Jerry Ji
commited on
Commit
·
37290e2
1
Parent(s):
824c9d3
Training in progress, epoch 0
Browse files- README.md +11 -13
- adapter_model.safetensors +1 -1
- all_results.json +16 -3
- eval_results.json +13 -13
- train_results.json +3 -3
- trainer_state.json +871 -871
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -15,15 +15,15 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) 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: 0.
|
| 23 |
-
- Logps/rejected: -
|
| 24 |
-
- Logps/chosen: -
|
| 25 |
-
- Logits/rejected: -
|
| 26 |
-
- Logits/chosen: -2.
|
| 27 |
|
| 28 |
## Model description
|
| 29 |
|
|
@@ -54,15 +54,13 @@ The following hyperparameters were used during training:
|
|
| 54 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 55 |
- lr_scheduler_type: linear
|
| 56 |
- lr_scheduler_warmup_ratio: 0.1
|
| 57 |
-
- num_epochs:
|
| 58 |
|
| 59 |
### Training results
|
| 60 |
|
| 61 |
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 62 |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 63 |
-
| 0.
|
| 64 |
-
| 0.545 | 2.0 | 1937 | 0.5313 | -0.1468 | -0.8623 | 0.7440 | 0.7156 | -227.9077 | -266.1287 | -1.9506 | -2.0746 |
|
| 65 |
-
| 0.5342 | 3.0 | 2904 | 0.5263 | -0.1493 | -0.8998 | 0.7480 | 0.7505 | -228.2820 | -266.1538 | -1.9412 | -2.0663 |
|
| 66 |
|
| 67 |
|
| 68 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.6642
|
| 19 |
+
- Rewards/chosen: 0.1042
|
| 20 |
+
- Rewards/rejected: 0.0401
|
| 21 |
+
- Rewards/accuracies: 0.6480
|
| 22 |
+
- Rewards/margins: 0.0641
|
| 23 |
+
- Logps/rejected: -230.4560
|
| 24 |
+
- Logps/chosen: -278.6917
|
| 25 |
+
- Logits/rejected: -2.3987
|
| 26 |
+
- Logits/chosen: -2.4597
|
| 27 |
|
| 28 |
## Model description
|
| 29 |
|
|
|
|
| 54 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 55 |
- lr_scheduler_type: linear
|
| 56 |
- lr_scheduler_warmup_ratio: 0.1
|
| 57 |
+
- num_epochs: 1
|
| 58 |
|
| 59 |
### Training results
|
| 60 |
|
| 61 |
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 62 |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 63 |
+
| 0.661 | 1.0 | 968 | 0.6642 | 0.1042 | 0.0401 | 0.6480 | 0.0641 | -230.4560 | -278.6917 | -2.3987 | -2.4597 |
|
|
|
|
|
|
|
| 64 |
|
| 65 |
|
| 66 |
### Framework versions
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 109086672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5355fdbd512d44198186b663cbaf8edc1b33b367aa093b0be23eee7161c1b84
|
| 3 |
size 109086672
|
all_results.json
CHANGED
|
@@ -1,8 +1,21 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"
|
| 4 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"train_samples": 61966,
|
| 6 |
-
"train_samples_per_second": 2.
|
| 7 |
"train_steps_per_second": 0.035
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.4597132205963135,
|
| 4 |
+
"eval_logits/rejected": -2.398695468902588,
|
| 5 |
+
"eval_logps/chosen": -278.69171142578125,
|
| 6 |
+
"eval_logps/rejected": -230.4560089111328,
|
| 7 |
+
"eval_loss": 0.6642152070999146,
|
| 8 |
+
"eval_rewards/accuracies": 0.6480000019073486,
|
| 9 |
+
"eval_rewards/chosen": 0.10415761172771454,
|
| 10 |
+
"eval_rewards/margins": 0.06405296921730042,
|
| 11 |
+
"eval_rewards/rejected": 0.04010463133454323,
|
| 12 |
+
"eval_runtime": 444.8959,
|
| 13 |
+
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 4.495,
|
| 15 |
+
"eval_steps_per_second": 0.281,
|
| 16 |
+
"train_loss": 0.6728762634529555,
|
| 17 |
+
"train_runtime": 27528.1814,
|
| 18 |
"train_samples": 61966,
|
| 19 |
+
"train_samples_per_second": 2.251,
|
| 20 |
"train_steps_per_second": 0.035
|
| 21 |
}
|
eval_results.json
CHANGED
|
@@ -1,16 +1,16 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_logits/chosen": -2.
|
| 4 |
-
"eval_logits/rejected": -
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_rewards/accuracies": 0.
|
| 9 |
-
"eval_rewards/chosen":
|
| 10 |
-
"eval_rewards/margins": 0.
|
| 11 |
-
"eval_rewards/rejected":
|
| 12 |
-
"eval_runtime":
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second": 4.
|
| 15 |
-
"eval_steps_per_second": 0.
|
| 16 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.4597132205963135,
|
| 4 |
+
"eval_logits/rejected": -2.398695468902588,
|
| 5 |
+
"eval_logps/chosen": -278.69171142578125,
|
| 6 |
+
"eval_logps/rejected": -230.4560089111328,
|
| 7 |
+
"eval_loss": 0.6642152070999146,
|
| 8 |
+
"eval_rewards/accuracies": 0.6480000019073486,
|
| 9 |
+
"eval_rewards/chosen": 0.10415761172771454,
|
| 10 |
+
"eval_rewards/margins": 0.06405296921730042,
|
| 11 |
+
"eval_rewards/rejected": 0.04010463133454323,
|
| 12 |
+
"eval_runtime": 444.8959,
|
| 13 |
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 4.495,
|
| 15 |
+
"eval_steps_per_second": 0.281
|
| 16 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 61966,
|
| 6 |
-
"train_samples_per_second": 2.
|
| 7 |
"train_steps_per_second": 0.035
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.6728762634529555,
|
| 4 |
+
"train_runtime": 27528.1814,
|
| 5 |
"train_samples": 61966,
|
| 6 |
+
"train_samples_per_second": 2.251,
|
| 7 |
"train_steps_per_second": 0.035
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -25,1370 +25,1370 @@
|
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
"learning_rate": 5.154639175257731e-08,
|
| 28 |
-
"logits/chosen": -2.
|
| 29 |
-
"logits/rejected": -2.
|
| 30 |
-
"logps/chosen": -284.
|
| 31 |
-
"logps/rejected": -205.
|
| 32 |
-
"loss": 0.
|
| 33 |
-
"rewards/accuracies": 0.
|
| 34 |
-
"rewards/chosen": -0.
|
| 35 |
-
"rewards/margins":
|
| 36 |
-
"rewards/rejected": -0.
|
| 37 |
"step": 10
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02,
|
| 41 |
"learning_rate": 1.0309278350515462e-07,
|
| 42 |
-
"logits/chosen": -2.
|
| 43 |
-
"logits/rejected": -2.
|
| 44 |
-
"logps/chosen": -320.
|
| 45 |
-
"logps/rejected": -248.
|
| 46 |
-
"loss": 0.
|
| 47 |
"rewards/accuracies": 0.512499988079071,
|
| 48 |
-
"rewards/chosen": 0.
|
| 49 |
-
"rewards/margins":
|
| 50 |
-
"rewards/rejected": 0.
|
| 51 |
"step": 20
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03,
|
| 55 |
"learning_rate": 1.5463917525773197e-07,
|
| 56 |
-
"logits/chosen": -2.
|
| 57 |
-
"logits/rejected": -2.
|
| 58 |
-
"logps/chosen": -268.
|
| 59 |
-
"logps/rejected": -227.
|
| 60 |
-
"loss": 0.
|
| 61 |
-
"rewards/accuracies": 0.
|
| 62 |
-
"rewards/chosen":
|
| 63 |
-
"rewards/margins": 0.
|
| 64 |
-
"rewards/rejected": -0.
|
| 65 |
"step": 30
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.04,
|
| 69 |
"learning_rate": 2.0618556701030925e-07,
|
| 70 |
-
"logits/chosen": -2.
|
| 71 |
-
"logits/rejected": -2.
|
| 72 |
-
"logps/chosen": -308.
|
| 73 |
-
"logps/rejected": -253.
|
| 74 |
-
"loss": 0.
|
| 75 |
-
"rewards/accuracies": 0.
|
| 76 |
-
"rewards/chosen": 0.
|
| 77 |
-
"rewards/margins": 0.
|
| 78 |
-
"rewards/rejected":
|
| 79 |
"step": 40
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05,
|
| 83 |
"learning_rate": 2.5773195876288655e-07,
|
| 84 |
-
"logits/chosen": -2.
|
| 85 |
-
"logits/rejected": -2.
|
| 86 |
-
"logps/chosen": -297.
|
| 87 |
-
"logps/rejected": -227.
|
| 88 |
-
"loss": 0.
|
| 89 |
-
"rewards/accuracies": 0.
|
| 90 |
-
"rewards/chosen": 0.
|
| 91 |
-
"rewards/margins": 0.
|
| 92 |
-
"rewards/rejected": -0.
|
| 93 |
"step": 50
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06,
|
| 97 |
"learning_rate": 3.0927835051546394e-07,
|
| 98 |
-
"logits/chosen": -2.
|
| 99 |
-
"logits/rejected": -2.
|
| 100 |
-
"logps/chosen": -256.
|
| 101 |
-
"logps/rejected": -229.
|
| 102 |
-
"loss": 0.
|
| 103 |
-
"rewards/accuracies": 0.
|
| 104 |
-
"rewards/chosen":
|
| 105 |
-
"rewards/margins": 0.
|
| 106 |
-
"rewards/rejected": -0.
|
| 107 |
"step": 60
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07,
|
| 111 |
"learning_rate": 3.608247422680412e-07,
|
| 112 |
-
"logits/chosen": -2.
|
| 113 |
-
"logits/rejected": -2.
|
| 114 |
-
"logps/chosen": -313.
|
| 115 |
-
"logps/rejected": -252.
|
| 116 |
-
"loss": 0.
|
| 117 |
-
"rewards/accuracies": 0.
|
| 118 |
-
"rewards/chosen":
|
| 119 |
-
"rewards/margins":
|
| 120 |
-
"rewards/rejected":
|
| 121 |
"step": 70
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08,
|
| 125 |
"learning_rate": 4.123711340206185e-07,
|
| 126 |
-
"logits/chosen": -2.
|
| 127 |
-
"logits/rejected": -2.
|
| 128 |
-
"logps/chosen": -302.
|
| 129 |
-
"logps/rejected": -243.
|
| 130 |
-
"loss": 0.
|
| 131 |
-
"rewards/accuracies": 0.
|
| 132 |
-
"rewards/chosen": 0.
|
| 133 |
-
"rewards/margins": -0.
|
| 134 |
-
"rewards/rejected": 0.
|
| 135 |
"step": 80
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09,
|
| 139 |
"learning_rate": 4.639175257731959e-07,
|
| 140 |
-
"logits/chosen": -2.
|
| 141 |
-
"logits/rejected": -2.
|
| 142 |
-
"logps/chosen": -270.
|
| 143 |
-
"logps/rejected": -216.
|
| 144 |
-
"loss": 0.
|
| 145 |
-
"rewards/accuracies": 0.
|
| 146 |
-
"rewards/chosen": 0.
|
| 147 |
-
"rewards/margins": 0.
|
| 148 |
-
"rewards/rejected": 0.
|
| 149 |
"step": 90
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.1,
|
| 153 |
"learning_rate": 4.982778415614236e-07,
|
| 154 |
-
"logits/chosen": -2.
|
| 155 |
-
"logits/rejected": -2.
|
| 156 |
-
"logps/chosen": -274.
|
| 157 |
-
"logps/rejected": -226.
|
| 158 |
-
"loss": 0.
|
| 159 |
-
"rewards/accuracies": 0.
|
| 160 |
-
"rewards/chosen": 0.
|
| 161 |
-
"rewards/margins": 0.
|
| 162 |
-
"rewards/rejected":
|
| 163 |
"step": 100
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.11,
|
| 167 |
"learning_rate": 4.925373134328357e-07,
|
| 168 |
-
"logits/chosen": -2.
|
| 169 |
-
"logits/rejected": -2.
|
| 170 |
-
"logps/chosen": -274.
|
| 171 |
-
"logps/rejected": -232.
|
| 172 |
-
"loss": 0.
|
| 173 |
-
"rewards/accuracies": 0.
|
| 174 |
-
"rewards/chosen": 0.
|
| 175 |
-
"rewards/margins": 0.
|
| 176 |
-
"rewards/rejected": -0.
|
| 177 |
"step": 110
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.12,
|
| 181 |
"learning_rate": 4.867967853042479e-07,
|
| 182 |
-
"logits/chosen": -2.
|
| 183 |
-
"logits/rejected": -2.
|
| 184 |
-
"logps/chosen": -319.
|
| 185 |
-
"logps/rejected": -235.
|
| 186 |
-
"loss": 0.
|
| 187 |
-
"rewards/accuracies": 0.
|
| 188 |
-
"rewards/chosen": 0.
|
| 189 |
-
"rewards/margins": 0.
|
| 190 |
-
"rewards/rejected": 0.
|
| 191 |
"step": 120
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.13,
|
| 195 |
"learning_rate": 4.810562571756601e-07,
|
| 196 |
-
"logits/chosen": -2.
|
| 197 |
-
"logits/rejected": -2.
|
| 198 |
-
"logps/chosen": -296.
|
| 199 |
-
"logps/rejected": -245.
|
| 200 |
-
"loss": 0.
|
| 201 |
-
"rewards/accuracies": 0.
|
| 202 |
-
"rewards/chosen": 0.
|
| 203 |
-
"rewards/margins": 0.
|
| 204 |
-
"rewards/rejected": 0.
|
| 205 |
"step": 130
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.14,
|
| 209 |
"learning_rate": 4.753157290470723e-07,
|
| 210 |
-
"logits/chosen": -2.
|
| 211 |
-
"logits/rejected": -2.
|
| 212 |
-
"logps/chosen": -301.
|
| 213 |
-
"logps/rejected": -239.
|
| 214 |
-
"loss": 0.
|
| 215 |
-
"rewards/accuracies": 0.
|
| 216 |
-
"rewards/chosen": 0.
|
| 217 |
-
"rewards/margins": 0.
|
| 218 |
-
"rewards/rejected": 0.
|
| 219 |
"step": 140
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.15,
|
| 223 |
"learning_rate": 4.6957520091848447e-07,
|
| 224 |
-
"logits/chosen": -2.
|
| 225 |
-
"logits/rejected": -2.
|
| 226 |
-
"logps/chosen": -284.
|
| 227 |
-
"logps/rejected": -259.
|
| 228 |
-
"loss": 0.
|
| 229 |
-
"rewards/accuracies": 0.
|
| 230 |
-
"rewards/chosen": 0.
|
| 231 |
-
"rewards/margins": 0.
|
| 232 |
-
"rewards/rejected": 0.
|
| 233 |
"step": 150
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.17,
|
| 237 |
"learning_rate": 4.6383467278989666e-07,
|
| 238 |
-
"logits/chosen": -2.
|
| 239 |
-
"logits/rejected": -2.
|
| 240 |
-
"logps/chosen": -286.
|
| 241 |
-
"logps/rejected": -221.
|
| 242 |
-
"loss": 0.
|
| 243 |
-
"rewards/accuracies": 0.
|
| 244 |
-
"rewards/chosen": 0.
|
| 245 |
-
"rewards/margins": 0.
|
| 246 |
-
"rewards/rejected": 0.
|
| 247 |
"step": 160
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.18,
|
| 251 |
"learning_rate": 4.580941446613088e-07,
|
| 252 |
-
"logits/chosen": -2.
|
| 253 |
-
"logits/rejected": -2.
|
| 254 |
-
"logps/chosen": -301.
|
| 255 |
-
"logps/rejected": -239.
|
| 256 |
-
"loss": 0.
|
| 257 |
-
"rewards/accuracies": 0.
|
| 258 |
-
"rewards/chosen": 0.
|
| 259 |
-
"rewards/margins": 0.
|
| 260 |
-
"rewards/rejected": 0.
|
| 261 |
"step": 170
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.19,
|
| 265 |
"learning_rate": 4.52353616532721e-07,
|
| 266 |
-
"logits/chosen": -2.
|
| 267 |
-
"logits/rejected": -2.
|
| 268 |
-
"logps/chosen": -257.
|
| 269 |
-
"logps/rejected": -214.
|
| 270 |
-
"loss": 0.
|
| 271 |
-
"rewards/accuracies": 0.
|
| 272 |
-
"rewards/chosen": 0.
|
| 273 |
-
"rewards/margins": 0.
|
| 274 |
-
"rewards/rejected": 0.
|
| 275 |
"step": 180
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2,
|
| 279 |
"learning_rate": 4.4661308840413316e-07,
|
| 280 |
-
"logits/chosen": -2.
