Model save
Browse files- README.md +12 -12
- adapter_model.safetensors +1 -1
- all_results.json +17 -17
- eval_results.json +14 -14
- train_results.json +3 -3
- trainer_state.json +1028 -1028
README.md
CHANGED
|
@@ -15,17 +15,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.
|
| 19 |
-
- Rewards/chosen: -
|
| 20 |
-
- Rewards/rejected: -
|
| 21 |
-
- Rewards/accuracies: 0.
|
| 22 |
-
- Rewards/margins:
|
| 23 |
-
- Logps/rejected: -
|
| 24 |
-
- Logps/chosen: -
|
| 25 |
-
- Logits/rejected:
|
| 26 |
-
- Logits/chosen:
|
| 27 |
-
- Use Label:
|
| 28 |
-
- Pred Label:
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
@@ -62,7 +62,7 @@ The following hyperparameters were used during training:
|
|
| 62 |
|
| 63 |
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Use Label | Pred Label |
|
| 64 |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:---------:|:----------:|
|
| 65 |
-
| 0.
|
| 66 |
|
| 67 |
|
| 68 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.0758
|
| 19 |
+
- Rewards/chosen: -15.6780
|
| 20 |
+
- Rewards/rejected: -27.9661
|
| 21 |
+
- Rewards/accuracies: 0.7080
|
| 22 |
+
- Rewards/margins: 12.2881
|
| 23 |
+
- Logps/rejected: -538.9609
|
| 24 |
+
- Logps/chosen: -441.0378
|
| 25 |
+
- Logits/rejected: -2.2748
|
| 26 |
+
- Logits/chosen: -2.3871
|
| 27 |
+
- Use Label: 2488.2161
|
| 28 |
+
- Pred Label: 13543.7842
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 62 |
|
| 63 |
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Use Label | Pred Label |
|
| 64 |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:---------:|:----------:|
|
| 65 |
+
| 0.069 | 1.0 | 955 | 0.0758 | -15.6780 | -27.9661 | 0.7080 | 12.2881 | -538.9609 | -441.0378 | -2.2748 | -2.3871 | 2447.2161 | 13084.7842 |
|
| 66 |
|
| 67 |
|
| 68 |
### Framework versions
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 218138576
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eeeffd360fdf6ce49d5baeca66af17570725e7e6e43a43b0a4910baa598329f4
|
| 3 |
size 218138576
|
all_results.json
CHANGED
|
@@ -1,23 +1,23 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_logits/chosen":
|
| 4 |
-
"eval_logits/rejected":
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_pred_label":
|
| 9 |
-
"eval_rewards/accuracies": 0.
|
| 10 |
-
"eval_rewards/chosen": -
|
| 11 |
-
"eval_rewards/margins":
|
| 12 |
-
"eval_rewards/rejected": -
|
| 13 |
-
"eval_runtime":
|
| 14 |
"eval_samples": 2000,
|
| 15 |
-
"eval_samples_per_second": 4.
|
| 16 |
-
"eval_steps_per_second": 0.
|
| 17 |
-
"eval_use_label":
|
| 18 |
-
"train_loss": 0.
|
| 19 |
-
"train_runtime":
|
| 20 |
"train_samples": 61135,
|
| 21 |
-
"train_samples_per_second": 2.
|
| 22 |
"train_steps_per_second": 0.038
|
| 23 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.3870656490325928,
|
| 4 |
+
"eval_logits/rejected": -2.274824380874634,
|
| 5 |
+
"eval_logps/chosen": -441.0377502441406,
|
| 6 |
+
"eval_logps/rejected": -538.9608764648438,
|
| 7 |
+
"eval_loss": 0.07583338022232056,
|
| 8 |
+
"eval_pred_label": 13543.7841796875,
|
| 9 |
+
"eval_rewards/accuracies": 0.7080000042915344,
|
| 10 |
+
"eval_rewards/chosen": -15.678034782409668,
|
| 11 |
+
"eval_rewards/margins": 12.288079261779785,
|
| 12 |
+
"eval_rewards/rejected": -27.966114044189453,
|
| 13 |
+
"eval_runtime": 449.5661,
|
| 14 |
"eval_samples": 2000,
|
| 15 |
+
"eval_samples_per_second": 4.449,
|
| 16 |
+
"eval_steps_per_second": 0.278,
|
| 17 |
+
"eval_use_label": 2488.216064453125,
|
| 18 |
+
"train_loss": 0.12920980815488006,
|
| 19 |
+
"train_runtime": 25162.7962,
|
| 20 |
"train_samples": 61135,
|
| 21 |
+
"train_samples_per_second": 2.43,
|
| 22 |
"train_steps_per_second": 0.038
|
| 23 |
}
|
eval_results.json
CHANGED
|
@@ -1,18 +1,18 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_logits/chosen":
|
| 4 |
-
"eval_logits/rejected":
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_pred_label":
|
| 9 |
-
"eval_rewards/accuracies": 0.
|
| 10 |
-
"eval_rewards/chosen": -
|
| 11 |
-
"eval_rewards/margins":
|
| 12 |
-
"eval_rewards/rejected": -
|
| 13 |
-
"eval_runtime":
|
| 14 |
"eval_samples": 2000,
|
| 15 |
-
"eval_samples_per_second": 4.
|
| 16 |
-
"eval_steps_per_second": 0.
|
| 17 |
-
"eval_use_label":
|
| 18 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.3870656490325928,
|
| 4 |
+
"eval_logits/rejected": -2.274824380874634,
|
| 5 |
+
"eval_logps/chosen": -441.0377502441406,
|
| 6 |
+
"eval_logps/rejected": -538.9608764648438,
|
| 7 |
+
"eval_loss": 0.07583338022232056,
|
| 8 |
+
"eval_pred_label": 13543.7841796875,
|
| 9 |
+
"eval_rewards/accuracies": 0.7080000042915344,
|
| 10 |
+
"eval_rewards/chosen": -15.678034782409668,
|
| 11 |
+
"eval_rewards/margins": 12.288079261779785,
|
| 12 |
+
"eval_rewards/rejected": -27.966114044189453,
|
| 13 |
+
"eval_runtime": 449.5661,
|
| 14 |
"eval_samples": 2000,
|
| 15 |
+
"eval_samples_per_second": 4.449,
|
| 16 |
+
"eval_steps_per_second": 0.278,
|
| 17 |
+
"eval_use_label": 2488.216064453125
|
| 18 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 61135,
|
| 6 |
-
"train_samples_per_second": 2.
|
| 7 |
"train_steps_per_second": 0.038
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.12920980815488006,
|
| 4 |
+
"train_runtime": 25162.7962,
|
| 5 |
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 2.43,
|
| 7 |
"train_steps_per_second": 0.038
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -75,1500 +75,1500 @@
|
|
| 75 |
{
|
| 76 |
"epoch": 0.04,
|
| 77 |
"learning_rate": 2.0833333333333336e-05,
|
| 78 |
-
"logits/chosen": -2.
|
| 79 |
-
"logits/rejected": -2.
|
| 80 |
-
"logps/chosen": -281.
|
| 81 |
-
"logps/rejected": -277.
|
| 82 |
-
"loss": 0.
|
| 83 |
-
"pred_label": 0.
|
| 84 |
-
"rewards/accuracies": 0.
|
| 85 |
-
"rewards/chosen": 0.
|
| 86 |
-
"rewards/margins": 0.
|
| 87 |
-
"rewards/rejected": -0.
|
| 88 |
"step": 40,
|
| 89 |
-
"use_label": 561.
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 0.05,
|
| 93 |
"learning_rate": 2.604166666666667e-05,
|
| 94 |
-
"logits/chosen": -2.
|
| 95 |
-
"logits/rejected": -2.
|
| 96 |
-
"logps/chosen": -
|
| 97 |
-
"logps/rejected": -
|
| 98 |
-
"loss": 0.
|
| 99 |
-
"pred_label":
|
| 100 |
-
"rewards/accuracies": 0.
|
| 101 |
-
"rewards/chosen": -0.
|
| 102 |
-
"rewards/margins": 0.
|
| 103 |
-
"rewards/rejected": -0.
|
| 104 |
"step": 50,
|
| 105 |
-
"use_label":
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.06,
|
| 109 |
"learning_rate": 3.125e-05,
|
| 110 |
-
"logits/chosen": -2.
|
| 111 |
-
"logits/rejected": -2.
|
| 112 |
-
"logps/chosen": -
|
| 113 |
-
"logps/rejected": -
|
| 114 |
-
"loss": 0.
|
| 115 |
-
"pred_label":
|
| 116 |
-
"rewards/accuracies": 0.
|
| 117 |
-
"rewards/chosen": -0.
|
| 118 |
-
"rewards/margins": 0.
|
| 119 |
-
"rewards/rejected": -0.
|
| 120 |
"step": 60,
|
| 121 |
-
"use_label":
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07,
|
| 125 |
"learning_rate": 3.6458333333333336e-05,
|
| 126 |
-
"logits/chosen": -2.
|
| 127 |
-
"logits/rejected": -2.
|
| 128 |
-
"logps/chosen": -
|
| 129 |
-
"logps/rejected": -
|
| 130 |
-
"loss": 0.
|
| 131 |
-
"pred_label":
|
| 132 |
-
"rewards/accuracies": 0.
|
| 133 |
-
"rewards/chosen": -0.
|
| 134 |
-
"rewards/margins": 0.
|
| 135 |
-
"rewards/rejected": -
|
| 136 |
"step": 70,
|
| 137 |
-
"use_label":
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 0.08,
|
| 141 |
"learning_rate": 4.166666666666667e-05,
|
| 142 |
-
"logits/chosen": -2.
|
| 143 |
-
"logits/rejected": -2.
|
| 144 |
-
"logps/chosen": -
|
| 145 |
-
"logps/rejected": -
|
| 146 |
-
"loss": 0.
|
| 147 |
-
"pred_label":
|
| 148 |
-
"rewards/accuracies": 0.
|
| 149 |
-
"rewards/chosen": -
|
| 150 |
-
"rewards/margins": 1.
|
| 151 |
-
"rewards/rejected": -
|
| 152 |
"step": 80,
|
| 153 |
-
"use_label":
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.09,
|
| 157 |
"learning_rate": 4.6875e-05,
|
| 158 |
-
"logits/chosen": -2.
|
| 159 |
-
"logits/rejected": -2.
|
| 160 |
-
"logps/chosen": -
|
| 161 |
-
"logps/rejected": -
|
| 162 |
-
"loss": 0.
|
| 163 |
-
"pred_label":
|
| 164 |
-
"rewards/accuracies": 0.
|
| 165 |
-
"rewards/chosen": -
|
| 166 |
-
"rewards/margins":
|
| 167 |
-
"rewards/rejected": -
|
| 168 |
"step": 90,
|
| 169 |
-
"use_label":
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 0.1,
|
| 173 |
"learning_rate": 4.976717112922003e-05,
|
| 174 |
-
"logits/chosen": -2.
|
| 175 |
-
"logits/rejected": -2.
|
| 176 |
-
"logps/chosen": -
|
| 177 |
-
"logps/rejected": -
|
| 178 |
-
"loss": 0.
|
| 179 |
-
"pred_label":
|
| 180 |
-
"rewards/accuracies": 0.
|
| 181 |
-
"rewards/chosen": -
|
| 182 |
-
"rewards/margins":
|
| 183 |
-
"rewards/rejected": -
|
| 184 |
"step": 100,
|
| 185 |
-
"use_label":
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.12,
|
| 189 |
"learning_rate": 4.918509895227008e-05,
|
| 190 |
-
"logits/chosen": -2.
|
| 191 |
-
"logits/rejected": -2.
|
| 192 |
-
"logps/chosen": -
|
| 193 |
-
"logps/rejected": -
|
| 194 |
-
"loss": 0.
|
| 195 |
-
"pred_label":
|
| 196 |
"rewards/accuracies": 0.675000011920929,
|
| 197 |
-
"rewards/chosen": -
|
| 198 |
-
"rewards/margins":
|
| 199 |
-
"rewards/rejected": -
|
| 200 |
"step": 110,
|
| 201 |
-
"use_label":
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 0.13,
|
| 205 |
"learning_rate": 4.860302677532014e-05,
|
| 206 |
-
"logits/chosen": -2.
|
| 207 |
-
"logits/rejected": -2.
|
| 208 |
-
"logps/chosen": -
|
| 209 |
-
"logps/rejected": -
|
| 210 |
-
"loss": 0.
|
| 211 |
-
"pred_label":
|
| 212 |
-
"rewards/accuracies": 0.
|
| 213 |
-
"rewards/chosen": -
|
| 214 |
-
"rewards/margins":
|
| 215 |
-
"rewards/rejected": -
|
| 216 |
"step": 120,
|
| 217 |
-
"use_label":
|
| 218 |
},
|
| 219 |
{
|
| 220 |
"epoch": 0.14,
|
| 221 |
"learning_rate": 4.80209545983702e-05,
|
| 222 |
-
"logits/chosen": -2.
|
| 223 |
-
"logits/rejected": -2.
|
| 224 |
-
"logps/chosen": -
|
| 225 |
-
"logps/rejected": -
|
| 226 |
-
"loss": 0.
|
| 227 |
-
"pred_label":
|
| 228 |
-
"rewards/accuracies": 0.
|
| 229 |
-
"rewards/chosen": -
|
| 230 |
-
"rewards/margins":
|
| 231 |
-
"rewards/rejected": -
|
| 232 |
"step": 130,
|
| 233 |
-
"use_label":
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.15,
|
| 237 |
"learning_rate": 4.743888242142026e-05,
|
| 238 |
-
"logits/chosen": -2.
|
| 239 |
-
"logits/rejected": -2.
|
| 240 |
-
"logps/chosen": -
|
| 241 |
-
"logps/rejected": -
|
| 242 |
-
"loss": 0.
|
| 243 |
-
"pred_label":
|
| 244 |
-
"rewards/accuracies": 0.
|
| 245 |
-
"rewards/chosen": -
|
| 246 |
-
"rewards/margins":
|
| 247 |
-
"rewards/rejected": -
|
| 248 |
"step": 140,
|
| 249 |
-
"use_label":
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.16,
|
| 253 |
"learning_rate": 4.685681024447032e-05,
|
| 254 |
-
"logits/chosen": -2.
|
| 255 |
-
"logits/rejected": -2.
|
| 256 |
-
"logps/chosen": -
|
| 257 |
-
"logps/rejected": -
|
| 258 |
-
"loss": 0.
|
| 259 |
-
"pred_label":
|
| 260 |
-
"rewards/accuracies": 0.
|
| 261 |
-
"rewards/chosen": -
|
| 262 |
-
"rewards/margins":
|
| 263 |
-
"rewards/rejected": -
|
| 264 |
"step": 150,
|
| 265 |
-
"use_label":
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"epoch": 0.17,
|
| 269 |
"learning_rate": 4.6274738067520374e-05,
|
| 270 |
-
"logits/chosen": -
|
| 271 |
-
"logits/rejected": -
|
| 272 |
-
"logps/chosen": -
|
| 273 |
-
"logps/rejected": -
|
| 274 |
-
"loss": 0.
|
| 275 |
-
"pred_label":
|
| 276 |
-
"rewards/accuracies": 0.
|
| 277 |
-
"rewards/chosen": -
|
| 278 |
-
"rewards/margins":
|
| 279 |
-
"rewards/rejected": -
|
| 280 |
"step": 160,
|
| 281 |
-
"use_label":
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 0.18,
|
| 285 |
"learning_rate": 4.5692665890570435e-05,
|
| 286 |
-
"logits/chosen": -
|
| 287 |
-
"logits/rejected": -
|
| 288 |
-
"logps/chosen": -
|
| 289 |
-
"logps/rejected": -
|
| 290 |
-
"loss": 0.
|
| 291 |
-
"pred_label":
|
| 292 |
-
"rewards/accuracies": 0.
|
| 293 |
-
"rewards/chosen": -
|
| 294 |
-
"rewards/margins":
|
| 295 |
-
"rewards/rejected": -
|
| 296 |
"step": 170,
|
| 297 |
-
"use_label":
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.19,
|
| 301 |
"learning_rate": 4.511059371362049e-05,
|
| 302 |
-
"logits/chosen": -
|
| 303 |
-
"logits/rejected": -
|
| 304 |
-
"logps/chosen": -
|
| 305 |
-
"logps/rejected": -
|
| 306 |
-
"loss": 0.
|
| 307 |
-
"pred_label":
|
| 308 |
-
"rewards/accuracies": 0.
|
| 309 |
-
"rewards/chosen": -
|
| 310 |
-
"rewards/margins":
|
| 311 |
-
"rewards/rejected": -
|
| 312 |
"step": 180,
|
| 313 |
-
"use_label":
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"epoch": 0.2,
|
| 317 |
"learning_rate": 4.452852153667055e-05,
|
| 318 |
-
"logits/chosen": -
|
| 319 |
-
"logits/rejected": -
|
| 320 |
-
"logps/chosen": -
|
| 321 |
-
"logps/rejected": -
|
| 322 |
-
"loss": 0.
|
| 323 |
-
"pred_label":
|
| 324 |
-
"rewards/accuracies": 0.
|
| 325 |
-
"rewards/chosen": -
|
| 326 |
-
"rewards/margins":
|
| 327 |
-
"rewards/rejected": -
|
| 328 |
"step": 190,
|
| 329 |
-
"use_label":
|
| 330 |
},
|
| 331 |
{
|
| 332 |
"epoch": 0.21,
|
| 333 |
"learning_rate": 4.394644935972061e-05,
|
| 334 |
-
"logits/chosen": -
|
| 335 |
-
"logits/rejected": -
|
| 336 |
-
"logps/chosen": -
|
| 337 |
-
"logps/rejected": -
|
| 338 |
-
"loss": 0.
|
| 339 |
-
"pred_label":
|
| 340 |
-
"rewards/accuracies": 0.
|
| 341 |
-
"rewards/chosen": -
|
| 342 |
-
"rewards/margins":
|
| 343 |
-
"rewards/rejected": -
|
| 344 |
"step": 200,
|
| 345 |
-
"use_label":
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.22,
|
| 349 |
"learning_rate": 4.336437718277067e-05,
|
| 350 |
-
"logits/chosen": -
|
| 351 |
-
"logits/rejected": -
|
| 352 |
-
"logps/chosen": -
|
| 353 |
-
"logps/rejected": -
|
| 354 |
-
"loss": 0.
