071,
|
| 906 |
+
"rewards/chosen": -0.2624071538448334,
|
| 907 |
+
"rewards/margins": 0.10756198316812515,
|
| 908 |
+
"rewards/rejected": -0.3699691891670227,
|
| 909 |
"step": 590
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 0.6775832862789385,
|
| 913 |
+
"grad_norm": 11.125,
|
| 914 |
+
"learning_rate": 2.8432894646616886e-07,
|
| 915 |
+
"logits/chosen": -2.9047579765319824,
|
| 916 |
+
"logits/rejected": -2.882847547531128,
|
| 917 |
+
"logps/chosen": -231.67147827148438,
|
| 918 |
+
"logps/rejected": -236.0974884033203,
|
| 919 |
+
"loss": 0.6688,
|
| 920 |
+
"rewards/accuracies": 0.574999988079071,
|
| 921 |
+
"rewards/chosen": -0.3304700553417206,
|
| 922 |
+
"rewards/margins": 0.07191035896539688,
|
| 923 |
+
"rewards/rejected": -0.40238040685653687,
|
| 924 |
"step": 600
|
| 925 |
},
|
| 926 |
{
|
| 927 |
"epoch": 0.6888763410502541,
|
| 928 |
+
"grad_norm": 7.4375,
|
| 929 |
+
"learning_rate": 2.6669807183392105e-07,
|
| 930 |
+
"logits/chosen": -2.92600679397583,
|
| 931 |
+
"logits/rejected": -2.8905367851257324,
|
| 932 |
+
"logps/chosen": -236.2364044189453,
|
| 933 |
+
"logps/rejected": -206.38174438476562,
|
| 934 |
+
"loss": 0.6684,
|
| 935 |
+
"rewards/accuracies": 0.550000011920929,
|
| 936 |
+
"rewards/chosen": -0.28464430570602417,
|
| 937 |
+
"rewards/margins": 0.07026463001966476,
|
| 938 |
+
"rewards/rejected": -0.35490894317626953,
|
| 939 |
"step": 610
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 0.7001693958215698,
|
| 943 |
+
"grad_norm": 10.8125,
|
| 944 |
+
"learning_rate": 2.494305559020822e-07,
|
| 945 |
+
"logits/chosen": -2.9433751106262207,
|
| 946 |
+
"logits/rejected": -2.907177686691284,
|
| 947 |
+
"logps/chosen": -240.6074676513672,
|
| 948 |
+
"logps/rejected": -226.85989379882812,
|
| 949 |
+
"loss": 0.6727,
|
| 950 |
+
"rewards/accuracies": 0.518750011920929,
|
| 951 |
+
"rewards/chosen": -0.3261820673942566,
|
| 952 |
+
"rewards/margins": 0.06915761530399323,
|
| 953 |
+
"rewards/rejected": -0.3953397274017334,
|
| 954 |
"step": 620
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.7114624505928854,
|
| 958 |
+
"grad_norm": 11.0,
|
| 959 |
+
"learning_rate": 2.3255329215669184e-07,
|
| 960 |
+
"logits/chosen": -2.950742244720459,
|
| 961 |
+
"logits/rejected": -2.920048952102661,
|
| 962 |
+
"logps/chosen": -244.8520965576172,
|
| 963 |
+
"logps/rejected": -253.5353546142578,
|
| 964 |
+
"loss": 0.6715,
|
| 965 |
+
"rewards/accuracies": 0.5625,
|
| 966 |
+
"rewards/chosen": -0.2790936529636383,
|
| 967 |
+
"rewards/margins": 0.06808016449213028,
|
| 968 |
+
"rewards/rejected": -0.3471737802028656,
|
| 969 |
"step": 630
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.722755505364201,
|
| 973 |
+
"grad_norm": 11.0625,
|
| 974 |
+
"learning_rate": 2.1609256628115312e-07,
|
| 975 |
+
"logits/chosen": -2.9226107597351074,
|
| 976 |
+
"logits/rejected": -2.8994510173797607,
|
| 977 |
+
"logps/chosen": -228.3145751953125,
|
| 978 |
+
"logps/rejected": -198.56753540039062,
|
| 979 |
+
"loss": 0.6619,
|
| 980 |
"rewards/accuracies": 0.5687500238418579,
|
| 981 |
+
"rewards/chosen": -0.284686803817749,