|
| 281 |
-
"logits/rejected": -2.
|
| 282 |
-
"logps/chosen": -253.
|
| 283 |
-
"logps/rejected": -206.
|
| 284 |
-
"loss": 0.
|
| 285 |
-
"rewards/accuracies": 0.
|
| 286 |
-
"rewards/chosen": 0.
|
| 287 |
-
"rewards/margins": 0.
|
| 288 |
-
"rewards/rejected": 0.
|
| 289 |
"step": 190
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.21,
|
| 293 |
"learning_rate": 4.408725602755453e-07,
|
| 294 |
-
"logits/chosen": -2.
|
| 295 |
-
"logits/rejected": -2.
|
| 296 |
-
"logps/chosen": -261.
|
| 297 |
-
"logps/rejected": -195.
|
| 298 |
-
"loss": 0.
|
| 299 |
-
"rewards/accuracies": 0.
|
| 300 |
-
"rewards/chosen": 0.
|
| 301 |
-
"rewards/margins": 0.
|
| 302 |
-
"rewards/rejected": 0.
|
| 303 |
"step": 200
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.22,
|
| 307 |
"learning_rate": 4.351320321469575e-07,
|
| 308 |
-
"logits/chosen": -2.
|
| 309 |
-
"logits/rejected": -2.
|
| 310 |
-
"logps/chosen": -302.
|
| 311 |
-
"logps/rejected": -218.
|
| 312 |
-
"loss": 0.
|
| 313 |
-
"rewards/accuracies": 0.
|
| 314 |
-
"rewards/chosen": 0.
|
| 315 |
-
"rewards/margins": 0.
|
| 316 |
-
"rewards/rejected": 0.
|
| 317 |
"step": 210
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.23,
|
| 321 |
"learning_rate": 4.2939150401836967e-07,
|
| 322 |
-
"logits/chosen": -2.
|
| 323 |
-
"logits/rejected": -2.
|
| 324 |
-
"logps/chosen": -269.
|
| 325 |
-
"logps/rejected": -235.
|
| 326 |
-
"loss": 0.
|
| 327 |
-
"rewards/accuracies": 0.
|
| 328 |
-
"rewards/chosen": 0.
|
| 329 |
-
"rewards/margins": 0.
|
| 330 |
-
"rewards/rejected": 0.
|
| 331 |
"step": 220
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.24,
|
| 335 |
"learning_rate": 4.236509758897818e-07,
|
| 336 |
-
"logits/chosen": -2.
|
| 337 |
-
"logits/rejected": -2.
|
| 338 |
-
"logps/chosen": -271.
|
| 339 |
-
"logps/rejected": -242.
|
| 340 |
"loss": 0.6826,
|
| 341 |
-
"rewards/accuracies": 0.
|
| 342 |
-
"rewards/chosen": 0.
|
| 343 |
-
"rewards/margins": 0.
|
| 344 |
-
"rewards/rejected": 0.
|
| 345 |
"step": 230
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.25,
|
| 349 |
"learning_rate": 4.17910447761194e-07,
|
| 350 |
-
"logits/chosen": -2.
|
| 351 |
-
"logits/rejected": -2.
|
| 352 |
-
"logps/chosen": -309.
|
| 353 |
-
"logps/rejected": -221.
|
| 354 |
-
"loss": 0.
|
| 355 |
-
"rewards/accuracies": 0.
|
| 356 |
-
"rewards/chosen": 0.
|
| 357 |
-
"rewards/margins": 0.
|
| 358 |
-
"rewards/rejected": 0.
|
| 359 |
"step": 240
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.26,
|
| 363 |
"learning_rate": 4.121699196326062e-07,
|
| 364 |
-
"logits/chosen": -2.
|
| 365 |
-
"logits/rejected": -2.
|
| 366 |
-
"logps/chosen": -272.
|
| 367 |
-
"logps/rejected": -237.
|
| 368 |
-
"loss": 0.
|
| 369 |
-
"rewards/accuracies": 0.
|
| 370 |
-
"rewards/chosen": 0.
|
| 371 |
-
"rewards/margins": 0.
|
| 372 |
-
"rewards/rejected": 0.
|
| 373 |
"step": 250
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.27,
|
| 377 |
"learning_rate": 4.0642939150401836e-07,
|
| 378 |
-
"logits/chosen": -2.
|
| 379 |
-
"logits/rejected": -2.
|
| 380 |
-
"logps/chosen": -270.
|
| 381 |
-
"logps/rejected": -221.
|
| 382 |
-
"loss": 0.
|
| 383 |
-
"rewards/accuracies": 0.
|
| 384 |
-
"rewards/chosen": 0.
|
| 385 |
-
"rewards/margins": 0.
|
| 386 |
-
"rewards/rejected": 0.
|
| 387 |
"step": 260
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.28,
|
| 391 |
"learning_rate": 4.006888633754305e-07,
|
| 392 |
-
"logits/chosen": -2.
|
| 393 |
-
"logits/rejected": -2.
|
| 394 |
-
"logps/chosen": -284.
|
| 395 |
-
"logps/rejected": -232.
|
| 396 |
-
"loss": 0.
|
| 397 |
-
"rewards/accuracies": 0.
|
| 398 |
-
"rewards/chosen": 0.
|
| 399 |
-
"rewards/margins": 0.
|
| 400 |
-
"rewards/rejected": 0.
|
| 401 |
"step": 270
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.29,
|
| 405 |
"learning_rate": 3.949483352468427e-07,
|
| 406 |
-
"logits/chosen": -2.
|
| 407 |
-
"logits/rejected": -2.
|
| 408 |
-
"logps/chosen": -293.
|
| 409 |
-
"logps/rejected": -236.
|
| 410 |
-
"loss": 0.
|
| 411 |
-
"rewards/accuracies": 0.
|
| 412 |
-
"rewards/chosen": 0.
|
| 413 |
-
"rewards/margins": 0.
|
| 414 |
-
"rewards/rejected": 0.
|
| 415 |
"step": 280
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.3,
|
| 419 |
"learning_rate": 3.8920780711825487e-07,
|
| 420 |
-
"logits/chosen": -2.
|
| 421 |
-
"logits/rejected": -2.
|
| 422 |
-
"logps/chosen": -278.
|
| 423 |
-
"logps/rejected": -227.
|
| 424 |
-
"loss": 0.
|
| 425 |
-
"rewards/accuracies": 0.
|
| 426 |
-
"rewards/chosen": 0.
|
| 427 |
-
"rewards/margins": 0.
|
| 428 |
-
"rewards/rejected": 0.
|
| 429 |
"step": 290
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.31,
|
| 433 |
"learning_rate": 3.83467278989667e-07,
|
| 434 |
-
"logits/chosen": -2.
|
| 435 |
-
"logits/rejected": -2.
|
| 436 |
-
"logps/chosen": -254.
|
| 437 |
-
"logps/rejected": -221.
|
| 438 |
-
"loss": 0.
|
| 439 |
-
"rewards/accuracies": 0.
|
| 440 |
-
"rewards/chosen": 0.
|
| 441 |
-
"rewards/margins": 0.
|
| 442 |
-
"rewards/rejected": 0.
|
| 443 |
"step": 300
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.32,
|
| 447 |
"learning_rate": 3.777267508610792e-07,
|
| 448 |
-
"logits/chosen": -2.
|
| 449 |
-
"logits/rejected": -2.
|
| 450 |
-
"logps/chosen": -306.
|
| 451 |
-
"logps/rejected": -257.
|
| 452 |
-
"loss": 0.
|
| 453 |
-
"rewards/accuracies": 0.
|
| 454 |
-
"rewards/chosen": 0.
|
| 455 |
-
"rewards/margins": 0.
|
| 456 |
-
"rewards/rejected": 0.
|
| 457 |
"step": 310
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.33,
|
| 461 |
"learning_rate": 3.7198622273249137e-07,
|
| 462 |
-
"logits/chosen": -2.
|
| 463 |
-
"logits/rejected": -2.
|
| 464 |
-
"logps/chosen": -251.
|
| 465 |
-
"logps/rejected": -
|
| 466 |
-
"loss": 0.
|
| 467 |
-
"rewards/accuracies": 0.
|
| 468 |
-
"rewards/chosen": 0.
|
| 469 |
-
"rewards/margins": 0.
|
| 470 |
-
"rewards/rejected": 0.
|
| 471 |
"step": 320
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.34,
|
| 475 |
"learning_rate": 3.662456946039035e-07,
|
| 476 |
-
"logits/chosen": -2.
|
| 477 |
-
"logits/rejected": -2.
|
| 478 |
-
"logps/chosen": -312.
|
| 479 |
-
"logps/rejected": -239.
|
| 480 |
"loss": 0.6761,
|
| 481 |
-
"rewards/accuracies": 0.
|
| 482 |
-
"rewards/chosen": 0.
|
| 483 |
-
"rewards/margins": 0.
|
| 484 |
-
"rewards/rejected": 0.
|
| 485 |
"step": 330
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.35,
|
| 489 |
"learning_rate": 3.605051664753157e-07,
|
| 490 |
-
"logits/chosen": -2.
|
| 491 |
-
"logits/rejected": -2.
|
| 492 |
-
"logps/chosen": -244.
|
| 493 |
-
"logps/rejected": -238.
|
| 494 |
-
"loss": 0.
|
| 495 |
-
"rewards/accuracies": 0.
|
| 496 |
-
"rewards/chosen": 0.
|
| 497 |
-
"rewards/margins": 0.
|
| 498 |
-
"rewards/rejected": 0.
|
| 499 |
"step": 340
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.36,
|
| 503 |
"learning_rate": 3.547646383467279e-07,
|
| 504 |
-
"logits/chosen": -2.
|
| 505 |
-
"logits/rejected": -2.
|
| 506 |
-
"logps/chosen": -313.
|
| 507 |
-
"logps/rejected": -248.
|
| 508 |
-
"loss": 0.
|
| 509 |
-
"rewards/accuracies": 0.
|
| 510 |
-
"rewards/chosen": 0.
|
| 511 |
-
"rewards/margins": 0.
|
| 512 |
-
"rewards/rejected": 0.
|
| 513 |
"step": 350
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.37,
|
| 517 |
"learning_rate": 3.4902411021814007e-07,
|
| 518 |
-
"logits/chosen": -2.
|
| 519 |
-
"logits/rejected": -2.
|
| 520 |
-
"logps/chosen": -303.
|
| 521 |
-
"logps/rejected": -249.
|
| 522 |
-
"loss": 0.
|
| 523 |
-
"rewards/accuracies": 0.
|
| 524 |
-
"rewards/chosen": 0.
|
| 525 |
-
"rewards/margins": 0.
|
| 526 |
-
"rewards/rejected": 0.
|
| 527 |
"step": 360
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.38,
|
| 531 |
"learning_rate": 3.432835820895522e-07,
|
| 532 |
-
"logits/chosen": -2.
|
| 533 |
-
"logits/rejected": -2.
|
| 534 |
-
"logps/chosen": -314.
|
| 535 |
-
"logps/rejected": -270.
|
| 536 |
"loss": 0.6759,
|
| 537 |
-
"rewards/accuracies": 0.
|
| 538 |
-
"rewards/chosen": 0.
|
| 539 |
-
"rewards/margins": 0.
|
| 540 |
-
"rewards/rejected": 0.
|
| 541 |
"step": 370
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.39,
|
| 545 |
"learning_rate": 3.375430539609644e-07,
|
| 546 |
-
"logits/chosen": -2.
|
| 547 |
-
"logits/rejected": -2.
|
| 548 |
-
"logps/chosen": -291.
|
| 549 |
-
"logps/rejected": -239.
|
| 550 |
-
"loss": 0.
|
| 551 |
-
"rewards/accuracies": 0.
|
| 552 |
-
"rewards/chosen": 0.
|
| 553 |
-
"rewards/margins": 0.
|
| 554 |
-
"rewards/rejected": 0.
|
| 555 |
"step": 380
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.4,
|
| 559 |
"learning_rate": 3.3180252583237657e-07,
|
| 560 |
-
"logits/chosen": -2.
|
| 561 |
-
"logits/rejected": -2.
|
| 562 |
-
"logps/chosen": -278.
|
| 563 |
-
"logps/rejected": -237.
|
| 564 |
-
"loss": 0.
|
| 565 |
-
"rewards/accuracies": 0.
|
| 566 |
-
"rewards/chosen": 0.
|
| 567 |
-
"rewards/margins": 0.
|
| 568 |
-
"rewards/rejected": 0.
|
| 569 |
"step": 390
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.41,
|
| 573 |
"learning_rate": 3.260619977037887e-07,
|
| 574 |
-
"logits/chosen": -2.
|
| 575 |
-
"logits/rejected": -2.
|
| 576 |
-
"logps/chosen": -263.
|
| 577 |
-
"logps/rejected": -213.
|
| 578 |
-
"loss": 0.
|
| 579 |
-
"rewards/accuracies": 0.
|
| 580 |
-
"rewards/chosen": 0.
|
| 581 |
-
"rewards/margins": 0.
|
| 582 |
-
"rewards/rejected": 0.
|
| 583 |
"step": 400
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.42,
|
| 587 |
"learning_rate": 3.203214695752009e-07,
|
| 588 |
-
"logits/chosen": -2.
|
| 589 |
-
"logits/rejected": -2.
|
| 590 |
-
"logps/chosen": -268.
|
| 591 |
-
"logps/rejected": -252.
|
| 592 |
-
"loss": 0.
|
| 593 |
-
"rewards/accuracies": 0.
|
| 594 |
-
"rewards/chosen": 0.
|
| 595 |
-
"rewards/margins": 0.
|
| 596 |
-
"rewards/rejected": 0.
|
| 597 |
"step": 410
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.43,
|
| 601 |
"learning_rate": 3.145809414466131e-07,
|
| 602 |
-
"logits/chosen": -2.
|
| 603 |
-
"logits/rejected": -2.
|
| 604 |
-
"logps/chosen": -252.
|
| 605 |
-
"logps/rejected": -204.
|
| 606 |
-
"loss": 0.
|
| 607 |
-
"rewards/accuracies": 0.
|
| 608 |
-
"rewards/chosen": 0.
|
| 609 |
-
"rewards/margins": 0.
|
| 610 |
-
"rewards/rejected": 0.
|
| 611 |
"step": 420
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.44,
|
| 615 |
"learning_rate": 3.0884041331802526e-07,
|
| 616 |
-
"logits/chosen": -2.
|
| 617 |
-
"logits/rejected": -2.
|
| 618 |
-
"logps/chosen": -263.
|
| 619 |
-
"logps/rejected": -241.
|
| 620 |
-
"loss": 0.
|
| 621 |
-
"rewards/accuracies": 0.
|
| 622 |
-
"rewards/chosen": 0.
|
| 623 |
-
"rewards/margins": 0.
|
| 624 |
-
"rewards/rejected": 0.
|
| 625 |
"step": 430
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.45,
|
| 629 |
"learning_rate": 3.030998851894374e-07,
|
| 630 |
-
"logits/chosen": -2.
|
| 631 |
-
"logits/rejected": -2.
|
| 632 |
-
"logps/chosen": -286.
|
| 633 |
-
"logps/rejected": -247.
|
| 634 |
-
"loss": 0.
|
| 635 |
-
"rewards/accuracies": 0.
|
| 636 |
-
"rewards/chosen": 0.
|
| 637 |
-
"rewards/margins": 0.
|
| 638 |
-
"rewards/rejected": 0.
|
| 639 |
"step": 440
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.46,
|
| 643 |
"learning_rate": 2.973593570608496e-07,
|
| 644 |
-
"logits/chosen": -2.
|
| 645 |
-
"logits/rejected": -2.
|
| 646 |
-
"logps/chosen": -276.
|
| 647 |
-
"logps/rejected": -230.
|
| 648 |
-
"loss": 0.
|
| 649 |
-
"rewards/accuracies": 0.
|
| 650 |
-
"rewards/chosen": 0.
|
| 651 |
-
"rewards/margins": 0.
|
| 652 |
-
"rewards/rejected": 0.
|
| 653 |
"step": 450
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.47,
|
| 657 |
"learning_rate": 2.9161882893226177e-07,
|
| 658 |
-
"logits/chosen": -2.
|
| 659 |
-
"logits/rejected": -2.
|
| 660 |
-
"logps/chosen": -273.
|
| 661 |
-
"logps/rejected": -222.
|
| 662 |
-
"loss": 0.
|
| 663 |
-
"rewards/accuracies": 0.
|
| 664 |
-
"rewards/chosen": 0.
|
| 665 |
-
"rewards/margins": 0.
|
| 666 |
-
"rewards/rejected": 0.
|
| 667 |
"step": 460
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.49,
|
| 671 |
"learning_rate": 2.858783008036739e-07,
|
| 672 |
-
"logits/chosen": -2.
|
| 673 |
-
"logits/rejected": -2.
