Model save
Browse files- README.md +17 -15
- all_results.json +18 -16
- eval_results.json +14 -12
- runs/Jan15_16-31-35_uclaml03.cs.ucla.edu/events.out.tfevents.1705390086.uclaml03.cs.ucla.edu.1514934.1 +3 -0
- train_results.json +4 -4
- trainer_state.json +1164 -970
README.md
CHANGED
|
@@ -15,15 +15,17 @@ 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: 0.
|
| 20 |
-
- Rewards/rejected: -0.
|
| 21 |
-
- Rewards/accuracies: 0.
|
| 22 |
-
- Rewards/margins: 0.
|
| 23 |
-
- Logps/rejected: -
|
| 24 |
-
- Logps/chosen: -
|
| 25 |
-
- Logits/rejected: -2.
|
| 26 |
-
- Logits/chosen: -2.
|
|
|
|
|
|
|
| 27 |
|
| 28 |
## Model description
|
| 29 |
|
|
@@ -47,10 +49,10 @@ The following hyperparameters were used during training:
|
|
| 47 |
- eval_batch_size: 4
|
| 48 |
- seed: 42
|
| 49 |
- distributed_type: multi-GPU
|
| 50 |
-
- num_devices:
|
| 51 |
-
- gradient_accumulation_steps:
|
| 52 |
- total_train_batch_size: 64
|
| 53 |
-
- total_eval_batch_size:
|
| 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
|
|
@@ -58,9 +60,9 @@ The following hyperparameters were used during training:
|
|
| 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 |
|
| 65 |
|
| 66 |
### 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: 0.6351
|
| 19 |
+
- Rewards/chosen: 0.0300
|
| 20 |
+
- Rewards/rejected: -0.0335
|
| 21 |
+
- Rewards/accuracies: 0.6200
|
| 22 |
+
- Rewards/margins: 0.0635
|
| 23 |
+
- Logps/rejected: -250.2864
|
| 24 |
+
- Logps/chosen: -272.9344
|
| 25 |
+
- Logits/rejected: -2.5200
|
| 26 |
+
- Logits/chosen: -2.5063
|
| 27 |
+
- Use Label: 7566.9282
|
| 28 |
+
- Pred Label: 8465.0723
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 49 |
- eval_batch_size: 4
|
| 50 |
- seed: 42
|
| 51 |
- distributed_type: multi-GPU
|
| 52 |
+
- num_devices: 4
|
| 53 |
+
- gradient_accumulation_steps: 4
|
| 54 |
- total_train_batch_size: 64
|
| 55 |
+
- total_eval_batch_size: 16
|
| 56 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 57 |
- lr_scheduler_type: linear
|
| 58 |
- lr_scheduler_warmup_ratio: 0.1
|
|
|
|
| 60 |
|
| 61 |
### Training results
|
| 62 |
|
| 63 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Use Label | Pred Label |
|
| 64 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:---------:|:----------:|
|
| 65 |
+
| 0.6403 | 1.0 | 955 | 0.6351 | 0.0300 | -0.0335 | 0.6200 | 0.0635 | -250.2864 | -272.9344 | -2.5200 | -2.5063 | 7400.9282 | 8131.0718 |
|
| 66 |
|
| 67 |
|
| 68 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,21 +1,23 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_logits/chosen": -2.
|
| 4 |
-
"eval_logits/rejected": -2.
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"
|
| 9 |
-
"eval_rewards/
|
| 10 |
-
"eval_rewards/
|
| 11 |
-
"eval_rewards/
|
| 12 |
-
"
|
|
|
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second":
|
| 15 |
-
"eval_steps_per_second": 0.
|
| 16 |
-
"
|
| 17 |
-
"
|
|
|
|
| 18 |
"train_samples": 61135,
|
| 19 |
-
"train_samples_per_second":
|
| 20 |
-
"train_steps_per_second": 0.
|
| 21 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.5062618255615234,
|
| 4 |
+
"eval_logits/rejected": -2.5199859142303467,
|
| 5 |
+
"eval_logps/chosen": -272.93438720703125,
|
| 6 |
+
"eval_logps/rejected": -250.28643798828125,
|
| 7 |
+
"eval_loss": 0.6350826025009155,
|
| 8 |
+
"eval_pred_label": 8465.072265625,
|
| 9 |
+
"eval_rewards/accuracies": 0.6200000047683716,
|
| 10 |
+
"eval_rewards/chosen": 0.029984984546899796,
|
| 11 |
+
"eval_rewards/margins": 0.06345725804567337,
|
| 12 |
+
"eval_rewards/rejected": -0.033472273498773575,
|
| 13 |
+
"eval_runtime": 446.4868,
|
| 14 |
"eval_samples": 2000,
|
| 15 |
+
"eval_samples_per_second": 4.479,
|
| 16 |
+
"eval_steps_per_second": 0.28,
|
| 17 |
+
"eval_use_label": 7566.92822265625,
|
| 18 |
+
"train_loss": 0.6552608817035616,
|
| 19 |
+
"train_runtime": 24261.6882,
|
| 20 |
"train_samples": 61135,
|
| 21 |
+
"train_samples_per_second": 2.52,
|
| 22 |
+
"train_steps_per_second": 0.039
|
| 23 |
}
|
eval_results.json
CHANGED
|
@@ -1,16 +1,18 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_logits/chosen": -2.
|
| 4 |
-
"eval_logits/rejected": -2.