|
| 355 |
-
"pred_label":
|
| 356 |
-
"rewards/accuracies": 0.
|
| 357 |
-
"rewards/chosen": -
|
| 358 |
-
"rewards/margins":
|
| 359 |
-
"rewards/rejected": -
|
| 360 |
"step": 210,
|
| 361 |
-
"use_label":
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.23,
|
| 365 |
"learning_rate": 4.278230500582072e-05,
|
| 366 |
-
"logits/chosen": -
|
| 367 |
-
"logits/rejected": -
|
| 368 |
-
"logps/chosen": -
|
| 369 |
-
"logps/rejected": -
|
| 370 |
-
"loss": 0.
|
| 371 |
-
"pred_label":
|
| 372 |
-
"rewards/accuracies": 0.
|
| 373 |
-
"rewards/chosen": -
|
| 374 |
-
"rewards/margins":
|
| 375 |
-
"rewards/rejected": -
|
| 376 |
"step": 220,
|
| 377 |
-
"use_label":
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.24,
|
| 381 |
"learning_rate": 4.220023282887078e-05,
|
| 382 |
-
"logits/chosen": -
|
| 383 |
-
"logits/rejected": -
|
| 384 |
-
"logps/chosen": -
|
| 385 |
-
"logps/rejected": -
|
| 386 |
-
"loss": 0.
|
| 387 |
-
"pred_label":
|
| 388 |
-
"rewards/accuracies": 0.
|
| 389 |
-
"rewards/chosen": -
|
| 390 |
-
"rewards/margins":
|
| 391 |
-
"rewards/rejected": -
|
| 392 |
"step": 230,
|
| 393 |
-
"use_label":
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 0.25,
|
| 397 |
"learning_rate": 4.161816065192084e-05,
|
| 398 |
-
"logits/chosen": -
|
| 399 |
-
"logits/rejected": -
|
| 400 |
-
"logps/chosen": -
|
| 401 |
-
"logps/rejected": -
|
| 402 |
-
"loss": 0.
|
| 403 |
-
"pred_label":
|
| 404 |
-
"rewards/accuracies": 0.
|
| 405 |
-
"rewards/chosen": -
|
| 406 |
-
"rewards/margins":
|
| 407 |
-
"rewards/rejected": -
|
| 408 |
"step": 240,
|
| 409 |
-
"use_label":
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.26,
|
| 413 |
"learning_rate": 4.10360884749709e-05,
|
| 414 |
-
"logits/chosen": -
|
| 415 |
-
"logits/rejected": -
|
| 416 |
-
"logps/chosen": -
|
| 417 |
-
"logps/rejected": -
|
| 418 |
-
"loss": 0.
|
| 419 |
-
"pred_label":
|
| 420 |
-
"rewards/accuracies": 0.
|
| 421 |
-
"rewards/chosen": -
|
| 422 |
-
"rewards/margins":
|
| 423 |
-
"rewards/rejected": -
|
| 424 |
"step": 250,
|
| 425 |
-
"use_label":
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"epoch": 0.27,
|
| 429 |
"learning_rate": 4.045401629802096e-05,
|
| 430 |
-
"logits/chosen": -
|
| 431 |
-
"logits/rejected": -
|
| 432 |
-
"logps/chosen": -
|
| 433 |
-
"logps/rejected": -
|
| 434 |
-
"loss": 0.
|
| 435 |
-
"pred_label":
|
| 436 |
-
"rewards/accuracies": 0.
|
| 437 |
-
"rewards/chosen": -
|
| 438 |
-
"rewards/margins":
|
| 439 |
-
"rewards/rejected": -
|
| 440 |
"step": 260,
|
| 441 |
-
"use_label":
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 0.28,
|
| 445 |
"learning_rate": 3.9871944121071014e-05,
|
| 446 |
-
"logits/chosen": -
|
| 447 |
-
"logits/rejected": -
|
| 448 |
-
"logps/chosen": -
|
| 449 |
-
"logps/rejected": -
|
| 450 |
-
"loss": 0.
|
| 451 |
-
"pred_label":
|
| 452 |
-
"rewards/accuracies": 0.
|
| 453 |
-
"rewards/chosen": -
|
| 454 |
-
"rewards/margins":
|
| 455 |
-
"rewards/rejected": -
|
| 456 |
"step": 270,
|
| 457 |
-
"use_label":
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.29,
|
| 461 |
"learning_rate": 3.928987194412107e-05,
|
| 462 |
-
"logits/chosen": -
|
| 463 |
-
"logits/rejected": -
|
| 464 |
-
"logps/chosen": -
|
| 465 |
-
"logps/rejected": -
|
| 466 |
-
"loss": 0.
|
| 467 |
-
"pred_label":
|
| 468 |
-
"rewards/accuracies": 0.
|
| 469 |
-
"rewards/chosen": -
|
| 470 |
-
"rewards/margins":
|
| 471 |
-
"rewards/rejected": -
|
| 472 |
"step": 280,
|
| 473 |
-
"use_label":
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 0.3,
|
| 477 |
"learning_rate": 3.870779976717113e-05,
|
| 478 |
-
"logits/chosen": -
|
| 479 |
-
"logits/rejected": -
|
| 480 |
-
"logps/chosen": -
|
| 481 |
-
"logps/rejected": -
|
| 482 |
-
"loss": 0.
|
| 483 |
-
"pred_label":
|
| 484 |
-
"rewards/accuracies": 0.
|
| 485 |
-
"rewards/chosen": -
|
| 486 |
-
"rewards/margins":
|
| 487 |
-
"rewards/rejected": -
|
| 488 |
"step": 290,
|
| 489 |
-
"use_label":
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.31,
|
| 493 |
"learning_rate": 3.812572759022119e-05,
|
| 494 |
-
"logits/chosen": -
|
| 495 |
-
"logits/rejected": -
|
| 496 |
-
"logps/chosen": -
|
| 497 |
-
"logps/rejected": -
|
| 498 |
-
"loss": 0.
|
| 499 |
-
"pred_label":
|
| 500 |
-
"rewards/accuracies": 0.
|
| 501 |
-
"rewards/chosen": -
|
| 502 |
-
"rewards/margins":
|
| 503 |
-
"rewards/rejected": -
|
| 504 |
"step": 300,
|
| 505 |
-
"use_label":
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.32,
|
| 509 |
"learning_rate": 3.7543655413271246e-05,
|
| 510 |
-
"logits/chosen": -
|
| 511 |
-
"logits/rejected": -
|
| 512 |
-
"logps/chosen": -
|
| 513 |
-
"logps/rejected": -
|
| 514 |
-
"loss": 0.
|
| 515 |
-
"pred_label":
|
| 516 |
-
"rewards/accuracies": 0.
|
| 517 |
-
"rewards/chosen": -
|
| 518 |
-
"rewards/margins":
|
| 519 |
-
"rewards/rejected": -
|
| 520 |
"step": 310,
|
| 521 |
-
"use_label":
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.33,
|
| 525 |
"learning_rate": 3.696158323632131e-05,
|
| 526 |
-
"logits/chosen": -
|
| 527 |
-
"logits/rejected": -
|
| 528 |
-
"logps/chosen": -
|
| 529 |
-
"logps/rejected": -
|
| 530 |
-
"loss": 0.
|
| 531 |
-
"pred_label":
|
| 532 |
-
"rewards/accuracies": 0.
|
| 533 |
-
"rewards/chosen": -
|
| 534 |
-
"rewards/margins":
|
| 535 |
-
"rewards/rejected": -
|
| 536 |
"step": 320,
|
| 537 |
-
"use_label":
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.35,
|
| 541 |
"learning_rate": 3.637951105937136e-05,
|
| 542 |
-
"logits/chosen": -
|
| 543 |
-
"logits/rejected": -
|
| 544 |
-
"logps/chosen": -
|
| 545 |
-
"logps/rejected": -
|
| 546 |
-
"loss": 0.
|
| 547 |
-
"pred_label":
|
| 548 |
-
"rewards/accuracies": 0.
|
| 549 |
-
"rewards/chosen": -
|
| 550 |
-
"rewards/margins":
|
| 551 |
-
"rewards/rejected": -
|
| 552 |
"step": 330,
|
| 553 |
-
"use_label":
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"epoch": 0.36,
|
| 557 |
"learning_rate": 3.579743888242142e-05,
|
| 558 |
-
"logits/chosen": -
|
| 559 |
-
"logits/rejected": -
|
| 560 |
-
"logps/chosen": -
|
| 561 |
-
"logps/rejected": -
|
| 562 |
-
"loss": 0.
|
| 563 |
-
"pred_label":
|
| 564 |
-
"rewards/accuracies": 0.
|
| 565 |
-
"rewards/chosen": -
|
| 566 |
-
"rewards/margins":
|
| 567 |
-
"rewards/rejected": -
|
| 568 |
"step": 340,
|
| 569 |
-
"use_label":
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.37,
|
| 573 |
"learning_rate": 3.5215366705471484e-05,
|
| 574 |
-
"logits/chosen": -
|
| 575 |
-
"logits/rejected": -
|
| 576 |
-
"logps/chosen": -
|
| 577 |
-
"logps/rejected": -
|
| 578 |
-
"loss": 0.
|
| 579 |
-
"pred_label":
|
| 580 |
-
"rewards/accuracies": 0.
|
| 581 |
-
"rewards/chosen": -
|
| 582 |
-
"rewards/margins":
|
| 583 |
-
"rewards/rejected": -
|
| 584 |
"step": 350,
|
| 585 |
-
"use_label":
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.38,
|
| 589 |
"learning_rate": 3.463329452852154e-05,
|
| 590 |
-
"logits/chosen": -
|
| 591 |
-
"logits/rejected": -
|
| 592 |
-
"logps/chosen": -
|
| 593 |
-
"logps/rejected": -
|
| 594 |
-
"loss": 0.
|
| 595 |
-
"pred_label":
|
| 596 |
-
"rewards/accuracies": 0.
|
| 597 |
-
"rewards/chosen": -
|
| 598 |
-
"rewards/margins":
|
| 599 |
-
"rewards/rejected": -
|
| 600 |
"step": 360,
|
| 601 |
-
"use_label":
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"epoch": 0.39,
|
| 605 |
"learning_rate": 3.40512223515716e-05,
|
| 606 |
-
"logits/chosen": -
|
| 607 |
-
"logits/rejected": -
|
| 608 |
-
"logps/chosen": -
|
| 609 |
-
"logps/rejected": -
|
| 610 |
-
"loss": 0.
|
| 611 |
-
"pred_label":
|
| 612 |
-
"rewards/accuracies": 0.
|
| 613 |
-
"rewards/chosen": -
|
| 614 |
-
"rewards/margins":
|
| 615 |
-
"rewards/rejected": -
|
| 616 |
"step": 370,
|
| 617 |
-
"use_label":
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 0.4,
|
| 621 |
"learning_rate": 3.3469150174621654e-05,
|
| 622 |
-
"logits/chosen": -
|
| 623 |
-
"logits/rejected": -
|
| 624 |
-
"logps/chosen": -
|
| 625 |
-
"logps/rejected": -
|
| 626 |
-
"loss": 0.
|
| 627 |
-
"pred_label":
|
| 628 |
-
"rewards/accuracies": 0.
|
| 629 |
-
"rewards/chosen": -
|
| 630 |
-
"rewards/margins":
|
| 631 |
-
"rewards/rejected": -
|
| 632 |
"step": 380,
|
| 633 |
-
"use_label":
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.41,
|
| 637 |
"learning_rate": 3.288707799767171e-05,
|
| 638 |
-
"logits/chosen": -
|
| 639 |
-
"logits/rejected": -
|
| 640 |
-
"logps/chosen": -
|
| 641 |
-
"logps/rejected": -
|
| 642 |
-
"loss": 0.
|
| 643 |
-
"pred_label":
|
| 644 |
-
"rewards/accuracies": 0.
|
| 645 |
-
"rewards/chosen": -
|
| 646 |
-
"rewards/margins":
|
| 647 |
-
"rewards/rejected": -
|
| 648 |
"step": 390,
|
| 649 |
-
"use_label":
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"epoch": 0.42,
|
| 653 |
"learning_rate": 3.2305005820721776e-05,
|
| 654 |
-
"logits/chosen": -
|
| 655 |
-
"logits/rejected": -
|
| 656 |
-
"logps/chosen": -
|
| 657 |
-
"logps/rejected": -
|
| 658 |
-
"loss": 0.
|
| 659 |
-
"pred_label":
|
| 660 |
-
"rewards/accuracies": 0.
|
| 661 |
-
"rewards/chosen": -
|
| 662 |
-
"rewards/margins":
|
| 663 |
-
"rewards/rejected": -
|
| 664 |
"step": 400,
|
| 665 |
-
"use_label":
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"epoch": 0.43,
|
| 669 |
"learning_rate": 3.172293364377183e-05,
|
| 670 |
-
"logits/chosen": -
|
| 671 |
-
"logits/rejected": -
|
| 672 |
-
"logps/chosen": -
|
| 673 |
-
"logps/rejected": -
|
| 674 |
-
"loss": 0.
|
| 675 |
-
"pred_label":
|
| 676 |
-
"rewards/accuracies": 0.
|
| 677 |
-
"rewards/chosen": -
|
| 678 |
-
"rewards/margins":
|
| 679 |
-
"rewards/rejected": -
|
| 680 |
"step": 410,
|
| 681 |
-
"use_label":
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.44,
|
| 685 |
"learning_rate": 3.1140861466821885e-05,
|
| 686 |
-
"logits/chosen": -
|
| 687 |
-
"logits/rejected": -
|
| 688 |
-
"logps/chosen": -
|
| 689 |
-
"logps/rejected": -
|
| 690 |
-
"loss": 0.
|
| 691 |
-
"pred_label":
|
| 692 |
-
"rewards/accuracies": 0.
|
| 693 |
-
"rewards/chosen": -
|
| 694 |
-
"rewards/margins":
|
| 695 |
-
"rewards/rejected": -
|
| 696 |
"step": 420,
|
| 697 |
-
"use_label":
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.45,
|
| 701 |
"learning_rate": 3.055878928987195e-05,
|
| 702 |
-
"logits/chosen": -
|
| 703 |
-
"logits/rejected": -
|
| 704 |
-
"logps/chosen": -
|
| 705 |
-
"logps/rejected": -
|
| 706 |
-
"loss": 0.
|
| 707 |
-
"pred_label":
|
| 708 |
-
"rewards/accuracies": 0.
|
| 709 |
-
"rewards/chosen": -
|
| 710 |
-
"rewards/margins":
|
| 711 |
-
"rewards/rejected": -
|
| 712 |
"step": 430,
|
| 713 |
-
"use_label":
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.46,
|
| 717 |
"learning_rate": 2.9976717112922005e-05,
|
| 718 |
-
"logits/chosen": -
|
| 719 |
-
"logits/rejected": -
|
| 720 |
-
"logps/chosen": -
|
| 721 |
-
"logps/rejected": -
|
| 722 |
-
"loss": 0.
|
| 723 |
-
"pred_label":
|
| 724 |
-
"rewards/accuracies": 0.
|
| 725 |
-
"rewards/chosen": -
|
| 726 |
-
"rewards/margins":
|
| 727 |
-
"rewards/rejected": -
|
| 728 |
"step": 440,
|
| 729 |
-
"use_label":
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.47,
|
| 733 |
"learning_rate": 2.939464493597206e-05,
|
| 734 |
-
"logits/chosen": -
|
| 735 |
-
"logits/rejected": -
|
| 736 |
-
"logps/chosen": -
|
| 737 |
-
"logps/rejected": -
|
| 738 |
-
"loss": 0.
|
| 739 |
-
"pred_label":
|
| 740 |
-
"rewards/accuracies": 0.
|
| 741 |
-
"rewards/chosen": -
|
| 742 |
-
"rewards/margins":
|
| 743 |
-
"rewards/rejected": -
|
| 744 |
"step": 450,
|
| 745 |
-
"use_label":
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.48,
|
| 749 |
"learning_rate": 2.881257275902212e-05,
|
| 750 |
-
"logits/chosen": -
|
| 751 |
-
"logits/rejected": -
|
| 752 |
-
"logps/chosen": -
|
| 753 |
-
"logps/rejected": -
|
| 754 |
-
"loss": 0.
|
| 755 |
-
"pred_label":
|
| 756 |
-
"rewards/accuracies": 0.
|
| 757 |
-
"rewards/chosen": -
|
| 758 |
-
"rewards/margins":
|
| 759 |
-
"rewards/rejected": -
|
| 760 |
"step": 460,
|
| 761 |
-
"use_label":
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"epoch": 0.49,
|
| 765 |
"learning_rate": 2.8230500582072178e-05,
|
| 766 |
-
"logits/chosen": -
|
| 767 |
-
"logits/rejected": -
|
| 768 |
-
"logps/chosen": -
|
| 769 |
-
"logps/rejected": -
|
| 770 |
-
"loss": 0.
|
| 771 |
-
"pred_label":
|
| 772 |
-
"rewards/accuracies": 0.
|
| 773 |
-
"rewards/chosen": -
|
| 774 |
-
"rewards/margins":
|
| 775 |
-
"rewards/rejected": -
|
| 776 |
"step": 470,
|
| 777 |
-
"use_label":
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.5,
|
| 781 |
"learning_rate": 2.7648428405122233e-05,
|
| 782 |
-
"logits/chosen": -
|
| 783 |
-
"logits/rejected": -
|
| 784 |
-
"logps/chosen": -
|
| 785 |
-
"logps/rejected": -
|
| 786 |
-
"loss": 0.
|
| 787 |
-
"pred_label":
|
| 788 |
-
"rewards/accuracies": 0.
|
| 789 |
-
"rewards/chosen": -
|
| 790 |
-
"rewards/margins":
|
| 791 |
-
"rewards/rejected": -
|
| 792 |
"step": 480,
|
| 793 |
-
"use_label":
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.51,
|
| 797 |
"learning_rate": 2.7066356228172297e-05,
|
| 798 |
-
"logits/chosen": -
|
| 799 |
-
"logits/rejected": -
|
| 800 |
-
"logps/chosen": -
|
| 801 |
-
"logps/rejected": -
|
| 802 |
-
"loss": 0.
|
| 803 |
-
"pred_label":
|
| 804 |
-
"rewards/accuracies": 0.