|
| 674 |
-
"logps/chosen": -
|
| 675 |
-
"logps/rejected": -215.
|
| 676 |
-
"loss": 0.
|
| 677 |
-
"rewards/accuracies": 0.
|
| 678 |
-
"rewards/chosen": 0.
|
| 679 |
-
"rewards/margins": 0.
|
| 680 |
-
"rewards/rejected": 0.
|
| 681 |
"step": 470
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.5,
|
| 685 |
"learning_rate": 2.801377726750861e-07,
|
| 686 |
-
"logits/chosen": -2.
|
| 687 |
-
"logits/rejected": -2.
|
| 688 |
-
"logps/chosen": -289.
|
| 689 |
-
"logps/rejected": -231.
|
| 690 |
-
"loss": 0.
|
| 691 |
-
"rewards/accuracies": 0.
|
| 692 |
-
"rewards/chosen": 0.
|
| 693 |
-
"rewards/margins": 0.
|
| 694 |
-
"rewards/rejected": 0.
|
| 695 |
"step": 480
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.51,
|
| 699 |
"learning_rate": 2.743972445464983e-07,
|
| 700 |
-
"logits/chosen": -2.
|
| 701 |
-
"logits/rejected": -2.
|
| 702 |
-
"logps/chosen": -293.
|
| 703 |
-
"logps/rejected": -243.
|
| 704 |
-
"loss": 0.
|
| 705 |
-
"rewards/accuracies": 0.
|
| 706 |
-
"rewards/chosen": 0.
|
| 707 |
-
"rewards/margins": 0.
|
| 708 |
-
"rewards/rejected": 0.
|
| 709 |
"step": 490
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.52,
|
| 713 |
"learning_rate": 2.686567164179104e-07,
|
| 714 |
-
"logits/chosen": -2.
|
| 715 |
-
"logits/rejected": -2.
|
| 716 |
-
"logps/chosen": -254.
|
| 717 |
-
"logps/rejected": -221.
|
| 718 |
-
"loss": 0.
|
| 719 |
-
"rewards/accuracies": 0.
|
| 720 |
-
"rewards/chosen": 0.
|
| 721 |
-
"rewards/margins": 0.
|
| 722 |
-
"rewards/rejected": 0.
|
| 723 |
"step": 500
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.53,
|
| 727 |
"learning_rate": 2.629161882893226e-07,
|
| 728 |
-
"logits/chosen": -2.
|
| 729 |
-
"logits/rejected": -2.
|
| 730 |
-
"logps/chosen": -310.
|
| 731 |
-
"logps/rejected": -256.
|
| 732 |
-
"loss": 0.
|
| 733 |
-
"rewards/accuracies": 0.
|
| 734 |
-
"rewards/chosen": 0.
|
| 735 |
-
"rewards/margins": 0.
|
| 736 |
-
"rewards/rejected": 0.
|
| 737 |
"step": 510
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.54,
|
| 741 |
"learning_rate": 2.571756601607348e-07,
|
| 742 |
-
"logits/chosen": -2.
|
| 743 |
-
"logits/rejected": -2.
|
| 744 |
-
"logps/chosen": -278.
|
| 745 |
-
"logps/rejected": -244.
|
| 746 |
-
"loss": 0.
|
| 747 |
"rewards/accuracies": 0.6499999761581421,
|
| 748 |
-
"rewards/chosen": 0.
|
| 749 |
-
"rewards/margins": 0.
|
| 750 |
-
"rewards/rejected": 0.
|
| 751 |
"step": 520
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.55,
|
| 755 |
"learning_rate": 2.5143513203214697e-07,
|
| 756 |
-
"logits/chosen": -2.
|
| 757 |
-
"logits/rejected": -2.
|
| 758 |
-
"logps/chosen": -242.
|
| 759 |
-
"logps/rejected": -224.
|
| 760 |
-
"loss": 0.
|
| 761 |
-
"rewards/accuracies": 0.
|
| 762 |
-
"rewards/chosen": 0.
|
| 763 |
-
"rewards/margins": 0.
|
| 764 |
-
"rewards/rejected": 0.
|
| 765 |
"step": 530
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.56,
|
| 769 |
"learning_rate": 2.456946039035591e-07,
|
| 770 |
-
"logits/chosen": -2.
|
| 771 |
-
"logits/rejected": -2.
|
| 772 |
-
"logps/chosen": -288.
|
| 773 |
-
"logps/rejected": -240.
|
| 774 |
-
"loss": 0.
|
| 775 |
-
"rewards/accuracies": 0.
|
| 776 |
-
"rewards/chosen": 0.
|
| 777 |
-
"rewards/margins": 0.
|
| 778 |
-
"rewards/rejected": 0.
|
| 779 |
"step": 540
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.57,
|
| 783 |
"learning_rate": 2.399540757749713e-07,
|
| 784 |
-
"logits/chosen": -2.
|
| 785 |
-
"logits/rejected": -2.
|
| 786 |
-
"logps/chosen": -265.
|
| 787 |
-
"logps/rejected": -230.
|
| 788 |
-
"loss": 0.
|
| 789 |
-
"rewards/accuracies": 0.
|
| 790 |
-
"rewards/chosen": 0.
|
| 791 |
-
"rewards/margins": 0.
|
| 792 |
-
"rewards/rejected": 0.
|
| 793 |
"step": 550
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.58,
|
| 797 |
"learning_rate": 2.3421354764638345e-07,
|
| 798 |
-
"logits/chosen": -2.
|
| 799 |
-
"logits/rejected": -2.
|
| 800 |
-
"logps/chosen": -302.
|
| 801 |
-
"logps/rejected": -
|
| 802 |
-
"loss": 0.
|
| 803 |
-
"rewards/accuracies": 0.
|
| 804 |
-
"rewards/chosen": 0.
|
| 805 |
-
"rewards/margins": 0.
|
| 806 |
-
"rewards/rejected": 0.
|
| 807 |
"step": 560
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.59,
|
| 811 |
"learning_rate": 2.2847301951779563e-07,
|
| 812 |
-
"logits/chosen": -2.
|
| 813 |
-
"logits/rejected": -2.
|
| 814 |
-
"logps/chosen": -268.
|
| 815 |
-
"logps/rejected": -223.
|
| 816 |
-
"loss": 0.
|
| 817 |
-
"rewards/accuracies": 0.
|
| 818 |
-
"rewards/chosen": 0.
|
| 819 |
-
"rewards/margins": 0.
|
| 820 |
-
"rewards/rejected": 0.
|
| 821 |
"step": 570
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.6,
|
| 825 |
"learning_rate": 2.227324913892078e-07,
|
| 826 |
-
"logits/chosen": -2.
|
| 827 |
-
"logits/rejected": -2.
|
| 828 |
-
"logps/chosen": -299.
|
| 829 |
-
"logps/rejected": -236.
|
| 830 |
-
"loss": 0.
|
| 831 |
-
"rewards/accuracies": 0.
|
| 832 |
-
"rewards/chosen": 0.
|
| 833 |
-
"rewards/margins": 0.
|
| 834 |
-
"rewards/rejected": 0.
|
| 835 |
"step": 580
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.61,
|
| 839 |
"learning_rate": 2.1699196326061998e-07,
|
| 840 |
-
"logits/chosen": -2.
|
| 841 |
-
"logits/rejected": -2.
|
| 842 |
-
"logps/chosen": -253.
|
| 843 |
-
"logps/rejected": -218.
|
| 844 |
-
"loss": 0.
|
| 845 |
-
"rewards/accuracies": 0.
|
| 846 |
-
"rewards/chosen": 0.
|
| 847 |
-
"rewards/margins": 0.
|
| 848 |
-
"rewards/rejected": 0.
|
| 849 |
"step": 590
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.62,
|
| 853 |
"learning_rate": 2.1125143513203214e-07,
|
| 854 |
-
"logits/chosen": -2.
|
| 855 |
-
"logits/rejected": -2.
|
| 856 |
-
"logps/chosen": -256.
|
| 857 |
-
"logps/rejected": -206.
|
| 858 |
-
"loss": 0.
|
| 859 |
-
"rewards/accuracies": 0.
|
| 860 |
-
"rewards/chosen": 0.
|
| 861 |
-
"rewards/margins": 0.
|
| 862 |
-
"rewards/rejected": 0.
|
| 863 |
"step": 600
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.63,
|
| 867 |
"learning_rate": 2.055109070034443e-07,
|
| 868 |
-
"logits/chosen": -2.
|
| 869 |
-
"logits/rejected": -2.
|
| 870 |
-
"logps/chosen": -266.
|
| 871 |
-
"logps/rejected": -223.
|
| 872 |
-
"loss": 0.
|
| 873 |
-
"rewards/accuracies": 0.
|
| 874 |
-
"rewards/chosen": 0.
|
| 875 |
-
"rewards/margins": 0.
|
| 876 |
-
"rewards/rejected": 0.
|
| 877 |
"step": 610
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.64,
|
| 881 |
"learning_rate": 1.997703788748565e-07,
|
| 882 |
-
"logits/chosen": -2.
|
| 883 |
-
"logits/rejected": -2.
|
| 884 |
-
"logps/chosen": -313.
|
| 885 |
-
"logps/rejected": -249.
|
| 886 |
-
"loss": 0.
|
| 887 |
-
"rewards/accuracies": 0.
|
| 888 |
-
"rewards/chosen": 0.
|
| 889 |
-
"rewards/margins": 0.
|
| 890 |
-
"rewards/rejected": 0.
|
| 891 |
"step": 620
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.65,
|
| 895 |
"learning_rate": 1.9402985074626865e-07,
|
| 896 |
-
"logits/chosen": -2.
|
| 897 |
-
"logits/rejected": -2.
|
| 898 |
-
"logps/chosen": -259.
|
| 899 |
-
"logps/rejected": -240.
|
| 900 |
-
"loss": 0.
|
| 901 |
-
"rewards/accuracies": 0.
|
| 902 |
-
"rewards/chosen": 0.
|
| 903 |
-
"rewards/margins": 0.
|
| 904 |
-
"rewards/rejected": 0.
|
| 905 |
"step": 630
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.66,
|
| 909 |
"learning_rate": 1.8828932261768083e-07,
|
| 910 |
-
"logits/chosen": -2.
|
| 911 |
-
"logits/rejected": -2.
|
| 912 |
-
"logps/chosen": -266.
|
| 913 |
-
"logps/rejected": -217.
|
| 914 |
-
"loss": 0.
|
| 915 |
-
"rewards/accuracies": 0.
|
| 916 |
-
"rewards/chosen": 0.
|
| 917 |
-
"rewards/margins": 0.
|
| 918 |
-
"rewards/rejected": 0.
|
| 919 |
"step": 640
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.67,
|
| 923 |
"learning_rate": 1.82548794489093e-07,
|
| 924 |
-
"logits/chosen": -2.
|
| 925 |
-
"logits/rejected": -2.
|
| 926 |
-
"logps/chosen": -284.
|
| 927 |
-
"logps/rejected": -232.
|
| 928 |
-
"loss": 0.
|
| 929 |
-
"rewards/accuracies": 0.
|
| 930 |
-
"rewards/chosen": 0.
|
| 931 |
-
"rewards/margins": 0.
|
| 932 |
-
"rewards/rejected": 0.
|
| 933 |
"step": 650
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.68,
|
| 937 |
"learning_rate": 1.7680826636050515e-07,
|
| 938 |
-
"logits/chosen": -2.
|
| 939 |
-
"logits/rejected": -2.
|
| 940 |
-
"logps/chosen": -279.
|
| 941 |
-
"logps/rejected": -233.
|
| 942 |
"loss": 0.6608,
|
| 943 |
-
"rewards/accuracies": 0.
|
| 944 |
-
"rewards/chosen": 0.
|
| 945 |
-
"rewards/margins": 0.
|
| 946 |
-
"rewards/rejected": 0.
|
| 947 |
"step": 660
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.69,
|
| 951 |
"learning_rate": 1.7106773823191734e-07,
|
| 952 |
-
"logits/chosen": -2.
|
| 953 |
-
"logits/rejected": -2.
|
| 954 |
-
"logps/chosen": -295.
|
| 955 |
-
"logps/rejected": -240.
|
| 956 |
-
"loss": 0.
|
| 957 |
-
"rewards/accuracies": 0.
|
| 958 |
-
"rewards/chosen": 0.
|
| 959 |
-
"rewards/margins": 0.
|
| 960 |
-
"rewards/rejected": 0.
|
| 961 |
"step": 670
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.7,
|
| 965 |
"learning_rate": 1.653272101033295e-07,
|
| 966 |
-
"logits/chosen": -2.
|
| 967 |
-
"logits/rejected": -2.
|
| 968 |
-
"logps/chosen": -289.
|
| 969 |
-
"logps/rejected": -230.
|
| 970 |
-
"loss": 0.
|
| 971 |
-
"rewards/accuracies": 0.
|
| 972 |
-
"rewards/chosen": 0.
|
| 973 |
-
"rewards/margins": 0.
|
| 974 |
-
"rewards/rejected": 0.
|
| 975 |
"step": 680
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.71,
|
| 979 |
"learning_rate": 1.5958668197474169e-07,
|
| 980 |
-
"logits/chosen": -2.
|
| 981 |
-
"logits/rejected": -2.
|
| 982 |
-
"logps/chosen": -268.
|
| 983 |
-
"logps/rejected": -229.
|
| 984 |
-
"loss": 0.
|
| 985 |
-
"rewards/accuracies": 0.
|
| 986 |
-
"rewards/chosen": 0.
|
| 987 |
-
"rewards/margins": 0.
|
| 988 |
-
"rewards/rejected": 0.
|
| 989 |
"step": 690
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.72,
|
| 993 |
"learning_rate": 1.5384615384615385e-07,
|
| 994 |
-
"logits/chosen": -2.
|
| 995 |
-
"logits/rejected": -2.
|
| 996 |
-
"logps/chosen": -282.
|
| 997 |
-
"logps/rejected": -222.
|
| 998 |
-
"loss": 0.
|
| 999 |
-
"rewards/accuracies": 0.
|
| 1000 |
-
"rewards/chosen": 0.
|
| 1001 |
-
"rewards/margins": 0.
|
| 1002 |
-
"rewards/rejected": 0.
|
| 1003 |
"step": 700
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.73,
|
| 1007 |
"learning_rate": 1.4810562571756603e-07,
|
| 1008 |
-
"logits/chosen": -2.
|
| 1009 |
-
"logits/rejected": -2.
|
| 1010 |
-
"logps/chosen": -272.
|
| 1011 |
-
"logps/rejected": -208.
|
| 1012 |
-
"loss": 0.
|
| 1013 |
-
"rewards/accuracies": 0.
|
| 1014 |
-
"rewards/chosen": 0.
|
| 1015 |
-
"rewards/margins": 0.
|
| 1016 |
-
"rewards/rejected": 0.
|
| 1017 |
"step": 710
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.74,
|
| 1021 |
"learning_rate": 1.423650975889782e-07,
|
| 1022 |
-
"logits/chosen": -2.
|
| 1023 |
-
"logits/rejected": -2.
|
| 1024 |
-
"logps/chosen": -303.
|
| 1025 |
-
"logps/rejected": -259.
|
| 1026 |
-
"loss": 0.
|
| 1027 |
-
"rewards/accuracies": 0.
|
| 1028 |
-
"rewards/chosen": 0.
|
| 1029 |
-
"rewards/margins": 0.
|
| 1030 |
-
"rewards/rejected": 0.
|
| 1031 |
"step": 720
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.75,
|
| 1035 |
"learning_rate": 1.3662456946039035e-07,
|
| 1036 |
-
"logits/chosen": -2.
|
| 1037 |
-
"logits/rejected": -2.
|
| 1038 |
-
"logps/chosen": -270.
|
| 1039 |
-
"logps/rejected": -252.
|
| 1040 |
-
"loss": 0.
|
| 1041 |
-
"rewards/accuracies": 0.
|
| 1042 |
-
"rewards/chosen": 0.
|
| 1043 |
-
"rewards/margins": 0.
|
| 1044 |
-
"rewards/rejected": 0.
|
| 1045 |
"step": 730
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.76,
|
| 1049 |
"learning_rate": 1.3088404133180254e-07,
|
| 1050 |
-
"logits/chosen": -2.
|
| 1051 |
-
"logits/rejected": -2.
|
| 1052 |
-
"logps/chosen": -276.
|
| 1053 |
-
"logps/rejected": -199.
|
| 1054 |
-
"loss": 0.
|
| 1055 |
-
"rewards/accuracies": 0.
|
| 1056 |
-
"rewards/chosen": 0.
|
| 1057 |
-
"rewards/margins": 0.
|
| 1058 |
-
"rewards/rejected": 0.
|
| 1059 |
"step": 740
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.77,
|
| 1063 |
"learning_rate": 1.251435132032147e-07,
|
| 1064 |
-
"logits/chosen": -2.
|
| 1065 |
-
"logits/rejected": -2.