|
| 982 |
+
"rewards/margins": 0.08213174343109131,
|
| 983 |
+
"rewards/rejected": -0.3668185770511627,
|
| 984 |
"step": 640
|
| 985 |
},
|
| 986 |
{
|
| 987 |
"epoch": 0.7340485601355167,
|
| 988 |
+
"grad_norm": 13.5625,
|
| 989 |
+
"learning_rate": 2.000740152172986e-07,
|
| 990 |
+
"logits/chosen": -2.9056789875030518,
|
| 991 |
+
"logits/rejected": -2.8870625495910645,
|
| 992 |
+
"logps/chosen": -245.57168579101562,
|
| 993 |
+
"logps/rejected": -241.86618041992188,
|
| 994 |
+
"loss": 0.6735,
|
| 995 |
"rewards/accuracies": 0.5562499761581421,
|
| 996 |
+
"rewards/chosen": -0.31977060437202454,
|
| 997 |
+
"rewards/margins": 0.060752321034669876,
|
| 998 |
+
"rewards/rejected": -0.3805229663848877,
|
| 999 |
"step": 650
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 0.7453416149068323,
|
| 1003 |
+
"grad_norm": 10.375,
|
| 1004 |
+
"learning_rate": 1.8452258723684995e-07,
|
| 1005 |
+
"logits/chosen": -2.9322152137756348,
|
| 1006 |
+
"logits/rejected": -2.8824894428253174,
|
| 1007 |
+
"logps/chosen": -228.6182098388672,
|
| 1008 |
+
"logps/rejected": -198.11158752441406,
|
| 1009 |
+
"loss": 0.6541,
|
| 1010 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 1011 |
+
"rewards/chosen": -0.30365481972694397,
|
| 1012 |
+
"rewards/margins": 0.10447897762060165,
|
| 1013 |
+
"rewards/rejected": -0.4081338346004486,
|
| 1014 |
"step": 660
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
"epoch": 0.756634669678148,
|
| 1018 |
+
"grad_norm": 9.8125,
|
| 1019 |
+
"learning_rate": 1.6946250308545124e-07,
|
| 1020 |
+
"logits/chosen": -2.9027180671691895,
|
| 1021 |
+
"logits/rejected": -2.8488707542419434,
|
| 1022 |
+
"logps/chosen": -238.1764678955078,
|
| 1023 |
+
"logps/rejected": -241.59326171875,
|
| 1024 |
+
"loss": 0.647,
|
| 1025 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1026 |
+
"rewards/chosen": -0.2668387293815613,
|
| 1027 |
+
"rewards/margins": 0.11640632152557373,
|
| 1028 |
+
"rewards/rejected": -0.383245050907135,
|
| 1029 |
"step": 670
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
"epoch": 0.7679277244494636,
|
| 1033 |
+
"grad_norm": 11.3125,
|
| 1034 |
+
"learning_rate": 1.54917218259799e-07,
|
| 1035 |
+
"logits/chosen": -2.941662549972534,
|
| 1036 |
+
"logits/rejected": -2.894787073135376,
|
| 1037 |
+
"logps/chosen": -256.3278503417969,
|
| 1038 |
+
"logps/rejected": -216.21945190429688,
|
| 1039 |
+
"loss": 0.667,
|
| 1040 |
+
"rewards/accuracies": 0.59375,
|
| 1041 |
+
"rewards/chosen": -0.3198787271976471,
|
| 1042 |
+
"rewards/margins": 0.07660157978534698,
|
| 1043 |
+
"rewards/rejected": -0.3964803218841553,
|
| 1044 |
"step": 680
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.7792207792207793,
|
| 1048 |
+
"grad_norm": 10.375,
|
| 1049 |
+
"learning_rate": 1.409093864766146e-07,
|
| 1050 |
+
"logits/chosen": -2.9099855422973633,
|
| 1051 |
+
"logits/rejected": -2.8785247802734375,
|
| 1052 |
+
"logps/chosen": -238.84249877929688,
|
| 1053 |
+
"logps/rejected": -240.7554473876953,
|
| 1054 |
+
"loss": 0.6672,
|
| 1055 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 1056 |
+