|
| 805 |
-
"rewards/chosen": -
|
| 806 |
-
"rewards/margins":
|
| 807 |
-
"rewards/rejected": -
|
| 808 |
"step": 490,
|
| 809 |
-
"use_label":
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 0.52,
|
| 813 |
"learning_rate": 2.6484284051222352e-05,
|
| 814 |
-
"logits/chosen": -
|
| 815 |
-
"logits/rejected": -
|
| 816 |
-
"logps/chosen": -
|
| 817 |
-
"logps/rejected": -
|
| 818 |
-
"loss": 0.
|
| 819 |
-
"pred_label":
|
| 820 |
-
"rewards/accuracies": 0.
|
| 821 |
-
"rewards/chosen": -
|
| 822 |
-
"rewards/margins":
|
| 823 |
-
"rewards/rejected": -
|
| 824 |
"step": 500,
|
| 825 |
-
"use_label":
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 0.53,
|
| 829 |
"learning_rate": 2.590221187427241e-05,
|
| 830 |
-
"logits/chosen": -
|
| 831 |
-
"logits/rejected": -
|
| 832 |
-
"logps/chosen": -
|
| 833 |
-
"logps/rejected": -
|
| 834 |
-
"loss": 0.
|
| 835 |
-
"pred_label":
|
| 836 |
-
"rewards/accuracies": 0.
|
| 837 |
-
"rewards/chosen": -
|
| 838 |
-
"rewards/margins":
|
| 839 |
-
"rewards/rejected": -
|
| 840 |
"step": 510,
|
| 841 |
-
"use_label":
|
| 842 |
},
|
| 843 |
{
|
| 844 |
"epoch": 0.54,
|
| 845 |
"learning_rate": 2.532013969732247e-05,
|
| 846 |
-
"logits/chosen": -
|
| 847 |
-
"logits/rejected": -
|
| 848 |
-
"logps/chosen": -
|
| 849 |
-
"logps/rejected": -
|
| 850 |
-
"loss": 0.
|
| 851 |
-
"pred_label":
|
| 852 |
-
"rewards/accuracies": 0.
|
| 853 |
-
"rewards/chosen": -
|
| 854 |
-
"rewards/margins":
|
| 855 |
-
"rewards/rejected": -
|
| 856 |
"step": 520,
|
| 857 |
-
"use_label":
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.55,
|
| 861 |
"learning_rate": 2.4738067520372525e-05,
|
| 862 |
-
"logits/chosen": -
|
| 863 |
-
"logits/rejected": -
|
| 864 |
-
"logps/chosen": -
|
| 865 |
-
"logps/rejected": -
|
| 866 |
-
"loss": 0.
|
| 867 |
-
"pred_label":
|
| 868 |
-
"rewards/accuracies": 0.
|
| 869 |
-
"rewards/chosen": -
|
| 870 |
-
"rewards/margins":
|
| 871 |
-
"rewards/rejected": -
|
| 872 |
"step": 530,
|
| 873 |
-
"use_label":
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.57,
|
| 877 |
"learning_rate": 2.4155995343422587e-05,
|
| 878 |
-
"logits/chosen": -
|
| 879 |
-
"logits/rejected": -
|
| 880 |
-
"logps/chosen": -
|
| 881 |
-
"logps/rejected": -
|
| 882 |
-
"loss": 0.
|
| 883 |
-
"pred_label":
|
| 884 |
-
"rewards/accuracies": 0.
|
| 885 |
-
"rewards/chosen": -
|
| 886 |
-
"rewards/margins":
|
| 887 |
-
"rewards/rejected": -
|
| 888 |
"step": 540,
|
| 889 |
-
"use_label":
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.58,
|
| 893 |
"learning_rate": 2.3573923166472644e-05,
|
| 894 |
-
"logits/chosen": -
|
| 895 |
-
"logits/rejected": -
|
| 896 |
-
"logps/chosen": -
|
| 897 |
-
"logps/rejected": -
|
| 898 |
-
"loss": 0.
|
| 899 |
-
"pred_label":
|
| 900 |
-
"rewards/accuracies": 0.
|
| 901 |
-
"rewards/chosen": -
|
| 902 |
-
"rewards/margins":
|
| 903 |
-
"rewards/rejected": -
|
| 904 |
"step": 550,
|
| 905 |
-
"use_label":
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.59,
|
| 909 |
"learning_rate": 2.2991850989522702e-05,
|
| 910 |
-
"logits/chosen": -
|
| 911 |
-
"logits/rejected": -
|
| 912 |
-
"logps/chosen": -
|
| 913 |
-
"logps/rejected": -
|
| 914 |
-
"loss": 0.
|
| 915 |
-
"pred_label":
|
| 916 |
-
"rewards/accuracies": 0.
|
| 917 |
-
"rewards/chosen": -
|
| 918 |
-
"rewards/margins":
|
| 919 |
-
"rewards/rejected": -
|
| 920 |
"step": 560,
|
| 921 |
-
"use_label":
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"epoch": 0.6,
|
| 925 |
"learning_rate": 2.240977881257276e-05,
|
| 926 |
-
"logits/chosen": -
|
| 927 |
-
"logits/rejected": -
|
| 928 |
-
"logps/chosen": -
|
| 929 |
-
"logps/rejected": -
|
| 930 |
-
"loss": 0.
|
| 931 |
-
"pred_label":
|
| 932 |
-
"rewards/accuracies": 0.
|
| 933 |
-
"rewards/chosen": -
|
| 934 |
-
"rewards/margins":
|
| 935 |
-
"rewards/rejected": -
|
| 936 |
"step": 570,
|
| 937 |
-
"use_label":
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 0.61,
|
| 941 |
"learning_rate": 2.1827706635622818e-05,
|
| 942 |
-
"logits/chosen": -
|
| 943 |
-
"logits/rejected": -
|
| 944 |
-
"logps/chosen": -
|
| 945 |
-
"logps/rejected": -
|
| 946 |
-
"loss": 0.
|
| 947 |
-
"pred_label":
|
| 948 |
-
"rewards/accuracies": 0.
|
| 949 |
-
"rewards/chosen": -
|
| 950 |
-
"rewards/margins":
|
| 951 |
-
"rewards/rejected": -
|
| 952 |
"step": 580,
|
| 953 |
-
"use_label":
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.62,
|
| 957 |
"learning_rate": 2.124563445867288e-05,
|
| 958 |
-
"logits/chosen": -
|
| 959 |
-
"logits/rejected": -
|
| 960 |
-
"logps/chosen": -
|
| 961 |
-
"logps/rejected": -
|
| 962 |
-
"loss": 0.
|
| 963 |
-
"pred_label":
|
| 964 |
-
"rewards/accuracies": 0.
|
| 965 |
-
"rewards/chosen": -
|
| 966 |
-
"rewards/margins":
|
| 967 |
-
"rewards/rejected": -
|
| 968 |
"step": 590,
|
| 969 |
-
"use_label":
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.63,
|
| 973 |
"learning_rate": 2.0663562281722934e-05,
|
| 974 |
-
"logits/chosen": -2.
|
| 975 |
-
"logits/rejected": -2.
|
| 976 |
-
"logps/chosen": -
|
| 977 |
-
"logps/rejected": -
|
| 978 |
-
"loss": 0.
|
| 979 |
-
"pred_label":
|
| 980 |
-
"rewards/accuracies": 0.
|
| 981 |
-
"rewards/chosen": -
|
| 982 |
-
"rewards/margins":
|
| 983 |
-
"rewards/rejected": -
|
| 984 |
"step": 600,
|
| 985 |
-
"use_label":
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 0.64,
|
| 989 |
"learning_rate": 2.0081490104772992e-05,
|
| 990 |
-
"logits/chosen":
|
| 991 |
-
"logits/rejected":
|
| 992 |
-
"logps/chosen": -
|
| 993 |
-
"logps/rejected": -
|
| 994 |
-
"loss": 0.
|
| 995 |
-
"pred_label":
|
| 996 |
-
"rewards/accuracies": 0.
|
| 997 |
-
"rewards/chosen": -
|
| 998 |
-
"rewards/margins":
|
| 999 |
-
"rewards/rejected": -
|
| 1000 |
"step": 610,
|
| 1001 |
-
"use_label":
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"epoch": 0.65,
|
| 1005 |
"learning_rate": 1.9499417927823053e-05,
|
| 1006 |
-
"logits/chosen":
|
| 1007 |
-
"logits/rejected":
|
| 1008 |
-
"logps/chosen": -
|
| 1009 |
-
"logps/rejected": -
|
| 1010 |
-
"loss": 0.
|
| 1011 |
-
"pred_label":
|
| 1012 |
-
"rewards/accuracies": 0.
|
| 1013 |
-
"rewards/chosen": -
|
| 1014 |
-
"rewards/margins":
|
| 1015 |
-
"rewards/rejected": -
|
| 1016 |
"step": 620,
|
| 1017 |
-
"use_label":
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.66,
|
| 1021 |
"learning_rate": 1.8917345750873107e-05,
|
| 1022 |
-
"logits/chosen":
|
| 1023 |
-
"logits/rejected":
|
| 1024 |
-
"logps/chosen": -
|
| 1025 |
-
"logps/rejected": -
|
| 1026 |
-
"loss": 0.
|
| 1027 |
-
"pred_label":
|
| 1028 |
-
"rewards/accuracies": 0.
|
| 1029 |
-
"rewards/chosen": -
|
| 1030 |
-
"rewards/margins":
|
| 1031 |
-
"rewards/rejected": -
|
| 1032 |
"step": 630,
|
| 1033 |
-
"use_label":
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 0.67,
|
| 1037 |
"learning_rate": 1.833527357392317e-05,
|
| 1038 |
-
"logits/chosen":
|
| 1039 |
-
"logits/rejected":
|
| 1040 |
-
"logps/chosen": -
|
| 1041 |
-
"logps/rejected": -
|
| 1042 |
-
"loss": 0.
|
| 1043 |
-
"pred_label":
|
| 1044 |
-
"rewards/accuracies": 0.
|
| 1045 |
-
"rewards/chosen": -
|
| 1046 |
-
"rewards/margins":
|
| 1047 |
-
"rewards/rejected": -
|
| 1048 |
"step": 640,
|
| 1049 |
-
"use_label":
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
"epoch": 0.68,
|
| 1053 |
"learning_rate": 1.7753201396973227e-05,
|
| 1054 |
-
"logits/chosen":
|
| 1055 |
-
"logits/rejected":
|
| 1056 |
-
"logps/chosen": -
|
| 1057 |
-
"logps/rejected": -
|
| 1058 |
-
"loss": 0.
|
| 1059 |
-
"pred_label":
|
| 1060 |
-
"rewards/accuracies": 0.
|
| 1061 |
-
"rewards/chosen": -
|
| 1062 |
-
"rewards/margins":
|
| 1063 |
-
"rewards/rejected": -
|
| 1064 |
"step": 650,
|
| 1065 |
-
"use_label":
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 0.69,
|
| 1069 |
"learning_rate": 1.717112922002328e-05,
|
| 1070 |
-
"logits/chosen":
|
| 1071 |
-
"logits/rejected":
|
| 1072 |
-
"logps/chosen": -
|
| 1073 |
-
"logps/rejected": -
|
| 1074 |
-
"loss": 0.
|
| 1075 |
-
"pred_label":
|
| 1076 |
-
"rewards/accuracies": 0.
|
| 1077 |
-
"rewards/chosen": -
|
| 1078 |
-
"rewards/margins":
|
| 1079 |
-
"rewards/rejected": -
|
| 1080 |
"step": 660,
|
| 1081 |
-
"use_label":
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.7,
|
| 1085 |
"learning_rate": 1.6589057043073342e-05,
|
| 1086 |
-
"logits/chosen":
|
| 1087 |
-
"logits/rejected":
|
| 1088 |
-
"logps/chosen": -
|
| 1089 |
-
"logps/rejected": -
|
| 1090 |
-
"loss": 0.
|
| 1091 |
-
"pred_label":
|
| 1092 |
-
"rewards/accuracies": 0.
|
| 1093 |
-
"rewards/chosen": -
|
| 1094 |
-
"rewards/margins":
|
| 1095 |
-
"rewards/rejected": -
|
| 1096 |
"step": 670,
|
| 1097 |
-
"use_label":
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 0.71,
|
| 1101 |
"learning_rate": 1.60069848661234e-05,
|
| 1102 |
-
"logits/chosen":
|
| 1103 |
-
"logits/rejected":
|
| 1104 |
-
"logps/chosen": -
|
| 1105 |
-
"logps/rejected": -
|
| 1106 |
-
"loss": 0.
|
| 1107 |
-
"pred_label":
|
| 1108 |
-
"rewards/accuracies": 0.
|
| 1109 |
-
"rewards/chosen": -
|
| 1110 |
-
"rewards/margins":
|
| 1111 |
-
"rewards/rejected": -
|
| 1112 |
"step": 680,
|
| 1113 |
-
"use_label":
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"epoch": 0.72,
|
| 1117 |
"learning_rate": 1.5424912689173458e-05,
|
| 1118 |
-
"logits/chosen":
|
| 1119 |
-
"logits/rejected":
|
| 1120 |
-
"logps/chosen": -
|
| 1121 |
-
"logps/rejected": -
|
| 1122 |
-
"loss": 0.
|
| 1123 |
-
"pred_label":
|
| 1124 |
-
"rewards/accuracies": 0.
|
| 1125 |
-
"rewards/chosen": -
|
| 1126 |
-
"rewards/margins":
|
| 1127 |
-
"rewards/rejected": -
|
| 1128 |
"step": 690,
|
| 1129 |
-
"use_label":
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.73,
|
| 1133 |
"learning_rate": 1.4842840512223516e-05,
|
| 1134 |
-
"logits/chosen":
|
| 1135 |
-
"logits/rejected":
|
| 1136 |
-
"logps/chosen": -
|
| 1137 |
-
"logps/rejected": -
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"pred_label":
|
| 1140 |
-
"rewards/accuracies": 0.
|
| 1141 |
-
"rewards/chosen": -
|
| 1142 |
-
"rewards/margins":
|
| 1143 |
-
"rewards/rejected": -
|
| 1144 |
"step": 700,
|
| 1145 |
-
"use_label":
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"epoch": 0.74,
|
| 1149 |
"learning_rate": 1.4260768335273575e-05,
|
| 1150 |
-
"logits/chosen":
|
| 1151 |
-
"logits/rejected":
|
| 1152 |
-
"logps/chosen": -
|
| 1153 |
-
"logps/rejected": -
|
| 1154 |
-
"loss": 0.
|
| 1155 |
-
"pred_label":
|
| 1156 |
-
"rewards/accuracies": 0.
|
| 1157 |
-
"rewards/chosen": -
|
| 1158 |
-
"rewards/margins":
|
| 1159 |
-
"rewards/rejected": -
|
| 1160 |
"step": 710,
|
| 1161 |
-
"use_label":
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.75,
|
| 1165 |
"learning_rate": 1.3678696158323633e-05,
|
| 1166 |
-
"logits/chosen":
|
| 1167 |
-
"logits/rejected":
|
| 1168 |
-
"logps/chosen": -
|
| 1169 |
-
"logps/rejected": -
|
| 1170 |
-
"loss": 0.
|
| 1171 |
-
"pred_label":
|
| 1172 |
-
"rewards/accuracies": 0.
|
| 1173 |
-
"rewards/chosen": -
|
| 1174 |
-
"rewards/margins":
|
| 1175 |
-
"rewards/rejected": -
|
| 1176 |
"step": 720,
|
| 1177 |
-
"use_label":
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"epoch": 0.76,
|
| 1181 |
"learning_rate": 1.309662398137369e-05,
|
| 1182 |
-
"logits/chosen":
|
| 1183 |
-
"logits/rejected":
|
| 1184 |
-
"logps/chosen": -
|
| 1185 |
-
"logps/rejected": -
|
| 1186 |
-
"loss": 0.
|
| 1187 |
-
"pred_label":
|
| 1188 |
-
"rewards/accuracies": 0.
|
| 1189 |
-
"rewards/chosen": -
|
| 1190 |
-
"rewards/margins":
|
| 1191 |
-
"rewards/rejected": -
|
| 1192 |
"step": 730,
|
| 1193 |
-
"use_label":
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.77,
|
| 1197 |
"learning_rate": 1.2514551804423749e-05,
|
| 1198 |
-
"logits/chosen":
|
| 1199 |
-
"logits/rejected":
|
| 1200 |
-
"logps/chosen": -
|
| 1201 |
-
"logps/rejected": -
|
| 1202 |
-
"loss": 0.
|
| 1203 |
-
"pred_label":
|
| 1204 |
-
"rewards/accuracies": 0.
|
| 1205 |
-
"rewards/chosen": -
|
| 1206 |
-
"rewards/margins":
|
| 1207 |
-
"rewards/rejected": -
|
| 1208 |
"step": 740,
|
| 1209 |
-
"use_label":
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 0.79,
|
| 1213 |
"learning_rate": 1.1932479627473807e-05,
|
| 1214 |
-
"logits/chosen":
|
| 1215 |
-
"logits/rejected":
|
| 1216 |
-
"logps/chosen": -
|
| 1217 |
-
"logps/rejected": -
|
| 1218 |
-
"loss": 0.
|
| 1219 |
-
"pred_label":
|
| 1220 |
-
"rewards/accuracies": 0.
|
| 1221 |
-
"rewards/chosen": -
|
| 1222 |
-
"rewards/margins":
|
| 1223 |
-
"rewards/rejected": -
|
| 1224 |
"step": 750,
|
| 1225 |
-
"use_label":
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"epoch": 0.8,
|
| 1229 |
"learning_rate": 1.1350407450523866e-05,
|
| 1230 |
-
"logits/chosen":
|
| 1231 |
-
"logits/rejected":
|
| 1232 |
-
"logps/chosen": -
|
| 1233 |
-
"logps/rejected": -
|
| 1234 |
-
"loss": 0.
|
| 1235 |
-
"pred_label":
|
| 1236 |
-
"rewards/accuracies": 0.
|
| 1237 |
-
"rewards/chosen": -
|
| 1238 |
-
"rewards/margins":
|
| 1239 |
-
"rewards/rejected": -
|
| 1240 |
"step": 760,
|
| 1241 |
-
"use_label":
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.81,
|
| 1245 |
"learning_rate": 1.0768335273573923e-05,
|
| 1246 |
-
"logits/chosen":
|
| 1247 |
-
"logits/rejected": 2.
|
| 1248 |
-
"logps/chosen": -
|
| 1249 |
-
"logps/rejected": -
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"pred_label":
|
| 1252 |
-
"rewards/accuracies": 0.