|
| 1066 |
-
"logps/chosen": -269.
|
| 1067 |
-
"logps/rejected": -220.
|
| 1068 |
-
"loss": 0.
|
| 1069 |
-
"rewards/accuracies": 0.
|
| 1070 |
-
"rewards/chosen": 0.
|
| 1071 |
-
"rewards/margins": 0.
|
| 1072 |
-
"rewards/rejected": 0.
|
| 1073 |
"step": 750
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.78,
|
| 1077 |
"learning_rate": 1.1940298507462686e-07,
|
| 1078 |
-
"logits/chosen": -2.
|
| 1079 |
-
"logits/rejected": -2.
|
| 1080 |
-
"logps/chosen": -267.
|
| 1081 |
-
"logps/rejected": -249.
|
| 1082 |
-
"loss": 0.
|
| 1083 |
-
"rewards/accuracies": 0.
|
| 1084 |
-
"rewards/chosen": 0.
|
| 1085 |
-
"rewards/margins": 0.
|
| 1086 |
-
"rewards/rejected": 0.
|
| 1087 |
"step": 760
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.8,
|
| 1091 |
"learning_rate": 1.1366245694603903e-07,
|
| 1092 |
-
"logits/chosen": -2.
|
| 1093 |
-
"logits/rejected": -2.
|
| 1094 |
-
"logps/chosen": -273.
|
| 1095 |
-
"logps/rejected": -238.
|
| 1096 |
-
"loss": 0.
|
| 1097 |
-
"rewards/accuracies": 0.
|
| 1098 |
-
"rewards/chosen": 0.
|
| 1099 |
-
"rewards/margins": 0.
|
| 1100 |
-
"rewards/rejected": 0.
|
| 1101 |
"step": 770
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.81,
|
| 1105 |
"learning_rate": 1.079219288174512e-07,
|
| 1106 |
-
"logits/chosen": -2.
|
| 1107 |
-
"logits/rejected": -2.
|
| 1108 |
-
"logps/chosen": -290.
|
| 1109 |
-
"logps/rejected": -236.
|
| 1110 |
-
"loss": 0.
|
| 1111 |
-
"rewards/accuracies": 0.
|
| 1112 |
-
"rewards/chosen": 0.
|
| 1113 |
-
"rewards/margins": 0.
|
| 1114 |
-
"rewards/rejected": 0.
|
| 1115 |
"step": 780
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.82,
|
| 1119 |
"learning_rate": 1.0218140068886336e-07,
|
| 1120 |
-
"logits/chosen": -2.
|
| 1121 |
-
"logits/rejected": -2.
|
| 1122 |
-
"logps/chosen": -270.
|
| 1123 |
-
"logps/rejected": -221.
|
| 1124 |
-
"loss": 0.
|
| 1125 |
-
"rewards/accuracies": 0.
|
| 1126 |
-
"rewards/chosen": 0.
|
| 1127 |
-
"rewards/margins": 0.
|
| 1128 |
-
"rewards/rejected": 0.
|
| 1129 |
"step": 790
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.83,
|
| 1133 |
"learning_rate": 9.644087256027554e-08,
|
| 1134 |
-
"logits/chosen": -2.
|
| 1135 |
-
"logits/rejected": -2.
|
| 1136 |
-
"logps/chosen": -284.
|
| 1137 |
-
"logps/rejected": -243.
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"rewards/accuracies": 0.
|
| 1140 |
-
"rewards/chosen": 0.
|
| 1141 |
-
"rewards/margins": 0.
|
| 1142 |
-
"rewards/rejected": 0.
|
| 1143 |
"step": 800
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.84,
|
| 1147 |
"learning_rate": 9.070034443168771e-08,
|
| 1148 |
-
"logits/chosen": -2.
|
| 1149 |
-
"logits/rejected": -2.
|
| 1150 |
-
"logps/chosen": -269.
|
| 1151 |
-
"logps/rejected": -230.
|
| 1152 |
-
"loss": 0.
|
| 1153 |
-
"rewards/accuracies": 0.
|
| 1154 |
-
"rewards/chosen": 0.
|
| 1155 |
-
"rewards/margins": 0.
|
| 1156 |
-
"rewards/rejected": 0.
|
| 1157 |
"step": 810
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.85,
|
| 1161 |
"learning_rate": 8.495981630309988e-08,
|
| 1162 |
-
"logits/chosen": -2.
|
| 1163 |
-
"logits/rejected": -2.
|
| 1164 |
-
"logps/chosen": -302.
|
| 1165 |
-
"logps/rejected": -228.
|
| 1166 |
-
"loss": 0.
|
| 1167 |
-
"rewards/accuracies": 0.
|
| 1168 |
-
"rewards/chosen": 0.
|
| 1169 |
-
"rewards/margins": 0.
|
| 1170 |
-
"rewards/rejected": 0.
|
| 1171 |
"step": 820
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.86,
|
| 1175 |
"learning_rate": 7.921928817451206e-08,
|
| 1176 |
-
"logits/chosen": -2.
|
| 1177 |
-
"logits/rejected": -2.
|
| 1178 |
-
"logps/chosen": -287.
|
| 1179 |
-
"logps/rejected": -222.
|
| 1180 |
-
"loss": 0.
|
| 1181 |
"rewards/accuracies": 0.675000011920929,
|
| 1182 |
-
"rewards/chosen": 0.
|
| 1183 |
-
"rewards/margins": 0.
|
| 1184 |
-
"rewards/rejected": 0.
|
| 1185 |
"step": 830
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.87,
|
| 1189 |
"learning_rate": 7.347876004592423e-08,
|
| 1190 |
-
"logits/chosen": -2.
|
| 1191 |
-
"logits/rejected": -2.
|
| 1192 |
-
"logps/chosen": -258.
|
| 1193 |
-
"logps/rejected": -
|
| 1194 |
-
"loss": 0.
|
| 1195 |
-
"rewards/accuracies": 0.
|
| 1196 |
-
"rewards/chosen": 0.
|
| 1197 |
-
"rewards/margins": 0.
|
| 1198 |
-
"rewards/rejected": 0.
|
| 1199 |
"step": 840
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.88,
|
| 1203 |
"learning_rate": 6.773823191733639e-08,
|
| 1204 |
-
"logits/chosen": -2.
|
| 1205 |
-
"logits/rejected": -2.
|
| 1206 |
-
"logps/chosen": -262.
|
| 1207 |
-
"logps/rejected": -231.
|
| 1208 |
-
"loss": 0.
|
| 1209 |
-
"rewards/accuracies": 0.
|
| 1210 |
-
"rewards/chosen": 0.
|
| 1211 |
-
"rewards/margins": 0.
|
| 1212 |
-
"rewards/rejected": 0.
|
| 1213 |
"step": 850
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.89,
|
| 1217 |
"learning_rate": 6.199770378874856e-08,
|
| 1218 |
-
"logits/chosen": -2.
|
| 1219 |
-
"logits/rejected": -2.
|
| 1220 |
-
"logps/chosen": -295.
|
| 1221 |
-
"logps/rejected": -270.
|
| 1222 |
-
"loss": 0.
|
| 1223 |
-
"rewards/accuracies": 0.
|
| 1224 |
-
"rewards/chosen": 0.
|
| 1225 |
-
"rewards/margins": 0.
|
| 1226 |
-
"rewards/rejected": 0.
|
| 1227 |
"step": 860
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.9,
|
| 1231 |
"learning_rate": 5.6257175660160735e-08,
|
| 1232 |
-
"logits/chosen": -2.
|
| 1233 |
-
"logits/rejected": -2.
|
| 1234 |
-
"logps/chosen": -312.
|
| 1235 |
-
"logps/rejected": -237.
|
| 1236 |
-
"loss": 0.
|
| 1237 |
-
"rewards/accuracies": 0.
|
| 1238 |
-
"rewards/chosen": 0.
|
| 1239 |
-
"rewards/margins": 0.
|
| 1240 |
-
"rewards/rejected": 0.
|
| 1241 |
"step": 870
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.91,
|
| 1245 |
"learning_rate": 5.05166475315729e-08,
|
| 1246 |
-
"logits/chosen": -2.
|
| 1247 |
-
"logits/rejected": -2.
|
| 1248 |
-
"logps/chosen": -291.
|
| 1249 |
-
"logps/rejected": -240.
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"rewards/accuracies": 0.
|
| 1252 |
-
"rewards/chosen": 0.
|
| 1253 |
-
"rewards/margins": 0.
|
| 1254 |
-
"rewards/rejected": 0.
|
| 1255 |
"step": 880
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.92,
|
| 1259 |
"learning_rate": 4.477611940298507e-08,
|
| 1260 |
-
"logits/chosen": -2.
|
| 1261 |
-
"logits/rejected": -2.
|
| 1262 |
-
"logps/chosen": -285.
|
| 1263 |
-
"logps/rejected": -235.
|
| 1264 |
-
"loss": 0.
|
| 1265 |
-
"rewards/accuracies": 0.
|
| 1266 |
-
"rewards/chosen": 0.
|
| 1267 |
-
"rewards/margins": 0.
|
| 1268 |
-
"rewards/rejected": 0.
|
| 1269 |
"step": 890
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.93,
|
| 1273 |
"learning_rate": 3.903559127439724e-08,
|
| 1274 |
-
"logits/chosen": -2.
|
| 1275 |
-
"logits/rejected": -2.
|
| 1276 |
-
"logps/chosen": -272.
|
| 1277 |
-
"logps/rejected": -211.
|
| 1278 |
-
"loss": 0.
|
| 1279 |
-
"rewards/accuracies": 0.
|
| 1280 |
-
"rewards/chosen": 0.
|
| 1281 |
-
"rewards/margins": 0.
|
| 1282 |
-
"rewards/rejected": 0.
|
| 1283 |
"step": 900
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.94,
|
| 1287 |
"learning_rate": 3.3295063145809414e-08,
|
| 1288 |
-
"logits/chosen": -2.
|
| 1289 |
-
"logits/rejected": -2.
|
| 1290 |
-
"logps/chosen": -238.
|
| 1291 |
-
"logps/rejected": -206.
|
| 1292 |
-
"loss": 0.
|
| 1293 |
-
"rewards/accuracies": 0.
|
| 1294 |
-
"rewards/chosen": 0.
|
| 1295 |
-
"rewards/margins": 0.
|
| 1296 |
-
"rewards/rejected": 0.
|
| 1297 |
"step": 910
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.95,
|
| 1301 |
"learning_rate": 2.755453501722158e-08,
|
| 1302 |
-
"logits/chosen": -2.
|
| 1303 |
-
"logits/rejected": -2.
|
| 1304 |
-
"logps/chosen": -281.
|
| 1305 |
-
"logps/rejected": -225.
|
| 1306 |
-
"loss": 0.
|
| 1307 |
-
"rewards/accuracies": 0.
|
| 1308 |
-
"rewards/chosen": 0.
|
| 1309 |
-
"rewards/margins": 0.
|
| 1310 |
-
"rewards/rejected": 0.
|
| 1311 |
"step": 920
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.96,
|
| 1315 |
"learning_rate": 2.1814006888633754e-08,
|
| 1316 |
-
"logits/chosen": -2.
|
| 1317 |
-
"logits/rejected": -2.
|
| 1318 |
-
"logps/chosen": -256.
|
| 1319 |
-
"logps/rejected": -203.
|
| 1320 |
-
"loss": 0.
|
| 1321 |
-
"rewards/accuracies": 0.
|
| 1322 |
-
"rewards/chosen": 0.
|
| 1323 |
-
"rewards/margins": 0.
|
| 1324 |
-
"rewards/rejected": 0.
|
| 1325 |
"step": 930
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.97,
|
| 1329 |
"learning_rate": 1.6073478760045924e-08,
|
| 1330 |
-
"logits/chosen": -2.
|
| 1331 |
-
"logits/rejected": -2.
|
| 1332 |
-
"logps/chosen": -271.
|
| 1333 |
-
"logps/rejected": -231.
|
| 1334 |
-
"loss": 0.
|
| 1335 |
-
"rewards/accuracies": 0.
|
| 1336 |
-
"rewards/chosen": 0.
|
| 1337 |
-
"rewards/margins": 0.
|
| 1338 |
-
"rewards/rejected": 0.
|
| 1339 |
"step": 940
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 0.98,
|
| 1343 |
"learning_rate": 1.0332950631458094e-08,
|
| 1344 |
-
"logits/chosen": -2.
|
| 1345 |
-
"logits/rejected": -2.
|
| 1346 |
-
"logps/chosen": -282.
|
| 1347 |
-
"logps/rejected": -233.
|
| 1348 |
-
"loss": 0.
|
| 1349 |
"rewards/accuracies": 0.6812499761581421,
|
| 1350 |
-
"rewards/chosen": 0.
|
| 1351 |
-
"rewards/margins": 0.
|
| 1352 |
-
"rewards/rejected": 0.
|
| 1353 |
"step": 950
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.99,
|
| 1357 |
"learning_rate": 4.592422502870264e-09,
|
| 1358 |
-
"logits/chosen": -2.
|
| 1359 |
-
"logits/rejected": -2.
|
| 1360 |
-
"logps/chosen": -281.
|
| 1361 |
-
"logps/rejected": -
|
| 1362 |
-
"loss": 0.
|
| 1363 |
-
"rewards/accuracies": 0.
|
| 1364 |
-
"rewards/chosen": 0.
|
| 1365 |
-
"rewards/margins": 0.
|
| 1366 |
-
"rewards/rejected": 0.
|
| 1367 |
"step": 960
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
"epoch": 1.0,
|
| 1371 |
-
"eval_logits/chosen": -2.
|
| 1372 |
-
"eval_logits/rejected": -2.
|
| 1373 |
-
"eval_logps/chosen": -278.
|
| 1374 |
-
"eval_logps/rejected": -230.
|
| 1375 |
-
"eval_loss": 0.
|
| 1376 |
-
"eval_rewards/accuracies": 0.
|
| 1377 |
-
"eval_rewards/chosen": 0.
|
| 1378 |
-
"eval_rewards/margins": 0.
|
| 1379 |
-
"eval_rewards/rejected": 0.
|
| 1380 |
-
"eval_runtime":
|
| 1381 |
-
"eval_samples_per_second": 4.
|
| 1382 |
-
"eval_steps_per_second": 0.
|
| 1383 |
"step": 968
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 1.0,
|
| 1387 |
"step": 968,
|
| 1388 |
"total_flos": 0.0,
|
| 1389 |
-
"train_loss": 0.
|
| 1390 |
-
"train_runtime":
|
| 1391 |
-
"train_samples_per_second": 2.