"rewards/chosen": -0.3247315585613251,
|
| 1057 |
+
"rewards/margins": 0.06632229685783386,
|
| 1058 |
+
"rewards/rejected": -0.3910538852214813,
|
| 1059 |
"step": 690
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.7905138339920948,
|
| 1063 |
+
"grad_norm": 9.0625,
|
| 1064 |
+
"learning_rate": 1.2746082439036114e-07,
|
| 1065 |
+
"logits/chosen": -2.9594428539276123,
|
| 1066 |
+
"logits/rejected": -2.939582347869873,
|
| 1067 |
+
"logps/chosen": -244.93490600585938,
|
| 1068 |
+
"logps/rejected": -220.2227020263672,
|
| 1069 |
+
"loss": 0.6717,
|
| 1070 |
+
"rewards/accuracies": 0.518750011920929,
|
| 1071 |
+
"rewards/chosen": -0.2836626172065735,
|
| 1072 |
+
"rewards/margins": 0.060529135167598724,
|
| 1073 |
+
"rewards/rejected": -0.3441917896270752,
|
| 1074 |
"step": 700
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 0.8018068887634106,
|
| 1078 |
"grad_norm": 9.5625,
|
| 1079 |
+
"learning_rate": 1.1459247761464907e-07,
|
| 1080 |
+
"logits/chosen": -2.9392457008361816,
|
| 1081 |
+
"logits/rejected": -2.9211068153381348,
|
| 1082 |
+
"logps/chosen": -238.0613555908203,
|
| 1083 |
+
"logps/rejected": -216.3132781982422,
|
| 1084 |
+
"loss": 0.6608,
|
| 1085 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 1086 |
+
"rewards/chosen": -0.28733736276626587,
|
| 1087 |
+
"rewards/margins": 0.0887908786535263,
|
| 1088 |
+
"rewards/rejected": -0.37612825632095337,
|
| 1089 |
"step": 710
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 0.8130999435347261,
|
| 1093 |
+
"grad_norm": 8.125,
|
| 1094 |
+
"learning_rate": 1.0232438810025728e-07,
|
| 1095 |
+
"logits/chosen": -2.929105043411255,
|
| 1096 |
+
"logits/rejected": -2.8989386558532715,
|
| 1097 |
+
"logps/chosen": -238.7790985107422,
|
| 1098 |
+
"logps/rejected": -196.06771850585938,
|
| 1099 |
+
"loss": 0.6614,
|
| 1100 |
"rewards/accuracies": 0.581250011920929,
|
| 1101 |
+
"rewards/chosen": -0.2947736084461212,
|
| 1102 |
+
"rewards/margins": 0.08014187216758728,
|
| 1103 |
+
"rewards/rejected": -0.3749154806137085,
|
| 1104 |
"step": 720
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 0.8243929983060417,
|
| 1108 |
+
"grad_norm": 8.25,
|
| 1109 |
+
"learning_rate": 9.067566292057083e-08,
|
| 1110 |
+
"logits/chosen": -2.927602767944336,
|
| 1111 |
+
"logits/rejected": -2.899686098098755,
|
| 1112 |
+
"logps/chosen": -226.2371826171875,
|
| 1113 |
+
"logps/rejected": -250.6548309326172,
|
| 1114 |
+
"loss": 0.6538,
|
| 1115 |
+
"rewards/accuracies": 0.59375,
|
| 1116 |
+
"rewards/chosen": -0.2974669933319092,
|
| 1117 |
+
"rewards/margins": 0.09935982525348663,
|
| 1118 |
+
"rewards/rejected": -0.3968268036842346,
|
| 1119 |
"step": 730
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
"epoch": 0.8356860530773574,
|
| 1123 |
+
"grad_norm": 8.5625,
|
| 1124 |
+
"learning_rate": 7.966444451305727e-08,
|
| 1125 |
+
"logits/chosen": -2.9271061420440674,
|
| 1126 |
+
"logits/rejected": -2.8855221271514893,
|
| 1127 |
+
"logps/chosen": -230.8705596923828,
|
| 1128 |
+
"logps/rejected": -248.7850341796875,
|
| 1129 |
+
"loss": 0.655,
|
| 1130 |
+
"rewards/accuracies": 0.574999988079071,