|
| 1253 |
-
"rewards/chosen": -
|
| 1254 |
-
"rewards/margins":
|
| 1255 |
-
"rewards/rejected": -
|
| 1256 |
"step": 770,
|
| 1257 |
-
"use_label":
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"epoch": 0.82,
|
| 1261 |
"learning_rate": 1.0186263096623982e-05,
|
| 1262 |
-
"logits/chosen": -
|
| 1263 |
-
"logits/rejected": -
|
| 1264 |
-
"logps/chosen": -
|
| 1265 |
-
"logps/rejected": -
|
| 1266 |
-
"loss": 0.
|
| 1267 |
-
"pred_label":
|
| 1268 |
-
"rewards/accuracies": 0.
|
| 1269 |
-
"rewards/chosen": -
|
| 1270 |
-
"rewards/margins":
|
| 1271 |
-
"rewards/rejected": -
|
| 1272 |
"step": 780,
|
| 1273 |
-
"use_label":
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 0.83,
|
| 1277 |
"learning_rate": 9.60419091967404e-06,
|
| 1278 |
-
"logits/chosen": -
|
| 1279 |
-
"logits/rejected": -
|
| 1280 |
-
"logps/chosen": -
|
| 1281 |
-
"logps/rejected": -
|
| 1282 |
-
"loss": 0.
|
| 1283 |
-
"pred_label":
|
| 1284 |
-
"rewards/accuracies": 0.
|
| 1285 |
-
"rewards/chosen": -
|
| 1286 |
-
"rewards/margins":
|
| 1287 |
-
"rewards/rejected": -
|
| 1288 |
"step": 790,
|
| 1289 |
-
"use_label":
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
"epoch": 0.84,
|
| 1293 |
"learning_rate": 9.022118742724098e-06,
|
| 1294 |
-
"logits/chosen": -2.
|
| 1295 |
-
"logits/rejected": -2.
|
| 1296 |
-
"logps/chosen": -
|
| 1297 |
-
"logps/rejected": -
|
| 1298 |
-
"loss": 0.
|
| 1299 |
-
"pred_label":
|
| 1300 |
-
"rewards/accuracies": 0.
|
| 1301 |
-
"rewards/chosen": -
|
| 1302 |
-
"rewards/margins":
|
| 1303 |
-
"rewards/rejected": -
|
| 1304 |
"step": 800,
|
| 1305 |
-
"use_label":
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.85,
|
| 1309 |
"learning_rate": 8.440046565774158e-06,
|
| 1310 |
-
"logits/chosen": -
|
| 1311 |
-
"logits/rejected": -
|
| 1312 |
-
"logps/chosen": -
|
| 1313 |
-
"logps/rejected": -
|
| 1314 |
-
"loss": 0.
|
| 1315 |
-
"pred_label":
|
| 1316 |
-
"rewards/accuracies": 0.
|
| 1317 |
-
"rewards/chosen": -
|
| 1318 |
-
"rewards/margins":
|
| 1319 |
-
"rewards/rejected": -
|
| 1320 |
"step": 810,
|
| 1321 |
-
"use_label":
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"epoch": 0.86,
|
| 1325 |
"learning_rate": 7.857974388824214e-06,
|
| 1326 |
-
"logits/chosen": -
|
| 1327 |
-
"logits/rejected": -
|
| 1328 |
-
"logps/chosen": -
|
| 1329 |
-
"logps/rejected": -
|
| 1330 |
-
"loss": 0.
|
| 1331 |
-
"pred_label":
|
| 1332 |
-
"rewards/accuracies": 0.
|
| 1333 |
-
"rewards/chosen": -
|
| 1334 |
-
"rewards/margins":
|
| 1335 |
-
"rewards/rejected": -
|
| 1336 |
"step": 820,
|
| 1337 |
-
"use_label":
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"epoch": 0.87,
|
| 1341 |
"learning_rate": 7.275902211874273e-06,
|
| 1342 |
-
"logits/chosen":
|
| 1343 |
-
"logits/rejected":
|
| 1344 |
-
"logps/chosen": -
|
| 1345 |
-
"logps/rejected": -
|
| 1346 |
-
"loss": 0.
|
| 1347 |
-
"pred_label":
|
| 1348 |
-
"rewards/accuracies": 0.
|
| 1349 |
-
"rewards/chosen": -
|
| 1350 |
-
"rewards/margins":
|
| 1351 |
-
"rewards/rejected": -
|
| 1352 |
"step": 830,
|
| 1353 |
-
"use_label":
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.88,
|
| 1357 |
"learning_rate": 6.693830034924331e-06,
|
| 1358 |
-
"logits/chosen":
|
| 1359 |
-
"logits/rejected":
|
| 1360 |
-
"logps/chosen": -
|
| 1361 |
-
"logps/rejected": -
|
| 1362 |
-
"loss": 0.
|
| 1363 |
-
"pred_label":
|
| 1364 |
-
"rewards/accuracies": 0.
|
| 1365 |
-
"rewards/chosen": -
|
| 1366 |
-
"rewards/margins":
|
| 1367 |
-
"rewards/rejected": -
|
| 1368 |
"step": 840,
|
| 1369 |
-
"use_label":
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.89,
|
| 1373 |
"learning_rate": 6.111757857974389e-06,
|
| 1374 |
-
"logits/chosen":
|
| 1375 |
-
"logits/rejected":
|
| 1376 |
-
"logps/chosen": -
|
| 1377 |
-
"logps/rejected": -
|
| 1378 |
-
"loss": 0.
|
| 1379 |
-
"pred_label":
|
| 1380 |
-
"rewards/accuracies": 0.
|
| 1381 |
-
"rewards/chosen": -
|
| 1382 |
-
"rewards/margins":
|
| 1383 |
-
"rewards/rejected": -
|
| 1384 |
"step": 850,
|
| 1385 |
-
"use_label":
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 0.9,
|
| 1389 |
"learning_rate": 5.529685681024447e-06,
|
| 1390 |
-
"logits/chosen":
|
| 1391 |
-
"logits/rejected":
|
| 1392 |
-
"logps/chosen": -
|
| 1393 |
-
"logps/rejected": -
|
| 1394 |
-
"loss": 0.
|
| 1395 |
-
"pred_label":
|
| 1396 |
-
"rewards/accuracies": 0.
|
| 1397 |
-
"rewards/chosen": -
|
| 1398 |
-
"rewards/margins":
|
| 1399 |
-
"rewards/rejected": -
|
| 1400 |
"step": 860,
|
| 1401 |
-
"use_label":
|
| 1402 |
},
|
| 1403 |
{
|
| 1404 |
"epoch": 0.91,
|
| 1405 |
"learning_rate": 4.947613504074506e-06,
|
| 1406 |
-
"logits/chosen":
|
| 1407 |
-
"logits/rejected":
|
| 1408 |
-
"logps/chosen": -
|
| 1409 |
-
"logps/rejected": -
|
| 1410 |
-
"loss": 0.
|
| 1411 |
-
"pred_label":
|
| 1412 |
-
"rewards/accuracies": 0.
|
| 1413 |
-
"rewards/chosen": -
|
| 1414 |
-
"rewards/margins":
|
| 1415 |
-
"rewards/rejected": -
|
| 1416 |
"step": 870,
|
| 1417 |
-
"use_label":
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 0.92,
|
| 1421 |
"learning_rate": 4.3655413271245635e-06,
|
| 1422 |
-
"logits/chosen":
|
| 1423 |
-
"logits/rejected":
|
| 1424 |
-
"logps/chosen": -
|
| 1425 |
-
"logps/rejected": -
|
| 1426 |
-
"loss": 0.
|
| 1427 |
-
"pred_label":
|
| 1428 |
-
"rewards/accuracies": 0.
|
| 1429 |
-
"rewards/chosen": -
|
| 1430 |
-
"rewards/margins":
|
| 1431 |
-
"rewards/rejected": -
|
| 1432 |
"step": 880,
|
| 1433 |
-
"use_label":
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
"epoch": 0.93,
|
| 1437 |
"learning_rate": 3.7834691501746217e-06,
|
| 1438 |
-
"logits/chosen":
|
| 1439 |
-
"logits/rejected":
|
| 1440 |
-
"logps/chosen": -
|
| 1441 |
-
"logps/rejected": -
|
| 1442 |
-
"loss": 0.
|
| 1443 |
-
"pred_label":
|
| 1444 |
-
"rewards/accuracies": 0.
|
| 1445 |
-
"rewards/chosen": -
|
| 1446 |
-
"rewards/margins":
|
| 1447 |
-
"rewards/rejected": -
|
| 1448 |
"step": 890,
|
| 1449 |
-
"use_label":
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.94,
|
| 1453 |
"learning_rate": 3.2013969732246805e-06,
|
| 1454 |
-
"logits/chosen":
|
| 1455 |
-
"logits/rejected":
|
| 1456 |
-
"logps/chosen": -
|
| 1457 |
-
"logps/rejected": -
|
| 1458 |
-
"loss": 0.
|
| 1459 |
-
"pred_label":
|
| 1460 |
-
"rewards/accuracies": 0.
|
| 1461 |
-
"rewards/chosen": -
|
| 1462 |
-
"rewards/margins":
|
| 1463 |
-
"rewards/rejected": -
|
| 1464 |
"step": 900,
|
| 1465 |
-
"use_label":
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 0.95,
|
| 1469 |
"learning_rate": 2.6193247962747383e-06,
|
| 1470 |
-
"logits/chosen":
|
| 1471 |
-
"logits/rejected":
|
| 1472 |
-
"logps/chosen": -
|
| 1473 |
-
"logps/rejected": -
|
| 1474 |
-
"loss": 0.
|
| 1475 |
-
"pred_label":
|
| 1476 |
-
"rewards/accuracies": 0.
|
| 1477 |
-
"rewards/chosen": -
|
| 1478 |
-
"rewards/margins":
|
| 1479 |
-
"rewards/rejected": -
|
| 1480 |
"step": 910,
|
| 1481 |
-
"use_label":
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.96,
|
| 1485 |
"learning_rate": 2.037252619324796e-06,
|
| 1486 |
-
"logits/chosen":
|
| 1487 |
-
"logits/rejected":
|
| 1488 |
-
"logps/chosen": -
|
| 1489 |
-
"logps/rejected": -
|
| 1490 |
-
"loss": 0.
|
| 1491 |
-
"pred_label":
|
| 1492 |
-
"rewards/accuracies": 0.
|
| 1493 |
-
"rewards/chosen": -
|
| 1494 |
-
"rewards/margins":
|
| 1495 |
-
"rewards/rejected": -
|
| 1496 |
"step": 920,
|
| 1497 |
-
"use_label":
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 0.97,
|
| 1501 |
"learning_rate": 1.4551804423748545e-06,
|
| 1502 |
-
"logits/chosen":
|
| 1503 |
-
"logits/rejected":
|
| 1504 |
-
"logps/chosen": -
|
| 1505 |
-
"logps/rejected": -
|
| 1506 |
-
"loss": 0.
|
| 1507 |
-
"pred_label":
|
| 1508 |
-
"rewards/accuracies": 0.
|
| 1509 |
-
"rewards/chosen": -
|
| 1510 |
-
"rewards/margins":
|
| 1511 |
-
"rewards/rejected": -
|
| 1512 |
"step": 930,
|
| 1513 |
-
"use_label":
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
"epoch": 0.98,
|
| 1517 |
"learning_rate": 8.731082654249127e-07,
|
| 1518 |
-
"logits/chosen":
|
| 1519 |
-
"logits/rejected":
|
| 1520 |
-
"logps/chosen": -
|
| 1521 |
-
"logps/rejected": -
|
| 1522 |
-
"loss": 0.
|
| 1523 |
-
"pred_label":
|
| 1524 |
-
"rewards/accuracies": 0.
|
| 1525 |
-
"rewards/chosen": -
|
| 1526 |
-
"rewards/margins":
|
| 1527 |
-
"rewards/rejected": -
|
| 1528 |
"step": 940,
|
| 1529 |
-
"use_label":
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 0.99,
|
| 1533 |
"learning_rate": 2.910360884749709e-07,
|
| 1534 |
-
"logits/chosen":
|
| 1535 |
-
"logits/rejected":
|
| 1536 |
-
"logps/chosen": -
|
| 1537 |
-
"logps/rejected": -
|
| 1538 |
-
"loss": 0.
|
| 1539 |
-
"pred_label":
|
| 1540 |
-
"rewards/accuracies": 0.
|
| 1541 |
-
"rewards/chosen": -
|
| 1542 |
-
"rewards/margins":
|
| 1543 |
-
"rewards/rejected": -
|
| 1544 |
"step": 950,
|
| 1545 |
-
"use_label":
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 1.0,
|
| 1549 |
-
"eval_logits/chosen":
|
| 1550 |
-
"eval_logits/rejected":
|
| 1551 |
-
"eval_logps/chosen": -
|
| 1552 |
-
"eval_logps/rejected": -
|
| 1553 |
-
"eval_loss": 0.
|
| 1554 |
-
"eval_pred_label":
|
| 1555 |
-
"eval_rewards/accuracies": 0.
|
| 1556 |
-
"eval_rewards/chosen": -
|
| 1557 |
-
"eval_rewards/margins":
|
| 1558 |
-
"eval_rewards/rejected": -
|
| 1559 |
-
"eval_runtime":
|
| 1560 |
-
"eval_samples_per_second": 4.
|
| 1561 |
-
"eval_steps_per_second": 0.
|
| 1562 |
-
"eval_use_label":
|
| 1563 |
"step": 955
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 1.0,
|
| 1567 |
"step": 955,
|
| 1568 |
"total_flos": 0.0,
|
| 1569 |
-
"train_loss": 0.
|
| 1570 |
-
"train_runtime":
|
| 1571 |
-
"train_samples_per_second": 2.