|
| 1392 |
"train_steps_per_second": 0.035
|
| 1393 |
}
|
| 1394 |
],
|
|
|
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
"learning_rate": 5.154639175257731e-08,
|
| 28 |
+
"logits/chosen": -2.223740339279175,
|
| 29 |
+
"logits/rejected": -2.180643081665039,
|
| 30 |
+
"logps/chosen": -284.7340087890625,
|
| 31 |
+
"logps/rejected": -205.98194885253906,
|
| 32 |
+
"loss": 0.694,
|
| 33 |
+
"rewards/accuracies": 0.4305555522441864,
|
| 34 |
+
"rewards/chosen": -0.0006893649115227163,
|
| 35 |
+
"rewards/margins": 0.0007374237175099552,
|
| 36 |
+
"rewards/rejected": -0.0014267880469560623,
|
| 37 |
"step": 10
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02,
|
| 41 |
"learning_rate": 1.0309278350515462e-07,
|
| 42 |
+
"logits/chosen": -2.33476185798645,
|
| 43 |
+
"logits/rejected": -2.2125375270843506,
|
| 44 |
+
"logps/chosen": -320.8204040527344,
|
| 45 |
+
"logps/rejected": -248.4267120361328,
|
| 46 |
+
"loss": 0.692,
|
| 47 |
"rewards/accuracies": 0.512499988079071,
|
| 48 |
+
"rewards/chosen": 0.0003039050498045981,
|
| 49 |
+
"rewards/margins": 0.0023796656168997288,
|
| 50 |
+
"rewards/rejected": -0.0020757606253027916,
|
| 51 |
"step": 20
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03,
|
| 55 |
"learning_rate": 1.5463917525773197e-07,
|
| 56 |
+
"logits/chosen": -2.339370012283325,
|
| 57 |
+
"logits/rejected": -2.304020404815674,
|
| 58 |
+
"logps/chosen": -268.95074462890625,
|
| 59 |
+
"logps/rejected": -227.067626953125,
|
| 60 |
+
"loss": 0.6921,
|
| 61 |
+
"rewards/accuracies": 0.46875,
|
| 62 |
+
"rewards/chosen": 0.0005883350968360901,
|
| 63 |
+
"rewards/margins": 0.002594549907371402,
|
| 64 |
+
"rewards/rejected": -0.0020062148105353117,
|
| 65 |
"step": 30
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.04,
|
| 69 |
"learning_rate": 2.0618556701030925e-07,
|
| 70 |
+
"logits/chosen": -2.3392791748046875,
|
| 71 |
+
"logits/rejected": -2.3300938606262207,
|
| 72 |
+
"logps/chosen": -308.5113220214844,
|
| 73 |
+
"logps/rejected": -253.8385467529297,
|
| 74 |
+
"loss": 0.6945,
|
| 75 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 76 |
+
"rewards/chosen": 0.0024464379530400038,
|
| 77 |
+
"rewards/margins": -0.00025889737298712134,
|
| 78 |
+
"rewards/rejected": 0.0027053358498960733,
|
| 79 |
"step": 40
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05,
|
| 83 |
"learning_rate": 2.5773195876288655e-07,
|
| 84 |
+
"logits/chosen": -2.251412868499756,
|
| 85 |
+
"logits/rejected": -2.2359275817871094,
|
| 86 |
+
"logps/chosen": -297.78375244140625,
|
| 87 |
+
"logps/rejected": -227.23556518554688,
|
| 88 |
+
"loss": 0.6922,
|
| 89 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 90 |
+
"rewards/chosen": 0.0033915191888809204,
|
| 91 |
+
"rewards/margins": 0.0055986023508012295,
|
| 92 |
+
"rewards/rejected": -0.0022070836275815964,
|
| 93 |
"step": 50
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06,
|
| 97 |
"learning_rate": 3.0927835051546394e-07,
|
| 98 |
+
"logits/chosen": -2.167163848876953,
|
| 99 |
+
"logits/rejected": -2.3376193046569824,
|
| 100 |
+
"logps/chosen": -256.54510498046875,
|
| 101 |
+
"logps/rejected": -229.5459747314453,
|
| 102 |
+
"loss": 0.6917,
|
| 103 |
+
"rewards/accuracies": 0.4937500059604645,
|
| 104 |
+
"rewards/chosen": 0.000388039683457464,
|
| 105 |
+
"rewards/margins": 0.007883811369538307,
|
| 106 |
+
"rewards/rejected": -0.0074957734905183315,
|
| 107 |
"step": 60
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07,
|
| 111 |
"learning_rate": 3.608247422680412e-07,
|
| 112 |
+
"logits/chosen": -2.3430614471435547,
|
| 113 |
+
"logits/rejected": -2.281782627105713,
|
| 114 |
+
"logps/chosen": -313.92608642578125,
|
| 115 |
+
"logps/rejected": -252.57284545898438,
|
| 116 |
+
"loss": 0.6924,
|
| 117 |
+
"rewards/accuracies": 0.46875,
|
| 118 |
+
"rewards/chosen": 0.0012417413527145982,
|
| 119 |
+
"rewards/margins": 0.0001173208438558504,
|
| 120 |
+
"rewards/rejected": 0.0011244199704378843,
|
| 121 |
"step": 70
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08,
|
| 125 |
"learning_rate": 4.123711340206185e-07,
|
| 126 |
+
"logits/chosen": -2.337070941925049,
|
| 127 |
+
"logits/rejected": -2.3018112182617188,
|
| 128 |
+
"logps/chosen": -302.9524841308594,
|
| 129 |
+
"logps/rejected": -243.9047088623047,
|
| 130 |
+
"loss": 0.6916,
|
| 131 |
+
"rewards/accuracies": 0.518750011920929,
|
| 132 |
+
"rewards/chosen": 0.0021400884725153446,
|
| 133 |
+
"rewards/margins": -0.0002812549355439842,
|
| 134 |
+
"rewards/rejected": 0.002421343233436346,
|
| 135 |
"step": 80
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09,
|
| 139 |
"learning_rate": 4.639175257731959e-07,
|
| 140 |
+
"logits/chosen": -2.259251356124878,
|
| 141 |
+
"logits/rejected": -2.2963995933532715,
|
| 142 |
+
"logps/chosen": -270.1668395996094,
|
| 143 |
+
"logps/rejected": -216.64822387695312,
|
| 144 |
+
"loss": 0.6913,
|
| 145 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 146 |
+
"rewards/chosen": 0.009941437281668186,
|
| 147 |
+
"rewards/margins": 0.010241752490401268,
|
| 148 |
+
"rewards/rejected": -0.00030031436472199857,
|
| 149 |
"step": 90
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.1,
|
| 153 |
"learning_rate": 4.982778415614236e-07,
|
| 154 |
+
"logits/chosen": -2.1677582263946533,
|
| 155 |
+
"logits/rejected": -2.2741990089416504,
|
| 156 |
+
"logps/chosen": -274.75836181640625,
|
| 157 |
+
"logps/rejected": -226.3966064453125,
|
| 158 |
+
"loss": 0.6901,
|
| 159 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 160 |
+
"rewards/chosen": 0.006115993484854698,
|
| 161 |
+
"rewards/margins": 0.0013887921813875437,
|
| 162 |
+
"rewards/rejected": 0.0047272020019590855,
|
| 163 |
"step": 100
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.11,
|
| 167 |
"learning_rate": 4.925373134328357e-07,
|
| 168 |
+
"logits/chosen": -2.271916389465332,
|
| 169 |
+
"logits/rejected": -2.197857141494751,
|
| 170 |
+
"logps/chosen": -274.72113037109375,
|
| 171 |
+
"logps/rejected": -232.5464324951172,
|
| 172 |
+
"loss": 0.6886,
|
| 173 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 174 |
+
"rewards/chosen": 0.005831545684486628,
|
| 175 |
+
"rewards/margins": 0.0067709460854530334,
|
| 176 |
+
"rewards/rejected": -0.000939400284551084,
|
| 177 |
"step": 110
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.12,
|
| 181 |
"learning_rate": 4.867967853042479e-07,
|
| 182 |
+
"logits/chosen": -2.2548232078552246,
|
| 183 |
+
"logits/rejected": -2.322075366973877,
|
| 184 |
+
"logps/chosen": -319.34521484375,
|
| 185 |
+
"logps/rejected": -235.76535034179688,
|
| 186 |
+
"loss": 0.689,
|
| 187 |
+
"rewards/accuracies": 0.574999988079071,
|
| 188 |
+
"rewards/chosen": 0.013832703232765198,
|
| 189 |
+
"rewards/margins": 0.01176449190825224,
|
| 190 |
+
"rewards/rejected": 0.002068211790174246,
|
| 191 |
"step": 120
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.13,
|
| 195 |
"learning_rate": 4.810562571756601e-07,
|
| 196 |
+
"logits/chosen": -2.32174015045166,
|
| 197 |
+
"logits/rejected": -2.3775150775909424,
|
| 198 |
+
"logps/chosen": -296.20733642578125,
|
| 199 |
+
"logps/rejected": -245.56655883789062,
|
| 200 |
+
"loss": 0.6875,
|
| 201 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 202 |
+
"rewards/chosen": 0.017552796751260757,
|
| 203 |
+
"rewards/margins": 0.013545483350753784,
|
| 204 |
+
"rewards/rejected": 0.004007314797490835,
|
| 205 |
"step": 130
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.14,
|
| 209 |
"learning_rate": 4.753157290470723e-07,
|
| 210 |
+
"logits/chosen": -2.3627283573150635,
|
| 211 |
+
"logits/rejected": -2.310948133468628,
|
| 212 |
+
"logps/chosen": -301.9321594238281,
|
| 213 |
+
"logps/rejected": -239.2898406982422,
|
| 214 |
+
"loss": 0.688,
|
| 215 |
+
"rewards/accuracies": 0.46875,
|
| 216 |
+
"rewards/chosen": 0.011156091466546059,
|
| 217 |
+
"rewards/margins": 0.009668431244790554,
|
| 218 |
+
"rewards/rejected": 0.0014876595232635736,
|
| 219 |
"step": 140
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.15,
|
| 223 |
"learning_rate": 4.6957520091848447e-07,
|
| 224 |
+
"logits/chosen": -2.2531113624572754,
|
| 225 |
+
"logits/rejected": -2.348215341567993,
|
| 226 |
+
"logps/chosen": -284.4292907714844,
|
| 227 |
+
"logps/rejected": -259.6882019042969,
|
| 228 |
+
"loss": 0.6858,
|
| 229 |
+
"rewards/accuracies": 0.5625,
|
| 230 |
+
"rewards/chosen": 0.017186133190989494,
|
| 231 |
+
"rewards/margins": 0.011862866580486298,
|
| 232 |
+
"rewards/rejected": 0.005323265679180622,
|
| 233 |
"step": 150
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.17,
|
| 237 |
"learning_rate": 4.6383467278989666e-07,
|
| 238 |
+
"logits/chosen": -2.361238956451416,
|
| 239 |
+
"logits/rejected": -2.4430744647979736,
|
| 240 |
+
"logps/chosen": -286.7644348144531,
|
| 241 |
+
"logps/rejected": -221.6837158203125,
|
| 242 |
+
"loss": 0.6857,
|
| 243 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 244 |
+
"rewards/chosen": 0.01911218836903572,
|
| 245 |
+
"rewards/margins": 0.014816234819591045,
|
| 246 |
+
"rewards/rejected": 0.00429595448076725,
|
| 247 |
"step": 160
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.18,
|
| 251 |
"learning_rate": 4.580941446613088e-07,
|
| 252 |
+
"logits/chosen": -2.32244610786438,
|
| 253 |
+
"logits/rejected": -2.3339757919311523,
|
| 254 |
+
"logps/chosen": -301.54693603515625,
|
| 255 |
+
"logps/rejected": -239.26095581054688,
|
| 256 |
+
"loss": 0.6839,
|
| 257 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 258 |
+
"rewards/chosen": 0.023171866312623024,
|
| 259 |
+
"rewards/margins": 0.014685508795082569,
|
| 260 |
+
"rewards/rejected": 0.00848635844886303,
|
| 261 |
"step": 170
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.19,
|
| 265 |
"learning_rate": 4.52353616532721e-07,
|
| 266 |
+
"logits/chosen": -2.347285032272339,
|
| 267 |
+
"logits/rejected": -2.3244121074676514,
|
| 268 |
+
"logps/chosen": -257.841552734375,
|
| 269 |
+
"logps/rejected": -214.5565643310547,
|
| 270 |
+
"loss": 0.6864,
|
| 271 |
+
"rewards/accuracies": 0.5625,
|
| 272 |
+
"rewards/chosen": 0.019994111731648445,
|
| 273 |
+
"rewards/margins": 0.01520625315606594,
|
| 274 |
+
"rewards/rejected": 0.004787858575582504,
|
| 275 |
"step": 180
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2,
|
| 279 |
"learning_rate": 4.4661308840413316e-07,
|
| 280 |
+
"logits/chosen": -2.2657313346862793,
|
| 281 |
+
"logits/rejected": -2.201254367828369,
|
| 282 |
+
"logps/chosen": -253.98916625976562,
|
| 283 |
+
"logps/rejected": -206.3340301513672,
|
| 284 |
+
"loss": 0.6833,
|
| 285 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 286 |
+
"rewards/chosen": 0.0245305635035038,
|
| 287 |
+
"rewards/margins": 0.017677443102002144,
|
| 288 |
+
"rewards/rejected": 0.006853120867162943,
|
| 289 |
"step": 190
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.21,
|
| 293 |
"learning_rate": 4.408725602755453e-07,
|
| 294 |
+
"logits/chosen": -2.284461498260498,
|
| 295 |
+
"logits/rejected": -2.2873706817626953,
|
| 296 |
+
"logps/chosen": -261.44427490234375,
|
| 297 |
+
"logps/rejected": -195.59422302246094,
|
| 298 |
+
"loss": 0.6835,
|
| 299 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 300 |
+
"rewards/chosen": 0.03187788277864456,
|
| 301 |
+
"rewards/margins": 0.024095263332128525,
|
| 302 |
+
"rewards/rejected": 0.007782619446516037,
|
| 303 |
"step": 200
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.22,
|
| 307 |
"learning_rate": 4.351320321469575e-07,
|
| 308 |
+
"logits/chosen": -2.18426513671875,
|
| 309 |
+
"logits/rejected": -2.1963071823120117,
|
| 310 |
+
"logps/chosen": -302.31195068359375,
|
| 311 |
+
"logps/rejected": -218.6005401611328,
|
| 312 |
+
"loss": 0.6815,
|
| 313 |
+
"rewards/accuracies": 0.612500011920929,
|
| 314 |
+
"rewards/chosen": 0.03718667849898338,
|
| 315 |
+
"rewards/margins": 0.026892077177762985,
|
| 316 |
+
"rewards/rejected": 0.010294605046510696,
|
| 317 |
"step": 210
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.23,
|
| 321 |
"learning_rate": 4.2939150401836967e-07,
|
| 322 |
+
"logits/chosen": -2.2150394916534424,
|
| 323 |
+
"logits/rejected": -2.2160990238189697,
|
| 324 |
+
"logps/chosen": -269.44769287109375,
|
| 325 |
+
"logps/rejected": -235.6748504638672,
|
| 326 |
+
"loss": 0.6801,
|
| 327 |
+
"rewards/accuracies": 0.59375,
|
| 328 |
+
"rewards/chosen": 0.038056183606386185,
|
| 329 |
+
"rewards/margins": 0.023441683501005173,
|
| 330 |
+
"rewards/rejected": 0.014614498242735863,
|
| 331 |
"step": 220
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.24,
|
| 335 |
"learning_rate": 4.236509758897818e-07,
|
| 336 |
+
"logits/chosen": -2.2152469158172607,
|
| 337 |
+
"logits/rejected": -2.1862380504608154,
|
| 338 |
+
"logps/chosen": -271.4049377441406,
|
| 339 |
+
"logps/rejected": -242.6397247314453,
|
| 340 |
"loss": 0.6826,
|
| 341 |
+
"rewards/accuracies": 0.637499988079071,
|
| 342 |
+
"rewards/chosen": 0.03268683701753616,
|
| 343 |
+
"rewards/margins": 0.026912549510598183,
|
| 344 |
+
"rewards/rejected": 0.0057742842473089695,
|
| 345 |
"step": 230
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.25,
|
| 349 |
"learning_rate": 4.17910447761194e-07,
|
| 350 |
+
"logits/chosen": -2.3059380054473877,
|
| 351 |
+
"logits/rejected": -2.2681984901428223,
|
| 352 |
+
"logps/chosen": -309.55499267578125,
|
| 353 |
+
"logps/rejected": -221.61703491210938,
|
| 354 |
+
"loss": 0.6827,
|
| 355 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 356 |
+
"rewards/chosen": 0.03509462997317314,
|
| 357 |
+
"rewards/margins": 0.012767216190695763,
|
| 358 |
+
"rewards/rejected": 0.02232741378247738,
|
| 359 |
"step": 240
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.26,
|
| 363 |
"learning_rate": 4.121699196326062e-07,
|
| 364 |
+
"logits/chosen": -2.307035446166992,
|
| 365 |
+
"logits/rejected": -2.2920923233032227,
|
| 366 |
+
"logps/chosen": -272.9412841796875,
|
| 367 |
+
"logps/rejected": -237.314208984375,
|
| 368 |
+