|
| 1131 |
+
"rewards/chosen": -0.32860639691352844,
|
| 1132 |
+
"rewards/margins": 0.10086989402770996,
|
| 1133 |
+
"rewards/rejected": -0.429476261138916,
|
| 1134 |
"step": 740
|
| 1135 |
},
|
| 1136 |
{
|
| 1137 |
"epoch": 0.846979107848673,
|
| 1138 |
+
"grad_norm": 7.84375,
|
| 1139 |
+
"learning_rate": 6.930788242312252e-08,
|
| 1140 |
+
"logits/chosen": -2.9216761589050293,
|
| 1141 |
+
"logits/rejected": -2.8618099689483643,
|
| 1142 |
+
"logps/chosen": -219.88839721679688,
|
| 1143 |
+
"logps/rejected": -189.08688354492188,
|
| 1144 |
+
"loss": 0.644,
|
| 1145 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1146 |
+
"rewards/chosen": -0.31415051221847534,
|
| 1147 |
+
"rewards/margins": 0.1270967572927475,
|
| 1148 |
+
"rewards/rejected": -0.44124728441238403,
|
| 1149 |
"step": 750
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
"epoch": 0.8582721626199887,
|
| 1153 |
+
"grad_norm": 10.4375,
|
| 1154 |
+
"learning_rate": 5.96221065943609e-08,
|
| 1155 |
+
"logits/chosen": -2.9630675315856934,
|
| 1156 |
+
"logits/rejected": -2.941108226776123,
|
| 1157 |
+
"logps/chosen": -223.5160369873047,
|
| 1158 |
+
"logps/rejected": -232.03451538085938,
|
| 1159 |
+
"loss": 0.6592,
|
| 1160 |
+
"rewards/accuracies": 0.581250011920929,
|
| 1161 |
+
"rewards/chosen": -0.30072498321533203,
|
| 1162 |
+
"rewards/margins": 0.09193964302539825,
|
| 1163 |
+
"rewards/rejected": -0.3926646113395691,
|
| 1164 |
"step": 760
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.8695652173913043,
|
| 1168 |
"grad_norm": 9.6875,
|
| 1169 |
+
"learning_rate": 5.062220224679276e-08,
|
| 1170 |
+
"logits/chosen": -2.936577320098877,
|
| 1171 |
+
"logits/rejected": -2.9232354164123535,
|
| 1172 |
+
"logps/chosen": -229.6407012939453,
|
| 1173 |
+
"logps/rejected": -229.1260223388672,
|
| 1174 |
+
"loss": 0.6702,
|
| 1175 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 1176 |
+
"rewards/chosen": -0.27426886558532715,
|
| 1177 |
+
"rewards/margins": 0.06261973083019257,
|
| 1178 |
+
"rewards/rejected": -0.3368886113166809,
|
| 1179 |
"step": 770
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.88085827216262,
|
| 1183 |
+
"grad_norm": 10.375,
|
| 1184 |
+
"learning_rate": 4.2322186382220295e-08,
|
| 1185 |
+
"logits/chosen": -2.913611650466919,
|
| 1186 |
+
"logits/rejected": -2.8943240642547607,
|
| 1187 |
+
"logps/chosen": -227.52316284179688,
|
| 1188 |
+
"logps/rejected": -237.62997436523438,
|
| 1189 |
+
"loss": 0.6853,
|
| 1190 |
+
"rewards/accuracies": 0.53125,
|
| 1191 |
+
"rewards/chosen": -0.3313713073730469,
|
| 1192 |
+
"rewards/margins": 0.03858897089958191,
|
| 1193 |
+
"rewards/rejected": -0.3699602782726288,
|
| 1194 |
"step": 780
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
"epoch": 0.8921513269339356,
|
| 1198 |
+
"grad_norm": 8.1875,
|
| 1199 |
+
"learning_rate": 3.4734985953290774e-08,
|
| 1200 |
+
"logits/chosen": -2.9326188564300537,
|
| 1201 |
+
"logits/rejected": -2.8966469764709473,
|
| 1202 |
+
"logps/chosen": -255.98684692382812,
|
| 1203 |
+
"logps/rejected": -265.226806640625,
|
| 1204 |
+
"loss": 0.644,
|