|
| 1572 |
"train_steps_per_second": 0.038
|
| 1573 |
}
|
| 1574 |
],
|
|
|
|
| 75 |
{
|
| 76 |
"epoch": 0.04,
|
| 77 |
"learning_rate": 2.0833333333333336e-05,
|
| 78 |
+
"logits/chosen": -2.841322422027588,
|
| 79 |
+
"logits/rejected": -2.84962797164917,
|
| 80 |
+
"logps/chosen": -281.39862060546875,
|
| 81 |
+
"logps/rejected": -277.9919738769531,
|
| 82 |
+
"loss": 0.6236,
|
| 83 |
+
"pred_label": 0.875,
|
| 84 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 85 |
+
"rewards/chosen": 0.01947467401623726,
|
| 86 |
+
"rewards/margins": 0.21414700150489807,
|
| 87 |
+
"rewards/rejected": -0.19467231631278992,
|
| 88 |
"step": 40,
|
| 89 |
+
"use_label": 561.125
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 0.05,
|
| 93 |
"learning_rate": 2.604166666666667e-05,
|
| 94 |
+
"logits/chosen": -2.8523213863372803,
|
| 95 |
+
"logits/rejected": -2.836536169052124,
|
| 96 |
+
"logps/chosen": -267.1900329589844,
|
| 97 |
+
"logps/rejected": -263.0260925292969,
|
| 98 |
+
"loss": 0.5468,
|
| 99 |
+
"pred_label": 18.174999237060547,
|
| 100 |
+
"rewards/accuracies": 0.637499988079071,
|
| 101 |
+
"rewards/chosen": -0.10816816240549088,
|
| 102 |
+
"rewards/margins": 0.4027964472770691,
|
| 103 |
+
"rewards/rejected": -0.5109646320343018,
|
| 104 |
"step": 50,
|
| 105 |
+
"use_label": 703.8250122070312
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.06,
|
| 109 |
"learning_rate": 3.125e-05,
|
| 110 |
+
"logits/chosen": -2.8082404136657715,
|
| 111 |
+
"logits/rejected": -2.7957923412323,
|
| 112 |
+
"logps/chosen": -302.7262878417969,
|
| 113 |
+
"logps/rejected": -294.58624267578125,
|
| 114 |
+
"loss": 0.4674,
|
| 115 |
+
"pred_label": 66.9000015258789,
|
| 116 |
+
"rewards/accuracies": 0.706250011920929,
|
| 117 |
+
"rewards/chosen": -0.2311042845249176,
|
| 118 |
+
"rewards/margins": 0.616974413394928,
|
| 119 |
+
"rewards/rejected": -0.8480786085128784,
|
| 120 |
"step": 60,
|
| 121 |
+
"use_label": 815.0999755859375
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07,
|
| 125 |
"learning_rate": 3.6458333333333336e-05,
|
| 126 |
+
"logits/chosen": -2.8012547492980957,
|
| 127 |
+
"logits/rejected": -2.788729429244995,
|
| 128 |
+
"logps/chosen": -299.8072204589844,
|
| 129 |
+
"logps/rejected": -288.31036376953125,
|
| 130 |
+
"loss": 0.3802,
|
| 131 |
+
"pred_label": 144.4499969482422,
|
| 132 |
+
"rewards/accuracies": 0.737500011920929,
|
| 133 |
+
"rewards/chosen": -0.5493733286857605,
|
| 134 |
+
"rewards/margins": 0.9851129651069641,
|
| 135 |
+
"rewards/rejected": -1.534486174583435,
|
| 136 |
"step": 70,
|
| 137 |
+
"use_label": 897.5499877929688
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 0.08,
|
| 141 |
"learning_rate": 4.166666666666667e-05,
|
| 142 |
+
"logits/chosen": -2.737917423248291,
|
| 143 |
+
"logits/rejected": -2.7173783779144287,
|
| 144 |
+
"logps/chosen": -318.4990234375,
|
| 145 |
+
"logps/rejected": -307.38311767578125,
|
| 146 |
+
"loss": 0.2535,
|
| 147 |
+
"pred_label": 244.1999969482422,
|
| 148 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 149 |
+
"rewards/chosen": -1.3193671703338623,
|
| 150 |
+
"rewards/margins": 1.5176814794540405,
|
| 151 |
+
"rewards/rejected": -2.8370487689971924,
|
| 152 |
"step": 80,
|
| 153 |
+
"use_label": 957.7999877929688
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.09,
|
| 157 |
"learning_rate": 4.6875e-05,
|
| 158 |
+
"logits/chosen": -2.650569200515747,
|
| 159 |
+
"logits/rejected": -2.7028753757476807,
|
| 160 |
+
"logps/chosen": -316.5365295410156,
|
| 161 |
+
"logps/rejected": -307.62799072265625,
|
| 162 |
+
"loss": 0.2342,
|
| 163 |
+
"pred_label": 363.79998779296875,
|
| 164 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 165 |
+
"rewards/chosen": -2.0753350257873535,
|
| 166 |
+
"rewards/margins": 2.060016632080078,
|
| 167 |
+
"rewards/rejected": -4.135351657867432,
|
| 168 |
"step": 90,
|
| 169 |
+
"use_label": 998.2000122070312
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 0.1,
|
| 173 |
"learning_rate": 4.976717112922003e-05,
|
| 174 |
+
"logits/chosen": -2.664170265197754,
|
| 175 |
+
"logits/rejected": -2.6609156131744385,
|
| 176 |
+
"logps/chosen": -301.24493408203125,
|
| 177 |
+
"logps/rejected": -345.8691101074219,
|
| 178 |
+
"loss": 0.2337,
|
| 179 |
+
"pred_label": 482.125,
|
| 180 |
+
"rewards/accuracies": 0.6875,
|
| 181 |
+
"rewards/chosen": -2.711317300796509,
|
| 182 |
+
"rewards/margins": 2.521953582763672,
|
| 183 |
+
"rewards/rejected": -5.23327112197876,
|
| 184 |
"step": 100,
|
| 185 |
+
"use_label": 1039.875
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.12,
|
| 189 |
"learning_rate": 4.918509895227008e-05,
|
| 190 |
+
"logits/chosen": -2.6410791873931885,
|
| 191 |
+
"logits/rejected": -2.587087631225586,
|
| 192 |
+
"logps/chosen": -328.8195495605469,
|
| 193 |
+
"logps/rejected": -354.773193359375,
|
| 194 |
+
"loss": 0.1381,
|
| 195 |
+
"pred_label": 612.7750244140625,
|
| 196 |
"rewards/accuracies": 0.675000011920929,
|
| 197 |
+
"rewards/chosen": -6.414492607116699,
|
| 198 |
+
"rewards/margins": 4.145485877990723,
|
| 199 |
+
"rewards/rejected": -10.559977531433105,
|
| 200 |
"step": 110,
|
| 201 |
+
"use_label": 1069.2249755859375
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 0.13,
|
| 205 |
"learning_rate": 4.860302677532014e-05,
|
| 206 |
+
"logits/chosen": -2.485827684402466,
|
| 207 |
+
"logits/rejected": -2.4543118476867676,
|
| 208 |
+
"logps/chosen": -401.1512756347656,
|
| 209 |
+
"logps/rejected": -415.895263671875,
|
| 210 |
+
"loss": 0.1156,
|
| 211 |
+
"pred_label": 747.8499755859375,
|
| 212 |
+
"rewards/accuracies": 0.699999988079071,
|
| 213 |
+
"rewards/chosen": -12.163839340209961,
|
| 214 |
+
"rewards/margins": 4.657315254211426,
|
| 215 |
+
"rewards/rejected": -16.821155548095703,
|
| 216 |
"step": 120,
|
| 217 |
+
"use_label": 1094.1500244140625
|
| 218 |
},
|
| 219 |
{
|
| 220 |
"epoch": 0.14,
|
| 221 |
"learning_rate": 4.80209545983702e-05,
|
| 222 |
+
"logits/chosen": -2.586129665374756,
|
| 223 |
+
"logits/rejected": -2.4943180084228516,
|
| 224 |
+
"logps/chosen": -397.7897033691406,
|
| 225 |
+
"logps/rejected": -390.68902587890625,
|
| 226 |
+
"loss": 0.181,
|
| 227 |
+
"pred_label": 880.6749877929688,
|
| 228 |
+
"rewards/accuracies": 0.625,
|
| 229 |
+
"rewards/chosen": -9.052229881286621,
|
| 230 |
+
"rewards/margins": 2.705533742904663,
|
| 231 |
+
"rewards/rejected": -11.757763862609863,
|
| 232 |
"step": 130,
|
| 233 |
+
"use_label": 1121.324951171875
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.15,
|
| 237 |
"learning_rate": 4.743888242142026e-05,
|
| 238 |
+
"logits/chosen": -2.4943366050720215,
|
| 239 |
+
"logits/rejected": -2.467221736907959,
|
| 240 |
+
"logps/chosen": -360.2618103027344,
|
| 241 |
+
"logps/rejected": -391.42254638671875,
|
| 242 |
+
"loss": 0.1307,
|
| 243 |
+
"pred_label": 1014.9000244140625,
|
| 244 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 245 |
+
"rewards/chosen": -10.215502738952637,
|
| 246 |
+
"rewards/margins": 5.424862384796143,
|
| 247 |
+
"rewards/rejected": -15.640365600585938,
|
| 248 |
"step": 140,
|
| 249 |
+
"use_label": 1147.0999755859375
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.16,
|
| 253 |
"learning_rate": 4.685681024447032e-05,
|
| 254 |
+
"logits/chosen": -2.448756456375122,
|
| 255 |
+
"logits/rejected": -2.4686672687530518,
|
| 256 |
+
"logps/chosen": -423.26348876953125,
|
| 257 |
+
"logps/rejected": -462.45587158203125,
|
| 258 |
+
"loss": 0.1291,
|
| 259 |
+
"pred_label": 1152.449951171875,
|
| 260 |
+
"rewards/accuracies": 0.612500011920929,
|
| 261 |
+
"rewards/chosen": -12.837747573852539,
|
| 262 |
+
"rewards/margins": 5.233094692230225,
|
| 263 |
+
"rewards/rejected": -18.070842742919922,
|
| 264 |
"step": 150,
|
| 265 |
+
"use_label": 1169.550048828125
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"epoch": 0.17,
|
| 269 |
"learning_rate": 4.6274738067520374e-05,
|
| 270 |
+
"logits/chosen": -2.5502519607543945,
|
| 271 |
+
"logits/rejected": -2.4940075874328613,
|
| 272 |
+
"logps/chosen": -338.2818908691406,
|
| 273 |
+
"logps/rejected": -399.33062744140625,
|
| 274 |
+
"loss": 0.1221,
|
| 275 |
+
"pred_label": 1286.300048828125,
|
| 276 |
+
"rewards/accuracies": 0.699999988079071,
|
| 277 |
+
"rewards/chosen": -7.589987754821777,
|
| 278 |
+
"rewards/margins": 5.438824653625488,
|
| 279 |
+
"rewards/rejected": -13.02881145477295,
|
| 280 |
"step": 160,
|
| 281 |
+
"use_label": 1195.699951171875
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 0.18,
|
| 285 |
"learning_rate": 4.5692665890570435e-05,
|
| 286 |
+
"logits/chosen": -2.3367486000061035,
|
| 287 |
+
"logits/rejected": -2.268026351928711,
|
| 288 |
+
"logps/chosen": -608.579833984375,
|
| 289 |
+
"logps/rejected": -566.7401123046875,
|
| 290 |
+
"loss": 0.0863,
|
| 291 |
+
"pred_label": 1421.699951171875,
|
| 292 |
+
"rewards/accuracies": 0.637499988079071,
|
| 293 |
+
"rewards/chosen": -31.210372924804688,
|
| 294 |
+
"rewards/margins": 0.4991304278373718,
|
| 295 |
+
"rewards/rejected": -31.70950698852539,
|
| 296 |
"step": 170,
|
| 297 |
+
"use_label": 1220.300048828125
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.19,
|
| 301 |
"learning_rate": 4.511059371362049e-05,
|
| 302 |
+
"logits/chosen": -2.5058579444885254,
|
| 303 |
+
"logits/rejected": -2.460585117340088,
|
| 304 |
+
"logps/chosen": -369.10260009765625,
|
| 305 |
+
"logps/rejected": -435.2815856933594,
|
| 306 |
+
"loss": 0.1377,
|
| 307 |
+
"pred_label": 1565.5,
|
| 308 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 309 |
+
"rewards/chosen": -11.317340850830078,
|
| 310 |
+
"rewards/margins": 5.142584800720215,
|
| 311 |
+
"rewards/rejected": -16.459924697875977,
|
| 312 |
"step": 180,
|
| 313 |
+
"use_label": 1236.5
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"epoch": 0.2,
|
| 317 |
"learning_rate": 4.452852153667055e-05,
|
| 318 |
+
"logits/chosen": -2.689570665359497,
|
| 319 |
+
"logits/rejected": -2.6224443912506104,
|
| 320 |
+
"logps/chosen": -352.5258483886719,
|
| 321 |
+
"logps/rejected": -360.037841796875,
|
| 322 |
+
"loss": 0.1658,
|
| 323 |
+
"pred_label": 1696.5,
|
| 324 |
+
"rewards/accuracies": 0.78125,
|
| 325 |
+
"rewards/chosen": -4.5269904136657715,
|
| 326 |
+
"rewards/margins": 5.3725738525390625,
|
| 327 |
+
"rewards/rejected": -9.899563789367676,
|
| 328 |
"step": 190,
|
| 329 |
+
"use_label": 1265.5
|
| 330 |
},
|
| 331 |
{
|
| 332 |
"epoch": 0.21,
|
| 333 |
"learning_rate": 4.394644935972061e-05,
|
| 334 |
+
"logits/chosen": -2.663055658340454,
|
| 335 |
+
"logits/rejected": -2.6471009254455566,
|
| 336 |
+
"logps/chosen": -329.58465576171875,
|
| 337 |
+
"logps/rejected": -377.0941467285156,
|
| 338 |
+
"loss": 0.1567,
|
| 339 |
+
"pred_label": 1831.4000244140625,
|
| 340 |
+
"rewards/accuracies": 0.668749988079071,
|
| 341 |
+
"rewards/chosen": -6.82558536529541,
|
| 342 |
+
"rewards/margins": 4.059788227081299,
|
| 343 |
+
"rewards/rejected": -10.8853759765625,
|
| 344 |
"step": 200,
|
| 345 |
+
"use_label": 1290.5999755859375
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.22,
|
| 349 |
"learning_rate": 4.336437718277067e-05,
|
| 350 |
+
"logits/chosen": -2.6221835613250732,
|
| 351 |
+
"logits/rejected": -2.6028318405151367,
|
| 352 |
+
"logps/chosen": -400.9311828613281,
|
| 353 |
+
"logps/rejected": -399.74774169921875,
|
| 354 |
+
"loss": 0.1038,
|
| 355 |
+
"pred_label": 1967.925048828125,
|
| 356 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 357 |
+
"rewards/chosen": -9.968328475952148,
|
| 358 |
+
"rewards/margins": 4.587479591369629,
|
| 359 |
+
"rewards/rejected": -14.555807113647461,
|
| 360 |
"step": 210,
|
| 361 |
+
"use_label": 1314.074951171875
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.23,
|
| 365 |
"learning_rate": 4.278230500582072e-05,
|
| 366 |
+
"logits/chosen": -2.5327601432800293,
|
| 367 |
+
"logits/rejected": -2.5170798301696777,
|
| 368 |
+
"logps/chosen": -431.26373291015625,
|
| 369 |
+
"logps/rejected": -491.82647705078125,
|
| 370 |
+
"loss": 0.0838,
|
| 371 |
+
"pred_label": 2115.125,
|
| 372 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 373 |
+
"rewards/chosen": -15.627206802368164,
|
| 374 |
+
"rewards/margins": 6.489752292633057,
|
| 375 |
+
"rewards/rejected": -22.116960525512695,
|
| 376 |
"step": 220,
|
| 377 |
+
"use_label": 1326.875
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.24,
|
| 381 |
"learning_rate": 4.220023282887078e-05,
|
| 382 |
+
"logits/chosen": -2.2676877975463867,
|
| 383 |
+
"logits/rejected": -2.275028944015503,
|
| 384 |
+
"logps/chosen": -759.7401123046875,
|
| 385 |
+
"logps/rejected": -772.6349487304688,
|
| 386 |
+
"loss": 0.0695,
|
| 387 |
+
"pred_label": 2261.175048828125,
|
| 388 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 389 |
+
"rewards/chosen": -45.156700134277344,
|
| 390 |
+
"rewards/margins": 5.6958770751953125,
|
| 391 |
+
"rewards/rejected": -50.852577209472656,
|
| 392 |
"step": 230,
|
| 393 |
+
"use_label": 1340.824951171875
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 0.25,
|
| 397 |
"learning_rate": 4.161816065192084e-05,
|
| 398 |
+
"logits/chosen": -2.720881700515747,
|
| 399 |
+
"logits/rejected": -2.667050361633301,
|
| 400 |
+
"logps/chosen": -346.90936279296875,
|
| 401 |
+
"logps/rejected": -372.1588439941406,
|
| 402 |
+
"loss": 0.1159,
|
| 403 |
+
"pred_label": 2400.875,
|
| 404 |
+
"rewards/accuracies": 0.706250011920929,
|
| 405 |
+
"rewards/chosen": -7.2331953048706055,
|
| 406 |
+
"rewards/margins": 7.680712699890137,
|
| 407 |
+
"rewards/rejected": -14.913908004760742,
|
| 408 |
"step": 240,
|
| 409 |
+
"use_label": 1361.125
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.26,
|
| 413 |
"learning_rate": 4.10360884749709e-05,
|
| 414 |
+
"logits/chosen": -2.7150168418884277,
|
| 415 |
+
"logits/rejected": -2.7007734775543213,
|
| 416 |
+
"logps/chosen": -368.1573486328125,
|
| 417 |
+
"logps/rejected": -409.0894470214844,
|
| 418 |
+
"loss": 0.1438,
|
| 419 |
+
"pred_label": 2528.425048828125,
|
| 420 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 421 |
+
"rewards/chosen": -7.751723289489746,
|
| 422 |
+
"rewards/margins": 6.2153401374816895,
|
| 423 |
+
"rewards/rejected": -13.967063903808594,
|
| 424 |
"step": 250,
|
| 425 |
+
"use_label": 1393.574951171875
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"epoch": 0.27,
|
| 429 |
"learning_rate": 4.045401629802096e-05,
|
| 430 |
+
"logits/chosen": -2.7338109016418457,
|
| 431 |
+
"logits/rejected": -2.7302541732788086,
|
| 432 |
+
"logps/chosen": -320.41326904296875,
|
| 433 |
+
"logps/rejected": -374.7262878417969,
|
| 434 |
+
"loss": 0.1538,
|
| 435 |
+
"pred_label": 2655.949951171875,
|
| 436 |
+
"rewards/accuracies": 0.65625,
|
| 437 |
+
"rewards/chosen": -7.180386543273926,
|
| 438 |
+
"rewards/margins": 5.263332843780518,
|
| 439 |
+
"rewards/rejected": -12.443718910217285,
|
| 440 |
"step": 260,
|
| 441 |
+
"use_label": 1426.050048828125
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 0.28,
|
| 445 |
"learning_rate": 3.9871944121071014e-05,
|
| 446 |
+
"logits/chosen": -2.729078769683838,
|
| 447 |
+
"logits/rejected": -2.7090542316436768,
|
| 448 |
+
"logps/chosen": -375.9404296875,
|
| 449 |
+
"logps/rejected": -433.7308654785156,
|
| 450 |
+
"loss": 0.1174,
|
| 451 |
+
"pred_label": 2794.300048828125,
|
| 452 |
+
"rewards/accuracies": 0.78125,