"loss": 0.6824,
|
| 369 |
+
"rewards/accuracies": 0.550000011920929,
|
| 370 |
+
"rewards/chosen": 0.03535359352827072,
|
| 371 |
+
"rewards/margins": 0.012216273695230484,
|
| 372 |
+
"rewards/rejected": 0.023137323558330536,
|
| 373 |
"step": 250
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.27,
|
| 377 |
"learning_rate": 4.0642939150401836e-07,
|
| 378 |
+
"logits/chosen": -2.3456673622131348,
|
| 379 |
+
"logits/rejected": -2.3194832801818848,
|
| 380 |
+
"logps/chosen": -270.475341796875,
|
| 381 |
+
"logps/rejected": -221.84536743164062,
|
| 382 |
+
"loss": 0.6805,
|
| 383 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 384 |
+
"rewards/chosen": 0.04569912329316139,
|
| 385 |
+
"rewards/margins": 0.029975151643157005,
|
| 386 |
+
"rewards/rejected": 0.015723969787359238,
|
| 387 |
"step": 260
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.28,
|
| 391 |
"learning_rate": 4.006888633754305e-07,
|
| 392 |
+
"logits/chosen": -2.385854721069336,
|
| 393 |
+
"logits/rejected": -2.3556528091430664,
|
| 394 |
+
"logps/chosen": -284.36029052734375,
|
| 395 |
+
"logps/rejected": -232.5426788330078,
|
| 396 |
+
"loss": 0.6793,
|
| 397 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 398 |
+
"rewards/chosen": 0.04998317360877991,
|
| 399 |
+
"rewards/margins": 0.032010577619075775,
|
| 400 |
+
"rewards/rejected": 0.017972594127058983,
|
| 401 |
"step": 270
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.29,
|
| 405 |
"learning_rate": 3.949483352468427e-07,
|
| 406 |
+
"logits/chosen": -2.308225154876709,
|
| 407 |
+
"logits/rejected": -2.259629726409912,
|
| 408 |
+
"logps/chosen": -293.1715087890625,
|
| 409 |
+
"logps/rejected": -236.4293975830078,
|
| 410 |
+
"loss": 0.6771,
|
| 411 |
+
"rewards/accuracies": 0.581250011920929,
|
| 412 |
+
"rewards/chosen": 0.057786036282777786,
|
| 413 |
+
"rewards/margins": 0.04149205610156059,
|
| 414 |
+
"rewards/rejected": 0.016293983906507492,
|
| 415 |
"step": 280
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.3,
|
| 419 |
"learning_rate": 3.8920780711825487e-07,
|
| 420 |
+
"logits/chosen": -2.278501033782959,
|
| 421 |
+
"logits/rejected": -2.369293689727783,
|
| 422 |
+
"logps/chosen": -278.4786376953125,
|
| 423 |
+
"logps/rejected": -227.40927124023438,
|
| 424 |
+
"loss": 0.6792,
|
| 425 |
+
"rewards/accuracies": 0.5625,
|
| 426 |
+
"rewards/chosen": 0.0485750176012516,
|
| 427 |
+
"rewards/margins": 0.02242155373096466,
|
| 428 |
+
"rewards/rejected": 0.02615346387028694,
|
| 429 |
"step": 290
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.31,
|
| 433 |
"learning_rate": 3.83467278989667e-07,
|
| 434 |
+
"logits/chosen": -2.2661235332489014,
|
| 435 |
+
"logits/rejected": -2.205644130706787,
|
| 436 |
+
"logps/chosen": -254.183837890625,
|
| 437 |
+
"logps/rejected": -221.9667510986328,
|
| 438 |
+
"loss": 0.6772,
|
| 439 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 440 |
+
"rewards/chosen": 0.05874975398182869,
|
| 441 |
+
"rewards/margins": 0.03965791314840317,
|
| 442 |
+
"rewards/rejected": 0.019091838970780373,
|
| 443 |
"step": 300
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.32,
|
| 447 |
"learning_rate": 3.777267508610792e-07,
|
| 448 |
+
"logits/chosen": -2.32353138923645,
|
| 449 |
+
"logits/rejected": -2.3743112087249756,
|
| 450 |
+
"logps/chosen": -306.22711181640625,
|
| 451 |
+
"logps/rejected": -257.60980224609375,
|
| 452 |
+
"loss": 0.6783,
|
| 453 |
+
"rewards/accuracies": 0.59375,
|
| 454 |
+
"rewards/chosen": 0.04823786020278931,
|
| 455 |
+
"rewards/margins": 0.017192820087075233,
|
| 456 |
+
"rewards/rejected": 0.03104504384100437,
|
| 457 |
"step": 310
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.33,
|
| 461 |
"learning_rate": 3.7198622273249137e-07,
|
| 462 |
+
"logits/chosen": -2.234679698944092,
|
| 463 |
+
"logits/rejected": -2.211430788040161,
|
| 464 |
+
"logps/chosen": -251.83053588867188,
|
| 465 |
+
"logps/rejected": -193.01544189453125,
|
| 466 |
+
"loss": 0.6739,
|
| 467 |
+
"rewards/accuracies": 0.637499988079071,
|
| 468 |
+
"rewards/chosen": 0.06583289802074432,
|
| 469 |
+
"rewards/margins": 0.047706056386232376,
|
| 470 |
+
"rewards/rejected": 0.018126841634511948,
|
| 471 |
"step": 320
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.34,
|
| 475 |
"learning_rate": 3.662456946039035e-07,
|
| 476 |
+
"logits/chosen": -2.259127140045166,
|
| 477 |
+
"logits/rejected": -2.287956714630127,
|
| 478 |
+
"logps/chosen": -312.1918029785156,
|
| 479 |
+
"logps/rejected": -239.03530883789062,
|
| 480 |
"loss": 0.6761,
|
| 481 |
+
"rewards/accuracies": 0.65625,
|
| 482 |
+
"rewards/chosen": 0.07068151980638504,
|
| 483 |
+
"rewards/margins": 0.051512353122234344,
|
| 484 |
+
"rewards/rejected": 0.0191691592335701,
|
| 485 |
"step": 330
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.35,
|
| 489 |
"learning_rate": 3.605051664753157e-07,
|
| 490 |
+
"logits/chosen": -2.197277784347534,
|
| 491 |
+
"logits/rejected": -2.13037109375,
|
| 492 |
+
"logps/chosen": -244.2609100341797,
|
| 493 |
+
"logps/rejected": -238.80953979492188,
|
| 494 |
+
"loss": 0.6788,
|
| 495 |
+
"rewards/accuracies": 0.581250011920929,
|
| 496 |
+
"rewards/chosen": 0.05732797831296921,
|
| 497 |
+
"rewards/margins": 0.030042264610528946,
|
| 498 |
+
"rewards/rejected": 0.027285713702440262,
|
| 499 |
"step": 340
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.36,
|
| 503 |
"learning_rate": 3.547646383467279e-07,
|
| 504 |
+
"logits/chosen": -2.365830421447754,
|
| 505 |
+
"logits/rejected": -2.3728528022766113,
|
| 506 |
+
"logps/chosen": -313.7022705078125,
|
| 507 |
+
"logps/rejected": -248.090087890625,
|
| 508 |
+
"loss": 0.6746,
|
| 509 |
+
"rewards/accuracies": 0.643750011920929,
|
| 510 |
+
"rewards/chosen": 0.08016298711299896,
|
| 511 |
+
"rewards/margins": 0.05509548634290695,
|
| 512 |
+
"rewards/rejected": 0.025067497044801712,
|
| 513 |
"step": 350
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.37,
|
| 517 |
"learning_rate": 3.4902411021814007e-07,
|
| 518 |
+
"logits/chosen": -2.22756290435791,
|
| 519 |
+
"logits/rejected": -2.259359121322632,
|
| 520 |
+
"logps/chosen": -303.25250244140625,
|
| 521 |
+
"logps/rejected": -249.8985595703125,
|
| 522 |
+
"loss": 0.6723,
|
| 523 |
+
"rewards/accuracies": 0.643750011920929,
|
| 524 |
+
"rewards/chosen": 0.06414168328046799,
|
| 525 |
+
"rewards/margins": 0.04363773763179779,
|
| 526 |
+
"rewards/rejected": 0.020503941923379898,
|
| 527 |
"step": 360
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.38,
|
| 531 |
"learning_rate": 3.432835820895522e-07,
|
| 532 |
+
"logits/chosen": -2.3700273036956787,
|
| 533 |
+
"logits/rejected": -2.3231639862060547,
|
| 534 |
+
"logps/chosen": -314.5257263183594,
|
| 535 |
+
"logps/rejected": -270.7105712890625,
|
| 536 |
"loss": 0.6759,
|
| 537 |
+
"rewards/accuracies": 0.612500011920929,
|
| 538 |
+
"rewards/chosen": 0.07061124593019485,
|
| 539 |
+
"rewards/margins": 0.03391130641102791,
|
| 540 |
+
"rewards/rejected": 0.03669993579387665,
|
| 541 |
"step": 370
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.39,
|
| 545 |
"learning_rate": 3.375430539609644e-07,
|
| 546 |
+
"logits/chosen": -2.3212878704071045,
|
| 547 |
+
"logits/rejected": -2.249602794647217,
|
| 548 |
+
"logps/chosen": -291.92474365234375,
|
| 549 |
+
"logps/rejected": -239.6724395751953,
|
| 550 |
+
"loss": 0.677,
|
| 551 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 552 |
+
"rewards/chosen": 0.07373902946710587,
|
| 553 |
+
"rewards/margins": 0.03367278352379799,
|
| 554 |
+
"rewards/rejected": 0.04006624594330788,
|
| 555 |
"step": 380
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.4,
|
| 559 |
"learning_rate": 3.3180252583237657e-07,
|
| 560 |
+
"logits/chosen": -2.297023057937622,
|
| 561 |
+
"logits/rejected": -2.264172077178955,
|
| 562 |
+
"logps/chosen": -278.0927734375,
|
| 563 |
+
"logps/rejected": -237.13436889648438,
|
| 564 |
+
"loss": 0.6722,
|
| 565 |
+
"rewards/accuracies": 0.668749988079071,
|
| 566 |
+
"rewards/chosen": 0.0686158686876297,
|
| 567 |
+
"rewards/margins": 0.051144860684871674,
|
| 568 |
+
"rewards/rejected": 0.01747100241482258,
|
| 569 |
"step": 390
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.41,
|
| 573 |
"learning_rate": 3.260619977037887e-07,
|
| 574 |
+
"logits/chosen": -2.237035036087036,
|
| 575 |
+
"logits/rejected": -2.2392399311065674,
|
| 576 |
+
"logps/chosen": -263.4399108886719,
|
| 577 |
+
"logps/rejected": -213.87451171875,
|
| 578 |
+
"loss": 0.6707,
|
| 579 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 580 |
+
"rewards/chosen": 0.06768475472927094,
|
| 581 |
+
"rewards/margins": 0.048441771417856216,
|
| 582 |
+
"rewards/rejected": 0.019242987036705017,
|
| 583 |
"step": 400
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.42,
|
| 587 |
"learning_rate": 3.203214695752009e-07,
|
| 588 |
+
"logits/chosen": -2.2776081562042236,
|
| 589 |
+
"logits/rejected": -2.2924447059631348,
|
| 590 |
+
"logps/chosen": -268.8953857421875,
|
| 591 |
+
"logps/rejected": -252.852294921875,
|
| 592 |
+
"loss": 0.6673,
|
| 593 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 594 |
+
"rewards/chosen": 0.08111406862735748,
|
| 595 |
+
"rewards/margins": 0.05318716913461685,
|
| 596 |
+
"rewards/rejected": 0.027926897630095482,
|
| 597 |
"step": 410
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.43,
|
| 601 |
"learning_rate": 3.145809414466131e-07,
|
| 602 |
+
"logits/chosen": -2.3054046630859375,
|
| 603 |
+
"logits/rejected": -2.2502362728118896,
|
| 604 |
+
"logps/chosen": -252.5205841064453,
|
| 605 |
+
"logps/rejected": -204.43344116210938,
|
| 606 |
+
"loss": 0.6749,
|
| 607 |
+
"rewards/accuracies": 0.606249988079071,
|
| 608 |
+
"rewards/chosen": 0.07272285223007202,
|
| 609 |
+
"rewards/margins": 0.04809904843568802,
|
| 610 |
+
"rewards/rejected": 0.024623800069093704,
|
| 611 |
"step": 420
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.44,
|
| 615 |
"learning_rate": 3.0884041331802526e-07,
|
| 616 |
+
"logits/chosen": -2.3482632637023926,
|
| 617 |
+
"logits/rejected": -2.3258707523345947,
|
| 618 |
+
"logps/chosen": -263.67095947265625,
|
| 619 |
+
"logps/rejected": -241.14047241210938,
|
| 620 |
+
"loss": 0.6741,
|
| 621 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 622 |
+
"rewards/chosen": 0.07022975385189056,
|
| 623 |
+
"rewards/margins": 0.04051927849650383,
|
| 624 |
+
"rewards/rejected": 0.029710477218031883,
|
| 625 |
"step": 430
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.45,
|
| 629 |
"learning_rate": 3.030998851894374e-07,
|
| 630 |
+
"logits/chosen": -2.286533832550049,
|
| 631 |
+
"logits/rejected": -2.320568084716797,
|
| 632 |
+
"logps/chosen": -286.72894287109375,
|
| 633 |
+
"logps/rejected": -247.65542602539062,
|
| 634 |
+
"loss": 0.6705,
|
| 635 |
+
"rewards/accuracies": 0.668749988079071,
|
| 636 |
+
"rewards/chosen": 0.07666246592998505,
|
| 637 |
+
"rewards/margins": 0.05972421169281006,
|
| 638 |
+
"rewards/rejected": 0.01693824864923954,
|
| 639 |
"step": 440
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.46,
|
| 643 |
"learning_rate": 2.973593570608496e-07,
|
| 644 |
+
"logits/chosen": -2.206477642059326,
|
| 645 |
+
"logits/rejected": -2.315464496612549,
|
| 646 |
+
"logps/chosen": -276.1682434082031,
|
| 647 |
+
"logps/rejected": -230.3959197998047,
|
| 648 |
+
"loss": 0.678,
|
| 649 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 650 |
+
"rewards/chosen": 0.0718303695321083,
|
| 651 |
+
"rewards/margins": 0.04074189439415932,
|
| 652 |
+
"rewards/rejected": 0.03108847141265869,
|
| 653 |
"step": 450
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.47,
|
| 657 |
"learning_rate": 2.9161882893226177e-07,
|
| 658 |
+
"logits/chosen": -2.277815103530884,
|
| 659 |
+
"logits/rejected": -2.342268705368042,
|
| 660 |
+
"logps/chosen": -273.23773193359375,
|
| 661 |
+
"logps/rejected": -222.5966796875,
|
| 662 |
+
"loss": 0.6662,
|
| 663 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 664 |
+
"rewards/chosen": 0.0802597850561142,
|
| 665 |
+
"rewards/margins": 0.050464123487472534,
|
| 666 |
+
"rewards/rejected": 0.029795657843351364,
|
| 667 |
"step": 460
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.49,
|
| 671 |
"learning_rate": 2.858783008036739e-07,
|
| 672 |
+
"logits/chosen": -2.2656216621398926,
|
| 673 |
+
"logits/rejected": -2.2778594493865967,
|
| 674 |
+
"logps/chosen": -248.9929656982422,
|
| 675 |
+
"logps/rejected": -215.5894012451172,
|
| 676 |
+
"loss": 0.6669,
|
| 677 |
+
"rewards/accuracies": 0.637499988079071,
|
| 678 |
+
"rewards/chosen": 0.08564073592424393,
|
| 679 |
+
"rewards/margins": 0.06490761041641235,
|
| 680 |
+
"rewards/rejected": 0.020733121782541275,
|
| 681 |
"step": 470
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.5,
|
| 685 |
"learning_rate": 2.801377726750861e-07,
|
| 686 |
+
"logits/chosen": -2.2962255477905273,
|
| 687 |
+
"logits/rejected": -2.27239727973938,
|
| 688 |
+
"logps/chosen": -289.5277404785156,
|
| 689 |
+
"logps/rejected": -231.601318359375,
|
| 690 |
+
"loss": 0.6713,
|
| 691 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 692 |
+
"rewards/chosen": 0.08144901692867279,
|
| 693 |
+
"rewards/margins": 0.05658548325300217,
|
| 694 |
+
"rewards/rejected": 0.024863524362444878,
|
| 695 |
"step": 480
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.51,
|
| 699 |
"learning_rate": 2.743972445464983e-07,
|
| 700 |
+
"logits/chosen": -2.445746660232544,
|
| 701 |
+
"logits/rejected": -2.267007827758789,
|
| 702 |
+
"logps/chosen": -293.1885986328125,
|
| 703 |
+
"logps/rejected": -243.8875274658203,
|
| 704 |
+
"loss": 0.6676,
|
| 705 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 706 |
+
"rewards/chosen": 0.10828351974487305,
|
| 707 |
+
"rewards/margins": 0.08175922185182571,
|
| 708 |
+
"rewards/rejected": 0.02652430161833763,
|
| 709 |
"step": 490
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.52,
|
| 713 |
"learning_rate": 2.686567164179104e-07,
|
| 714 |
+