| 1205 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1206 |
+
"rewards/chosen": -0.2588121294975281,
|
| 1207 |
+
"rewards/margins": 0.12674114108085632,
|
| 1208 |
+
"rewards/rejected": -0.3855533003807068,
|
| 1209 |
"step": 790
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 0.9034443817052513,
|
| 1213 |
+
"grad_norm": 9.5625,
|
| 1214 |
+
"learning_rate": 2.7872417730269327e-08,
|
| 1215 |
+
"logits/chosen": -2.933467388153076,
|
| 1216 |
+
"logits/rejected": -2.898324728012085,
|
| 1217 |
+
"logps/chosen": -246.2470245361328,
|
| 1218 |
+
"logps/rejected": -263.4474792480469,
|
| 1219 |
+
"loss": 0.6672,
|
| 1220 |
+
"rewards/accuracies": 0.543749988079071,
|
| 1221 |
+
"rewards/chosen": -0.29884645342826843,
|
| 1222 |
+
"rewards/margins": 0.07485932111740112,
|
| 1223 |
+
"rewards/rejected": -0.37370574474334717,
|
| 1224 |
"step": 800
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 0.9147374364765669,
|
| 1228 |
+
"grad_norm": 8.375,
|
| 1229 |
+
"learning_rate": 2.1745169896878412e-08,
|
| 1230 |
+
"logits/chosen": -2.971459150314331,
|
| 1231 |
+
"logits/rejected": -2.9422390460968018,
|
| 1232 |
+
"logps/chosen": -257.4516296386719,
|
| 1233 |
+
"logps/rejected": -244.69961547851562,
|
| 1234 |
+
"loss": 0.6585,
|
| 1235 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 1236 |
+
"rewards/chosen": -0.2682439386844635,
|
| 1237 |
+
"rewards/margins": 0.09004910290241241,
|
| 1238 |
+
"rewards/rejected": -0.3582930564880371,
|
| 1239 |
"step": 810
|
| 1240 |
},
|
| 1241 |
{
|
| 1242 |
"epoch": 0.9260304912478825,
|
| 1243 |
+
"grad_norm": 10.5625,
|
| 1244 |
+
"learning_rate": 1.6362785403865488e-08,
|
| 1245 |
+
"logits/chosen": -2.9577174186706543,
|
| 1246 |
+
"logits/rejected": -2.9316866397857666,
|
| 1247 |
+
"logps/chosen": -236.38302612304688,
|
| 1248 |
+
"logps/rejected": -231.70822143554688,
|
| 1249 |
+
"loss": 0.6696,
|
| 1250 |
"rewards/accuracies": 0.5375000238418579,
|
| 1251 |
+
"rewards/chosen": -0.3118794858455658,
|
| 1252 |
+
"rewards/margins": 0.07028525322675705,
|
| 1253 |
+
"rewards/rejected": -0.38216471672058105,
|
| 1254 |
"step": 820
|
| 1255 |
},
|
| 1256 |
{
|
| 1257 |
"epoch": 0.9373235460191982,
|
| 1258 |
+
"grad_norm": 12.0,
|
| 1259 |
+
"learning_rate": 1.1733647106228373e-08,
|
| 1260 |
+
"logits/chosen": -2.965700149536133,
|
| 1261 |
+
"logits/rejected": -2.935534715652466,
|
| 1262 |
+
"logps/chosen": -241.2149658203125,
|
| 1263 |
+
"logps/rejected": -219.77786254882812,
|
| 1264 |
+
"loss": 0.6709,
|
| 1265 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1266 |
+
"rewards/chosen": -0.3273731470108032,
|
| 1267 |
+
"rewards/margins": 0.060812730342149734,
|
| 1268 |
+
"rewards/rejected": -0.38818588852882385,
|
| 1269 |
"step": 830
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.9486166007905138,
|
| 1273 |
+
"grad_norm": 11.625,
|
| 1274 |
+
"learning_rate": 7.864964707243071e-09,
|
| 1275 |
+
"logits/chosen": -2.9139742851257324,
|
| 1276 |
+
"logits/rejected": -2.8788352012634277,
|
| 1277 |
+
"logps/chosen": -244.0455322265625,
|
| 1278 |
+
"logps/rejected": -233.3373565673828,
|