|
| 453 |
+
"rewards/chosen": -8.406911849975586,
|
| 454 |
+
"rewards/margins": 7.4909772872924805,
|
| 455 |
+
"rewards/rejected": -15.89788818359375,
|
| 456 |
"step": 270,
|
| 457 |
+
"use_label": 1447.699951171875
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.29,
|
| 461 |
"learning_rate": 3.928987194412107e-05,
|
| 462 |
+
"logits/chosen": -2.7120089530944824,
|
| 463 |
+
"logits/rejected": -2.690830707550049,
|
| 464 |
+
"logps/chosen": -403.83233642578125,
|
| 465 |
+
"logps/rejected": -437.89410400390625,
|
| 466 |
+
"loss": 0.0926,
|
| 467 |
+
"pred_label": 2935.449951171875,
|
| 468 |
+
"rewards/accuracies": 0.737500011920929,
|
| 469 |
+
"rewards/chosen": -10.986837387084961,
|
| 470 |
+
"rewards/margins": 6.579653263092041,
|
| 471 |
+
"rewards/rejected": -17.566490173339844,
|
| 472 |
"step": 280,
|
| 473 |
+
"use_label": 1466.550048828125
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 0.3,
|
| 477 |
"learning_rate": 3.870779976717113e-05,
|
| 478 |
+
"logits/chosen": -2.637876033782959,
|
| 479 |
+
"logits/rejected": -2.5659611225128174,
|
| 480 |
+
"logps/chosen": -455.518310546875,
|
| 481 |
+
"logps/rejected": -462.74267578125,
|
| 482 |
+
"loss": 0.0968,
|
| 483 |
+
"pred_label": 3076.324951171875,
|
| 484 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 485 |
+
"rewards/chosen": -16.259593963623047,
|
| 486 |
+
"rewards/margins": 6.993022918701172,
|
| 487 |
+
"rewards/rejected": -23.25261688232422,
|
| 488 |
"step": 290,
|
| 489 |
+
"use_label": 1485.675048828125
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.31,
|
| 493 |
"learning_rate": 3.812572759022119e-05,
|
| 494 |
+
"logits/chosen": -2.5933048725128174,
|
| 495 |
+
"logits/rejected": -2.5245001316070557,
|
| 496 |
+
"logps/chosen": -432.7098083496094,
|
| 497 |
+
"logps/rejected": -497.24554443359375,
|
| 498 |
+
"loss": 0.0835,
|
| 499 |
+
"pred_label": 3217.25,
|
| 500 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 501 |
+
"rewards/chosen": -16.903099060058594,
|
| 502 |
+
"rewards/margins": 8.837041854858398,
|
| 503 |
+
"rewards/rejected": -25.74013900756836,
|
| 504 |
"step": 300,
|
| 505 |
+
"use_label": 1504.75
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.32,
|
| 509 |
"learning_rate": 3.7543655413271246e-05,
|
| 510 |
+
"logits/chosen": -2.4902329444885254,
|
| 511 |
+
"logits/rejected": -2.3905444145202637,
|
| 512 |
+
"logps/chosen": -440.56085205078125,
|
| 513 |
+
"logps/rejected": -660.7030029296875,
|
| 514 |
+
"loss": 0.0981,
|
| 515 |
+
"pred_label": 3356.77490234375,
|
| 516 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 517 |
+
"rewards/chosen": -17.362918853759766,
|
| 518 |
+
"rewards/margins": 19.8530330657959,
|
| 519 |
+
"rewards/rejected": -37.21595764160156,
|
| 520 |
"step": 310,
|
| 521 |
+
"use_label": 1525.2249755859375
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.33,
|
| 525 |
"learning_rate": 3.696158323632131e-05,
|
| 526 |
+
"logits/chosen": -2.2354841232299805,
|
| 527 |
+
"logits/rejected": -2.1421380043029785,
|
| 528 |
+
"logps/chosen": -747.2213134765625,
|
| 529 |
+
"logps/rejected": -914.2220458984375,
|
| 530 |
+
"loss": 0.0656,
|
| 531 |
+
"pred_label": 3500.35009765625,
|
| 532 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 533 |
+
"rewards/chosen": -45.124141693115234,
|
| 534 |
+
"rewards/margins": 18.217256546020508,
|
| 535 |
+
"rewards/rejected": -63.341407775878906,
|
| 536 |
"step": 320,
|
| 537 |
+
"use_label": 1541.6500244140625
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.35,
|
| 541 |
"learning_rate": 3.637951105937136e-05,
|
| 542 |
+
"logits/chosen": -2.2868189811706543,
|
| 543 |
+
"logits/rejected": -2.2072067260742188,
|
| 544 |
+
"logps/chosen": -698.5914306640625,
|
| 545 |
+
"logps/rejected": -843.9619140625,
|
| 546 |
+
"loss": 0.0646,
|
| 547 |
+
"pred_label": 3651.77490234375,
|
| 548 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 549 |
+
"rewards/chosen": -42.7642936706543,
|
| 550 |
+
"rewards/margins": 18.20369529724121,
|
| 551 |
+
"rewards/rejected": -60.96799850463867,
|
| 552 |
"step": 330,
|
| 553 |
+
"use_label": 1550.2249755859375
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"epoch": 0.36,
|
| 557 |
"learning_rate": 3.579743888242142e-05,
|
| 558 |
+
"logits/chosen": -2.1754813194274902,
|
| 559 |
+
"logits/rejected": -2.1737990379333496,
|
| 560 |
+
"logps/chosen": -933.4059448242188,
|
| 561 |
+
"logps/rejected": -1077.489013671875,
|
| 562 |
+
"loss": 0.0609,
|
| 563 |
+
"pred_label": 3802.02490234375,
|
| 564 |
+
"rewards/accuracies": 0.581250011920929,
|
| 565 |
+
"rewards/chosen": -64.42259216308594,
|
| 566 |
+
"rewards/margins": 19.241281509399414,
|
| 567 |
+
"rewards/rejected": -83.66387176513672,
|
| 568 |
"step": 340,
|
| 569 |
+
"use_label": 1559.9749755859375
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.37,
|
| 573 |
"learning_rate": 3.5215366705471484e-05,
|
| 574 |
+
"logits/chosen": -2.3112852573394775,
|
| 575 |
+
"logits/rejected": -2.2864885330200195,
|
| 576 |
+
"logps/chosen": -736.1566162109375,
|
| 577 |
+
"logps/rejected": -973.3111572265625,
|
| 578 |
+
"loss": 0.0509,
|
| 579 |
+
"pred_label": 3953.425048828125,
|
| 580 |
+
"rewards/accuracies": 0.6875,
|
| 581 |
+
"rewards/chosen": -43.89619064331055,
|
| 582 |
+
"rewards/margins": 25.668701171875,
|
| 583 |
+
"rewards/rejected": -69.56489562988281,
|
| 584 |
"step": 350,
|
| 585 |
+
"use_label": 1568.574951171875
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.38,
|
| 589 |
"learning_rate": 3.463329452852154e-05,
|
| 590 |
+
"logits/chosen": -2.3861846923828125,
|
| 591 |
+
"logits/rejected": -2.348050117492676,
|
| 592 |
+
"logps/chosen": -612.4452514648438,
|
| 593 |
+
"logps/rejected": -722.9779663085938,
|
| 594 |
+
"loss": 0.0589,
|
| 595 |
+
"pred_label": 4104.22509765625,
|
| 596 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 597 |
+
"rewards/chosen": -33.965110778808594,
|
| 598 |
+
"rewards/margins": 11.812593460083008,
|
| 599 |
+
"rewards/rejected": -45.777706146240234,
|
| 600 |
"step": 360,
|
| 601 |
+
"use_label": 1577.7750244140625
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"epoch": 0.39,
|
| 605 |
"learning_rate": 3.40512223515716e-05,
|
| 606 |
+
"logits/chosen": -2.5135788917541504,
|
| 607 |
+
"logits/rejected": -2.431811571121216,
|
| 608 |
+
"logps/chosen": -424.55120849609375,
|
| 609 |
+
"logps/rejected": -567.4149780273438,
|
| 610 |
+
"loss": 0.0861,
|
| 611 |
+
"pred_label": 4248.5498046875,
|
| 612 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 613 |
+
"rewards/chosen": -14.834848403930664,
|
| 614 |
+
"rewards/margins": 14.3344144821167,
|
| 615 |
+
"rewards/rejected": -29.169261932373047,
|
| 616 |
"step": 370,
|
| 617 |
+
"use_label": 1593.449951171875
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 0.4,
|
| 621 |
"learning_rate": 3.3469150174621654e-05,
|
| 622 |
+
"logits/chosen": -2.5888447761535645,
|
| 623 |
+
"logits/rejected": -2.5503790378570557,
|
| 624 |
+
"logps/chosen": -432.38995361328125,
|
| 625 |
+
"logps/rejected": -449.0523986816406,
|
| 626 |
+
"loss": 0.1037,
|
| 627 |
+
"pred_label": 4388.47509765625,
|
| 628 |
+
"rewards/accuracies": 0.699999988079071,
|
| 629 |
+
"rewards/chosen": -14.715968132019043,
|
| 630 |
+
"rewards/margins": 4.131407260894775,
|
| 631 |
+
"rewards/rejected": -18.847375869750977,
|
| 632 |
"step": 380,
|
| 633 |
+
"use_label": 1613.5250244140625
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.41,
|
| 637 |
"learning_rate": 3.288707799767171e-05,
|
| 638 |
+
"logits/chosen": -2.4642863273620605,
|
| 639 |
+
"logits/rejected": -2.3922629356384277,
|
| 640 |
+
"logps/chosen": -390.8463439941406,
|
| 641 |
+
"logps/rejected": -461.17919921875,
|
| 642 |
+
"loss": 0.1182,
|
| 643 |
+
"pred_label": 4530.875,
|
| 644 |
+
"rewards/accuracies": 0.731249988079071,
|
| 645 |
+
"rewards/chosen": -10.805445671081543,
|
| 646 |
+
"rewards/margins": 10.884265899658203,
|
| 647 |
+
"rewards/rejected": -21.689708709716797,
|
| 648 |
"step": 390,
|
| 649 |
+
"use_label": 1631.125
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"epoch": 0.42,
|
| 653 |
"learning_rate": 3.2305005820721776e-05,
|
| 654 |
+
"logits/chosen": -2.708742141723633,
|
| 655 |
+
"logits/rejected": -2.668813943862915,
|
| 656 |
+
"logps/chosen": -360.8929138183594,
|
| 657 |
+
"logps/rejected": -357.1689758300781,
|
| 658 |
+
"loss": 0.1227,
|
| 659 |
+
"pred_label": 4662.125,
|
| 660 |
+
"rewards/accuracies": 0.675000011920929,
|
| 661 |
+
"rewards/chosen": -5.38419246673584,
|
| 662 |
+
"rewards/margins": 4.3987908363342285,
|
| 663 |
+
"rewards/rejected": -9.782983779907227,
|
| 664 |
"step": 400,
|
| 665 |
+
"use_label": 1659.875
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"epoch": 0.43,
|
| 669 |
"learning_rate": 3.172293364377183e-05,
|
| 670 |
+
"logits/chosen": -2.703439474105835,
|
| 671 |
+
"logits/rejected": -2.6409201622009277,
|
| 672 |
+
"logps/chosen": -355.69219970703125,
|
| 673 |
+
"logps/rejected": -346.83416748046875,
|
| 674 |
+
"loss": 0.1313,
|
| 675 |
+
"pred_label": 4792.97509765625,
|
| 676 |
+
"rewards/accuracies": 0.606249988079071,
|
| 677 |
+
"rewards/chosen": -5.759453773498535,
|
| 678 |
+
"rewards/margins": 3.0691254138946533,
|
| 679 |
+
"rewards/rejected": -8.828579902648926,
|
| 680 |
"step": 410,
|
| 681 |
+
"use_label": 1689.0250244140625
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.44,
|
| 685 |
"learning_rate": 3.1140861466821885e-05,
|
| 686 |
+
"logits/chosen": -2.617567539215088,
|
| 687 |
+
"logits/rejected": -2.578650951385498,
|
| 688 |
+
"logps/chosen": -307.09552001953125,
|
| 689 |
+
"logps/rejected": -360.4197998046875,
|
| 690 |
+
"loss": 0.1338,
|
| 691 |
+
"pred_label": 4930.9501953125,
|
| 692 |
+
"rewards/accuracies": 0.706250011920929,
|
| 693 |
+
"rewards/chosen": -6.651512145996094,
|
| 694 |
+
"rewards/margins": 4.47898006439209,
|
| 695 |
+
"rewards/rejected": -11.130491256713867,
|
| 696 |
"step": 420,
|
| 697 |
+
"use_label": 1711.050048828125
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.45,
|
| 701 |
"learning_rate": 3.055878928987195e-05,
|
| 702 |
+
"logits/chosen": -2.6378731727600098,
|
| 703 |
+
"logits/rejected": -2.5847997665405273,
|
| 704 |
+
"logps/chosen": -359.0061340332031,
|
| 705 |
+
"logps/rejected": -389.7403564453125,
|
| 706 |
+
"loss": 0.1136,
|
| 707 |
+
"pred_label": 5066.6748046875,
|
| 708 |
+
"rewards/accuracies": 0.668749988079071,
|
| 709 |
+
"rewards/chosen": -8.110578536987305,
|
| 710 |
+
"rewards/margins": 4.911756992340088,
|
| 711 |
+
"rewards/rejected": -13.02233600616455,
|
| 712 |
"step": 430,
|
| 713 |
+
"use_label": 1735.324951171875
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.46,
|
| 717 |
"learning_rate": 2.9976717112922005e-05,
|
| 718 |
+
"logits/chosen": -2.705080509185791,
|
| 719 |
+
"logits/rejected": -2.6582016944885254,
|
| 720 |
+
"logps/chosen": -361.77703857421875,
|
| 721 |
+
"logps/rejected": -369.16888427734375,
|
| 722 |
+
"loss": 0.1282,
|
| 723 |
+
"pred_label": 5203.0498046875,
|
| 724 |
+
"rewards/accuracies": 0.65625,
|
| 725 |
+
"rewards/chosen": -7.850671291351318,
|
| 726 |
+
"rewards/margins": 3.7665488719940186,
|
| 727 |
+
"rewards/rejected": -11.617219924926758,
|
| 728 |
"step": 440,
|
| 729 |
+
"use_label": 1758.949951171875
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.47,
|
| 733 |
"learning_rate": 2.939464493597206e-05,
|
| 734 |
+
"logits/chosen": -2.387315273284912,
|
| 735 |
+
"logits/rejected": -2.332853317260742,
|
| 736 |
+
"logps/chosen": -343.62481689453125,
|
| 737 |
+
"logps/rejected": -404.4623107910156,
|
| 738 |
+
"loss": 0.1146,
|
| 739 |
+
"pred_label": 5345.10009765625,
|
| 740 |
+
"rewards/accuracies": 0.637499988079071,
|
| 741 |
+
"rewards/chosen": -11.390606880187988,
|
| 742 |
+
"rewards/margins": 6.090449333190918,
|
| 743 |
+
"rewards/rejected": -17.481056213378906,
|
| 744 |
"step": 450,
|
| 745 |
+
"use_label": 1776.9000244140625
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.48,
|
| 749 |
"learning_rate": 2.881257275902212e-05,
|
| 750 |
+
"logits/chosen": -2.5489556789398193,
|
| 751 |
+
"logits/rejected": -2.5125844478607178,
|
| 752 |
+
"logps/chosen": -429.1163024902344,
|
| 753 |
+
"logps/rejected": -433.0736389160156,
|
| 754 |
+
"loss": 0.1077,
|
| 755 |
+
"pred_label": 5487.8251953125,
|
| 756 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 757 |
+
"rewards/chosen": -12.731410026550293,
|
| 758 |
+
"rewards/margins": 3.6760268211364746,
|
| 759 |
+
"rewards/rejected": -16.40743637084961,
|
| 760 |
"step": 460,
|
| 761 |
+
"use_label": 1794.175048828125
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"epoch": 0.49,
|
| 765 |
"learning_rate": 2.8230500582072178e-05,
|
| 766 |
+
"logits/chosen": -2.482365369796753,
|
| 767 |
+
"logits/rejected": -2.3831024169921875,
|
| 768 |
+
"logps/chosen": -389.73785400390625,
|
| 769 |
+
"logps/rejected": -430.03973388671875,
|
| 770 |
+
"loss": 0.0982,
|
| 771 |
+
"pred_label": 5632.7998046875,
|
| 772 |
+
"rewards/accuracies": 0.65625,
|
| 773 |
+
"rewards/chosen": -12.2637357711792,
|
| 774 |
+
"rewards/margins": 5.055985450744629,
|
| 775 |
+
"rewards/rejected": -17.319721221923828,
|
| 776 |
"step": 470,
|
| 777 |
+
"use_label": 1809.199951171875
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.5,
|
| 781 |
"learning_rate": 2.7648428405122233e-05,
|
| 782 |
+
"logits/chosen": -2.414053201675415,
|
| 783 |
+
"logits/rejected": -2.3774731159210205,
|
| 784 |
+
"logps/chosen": -391.9467468261719,
|
| 785 |
+
"logps/rejected": -501.77960205078125,
|
| 786 |
+
"loss": 0.094,
|
| 787 |
+
"pred_label": 5778.2001953125,
|
| 788 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 789 |
+
"rewards/chosen": -14.637247085571289,
|
| 790 |
+
"rewards/margins": 10.688766479492188,
|
| 791 |
+
"rewards/rejected": -25.32601547241211,
|
| 792 |
"step": 480,
|
| 793 |
+
"use_label": 1823.800048828125
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.51,
|
| 797 |
"learning_rate": 2.7066356228172297e-05,
|
| 798 |
+
"logits/chosen": -2.541294574737549,
|
| 799 |
+
"logits/rejected": -2.4439327716827393,
|
| 800 |
+
"logps/chosen": -361.71771240234375,
|
| 801 |
+
"logps/rejected": -378.765869140625,
|
| 802 |
+
"loss": 0.0927,
|
| 803 |
+
"pred_label": 5916.3251953125,
|
| 804 |
+
"rewards/accuracies": 0.699999988079071,
|
| 805 |
+
"rewards/chosen": -9.4688720703125,
|
| 806 |
+
"rewards/margins": 4.668297290802002,
|
| 807 |
+
"rewards/rejected": -14.137168884277344,
|
| 808 |
"step": 490,
|
| 809 |
+
"use_label": 1845.675048828125
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 0.52,
|
| 813 |
"learning_rate": 2.6484284051222352e-05,
|
| 814 |
+
"logits/chosen": -2.6095941066741943,
|
| 815 |
+
"logits/rejected": -2.4995357990264893,
|
| 816 |
+
"logps/chosen": -364.07427978515625,
|
| 817 |
+
"logps/rejected": -420.4052734375,
|
| 818 |
+
"loss": 0.1016,
|
| 819 |
+
"pred_label": 6059.4501953125,
|
| 820 |
+
"rewards/accuracies": 0.71875,
|
| 821 |
+
"rewards/chosen": -7.4617791175842285,
|
| 822 |
+
"rewards/margins": 8.752523422241211,
|
| 823 |
+
"rewards/rejected": -16.214303970336914,
|
| 824 |
"step": 500,
|
| 825 |
+
"use_label": 1862.550048828125
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 0.53,
|
| 829 |
"learning_rate": 2.590221187427241e-05,
|
| 830 |
+