"logits/chosen": -2.278276205062866,
|
| 715 |
+
"logits/rejected": -2.295633316040039,
|
| 716 |
+
"logps/chosen": -254.94760131835938,
|
| 717 |
+
"logps/rejected": -221.79452514648438,
|
| 718 |
+
"loss": 0.6672,
|
| 719 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 720 |
+
"rewards/chosen": 0.08227074891328812,
|
| 721 |
+
"rewards/margins": 0.055896710604429245,
|
| 722 |
+
"rewards/rejected": 0.026374032720923424,
|
| 723 |
"step": 500
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.53,
|
| 727 |
"learning_rate": 2.629161882893226e-07,
|
| 728 |
+
"logits/chosen": -2.202611207962036,
|
| 729 |
+
"logits/rejected": -2.2495861053466797,
|
| 730 |
+
"logps/chosen": -310.4443664550781,
|
| 731 |
+
"logps/rejected": -256.72406005859375,
|
| 732 |
+
"loss": 0.6666,
|
| 733 |
+
"rewards/accuracies": 0.606249988079071,
|
| 734 |
+
"rewards/chosen": 0.07021793723106384,
|
| 735 |
+
"rewards/margins": 0.040728576481342316,
|
| 736 |
+
"rewards/rejected": 0.02948935702443123,
|
| 737 |
"step": 510
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.54,
|
| 741 |
"learning_rate": 2.571756601607348e-07,
|
| 742 |
+
"logits/chosen": -2.3376307487487793,
|
| 743 |
+
"logits/rejected": -2.352074146270752,
|
| 744 |
+
"logps/chosen": -278.10504150390625,
|
| 745 |
+
"logps/rejected": -244.0722198486328,
|
| 746 |
+
"loss": 0.6697,
|
| 747 |
"rewards/accuracies": 0.6499999761581421,
|
| 748 |
+
"rewards/chosen": 0.0925985723733902,
|
| 749 |
+
"rewards/margins": 0.0637633204460144,
|
| 750 |
+
"rewards/rejected": 0.028835251927375793,
|
| 751 |
"step": 520
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.55,
|
| 755 |
"learning_rate": 2.5143513203214697e-07,
|
| 756 |
+
"logits/chosen": -2.243332624435425,
|
| 757 |
+
"logits/rejected": -2.2513413429260254,
|
| 758 |
+
"logps/chosen": -242.59439086914062,
|
| 759 |
+
"logps/rejected": -224.13259887695312,
|
| 760 |
+
"loss": 0.6716,
|
| 761 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 762 |
+
"rewards/chosen": 0.07866770029067993,
|
| 763 |
+
"rewards/margins": 0.057711243629455566,
|
| 764 |
+
"rewards/rejected": 0.020956454798579216,
|
| 765 |
"step": 530
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.56,
|
| 769 |
"learning_rate": 2.456946039035591e-07,
|
| 770 |
+
"logits/chosen": -2.300567150115967,
|
| 771 |
+
"logits/rejected": -2.271827220916748,
|
| 772 |
+
"logps/chosen": -288.2174377441406,
|
| 773 |
+
"logps/rejected": -240.34439086914062,
|
| 774 |
+
"loss": 0.6682,
|
| 775 |
+
"rewards/accuracies": 0.65625,
|
| 776 |
+
"rewards/chosen": 0.10411250591278076,
|
| 777 |
+
"rewards/margins": 0.05851038545370102,
|
| 778 |
+
"rewards/rejected": 0.04560210928320885,
|
| 779 |
"step": 540
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.57,
|
| 783 |
"learning_rate": 2.399540757749713e-07,
|
| 784 |
+
"logits/chosen": -2.3359756469726562,
|
| 785 |
+
"logits/rejected": -2.194058895111084,
|
| 786 |
+
"logps/chosen": -265.052001953125,
|
| 787 |
+
"logps/rejected": -230.23605346679688,
|
| 788 |
+
"loss": 0.6686,
|
| 789 |
+
"rewards/accuracies": 0.65625,
|
| 790 |
+
"rewards/chosen": 0.0775262787938118,
|
| 791 |
+
"rewards/margins": 0.05575944110751152,
|
| 792 |
+
"rewards/rejected": 0.021766824647784233,
|
| 793 |
"step": 550
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.58,
|
| 797 |
"learning_rate": 2.3421354764638345e-07,
|
| 798 |
+
"logits/chosen": -2.3195242881774902,
|
| 799 |
+
"logits/rejected": -2.283975124359131,
|
| 800 |
+
"logps/chosen": -302.0104064941406,
|
| 801 |
+
"logps/rejected": -252.0124053955078,
|
| 802 |
+
"loss": 0.6708,
|
| 803 |
+
"rewards/accuracies": 0.612500011920929,
|
| 804 |
+
"rewards/chosen": 0.10010389983654022,
|
| 805 |
+
"rewards/margins": 0.053703296929597855,
|
| 806 |
+
"rewards/rejected": 0.04640059918165207,
|
| 807 |
"step": 560
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.59,
|
| 811 |
"learning_rate": 2.2847301951779563e-07,
|
| 812 |
+
"logits/chosen": -2.2481091022491455,
|
| 813 |
+
"logits/rejected": -2.400871515274048,
|
| 814 |
+
"logps/chosen": -268.6519775390625,
|
| 815 |
+
"logps/rejected": -223.69882202148438,
|
| 816 |
+
"loss": 0.6654,
|
| 817 |
+
"rewards/accuracies": 0.65625,
|
| 818 |
+
"rewards/chosen": 0.0826568529009819,
|
| 819 |
+
"rewards/margins": 0.05431235954165459,
|
| 820 |
+
"rewards/rejected": 0.028344491496682167,
|
| 821 |
"step": 570
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.6,
|
| 825 |
"learning_rate": 2.227324913892078e-07,
|
| 826 |
+
"logits/chosen": -2.299408197402954,
|
| 827 |
+
"logits/rejected": -2.22338604927063,
|
| 828 |
+
"logps/chosen": -299.3912353515625,
|
| 829 |
+
"logps/rejected": -236.9815216064453,
|
| 830 |
+
"loss": 0.661,
|
| 831 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 832 |
+
"rewards/chosen": 0.10458721220493317,
|
| 833 |
+
"rewards/margins": 0.08465038239955902,
|
| 834 |
+
"rewards/rejected": 0.019936833530664444,
|
| 835 |
"step": 580
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.61,
|
| 839 |
"learning_rate": 2.1699196326061998e-07,
|
| 840 |
+
"logits/chosen": -2.2584633827209473,
|
| 841 |
+
"logits/rejected": -2.2311649322509766,
|
| 842 |
+
"logps/chosen": -253.76913452148438,
|
| 843 |
+
"logps/rejected": -218.6166534423828,
|
| 844 |
+
"loss": 0.6687,
|
| 845 |
+
"rewards/accuracies": 0.643750011920929,
|
| 846 |
+
"rewards/chosen": 0.07234074175357819,
|
| 847 |
+
"rewards/margins": 0.04758009687066078,
|
| 848 |
+
"rewards/rejected": 0.024760644882917404,
|
| 849 |
"step": 590
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.62,
|
| 853 |
"learning_rate": 2.1125143513203214e-07,
|
| 854 |
+
"logits/chosen": -2.318943738937378,
|
| 855 |
+
"logits/rejected": -2.2511682510375977,
|
| 856 |
+
"logps/chosen": -256.5652770996094,
|
| 857 |
+
"logps/rejected": -206.35586547851562,
|
| 858 |
+
"loss": 0.669,
|
| 859 |
+
"rewards/accuracies": 0.643750011920929,
|
| 860 |
+
"rewards/chosen": 0.07542125880718231,
|
| 861 |
+
"rewards/margins": 0.0553053617477417,
|
| 862 |
+
"rewards/rejected": 0.020115893334150314,
|
| 863 |
"step": 600
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.63,
|
| 867 |
"learning_rate": 2.055109070034443e-07,
|
| 868 |
+
"logits/chosen": -2.3058714866638184,
|
| 869 |
+
"logits/rejected": -2.304198741912842,
|
| 870 |
+
"logps/chosen": -266.4674987792969,
|
| 871 |
+
"logps/rejected": -223.82711791992188,
|
| 872 |
+
"loss": 0.6677,
|
| 873 |
+
"rewards/accuracies": 0.699999988079071,
|
| 874 |
+
"rewards/chosen": 0.09824246913194656,
|
| 875 |
+
"rewards/margins": 0.06738617271184921,
|
| 876 |
+
"rewards/rejected": 0.03085630014538765,
|
| 877 |
"step": 610
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.64,
|
| 881 |
"learning_rate": 1.997703788748565e-07,
|
| 882 |
+
"logits/chosen": -2.337787389755249,
|
| 883 |
+
"logits/rejected": -2.2819180488586426,
|
| 884 |
+
"logps/chosen": -313.7826232910156,
|
| 885 |
+
"logps/rejected": -249.5704803466797,
|
| 886 |
+
"loss": 0.6582,
|
| 887 |
+
"rewards/accuracies": 0.668749988079071,
|
| 888 |
+
"rewards/chosen": 0.10966908931732178,
|
| 889 |
+
"rewards/margins": 0.08016980439424515,
|
| 890 |
+
"rewards/rejected": 0.029499292373657227,
|
| 891 |
"step": 620
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.65,
|
| 895 |
"learning_rate": 1.9402985074626865e-07,
|
| 896 |
+
"logits/chosen": -2.2067112922668457,
|
| 897 |
+
"logits/rejected": -2.246953010559082,
|
| 898 |
+
"logps/chosen": -259.2144775390625,
|
| 899 |
+
"logps/rejected": -240.3810272216797,
|
| 900 |
+
"loss": 0.6653,
|
| 901 |
+
"rewards/accuracies": 0.6875,
|
| 902 |
+
"rewards/chosen": 0.09941162168979645,
|
| 903 |
+
"rewards/margins": 0.06417630612850189,
|
| 904 |
+
"rewards/rejected": 0.035235337913036346,
|
| 905 |
"step": 630
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.66,
|
| 909 |
"learning_rate": 1.8828932261768083e-07,
|
| 910 |
+
"logits/chosen": -2.2894420623779297,
|
| 911 |
+
"logits/rejected": -2.2385382652282715,
|
| 912 |
+
"logps/chosen": -266.48992919921875,
|
| 913 |
+
"logps/rejected": -217.8952178955078,
|
| 914 |
+
"loss": 0.661,
|
| 915 |
+
"rewards/accuracies": 0.65625,
|
| 916 |
+
"rewards/chosen": 0.095299132168293,
|
| 917 |
+
"rewards/margins": 0.07987986505031586,
|
| 918 |
+
"rewards/rejected": 0.01541926246136427,
|
| 919 |
"step": 640
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.67,
|
| 923 |
"learning_rate": 1.82548794489093e-07,
|
| 924 |
+
"logits/chosen": -2.33485746383667,
|
| 925 |
+
"logits/rejected": -2.3108019828796387,
|
| 926 |
+
"logps/chosen": -284.7020568847656,
|
| 927 |
+
"logps/rejected": -232.82080078125,
|
| 928 |
+
"loss": 0.664,
|
| 929 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 930 |
+
"rewards/chosen": 0.10341651737689972,
|
| 931 |
+
"rewards/margins": 0.07464977353811264,
|
| 932 |
+
"rewards/rejected": 0.028766745701432228,
|
| 933 |
"step": 650
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.68,
|
| 937 |
"learning_rate": 1.7680826636050515e-07,
|
| 938 |
+
"logits/chosen": -2.3347816467285156,
|
| 939 |
+
"logits/rejected": -2.2758853435516357,
|
| 940 |
+
"logps/chosen": -279.80059814453125,
|
| 941 |
+
"logps/rejected": -233.2425994873047,
|
| 942 |
"loss": 0.6608,
|
| 943 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 944 |
+
"rewards/chosen": 0.11068934202194214,
|
| 945 |
+
"rewards/margins": 0.07695071399211884,
|
| 946 |
+
"rewards/rejected": 0.0337386280298233,
|
| 947 |
"step": 660
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.69,
|
| 951 |
"learning_rate": 1.7106773823191734e-07,
|
| 952 |
+
"logits/chosen": -2.2854952812194824,
|
| 953 |
+
"logits/rejected": -2.273536205291748,
|
| 954 |
+
"logps/chosen": -295.6964416503906,
|
| 955 |
+
"logps/rejected": -240.4071502685547,
|
| 956 |
+
"loss": 0.6615,
|
| 957 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 958 |
+
"rewards/chosen": 0.1013779416680336,
|
| 959 |
+
"rewards/margins": 0.060683172196149826,
|
| 960 |
+
"rewards/rejected": 0.04069476202130318,
|
| 961 |
"step": 670
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.7,
|
| 965 |
"learning_rate": 1.653272101033295e-07,
|
| 966 |
+
"logits/chosen": -2.34243106842041,
|
| 967 |
+
"logits/rejected": -2.2720611095428467,
|
| 968 |
+
"logps/chosen": -289.71722412109375,
|
| 969 |
+
"logps/rejected": -230.321533203125,
|
| 970 |
+
"loss": 0.6729,
|
| 971 |
+
"rewards/accuracies": 0.606249988079071,
|
| 972 |
+
"rewards/chosen": 0.09767869859933853,
|
| 973 |
+
"rewards/margins": 0.039280109107494354,
|
| 974 |
+
"rewards/rejected": 0.05839858204126358,
|
| 975 |
"step": 680
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.71,
|
| 979 |
"learning_rate": 1.5958668197474169e-07,
|
| 980 |
+
"logits/chosen": -2.371598482131958,
|
| 981 |
+
"logits/rejected": -2.362656354904175,
|
| 982 |
+
"logps/chosen": -268.17828369140625,
|
| 983 |
+
"logps/rejected": -229.41232299804688,
|
| 984 |
+
"loss": 0.6659,
|
| 985 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 986 |
+
"rewards/chosen": 0.0969640463590622,
|
| 987 |
+
"rewards/margins": 0.06369610875844955,
|
| 988 |
+
"rewards/rejected": 0.033267926424741745,
|
| 989 |
"step": 690
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.72,
|
| 993 |
"learning_rate": 1.5384615384615385e-07,
|
| 994 |
+
"logits/chosen": -2.2588796615600586,
|
| 995 |
+
"logits/rejected": -2.2576823234558105,
|
| 996 |
+
"logps/chosen": -282.4342041015625,
|
| 997 |
+
"logps/rejected": -222.56381225585938,
|
| 998 |
+
"loss": 0.664,
|
| 999 |
+
"rewards/accuracies": 0.65625,
|
| 1000 |
+
"rewards/chosen": 0.10399500280618668,
|
| 1001 |
+
"rewards/margins": 0.08138440549373627,
|
| 1002 |
+
"rewards/rejected": 0.0226106159389019,
|
| 1003 |
"step": 700
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.73,
|
| 1007 |
"learning_rate": 1.4810562571756603e-07,
|
| 1008 |
+
"logits/chosen": -2.3341283798217773,
|
| 1009 |
+
"logits/rejected": -2.2046780586242676,
|
| 1010 |
+
"logps/chosen": -272.2647399902344,
|
| 1011 |
+
"logps/rejected": -208.01364135742188,
|
| 1012 |
+
"loss": 0.666,
|
| 1013 |
+
"rewards/accuracies": 0.699999988079071,
|
| 1014 |
+
"rewards/chosen": 0.10669133812189102,
|
| 1015 |
+
"rewards/margins": 0.08235933631658554,
|
| 1016 |
+
"rewards/rejected": 0.02433200553059578,
|
| 1017 |
"step": 710
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.74,
|
| 1021 |
"learning_rate": 1.423650975889782e-07,
|
| 1022 |
+
"logits/chosen": -2.323979139328003,
|
| 1023 |
+
"logits/rejected": -2.340238094329834,
|
| 1024 |
+
"logps/chosen": -303.2074279785156,
|
| 1025 |
+
"logps/rejected": -259.44268798828125,
|
| 1026 |
+
"loss": 0.6667,
|
| 1027 |
+
"rewards/accuracies": 0.625,
|
| 1028 |
+
"rewards/chosen": 0.11533965170383453,
|
| 1029 |
+
"rewards/margins": 0.047552816569805145,
|
| 1030 |
+
"rewards/rejected": 0.06778682768344879,
|
| 1031 |
"step": 720
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.75,
|
| 1035 |
"learning_rate": 1.3662456946039035e-07,
|
| 1036 |
+
"logits/chosen": -2.3031513690948486,
|
| 1037 |
+
"logits/rejected": -2.28584623336792,
|
| 1038 |
+
"logps/chosen": -270.1670837402344,
|
| 1039 |
+
"logps/rejected": -252.5519256591797,
|
| 1040 |
+
"loss": 0.6642,
|
| 1041 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1042 |
+
"rewards/chosen": 0.10461707413196564,
|
| 1043 |
+
"rewards/margins": 0.058367032557725906,
|
| 1044 |
+
"rewards/rejected": 0.04625004902482033,
|
| 1045 |
"step": 730
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.76,
|
| 1049 |
"learning_rate": 1.3088404133180254e-07,
|
| 1050 |
+
"logits/chosen": -2.2157022953033447,
|
| 1051 |
+
"logits/rejected": -2.2670745849609375,
|
| 1052 |
+
"logps/chosen": -276.71240234375,
|
| 1053 |
+
"logps/rejected": -199.2496795654297,
|
| 1054 |
+
"loss": 0.6635,
|
| 1055 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1056 |
+
"rewards/chosen": 0.11176248639822006,
|
| 1057 |
+
"rewards/margins": 0.08353973925113678,
|
| 1058 |
+