| 1279 |
+
"loss": 0.6784,
|
| 1280 |
+
"rewards/accuracies": 0.53125,
|
| 1281 |
+
"rewards/chosen": -0.3300473690032959,
|
| 1282 |
+
"rewards/margins": 0.05264363810420036,
|
| 1283 |
+
"rewards/rejected": -0.382690966129303,
|
| 1284 |
"step": 840
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.9599096555618295,
|
| 1288 |
+
"grad_norm": 9.75,
|
| 1289 |
+
"learning_rate": 4.762763529631342e-09,
|
| 1290 |
+
"logits/chosen": -2.956047773361206,
|
| 1291 |
+
"logits/rejected": -2.936790943145752,
|
| 1292 |
+
"logps/chosen": -240.4857177734375,
|
| 1293 |
+
"logps/rejected": -235.99301147460938,
|
| 1294 |
+
"loss": 0.6877,
|
| 1295 |
+
"rewards/accuracies": 0.550000011920929,
|
| 1296 |
+
"rewards/chosen": -0.31692594289779663,
|
| 1297 |
+
"rewards/margins": 0.042497653514146805,
|
| 1298 |
+
"rewards/rejected": -0.35942360758781433,
|
| 1299 |
"step": 850
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
"epoch": 0.9712027103331451,
|
| 1303 |
+
"grad_norm": 10.1875,
|
| 1304 |
+
"learning_rate": 2.431875131354011e-09,
|
| 1305 |
+
"logits/chosen": -2.970977544784546,
|
| 1306 |
+
"logits/rejected": -2.9481375217437744,
|
| 1307 |
+
"logps/chosen": -258.04437255859375,
|
| 1308 |
+
"logps/rejected": -255.94552612304688,
|
| 1309 |
+
"loss": 0.664,
|
| 1310 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 1311 |
+
"rewards/chosen": -0.3285280168056488,
|
| 1312 |
+
"rewards/margins": 0.08176346123218536,
|
| 1313 |
+
"rewards/rejected": -0.41029149293899536,
|
| 1314 |
"step": 860
|
| 1315 |
},
|
| 1316 |
{
|
| 1317 |
"epoch": 0.9824957651044608,
|
| 1318 |
+
"grad_norm": 9.9375,
|
| 1319 |
+
"learning_rate": 8.759297806469335e-10,
|
| 1320 |
+
"logits/chosen": -2.9342379570007324,
|
| 1321 |
+
"logits/rejected": -2.8985533714294434,
|
| 1322 |
+
"logps/chosen": -260.99530029296875,
|
| 1323 |
+
"logps/rejected": -243.7850799560547,
|
| 1324 |
+
"loss": 0.6639,
|
| 1325 |
+
"rewards/accuracies": 0.581250011920929,
|
| 1326 |
+
"rewards/chosen": -0.28514906764030457,
|
| 1327 |
+
"rewards/margins": 0.0856957957148552,
|
| 1328 |
+
"rewards/rejected": -0.37084484100341797,
|
| 1329 |
"step": 870
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 0.9937888198757764,
|
| 1333 |
+
"grad_norm": 9.25,
|
| 1334 |
+
"learning_rate": 9.735080201922486e-11,
|
| 1335 |
+
"logits/chosen": -2.933637857437134,
|
| 1336 |
+
"logits/rejected": -2.88521146774292,
|
| 1337 |
+
"logps/chosen": -285.87249755859375,
|
| 1338 |
+
"logps/rejected": -235.1495819091797,
|
| 1339 |
+
"loss": 0.6537,
|
| 1340 |
+
"rewards/accuracies": 0.574999988079071,
|
| 1341 |
+
"rewards/chosen": -0.32956260442733765,
|
| 1342 |
+
"rewards/margins": 0.10394410043954849,
|
| 1343 |
+
"rewards/rejected": -0.43350666761398315,
|
| 1344 |
"step": 880
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
"epoch": 0.9994353472614342,
|
| 1348 |
"step": 885,
|
| 1349 |
"total_flos": 0.0,
|
| 1350 |
+
"train_loss": 0.6718437324135991,
|
| 1351 |
+
"train_runtime": 9969.8232,
|
| 1352 |
+
"train_samples_per_second": 1.421,
|
| 1353 |
"train_steps_per_second": 0.089
|
| 1354 |
}
|
| 1355 |
],
|