"logits/chosen": -2.570039749145508,
|
| 831 |
+
"logits/rejected": -2.480402946472168,
|
| 832 |
+
"logps/chosen": -391.964111328125,
|
| 833 |
+
"logps/rejected": -434.7195739746094,
|
| 834 |
+
"loss": 0.1051,
|
| 835 |
+
"pred_label": 6202.64990234375,
|
| 836 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 837 |
+
"rewards/chosen": -11.662691116333008,
|
| 838 |
+
"rewards/margins": 8.459746360778809,
|
| 839 |
+
"rewards/rejected": -20.1224365234375,
|
| 840 |
"step": 510,
|
| 841 |
+
"use_label": 1879.3499755859375
|
| 842 |
},
|
| 843 |
{
|
| 844 |
"epoch": 0.54,
|
| 845 |
"learning_rate": 2.532013969732247e-05,
|
| 846 |
+
"logits/chosen": -2.4564194679260254,
|
| 847 |
+
"logits/rejected": -2.381622076034546,
|
| 848 |
+
"logps/chosen": -406.453125,
|
| 849 |
+
"logps/rejected": -494.87664794921875,
|
| 850 |
+
"loss": 0.068,
|
| 851 |
+
"pred_label": 6340.1748046875,
|
| 852 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 853 |
+
"rewards/chosen": -11.449905395507812,
|
| 854 |
+
"rewards/margins": 11.066286087036133,
|
| 855 |
+
"rewards/rejected": -22.516191482543945,
|
| 856 |
"step": 520,
|
| 857 |
+
"use_label": 1901.824951171875
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.55,
|
| 861 |
"learning_rate": 2.4738067520372525e-05,
|
| 862 |
+
"logits/chosen": -2.3373048305511475,
|
| 863 |
+
"logits/rejected": -2.183617115020752,
|
| 864 |
+
"logps/chosen": -547.97314453125,
|
| 865 |
+
"logps/rejected": -691.542236328125,
|
| 866 |
+
"loss": 0.0709,
|
| 867 |
+
"pred_label": 6487.5498046875,
|
| 868 |
+
"rewards/accuracies": 0.643750011920929,
|
| 869 |
+
"rewards/chosen": -24.765727996826172,
|
| 870 |
+
"rewards/margins": 14.99103832244873,
|
| 871 |
+
"rewards/rejected": -39.75676727294922,
|
| 872 |
"step": 530,
|
| 873 |
+
"use_label": 1914.449951171875
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.57,
|
| 877 |
"learning_rate": 2.4155995343422587e-05,
|
| 878 |
+
"logits/chosen": -2.211622714996338,
|
| 879 |
+
"logits/rejected": -2.0457851886749268,
|
| 880 |
+
"logps/chosen": -632.00634765625,
|
| 881 |
+
"logps/rejected": -867.4461059570312,
|
| 882 |
+
"loss": 0.05,
|
| 883 |
+
"pred_label": 6636.0498046875,
|
| 884 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 885 |
+
"rewards/chosen": -34.34697723388672,
|
| 886 |
+
"rewards/margins": 25.800724029541016,
|
| 887 |
+
"rewards/rejected": -60.14769744873047,
|
| 888 |
"step": 540,
|
| 889 |
+
"use_label": 1925.949951171875
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.58,
|
| 893 |
"learning_rate": 2.3573923166472644e-05,
|
| 894 |
+
"logits/chosen": -2.088925838470459,
|
| 895 |
+
"logits/rejected": -1.9701734781265259,
|
| 896 |
+
"logps/chosen": -796.8790893554688,
|
| 897 |
+
"logps/rejected": -983.8599853515625,
|
| 898 |
+
"loss": 0.045,
|
| 899 |
+
"pred_label": 6784.7998046875,
|
| 900 |
+
"rewards/accuracies": 0.675000011920929,
|
| 901 |
+
"rewards/chosen": -50.76759338378906,
|
| 902 |
+
"rewards/margins": 21.424175262451172,
|
| 903 |
+
"rewards/rejected": -72.19176483154297,
|
| 904 |
"step": 550,
|
| 905 |
+
"use_label": 1937.199951171875
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.59,
|
| 909 |
"learning_rate": 2.2991850989522702e-05,
|
| 910 |
+
"logits/chosen": -2.2610983848571777,
|
| 911 |
+
"logits/rejected": -2.1500396728515625,
|
| 912 |
+
"logps/chosen": -535.4710693359375,
|
| 913 |
+
"logps/rejected": -635.5113525390625,
|
| 914 |
+
"loss": 0.067,
|
| 915 |
+
"pred_label": 6936.47509765625,
|
| 916 |
+
"rewards/accuracies": 0.6875,
|
| 917 |
+
"rewards/chosen": -26.10666847229004,
|
| 918 |
+
"rewards/margins": 15.695854187011719,
|
| 919 |
+
"rewards/rejected": -41.802528381347656,
|
| 920 |
"step": 560,
|
| 921 |
+
"use_label": 1945.5250244140625
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"epoch": 0.6,
|
| 925 |
"learning_rate": 2.240977881257276e-05,
|
| 926 |
+
"logits/chosen": -2.3317577838897705,
|
| 927 |
+
"logits/rejected": -2.215259313583374,
|
| 928 |
+
"logps/chosen": -497.422119140625,
|
| 929 |
+
"logps/rejected": -631.938232421875,
|
| 930 |
+
"loss": 0.0596,
|
| 931 |
+
"pred_label": 7086.02490234375,
|
| 932 |
+
"rewards/accuracies": 0.706250011920929,
|
| 933 |
+
"rewards/chosen": -22.793350219726562,
|
| 934 |
+
"rewards/margins": 14.197186470031738,
|
| 935 |
+
"rewards/rejected": -36.99053955078125,
|
| 936 |
"step": 570,
|
| 937 |
+
"use_label": 1955.9749755859375
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 0.61,
|
| 941 |
"learning_rate": 2.1827706635622818e-05,
|
| 942 |
+
"logits/chosen": -2.399937152862549,
|
| 943 |
+
"logits/rejected": -2.2866618633270264,
|
| 944 |
+
"logps/chosen": -441.74090576171875,
|
| 945 |
+
"logps/rejected": -541.2008666992188,
|
| 946 |
+
"loss": 0.0784,
|
| 947 |
+
"pred_label": 7230.875,
|
| 948 |
+
"rewards/accuracies": 0.699999988079071,
|
| 949 |
+
"rewards/chosen": -16.363243103027344,
|
| 950 |
+
"rewards/margins": 11.310129165649414,
|
| 951 |
+
"rewards/rejected": -27.67337417602539,
|
| 952 |
"step": 580,
|
| 953 |
+
"use_label": 1971.125
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.62,
|
| 957 |
"learning_rate": 2.124563445867288e-05,
|
| 958 |
+
"logits/chosen": -2.3154327869415283,
|
| 959 |
+
"logits/rejected": -2.2939510345458984,
|
| 960 |
+
"logps/chosen": -400.68572998046875,
|
| 961 |
+
"logps/rejected": -483.30279541015625,
|
| 962 |
+
"loss": 0.089,
|
| 963 |
+
"pred_label": 7372.14990234375,
|
| 964 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 965 |
+
"rewards/chosen": -12.908785820007324,
|
| 966 |
+
"rewards/margins": 7.343487739562988,
|
| 967 |
+
"rewards/rejected": -20.252273559570312,
|
| 968 |
"step": 590,
|
| 969 |
+
"use_label": 1989.8499755859375
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.63,
|
| 973 |
"learning_rate": 2.0663562281722934e-05,
|
| 974 |
+
"logits/chosen": -2.463757276535034,
|
| 975 |
+
"logits/rejected": -2.3977725505828857,
|
| 976 |
+
"logps/chosen": -393.55133056640625,
|
| 977 |
+
"logps/rejected": -500.068359375,
|
| 978 |
+
"loss": 0.1022,
|
| 979 |
+
"pred_label": 7510.75,
|
| 980 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 981 |
+
"rewards/chosen": -8.018844604492188,
|
| 982 |
+
"rewards/margins": 11.711953163146973,
|
| 983 |
+
"rewards/rejected": -19.730796813964844,
|
| 984 |
"step": 600,
|
| 985 |
+
"use_label": 2011.25
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 0.64,
|
| 989 |
"learning_rate": 2.0081490104772992e-05,
|
| 990 |
+
"logits/chosen": -2.409186840057373,
|
| 991 |
+
"logits/rejected": -2.3435616493225098,
|
| 992 |
+
"logps/chosen": -412.2386779785156,
|
| 993 |
+
"logps/rejected": -426.6654357910156,
|
| 994 |
+
"loss": 0.0856,
|
| 995 |
+
"pred_label": 7652.7001953125,
|
| 996 |
+
"rewards/accuracies": 0.699999988079071,
|
| 997 |
+
"rewards/chosen": -10.632938385009766,
|
| 998 |
+
"rewards/margins": 8.070967674255371,
|
| 999 |
+
"rewards/rejected": -18.703907012939453,
|
| 1000 |
"step": 610,
|
| 1001 |
+
"use_label": 2029.300048828125
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"epoch": 0.65,
|
| 1005 |
"learning_rate": 1.9499417927823053e-05,
|
| 1006 |
+
"logits/chosen": -2.4093079566955566,
|
| 1007 |
+
"logits/rejected": -2.260230779647827,
|
| 1008 |
+
"logps/chosen": -429.3250427246094,
|
| 1009 |
+
"logps/rejected": -516.5551147460938,
|
| 1010 |
+
"loss": 0.0781,
|
| 1011 |
+
"pred_label": 7797.64990234375,
|
| 1012 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 1013 |
+
"rewards/chosen": -14.934354782104492,
|
| 1014 |
+
"rewards/margins": 11.502239227294922,
|
| 1015 |
+
"rewards/rejected": -26.436594009399414,
|
| 1016 |
"step": 620,
|
| 1017 |
+
"use_label": 2044.3499755859375
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.66,
|
| 1021 |
"learning_rate": 1.8917345750873107e-05,
|
| 1022 |
+
"logits/chosen": -2.2741165161132812,
|
| 1023 |
+
"logits/rejected": -2.1207118034362793,
|
| 1024 |
+
"logps/chosen": -464.8804626464844,
|
| 1025 |
+
"logps/rejected": -637.3003540039062,
|
| 1026 |
+
"loss": 0.0665,
|
| 1027 |
+
"pred_label": 7949.27490234375,
|
| 1028 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 1029 |
+
"rewards/chosen": -17.97607421875,
|
| 1030 |
+
"rewards/margins": 19.63665199279785,
|
| 1031 |
+
"rewards/rejected": -37.61272430419922,
|
| 1032 |
"step": 630,
|
| 1033 |
+
"use_label": 2052.72509765625
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 0.67,
|
| 1037 |
"learning_rate": 1.833527357392317e-05,
|
| 1038 |
+
"logits/chosen": -2.1046862602233887,
|
| 1039 |
+
"logits/rejected": -1.9926544427871704,
|
| 1040 |
+
"logps/chosen": -479.60479736328125,
|
| 1041 |
+
"logps/rejected": -621.544921875,
|
| 1042 |
+
"loss": 0.0622,
|
| 1043 |
+
"pred_label": 8098.72509765625,
|
| 1044 |
+
"rewards/accuracies": 0.699999988079071,
|
| 1045 |
+
"rewards/chosen": -22.641353607177734,
|
| 1046 |
+
"rewards/margins": 14.655824661254883,
|
| 1047 |
+
"rewards/rejected": -37.29717254638672,
|
| 1048 |
"step": 640,
|
| 1049 |
+
"use_label": 2063.27490234375
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
"epoch": 0.68,
|
| 1053 |
"learning_rate": 1.7753201396973227e-05,
|
| 1054 |
+
"logits/chosen": -2.304948329925537,
|
| 1055 |
+
"logits/rejected": -2.159106731414795,
|
| 1056 |
+
"logps/chosen": -497.8910217285156,
|
| 1057 |
+
"logps/rejected": -595.5516967773438,
|
| 1058 |
+
"loss": 0.0522,
|
| 1059 |
+
"pred_label": 8248.5,
|
| 1060 |
+
"rewards/accuracies": 0.737500011920929,
|
| 1061 |
+
"rewards/chosen": -19.594234466552734,
|
| 1062 |
+
"rewards/margins": 12.16260814666748,
|
| 1063 |
+
"rewards/rejected": -31.7568416595459,
|
| 1064 |
"step": 650,
|
| 1065 |
+
"use_label": 2073.5
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 0.69,
|
| 1069 |
"learning_rate": 1.717112922002328e-05,
|
| 1070 |
+
"logits/chosen": -2.3109021186828613,
|
| 1071 |
+
"logits/rejected": -2.2408225536346436,
|
| 1072 |
+
"logps/chosen": -490.71112060546875,
|
| 1073 |
+
"logps/rejected": -506.54766845703125,
|
| 1074 |
+
"loss": 0.087,
|
| 1075 |
+
"pred_label": 8398.224609375,
|
| 1076 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1077 |
+
"rewards/chosen": -19.534738540649414,
|
| 1078 |
+
"rewards/margins": 6.132468223571777,
|
| 1079 |
+
"rewards/rejected": -25.667205810546875,
|
| 1080 |
"step": 660,
|
| 1081 |
+
"use_label": 2083.77490234375
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.7,
|
| 1085 |
"learning_rate": 1.6589057043073342e-05,
|
| 1086 |
+
"logits/chosen": -2.2374701499938965,
|
| 1087 |
+
"logits/rejected": -2.11650013923645,
|
| 1088 |
+
"logps/chosen": -484.6102600097656,
|
| 1089 |
+
"logps/rejected": -569.1256103515625,
|
| 1090 |
+
"loss": 0.064,
|
| 1091 |
+
"pred_label": 8546.0,
|
| 1092 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 1093 |
+
"rewards/chosen": -20.35372543334961,
|
| 1094 |
+
"rewards/margins": 12.419096946716309,
|
| 1095 |
+
"rewards/rejected": -32.772823333740234,
|
| 1096 |
"step": 670,
|
| 1097 |
+
"use_label": 2096.0
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 0.71,
|
| 1101 |
"learning_rate": 1.60069848661234e-05,
|
| 1102 |
+
"logits/chosen": -2.3477349281311035,
|
| 1103 |
+
"logits/rejected": -2.2388532161712646,
|
| 1104 |
+
"logps/chosen": -479.6822814941406,
|
| 1105 |
+
"logps/rejected": -505.4498596191406,
|
| 1106 |
+
"loss": 0.0759,
|
| 1107 |
+
"pred_label": 8695.724609375,
|
| 1108 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1109 |
+
"rewards/chosen": -19.274688720703125,
|
| 1110 |
+
"rewards/margins": 5.44730281829834,
|
| 1111 |
+
"rewards/rejected": -24.72199058532715,
|
| 1112 |
"step": 680,
|
| 1113 |
+
"use_label": 2106.27490234375
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"epoch": 0.72,
|
| 1117 |
"learning_rate": 1.5424912689173458e-05,
|
| 1118 |
+
"logits/chosen": -2.179953098297119,
|
| 1119 |
+
"logits/rejected": -2.0111422538757324,
|
| 1120 |
+
"logps/chosen": -476.6915588378906,
|
| 1121 |
+
"logps/rejected": -548.3240966796875,
|
| 1122 |
+
"loss": 0.0593,
|
| 1123 |
+
"pred_label": 8846.5,
|
| 1124 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1125 |
+
"rewards/chosen": -22.6307430267334,
|
| 1126 |
+
"rewards/margins": 11.273223876953125,
|
| 1127 |
+
"rewards/rejected": -33.903968811035156,
|
| 1128 |
"step": 690,
|
| 1129 |
+
"use_label": 2115.5
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.73,
|
| 1133 |
"learning_rate": 1.4842840512223516e-05,
|
| 1134 |
+
"logits/chosen": -2.298417329788208,
|
| 1135 |
+
"logits/rejected": -2.1161131858825684,
|
| 1136 |
+
"logps/chosen": -477.685546875,
|
| 1137 |
+
"logps/rejected": -721.3785400390625,
|
| 1138 |
+
"loss": 0.0546,
|
| 1139 |
+
"pred_label": 8998.650390625,
|
| 1140 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 1141 |
+
"rewards/chosen": -20.19955062866211,
|
| 1142 |
+
"rewards/margins": 27.11043930053711,
|
| 1143 |
+
"rewards/rejected": -47.30998992919922,
|
| 1144 |
"step": 700,
|
| 1145 |
+
"use_label": 2123.35009765625
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"epoch": 0.74,
|
| 1149 |
"learning_rate": 1.4260768335273575e-05,
|
| 1150 |
+
"logits/chosen": -2.19868803024292,
|
| 1151 |
+
"logits/rejected": -2.0781731605529785,
|
| 1152 |
+
"logps/chosen": -629.6842041015625,
|
| 1153 |
+
"logps/rejected": -780.2570190429688,
|
| 1154 |
+
"loss": 0.0661,
|
| 1155 |
+
"pred_label": 9147.0498046875,
|
| 1156 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 1157 |
+
"rewards/chosen": -34.010963439941406,
|
| 1158 |
+
"rewards/margins": 19.887874603271484,
|
| 1159 |
+
"rewards/rejected": -53.898834228515625,
|
| 1160 |
"step": 710,
|
| 1161 |
+
"use_label": 2134.949951171875
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.75,
|
| 1165 |
"learning_rate": 1.3678696158323633e-05,
|
| 1166 |
+
"logits/chosen": -2.2372653484344482,
|
| 1167 |
+
"logits/rejected": -2.1191489696502686,
|
| 1168 |
+
"logps/chosen": -515.64208984375,
|
| 1169 |
+
"logps/rejected": -691.5008544921875,
|
| 1170 |
+
"loss": 0.0704,
|
| 1171 |
+
"pred_label": 9300.599609375,
|
| 1172 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 1173 |
+
"rewards/chosen": -24.41898536682129,
|
| 1174 |
+
"rewards/margins": 19.390825271606445,
|
| 1175 |
+
"rewards/rejected": -43.809810638427734,
|
| 1176 |
"step": 720,
|
| 1177 |
+
"use_label": 2141.39990234375
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"epoch": 0.76,
|
| 1181 |
"learning_rate": 1.309662398137369e-05,
|
| 1182 |
+
"logits/chosen": -2.2743449211120605,
|
| 1183 |
+
"logits/rejected": -2.1811537742614746,
|
| 1184 |
+
"logps/chosen": -521.6466064453125,
|
| 1185 |
+
"logps/rejected": -568.2872314453125,
|
| 1186 |
+
"loss": 0.0724,
|
| 1187 |
+
"pred_label": 9445.150390625,
|
| 1188 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1189 |
+
"rewards/chosen": -22.4196720123291,
|
| 1190 |
+
"rewards/margins": 10.242195129394531,
|
| 1191 |
+
"rewards/rejected": -32.661869049072266,
|
| 1192 |
"step": 730,
|
| 1193 |
+
"use_label": 2156.85009765625
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.77,
|
| 1197 |
"learning_rate": 1.2514551804423749e-05,
|
| 1198 |
+
"logits/chosen": -2.2904646396636963,
|
| 1199 |
+
"logits/rejected": -2.2049851417541504,
|
| 1200 |
+
"logps/chosen": -520.117919921875,
|
| 1201 |
+
"logps/rejected": -695.0534057617188,
|
| 1202 |
+
"loss": 0.0636,
|
| 1203 |