"rewards/rejected": 0.02822275087237358,
|
| 1059 |
"step": 740
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.77,
|
| 1063 |
"learning_rate": 1.251435132032147e-07,
|
| 1064 |
+
"logits/chosen": -2.2043914794921875,
|
| 1065 |
+
"logits/rejected": -2.221619129180908,
|
| 1066 |
+
"logps/chosen": -269.0702819824219,
|
| 1067 |
+
"logps/rejected": -220.8921356201172,
|
| 1068 |
+
"loss": 0.665,
|
| 1069 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1070 |
+
"rewards/chosen": 0.09922349452972412,
|
| 1071 |
+
"rewards/margins": 0.04318443313241005,
|
| 1072 |
+
"rewards/rejected": 0.05603905767202377,
|
| 1073 |
"step": 750
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.78,
|
| 1077 |
"learning_rate": 1.1940298507462686e-07,
|
| 1078 |
+
"logits/chosen": -2.232959270477295,
|
| 1079 |
+
"logits/rejected": -2.2529525756835938,
|
| 1080 |
+
"logps/chosen": -267.9338684082031,
|
| 1081 |
+
"logps/rejected": -249.4876251220703,
|
| 1082 |
+
"loss": 0.6684,
|
| 1083 |
+
"rewards/accuracies": 0.625,
|
| 1084 |
+
"rewards/chosen": 0.08004304021596909,
|
| 1085 |
+
"rewards/margins": 0.04949140548706055,
|
| 1086 |
+
"rewards/rejected": 0.030551627278327942,
|
| 1087 |
"step": 760
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.8,
|
| 1091 |
"learning_rate": 1.1366245694603903e-07,
|
| 1092 |
+
"logits/chosen": -2.293257236480713,
|
| 1093 |
+
"logits/rejected": -2.2078585624694824,
|
| 1094 |
+
"logps/chosen": -273.19671630859375,
|
| 1095 |
+
"logps/rejected": -238.57858276367188,
|
| 1096 |
+
"loss": 0.661,
|
| 1097 |
+
"rewards/accuracies": 0.625,
|
| 1098 |
+
"rewards/chosen": 0.11353409290313721,
|
| 1099 |
+
"rewards/margins": 0.06645722687244415,
|
| 1100 |
+
"rewards/rejected": 0.04707685858011246,
|
| 1101 |
"step": 770
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.81,
|
| 1105 |
"learning_rate": 1.079219288174512e-07,
|
| 1106 |
+
"logits/chosen": -2.3507869243621826,
|
| 1107 |
+
"logits/rejected": -2.325718879699707,
|
| 1108 |
+
"logps/chosen": -290.9693298339844,
|
| 1109 |
+
"logps/rejected": -236.1486358642578,
|
| 1110 |
+
"loss": 0.6633,
|
| 1111 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1112 |
+
"rewards/chosen": 0.0980958342552185,
|
| 1113 |
+
"rewards/margins": 0.07181811332702637,
|
| 1114 |
+
"rewards/rejected": 0.026277724653482437,
|
| 1115 |
"step": 780
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.82,
|
| 1119 |
"learning_rate": 1.0218140068886336e-07,
|
| 1120 |
+
"logits/chosen": -2.268038272857666,
|
| 1121 |
+
"logits/rejected": -2.286581516265869,
|
| 1122 |
+
"logps/chosen": -270.3387451171875,
|
| 1123 |
+
"logps/rejected": -221.06356811523438,
|
| 1124 |
+
"loss": 0.6564,
|
| 1125 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 1126 |
+
"rewards/chosen": 0.12088136374950409,
|
| 1127 |
+
"rewards/margins": 0.080001600086689,
|
| 1128 |
+
"rewards/rejected": 0.040879763662815094,
|
| 1129 |
"step": 790
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.83,
|
| 1133 |
"learning_rate": 9.644087256027554e-08,
|
| 1134 |
+
"logits/chosen": -2.272735118865967,
|
| 1135 |
+
"logits/rejected": -2.2941083908081055,
|
| 1136 |
+
"logps/chosen": -284.6488952636719,
|
| 1137 |
+
"logps/rejected": -243.56796264648438,
|
| 1138 |
+
"loss": 0.6639,
|
| 1139 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1140 |
+
"rewards/chosen": 0.1113913282752037,
|
| 1141 |
+
"rewards/margins": 0.05327050760388374,
|
| 1142 |
+
"rewards/rejected": 0.05812082439661026,
|
| 1143 |
"step": 800
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.84,
|
| 1147 |
"learning_rate": 9.070034443168771e-08,
|
| 1148 |
+
"logits/chosen": -2.2838375568389893,
|
| 1149 |
+
"logits/rejected": -2.289247751235962,
|
| 1150 |
+
"logps/chosen": -269.5845642089844,
|
| 1151 |
+
"logps/rejected": -230.6207275390625,
|
| 1152 |
+
"loss": 0.6617,
|
| 1153 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1154 |
+
"rewards/chosen": 0.09149408340454102,
|
| 1155 |
+
"rewards/margins": 0.06341233849525452,
|
| 1156 |
+
"rewards/rejected": 0.02808173932135105,
|
| 1157 |
"step": 810
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.85,
|
| 1161 |
"learning_rate": 8.495981630309988e-08,
|
| 1162 |
+
"logits/chosen": -2.365980863571167,
|
| 1163 |
+
"logits/rejected": -2.3436598777770996,
|
| 1164 |
+
"logps/chosen": -302.0718688964844,
|
| 1165 |
+
"logps/rejected": -228.1407470703125,
|
| 1166 |
+
"loss": 0.6623,
|
| 1167 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1168 |
+
"rewards/chosen": 0.13062262535095215,
|
| 1169 |
+
"rewards/margins": 0.08858474344015121,
|
| 1170 |
+
"rewards/rejected": 0.04203786700963974,
|
| 1171 |
"step": 820
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.86,
|
| 1175 |
"learning_rate": 7.921928817451206e-08,
|
| 1176 |
+
"logits/chosen": -2.342413902282715,
|
| 1177 |
+
"logits/rejected": -2.2254080772399902,
|
| 1178 |
+
"logps/chosen": -287.4922180175781,
|
| 1179 |
+
"logps/rejected": -222.5606231689453,
|
| 1180 |
+
"loss": 0.6565,
|
| 1181 |
"rewards/accuracies": 0.675000011920929,
|
| 1182 |
+
"rewards/chosen": 0.12904855608940125,
|
| 1183 |
+
"rewards/margins": 0.08615640550851822,
|
| 1184 |
+
"rewards/rejected": 0.04289213940501213,
|
| 1185 |
"step": 830
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.87,
|
| 1189 |
"learning_rate": 7.347876004592423e-08,
|
| 1190 |
+
"logits/chosen": -2.259397029876709,
|
| 1191 |
+
"logits/rejected": -2.227036476135254,
|
| 1192 |
+
"logps/chosen": -258.3423767089844,
|
| 1193 |
+
"logps/rejected": -216.99606323242188,
|
| 1194 |
+
"loss": 0.6714,
|
| 1195 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1196 |
+
"rewards/chosen": 0.10358164459466934,
|
| 1197 |
+
"rewards/margins": 0.06773830950260162,
|
| 1198 |
+
"rewards/rejected": 0.03584333881735802,
|
| 1199 |
"step": 840
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.88,
|
| 1203 |
"learning_rate": 6.773823191733639e-08,
|
| 1204 |
+
"logits/chosen": -2.2834537029266357,
|
| 1205 |
+
"logits/rejected": -2.3872971534729004,
|
| 1206 |
+
"logps/chosen": -262.05084228515625,
|
| 1207 |
+
"logps/rejected": -231.11306762695312,
|
| 1208 |
+
"loss": 0.6647,
|
| 1209 |
+
"rewards/accuracies": 0.625,
|
| 1210 |
+
"rewards/chosen": 0.09495140612125397,
|
| 1211 |
+
"rewards/margins": 0.055265575647354126,
|
| 1212 |
+
"rewards/rejected": 0.03968583419919014,
|
| 1213 |
"step": 850
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.89,
|
| 1217 |
"learning_rate": 6.199770378874856e-08,
|
| 1218 |
+
"logits/chosen": -2.4065003395080566,
|
| 1219 |
+
"logits/rejected": -2.3337345123291016,
|
| 1220 |
+
"logps/chosen": -295.71478271484375,
|
| 1221 |
+
"logps/rejected": -270.1822814941406,
|
| 1222 |
+
"loss": 0.6693,
|
| 1223 |
+
"rewards/accuracies": 0.65625,
|
| 1224 |
+
"rewards/chosen": 0.11348612606525421,
|
| 1225 |
+
"rewards/margins": 0.07466179132461548,
|
| 1226 |
+
"rewards/rejected": 0.03882431983947754,
|
| 1227 |
"step": 860
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.9,
|
| 1231 |
"learning_rate": 5.6257175660160735e-08,
|
| 1232 |
+
"logits/chosen": -2.2463555335998535,
|
| 1233 |
+
"logits/rejected": -2.2443947792053223,
|
| 1234 |
+
"logps/chosen": -312.9588317871094,
|
| 1235 |
+
"logps/rejected": -237.4109344482422,
|
| 1236 |
+
"loss": 0.6644,
|
| 1237 |
+
"rewards/accuracies": 0.59375,
|
| 1238 |
+
"rewards/chosen": 0.10128283500671387,
|
| 1239 |
+
"rewards/margins": 0.053178369998931885,
|
| 1240 |
+
"rewards/rejected": 0.04810447618365288,
|
| 1241 |
"step": 870
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.91,
|
| 1245 |
"learning_rate": 5.05166475315729e-08,
|
| 1246 |
+
"logits/chosen": -2.358501434326172,
|
| 1247 |
+
"logits/rejected": -2.313483715057373,
|
| 1248 |
+
"logps/chosen": -291.43377685546875,
|
| 1249 |
+
"logps/rejected": -240.09054565429688,
|
| 1250 |
+
"loss": 0.6632,
|
| 1251 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1252 |
+
"rewards/chosen": 0.10742716491222382,
|
| 1253 |
+
"rewards/margins": 0.07204015552997589,
|
| 1254 |
+
"rewards/rejected": 0.03538701683282852,
|
| 1255 |
"step": 880
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.92,
|
| 1259 |
"learning_rate": 4.477611940298507e-08,
|
| 1260 |
+
"logits/chosen": -2.313149929046631,
|
| 1261 |
+
"logits/rejected": -2.3558261394500732,
|
| 1262 |
+
"logps/chosen": -285.90643310546875,
|
| 1263 |
+
"logps/rejected": -235.43051147460938,
|
| 1264 |
+
"loss": 0.6666,
|
| 1265 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 1266 |
+
"rewards/chosen": 0.12259715795516968,
|
| 1267 |
+
"rewards/margins": 0.09698096662759781,
|
| 1268 |
+
"rewards/rejected": 0.02561618760228157,
|
| 1269 |
"step": 890
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.93,
|
| 1273 |
"learning_rate": 3.903559127439724e-08,
|
| 1274 |
+
"logits/chosen": -2.3278651237487793,
|
| 1275 |
+
"logits/rejected": -2.195068836212158,
|
| 1276 |
+
"logps/chosen": -272.7381896972656,
|
| 1277 |
+
"logps/rejected": -211.40640258789062,
|
| 1278 |
+
"loss": 0.658,
|
| 1279 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 1280 |
+
"rewards/chosen": 0.1207551583647728,
|
| 1281 |
+
"rewards/margins": 0.09316142648458481,
|
| 1282 |
+
"rewards/rejected": 0.027593741193413734,
|
| 1283 |
"step": 900
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.94,
|
| 1287 |
"learning_rate": 3.3295063145809414e-08,
|
| 1288 |
+
"logits/chosen": -2.290696859359741,
|
| 1289 |
+
"logits/rejected": -2.3440823554992676,
|
| 1290 |
+
"logps/chosen": -238.2651824951172,
|
| 1291 |
+
"logps/rejected": -206.77969360351562,
|
| 1292 |
+
"loss": 0.6616,
|
| 1293 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1294 |
+
"rewards/chosen": 0.09928463399410248,
|
| 1295 |
+
"rewards/margins": 0.07226204872131348,
|
| 1296 |
+
"rewards/rejected": 0.027022594586014748,
|
| 1297 |
"step": 910
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.95,
|
| 1301 |
"learning_rate": 2.755453501722158e-08,
|
| 1302 |
+
"logits/chosen": -2.375807762145996,
|
| 1303 |
+
"logits/rejected": -2.367743730545044,
|
| 1304 |
+
"logps/chosen": -281.56195068359375,
|
| 1305 |
+
"logps/rejected": -225.125244140625,
|
| 1306 |
+
"loss": 0.662,
|
| 1307 |
+
"rewards/accuracies": 0.581250011920929,
|
| 1308 |
+
"rewards/chosen": 0.1072310209274292,
|
| 1309 |
+
"rewards/margins": 0.056608647108078,
|
| 1310 |
+
"rewards/rejected": 0.050622373819351196,
|
| 1311 |
"step": 920
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.96,
|
| 1315 |
"learning_rate": 2.1814006888633754e-08,
|
| 1316 |
+
"logits/chosen": -2.281919002532959,
|
| 1317 |
+
"logits/rejected": -2.254122734069824,
|
| 1318 |
+
"logps/chosen": -256.39105224609375,
|
| 1319 |
+
"logps/rejected": -203.3081817626953,
|
| 1320 |
+
"loss": 0.6617,
|
| 1321 |
+
"rewards/accuracies": 0.643750011920929,
|
| 1322 |
+
"rewards/chosen": 0.11211923509836197,
|
| 1323 |
+
"rewards/margins": 0.07925260812044144,
|
| 1324 |
+
"rewards/rejected": 0.03286661207675934,
|
| 1325 |
"step": 930
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.97,
|
| 1329 |
"learning_rate": 1.6073478760045924e-08,
|
| 1330 |
+
"logits/chosen": -2.316282272338867,
|
| 1331 |
+
"logits/rejected": -2.3123340606689453,
|
| 1332 |
+
"logps/chosen": -271.6207580566406,
|
| 1333 |
+
"logps/rejected": -231.7317352294922,
|
| 1334 |
+
"loss": 0.6626,
|
| 1335 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1336 |
+
"rewards/chosen": 0.10637687146663666,
|
| 1337 |
+
"rewards/margins": 0.06768520176410675,
|
| 1338 |
+
"rewards/rejected": 0.0386916846036911,
|
| 1339 |
"step": 940
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 0.98,
|
| 1343 |
"learning_rate": 1.0332950631458094e-08,
|
| 1344 |
+
"logits/chosen": -2.3146958351135254,
|
| 1345 |
+
"logits/rejected": -2.2793381214141846,
|
| 1346 |
+
"logps/chosen": -282.83270263671875,
|
| 1347 |
+
"logps/rejected": -233.0804443359375,
|
| 1348 |
+
"loss": 0.6612,
|
| 1349 |
"rewards/accuracies": 0.6812499761581421,
|
| 1350 |
+
"rewards/chosen": 0.11455857753753662,
|
| 1351 |
+
"rewards/margins": 0.0838586837053299,
|
| 1352 |
+
"rewards/rejected": 0.030699897557497025,
|
| 1353 |
"step": 950
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.99,
|
| 1357 |
"learning_rate": 4.592422502870264e-09,
|
| 1358 |
+
"logits/chosen": -2.251638889312744,
|
| 1359 |
+
"logits/rejected": -2.234907627105713,
|
| 1360 |
+
"logps/chosen": -281.0075378417969,
|
| 1361 |
+
"logps/rejected": -239.98049926757812,
|
| 1362 |
+
"loss": 0.661,
|
| 1363 |
+
"rewards/accuracies": 0.65625,
|
| 1364 |
+
"rewards/chosen": 0.1062885969877243,
|
| 1365 |
+
"rewards/margins": 0.06708581745624542,
|
| 1366 |
+
"rewards/rejected": 0.03920278698205948,
|
| 1367 |
"step": 960
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
"epoch": 1.0,
|
| 1371 |
+
"eval_logits/chosen": -2.4597132205963135,
|
| 1372 |
+
"eval_logits/rejected": -2.398695468902588,
|
| 1373 |
+
"eval_logps/chosen": -278.69171142578125,
|
| 1374 |
+
"eval_logps/rejected": -230.4560089111328,
|
| 1375 |
+
"eval_loss": 0.6642152070999146,
|
| 1376 |
+
"eval_rewards/accuracies": 0.6480000019073486,
|
| 1377 |
+
"eval_rewards/chosen": 0.10415761172771454,
|
| 1378 |
+
"eval_rewards/margins": 0.06405296921730042,
|
| 1379 |
+
"eval_rewards/rejected": 0.04010463133454323,
|
| 1380 |
+
"eval_runtime": 443.9432,
|
| 1381 |
+
"eval_samples_per_second": 4.505,
|
| 1382 |
+
"eval_steps_per_second": 0.282,
|
| 1383 |
"step": 968
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 1.0,
|
| 1387 |
"step": 968,
|
| 1388 |
"total_flos": 0.0,
|
| 1389 |
+
"train_loss": 0.6728762634529555,
|
| 1390 |
+
"train_runtime": 27528.1814,
|
| 1391 |
+
"train_samples_per_second": 2.251,
|
| 1392 |
"train_steps_per_second": 0.035
|
| 1393 |
}
|
| 1394 |
],
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4792
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cfc6fa65238373edb8c038b73d0de99649ac0d248e697a0222bd24510217b308
|
| 3 |
size 4792
|