+
"pred_label": 9594.025390625,
|
| 1204 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1205 |
+
"rewards/chosen": -21.274608612060547,
|
| 1206 |
+
"rewards/margins": 19.581838607788086,
|
| 1207 |
+
"rewards/rejected": -40.8564453125,
|
| 1208 |
"step": 740,
|
| 1209 |
+
"use_label": 2167.97509765625
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 0.79,
|
| 1213 |
"learning_rate": 1.1932479627473807e-05,
|
| 1214 |
+
"logits/chosen": -2.25730562210083,
|
| 1215 |
+
"logits/rejected": -2.1175169944763184,
|
| 1216 |
+
"logps/chosen": -540.2493896484375,
|
| 1217 |
+
"logps/rejected": -738.2742919921875,
|
| 1218 |
+
"loss": 0.063,
|
| 1219 |
+
"pred_label": 9744.1748046875,
|
| 1220 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1221 |
+
"rewards/chosen": -27.976119995117188,
|
| 1222 |
+
"rewards/margins": 19.7985782623291,
|
| 1223 |
+
"rewards/rejected": -47.77469253540039,
|
| 1224 |
"step": 750,
|
| 1225 |
+
"use_label": 2177.824951171875
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"epoch": 0.8,
|
| 1229 |
"learning_rate": 1.1350407450523866e-05,
|
| 1230 |
+
"logits/chosen": -2.3123764991760254,
|
| 1231 |
+
"logits/rejected": -2.2530713081359863,
|
| 1232 |
+
"logps/chosen": -507.7060546875,
|
| 1233 |
+
"logps/rejected": -641.7669067382812,
|
| 1234 |
+
"loss": 0.0539,
|
| 1235 |
+
"pred_label": 9894.150390625,
|
| 1236 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1237 |
+
"rewards/chosen": -21.224315643310547,
|
| 1238 |
+
"rewards/margins": 16.741058349609375,
|
| 1239 |
+
"rewards/rejected": -37.96537780761719,
|
| 1240 |
"step": 760,
|
| 1241 |
+
"use_label": 2187.85009765625
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.81,
|
| 1245 |
"learning_rate": 1.0768335273573923e-05,
|
| 1246 |
+
"logits/chosen": -2.314664363861084,
|
| 1247 |
+
"logits/rejected": -2.259186029434204,
|
| 1248 |
+
"logps/chosen": -518.42333984375,
|
| 1249 |
+
"logps/rejected": -678.8192749023438,
|
| 1250 |
+
"loss": 0.0433,
|
| 1251 |
+
"pred_label": 10047.875,
|
| 1252 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1253 |
+
"rewards/chosen": -21.612756729125977,
|
| 1254 |
+
"rewards/margins": 17.79681396484375,
|
| 1255 |
+
"rewards/rejected": -39.409568786621094,
|
| 1256 |
"step": 770,
|
| 1257 |
+
"use_label": 2194.125
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"epoch": 0.82,
|
| 1261 |
"learning_rate": 1.0186263096623982e-05,
|
| 1262 |
+
"logits/chosen": -2.287078380584717,
|
| 1263 |
+
"logits/rejected": -2.199553966522217,
|
| 1264 |
+
"logps/chosen": -530.3980102539062,
|
| 1265 |
+
"logps/rejected": -631.2974243164062,
|
| 1266 |
+
"loss": 0.051,
|
| 1267 |
+
"pred_label": 10198.8251953125,
|
| 1268 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 1269 |
+
"rewards/chosen": -24.92042350769043,
|
| 1270 |
+
"rewards/margins": 13.495773315429688,
|
| 1271 |
+
"rewards/rejected": -38.41619873046875,
|
| 1272 |
"step": 780,
|
| 1273 |
+
"use_label": 2203.175048828125
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 0.83,
|
| 1277 |
"learning_rate": 9.60419091967404e-06,
|
| 1278 |
+
"logits/chosen": -2.3670148849487305,
|
| 1279 |
+
"logits/rejected": -2.2450852394104004,
|
| 1280 |
+
"logps/chosen": -533.3482666015625,
|
| 1281 |
+
"logps/rejected": -667.7590942382812,
|
| 1282 |
+
"loss": 0.0574,
|
| 1283 |
+
"pred_label": 10346.0,
|
| 1284 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1285 |
+
"rewards/chosen": -26.55388832092285,
|
| 1286 |
+
"rewards/margins": 14.149024963378906,
|
| 1287 |
+
"rewards/rejected": -40.702919006347656,
|
| 1288 |
"step": 790,
|
| 1289 |
+
"use_label": 2216.0
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
"epoch": 0.84,
|
| 1293 |
"learning_rate": 9.022118742724098e-06,
|
| 1294 |
+
"logits/chosen": -2.2852892875671387,
|
| 1295 |
+
"logits/rejected": -2.191178798675537,
|
| 1296 |
+
"logps/chosen": -515.4843139648438,
|
| 1297 |
+
"logps/rejected": -658.4345703125,
|
| 1298 |
+
"loss": 0.0732,
|
| 1299 |
+
"pred_label": 10492.0,
|
| 1300 |
+
"rewards/accuracies": 0.75,
|
| 1301 |
+
"rewards/chosen": -20.101245880126953,
|
| 1302 |
+
"rewards/margins": 18.136005401611328,
|
| 1303 |
+
"rewards/rejected": -38.237247467041016,
|
| 1304 |
"step": 800,
|
| 1305 |
+
"use_label": 2230.0
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.85,
|
| 1309 |
"learning_rate": 8.440046565774158e-06,
|
| 1310 |
+
"logits/chosen": -2.308079481124878,
|
| 1311 |
+
"logits/rejected": -2.2213778495788574,
|
| 1312 |
+
"logps/chosen": -485.36602783203125,
|
| 1313 |
+
"logps/rejected": -505.19189453125,
|
| 1314 |
+
"loss": 0.0907,
|
| 1315 |
+
"pred_label": 10638.150390625,
|
| 1316 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1317 |
+
"rewards/chosen": -20.434207916259766,
|
| 1318 |
+
"rewards/margins": 7.203681945800781,
|
| 1319 |
+
"rewards/rejected": -27.637889862060547,
|
| 1320 |
"step": 810,
|
| 1321 |
+
"use_label": 2243.85009765625
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"epoch": 0.86,
|
| 1325 |
"learning_rate": 7.857974388824214e-06,
|
| 1326 |
+
"logits/chosen": -2.3282344341278076,
|
| 1327 |
+
"logits/rejected": -2.243950366973877,
|
| 1328 |
+
"logps/chosen": -473.38262939453125,
|
| 1329 |
+
"logps/rejected": -566.6471557617188,
|
| 1330 |
+
"loss": 0.0883,
|
| 1331 |
+
"pred_label": 10780.900390625,
|
| 1332 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1333 |
+
"rewards/chosen": -18.006444931030273,
|
| 1334 |
+
"rewards/margins": 14.502110481262207,
|
| 1335 |
+
"rewards/rejected": -32.5085563659668,
|
| 1336 |
"step": 820,
|
| 1337 |
+
"use_label": 2261.10009765625
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"epoch": 0.87,
|
| 1341 |
"learning_rate": 7.275902211874273e-06,
|
| 1342 |
+
"logits/chosen": -2.345092296600342,
|
| 1343 |
+
"logits/rejected": -2.301511287689209,
|
| 1344 |
+
"logps/chosen": -431.64520263671875,
|
| 1345 |
+
"logps/rejected": -632.6573486328125,
|
| 1346 |
+
"loss": 0.0699,
|
| 1347 |
+
"pred_label": 10927.275390625,
|
| 1348 |
+
"rewards/accuracies": 0.75,
|
| 1349 |
+
"rewards/chosen": -16.88318634033203,
|
| 1350 |
+
"rewards/margins": 18.3623104095459,
|
| 1351 |
+
"rewards/rejected": -35.24549865722656,
|
| 1352 |
"step": 830,
|
| 1353 |
+
"use_label": 2274.72509765625
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.88,
|
| 1357 |
"learning_rate": 6.693830034924331e-06,
|
| 1358 |
+
"logits/chosen": -2.3140549659729004,
|
| 1359 |
+
"logits/rejected": -2.3112993240356445,
|
| 1360 |
+
"logps/chosen": -505.6224670410156,
|
| 1361 |
+
"logps/rejected": -544.6993408203125,
|
| 1362 |
+
"loss": 0.074,
|
| 1363 |
+
"pred_label": 11074.5498046875,
|
| 1364 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1365 |
+
"rewards/chosen": -23.76053237915039,
|
| 1366 |
+
"rewards/margins": 4.261009216308594,
|
| 1367 |
+
"rewards/rejected": -28.02153968811035,
|
| 1368 |
"step": 840,
|
| 1369 |
+
"use_label": 2287.449951171875
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.89,
|
| 1373 |
"learning_rate": 6.111757857974389e-06,
|
| 1374 |
+
"logits/chosen": -2.3638854026794434,
|
| 1375 |
+
"logits/rejected": -2.3075807094573975,
|
| 1376 |
+
"logps/chosen": -458.67181396484375,
|
| 1377 |
+
"logps/rejected": -516.2307739257812,
|
| 1378 |
+
"loss": 0.0849,
|
| 1379 |
+
"pred_label": 11220.900390625,
|
| 1380 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 1381 |
+
"rewards/chosen": -16.364498138427734,
|
| 1382 |
+
"rewards/margins": 10.922462463378906,
|
| 1383 |
+
"rewards/rejected": -27.28696060180664,
|
| 1384 |
"step": 850,
|
| 1385 |
+
"use_label": 2301.10009765625
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 0.9,
|
| 1389 |
"learning_rate": 5.529685681024447e-06,
|
| 1390 |
+
"logits/chosen": -2.3312666416168213,
|
| 1391 |
+
"logits/rejected": -2.2601587772369385,
|
| 1392 |
+
"logps/chosen": -464.119140625,
|
| 1393 |
+
"logps/rejected": -580.7103881835938,
|
| 1394 |
+
"loss": 0.0727,
|
| 1395 |
+
"pred_label": 11369.724609375,
|
| 1396 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1397 |
+
"rewards/chosen": -17.923574447631836,
|
| 1398 |
+
"rewards/margins": 12.89836597442627,
|
| 1399 |
+
"rewards/rejected": -30.821941375732422,
|
| 1400 |
"step": 860,
|
| 1401 |
+
"use_label": 2312.27490234375
|
| 1402 |
},
|
| 1403 |
{
|
| 1404 |
"epoch": 0.91,
|
| 1405 |
"learning_rate": 4.947613504074506e-06,
|
| 1406 |
+
"logits/chosen": -2.3686623573303223,
|
| 1407 |
+
"logits/rejected": -2.2975738048553467,
|
| 1408 |
+
"logps/chosen": -374.78192138671875,
|
| 1409 |
+
"logps/rejected": -505.87152099609375,
|
| 1410 |
+
"loss": 0.0739,
|
| 1411 |
+
"pred_label": 11518.5751953125,
|
| 1412 |
+
"rewards/accuracies": 0.731249988079071,
|
| 1413 |
+
"rewards/chosen": -13.179441452026367,
|
| 1414 |
+
"rewards/margins": 13.382822036743164,
|
| 1415 |
+
"rewards/rejected": -26.562265396118164,
|
| 1416 |
"step": 870,
|
| 1417 |
+
"use_label": 2323.425048828125
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 0.92,
|
| 1421 |
"learning_rate": 4.3655413271245635e-06,
|
| 1422 |
+
"logits/chosen": -2.3770673274993896,
|
| 1423 |
+
"logits/rejected": -2.2960658073425293,
|
| 1424 |
+
"logps/chosen": -470.25439453125,
|
| 1425 |
+
"logps/rejected": -548.5438842773438,
|
| 1426 |
+
"loss": 0.0668,
|
| 1427 |
+
"pred_label": 11666.25,
|
| 1428 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 1429 |
+
"rewards/chosen": -18.366519927978516,
|
| 1430 |
+
"rewards/margins": 10.322815895080566,
|
| 1431 |
+
"rewards/rejected": -28.689334869384766,
|
| 1432 |
"step": 880,
|
| 1433 |
+
"use_label": 2335.75
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
"epoch": 0.93,
|
| 1437 |
"learning_rate": 3.7834691501746217e-06,
|
| 1438 |
+
"logits/chosen": -2.3971574306488037,
|
| 1439 |
+
"logits/rejected": -2.3389389514923096,
|
| 1440 |
+
"logps/chosen": -445.2398986816406,
|
| 1441 |
+
"logps/rejected": -555.1053466796875,
|
| 1442 |
+
"loss": 0.0667,
|
| 1443 |
+
"pred_label": 11815.775390625,
|
| 1444 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1445 |
+
"rewards/chosen": -16.39553451538086,
|
| 1446 |
+
"rewards/margins": 11.912240982055664,
|
| 1447 |
+
"rewards/rejected": -28.30777931213379,
|
| 1448 |
"step": 890,
|
| 1449 |
+
"use_label": 2346.22509765625
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.94,
|
| 1453 |
"learning_rate": 3.2013969732246805e-06,
|
| 1454 |
+
"logits/chosen": -2.279033660888672,
|
| 1455 |
+
"logits/rejected": -2.161271572113037,
|
| 1456 |
+
"logps/chosen": -450.6705627441406,
|
| 1457 |
+
"logps/rejected": -566.3276977539062,
|
| 1458 |
+
"loss": 0.0876,
|
| 1459 |
+
"pred_label": 11961.5751953125,
|
| 1460 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1461 |
+
"rewards/chosen": -17.153247833251953,
|
| 1462 |
+
"rewards/margins": 13.627708435058594,
|
| 1463 |
+
"rewards/rejected": -30.780956268310547,
|
| 1464 |
"step": 900,
|
| 1465 |
+
"use_label": 2360.425048828125
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 0.95,
|
| 1469 |
"learning_rate": 2.6193247962747383e-06,
|
| 1470 |
+
"logits/chosen": -2.414407253265381,
|
| 1471 |
+
"logits/rejected": -2.305093288421631,
|
| 1472 |
+
"logps/chosen": -450.60595703125,
|
| 1473 |
+
"logps/rejected": -530.6917724609375,
|
| 1474 |
+
"loss": 0.0717,
|
| 1475 |
+
"pred_label": 12107.375,
|
| 1476 |
+
"rewards/accuracies": 0.731249988079071,
|
| 1477 |
+
"rewards/chosen": -15.418998718261719,
|
| 1478 |
+
"rewards/margins": 11.336746215820312,
|
| 1479 |
+
"rewards/rejected": -26.7557430267334,
|
| 1480 |
"step": 910,
|
| 1481 |
+
"use_label": 2374.625
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.96,
|
| 1485 |
"learning_rate": 2.037252619324796e-06,
|
| 1486 |
+
"logits/chosen": -2.346126079559326,
|
| 1487 |
+
"logits/rejected": -2.2459468841552734,
|
| 1488 |
+
"logps/chosen": -454.90740966796875,
|
| 1489 |
+
"logps/rejected": -556.7593994140625,
|
| 1490 |
+
"loss": 0.0626,
|
| 1491 |
+
"pred_label": 12256.9501953125,
|
| 1492 |
+
"rewards/accuracies": 0.699999988079071,
|
| 1493 |
+
"rewards/chosen": -15.801968574523926,
|
| 1494 |
+
"rewards/margins": 13.915499687194824,
|
| 1495 |
+
"rewards/rejected": -29.717464447021484,
|
| 1496 |
"step": 920,
|
| 1497 |
+
"use_label": 2385.050048828125
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 0.97,
|
| 1501 |
"learning_rate": 1.4551804423748545e-06,
|
| 1502 |
+
"logits/chosen": -2.400489091873169,
|
| 1503 |
+
"logits/rejected": -2.255152702331543,
|
| 1504 |
+
"logps/chosen": -392.2051086425781,
|
| 1505 |
+
"logps/rejected": -571.8855590820312,
|
| 1506 |
+
"loss": 0.0663,
|
| 1507 |
+
"pred_label": 12408.974609375,
|
| 1508 |
+
"rewards/accuracies": 0.8125,
|
| 1509 |
+
"rewards/chosen": -12.399200439453125,
|
| 1510 |
+
"rewards/margins": 20.609838485717773,
|
| 1511 |
+
"rewards/rejected": -33.00904083251953,
|
| 1512 |
"step": 930,
|
| 1513 |
+
"use_label": 2393.02490234375
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
"epoch": 0.98,
|
| 1517 |
"learning_rate": 8.731082654249127e-07,
|
| 1518 |
+
"logits/chosen": -2.398634195327759,
|
| 1519 |
+
"logits/rejected": -2.28908634185791,
|
| 1520 |
+
"logps/chosen": -459.11846923828125,
|
| 1521 |
+
"logps/rejected": -550.9254150390625,
|
| 1522 |
+
"loss": 0.0731,
|
| 1523 |
+
"pred_label": 12559.150390625,
|
| 1524 |
+
"rewards/accuracies": 0.71875,
|
| 1525 |
+
"rewards/chosen": -17.1726016998291,
|
| 1526 |
+
"rewards/margins": 11.231449127197266,
|
| 1527 |
+
"rewards/rejected": -28.404048919677734,
|
| 1528 |
"step": 940,
|
| 1529 |
+
"use_label": 2402.85009765625
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 0.99,
|
| 1533 |
"learning_rate": 2.910360884749709e-07,
|
| 1534 |
+
"logits/chosen": -2.373722791671753,
|
| 1535 |
+
"logits/rejected": -2.2480380535125732,
|
| 1536 |
+
"logps/chosen": -450.88623046875,
|
| 1537 |
+
"logps/rejected": -560.01220703125,
|
| 1538 |
+
"loss": 0.069,
|
| 1539 |
+
"pred_label": 12709.775390625,
|
| 1540 |
+
"rewards/accuracies": 0.6875,
|
| 1541 |
+
"rewards/chosen": -18.243160247802734,
|
| 1542 |
+
"rewards/margins": 12.46325397491455,
|
| 1543 |
+
"rewards/rejected": -30.7064151763916,
|
| 1544 |
"step": 950,
|
| 1545 |
+
"use_label": 2412.22509765625
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 1.0,
|
| 1549 |
+
"eval_logits/chosen": -2.3870656490325928,
|
| 1550 |
+
"eval_logits/rejected": -2.274824380874634,
|
| 1551 |
+
"eval_logps/chosen": -441.0377502441406,
|
| 1552 |
+
"eval_logps/rejected": -538.9608764648438,
|
| 1553 |
+
"eval_loss": 0.07583338022232056,
|
| 1554 |
+
"eval_pred_label": 13084.7841796875,
|
| 1555 |
+
"eval_rewards/accuracies": 0.7080000042915344,
|
| 1556 |
+
"eval_rewards/chosen": -15.678034782409668,
|
| 1557 |
+
"eval_rewards/margins": 12.288079261779785,
|
| 1558 |
+
"eval_rewards/rejected": -27.966114044189453,
|
| 1559 |
+
"eval_runtime": 450.8813,
|
| 1560 |
+
"eval_samples_per_second": 4.436,
|
| 1561 |
+
"eval_steps_per_second": 0.277,
|
| 1562 |
+
"eval_use_label": 2447.216064453125,
|
| 1563 |
"step": 955
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 1.0,
|
| 1567 |
"step": 955,
|
| 1568 |
"total_flos": 0.0,
|
| 1569 |
+
"train_loss": 0.12920980815488006,
|
| 1570 |
+
"train_runtime": 25162.7962,
|
| 1571 |
+
"train_samples_per_second": 2.43,
|
| 1572 |
"train_steps_per_second": 0.038
|
| 1573 |
}
|
| 1574 |
],
|