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"
|
| 9 |
-
"eval_rewards/
|
| 10 |
-
"eval_rewards/
|
| 11 |
-
"eval_rewards/
|
| 12 |
-
"
|
|
|
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second":
|
| 15 |
-
"eval_steps_per_second": 0.
|
|
|
|
| 16 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.5062618255615234,
|
| 4 |
+
"eval_logits/rejected": -2.5199859142303467,
|
| 5 |
+
"eval_logps/chosen": -272.93438720703125,
|
| 6 |
+
"eval_logps/rejected": -250.28643798828125,
|
| 7 |
+
"eval_loss": 0.6350826025009155,
|
| 8 |
+
"eval_pred_label": 8465.072265625,
|
| 9 |
+
"eval_rewards/accuracies": 0.6200000047683716,
|
| 10 |
+
"eval_rewards/chosen": 0.029984984546899796,
|
| 11 |
+
"eval_rewards/margins": 0.06345725804567337,
|
| 12 |
+
"eval_rewards/rejected": -0.033472273498773575,
|
| 13 |
+
"eval_runtime": 446.4868,
|
| 14 |
"eval_samples": 2000,
|
| 15 |
+
"eval_samples_per_second": 4.479,
|
| 16 |
+
"eval_steps_per_second": 0.28,
|
| 17 |
+
"eval_use_label": 7566.92822265625
|
| 18 |
}
|
runs/Jan15_16-31-35_uclaml03.cs.ucla.edu/events.out.tfevents.1705390086.uclaml03.cs.ucla.edu.1514934.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ae2d670c8fd9c570877d76918f04f157685c085f6f2daed5b6d0a504d569ac4
|
| 3 |
+
size 935
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 61135,
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.6552608817035616,
|
| 4 |
+
"train_runtime": 24261.6882,
|
| 5 |
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 2.52,
|
| 7 |
+
"train_steps_per_second": 0.039
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -11,1371 +11,1565 @@
|
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
"learning_rate": 5.208333333333333e-09,
|
| 14 |
-
"logits/chosen": -2.
|
| 15 |
-
"logits/rejected": -2.
|
| 16 |
-
"logps/chosen": -
|
| 17 |
-
"logps/rejected": -
|
| 18 |
"loss": 0.6931,
|
|
|
|
| 19 |
"rewards/accuracies": 0.0,
|
| 20 |
"rewards/chosen": 0.0,
|
| 21 |
"rewards/margins": 0.0,
|
| 22 |
"rewards/rejected": 0.0,
|
| 23 |
-
"step": 1
|
|
|
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
"learning_rate": 5.208333333333333e-08,
|
| 28 |
-
"logits/chosen": -2.
|
| 29 |
-
"logits/rejected": -2.
|
| 30 |
-
"logps/chosen": -
|
| 31 |
-
"logps/rejected": -
|
| 32 |
-
"loss": 0.
|
| 33 |
-
"
|
| 34 |
-
"rewards/
|
| 35 |
-
"rewards/
|
| 36 |
-
"rewards/
|
| 37 |
-
"
|
|
|
|
|
|
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02,
|
| 41 |
"learning_rate": 1.0416666666666667e-07,
|
| 42 |
-
"logits/chosen": -2.
|
| 43 |
-
"logits/rejected": -2.
|
| 44 |
-
"logps/chosen": -
|
| 45 |
-
"logps/rejected": -
|
| 46 |
-
"loss": 0.
|
| 47 |
-
"
|
| 48 |
-
"rewards/
|
| 49 |
-
"rewards/
|
| 50 |
-
"rewards/
|
| 51 |
-
"
|
|
|
|
|
|
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03,
|
| 55 |
"learning_rate": 1.5624999999999999e-07,
|
| 56 |
-
"logits/chosen": -2.
|
| 57 |
-
"logits/rejected": -2.
|
| 58 |
-
"logps/chosen": -
|
| 59 |
-
"logps/rejected": -244.
|
| 60 |
-
"loss": 0.
|
| 61 |
-
"
|
| 62 |
-
"rewards/
|
| 63 |
-
"rewards/
|
| 64 |
-
"rewards/
|
| 65 |
-
"
|
|
|
|
|
|
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.04,
|
| 69 |
"learning_rate": 2.0833333333333333e-07,
|
| 70 |
-
"logits/chosen": -2.
|
| 71 |
-
"logits/rejected": -2.
|
| 72 |
-
"logps/chosen": -
|
| 73 |
-
"logps/rejected": -
|
| 74 |
-
"loss": 0.
|
| 75 |
-
"
|
| 76 |
-
"rewards/
|
| 77 |
-
"rewards/
|
| 78 |
-
"rewards/
|
| 79 |
-
"
|
|
|
|
|
|
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05,
|
| 83 |
"learning_rate": 2.604166666666667e-07,
|
| 84 |
-
"logits/chosen": -2.
|
| 85 |
-
"logits/rejected": -2.
|
| 86 |
-
"logps/chosen": -255.
|
| 87 |
-
"logps/rejected": -249.
|
| 88 |
-
"loss": 0.
|
| 89 |
-
"
|
| 90 |
-
"rewards/
|
| 91 |
-
"rewards/
|
| 92 |
-
"rewards/
|
| 93 |
-
"
|
|
|
|
|
|
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06,
|
| 97 |
"learning_rate": 3.1249999999999997e-07,
|
| 98 |
-
"logits/chosen": -2.
|
| 99 |
-
"logits/rejected": -2.
|
| 100 |
-
"logps/chosen": -
|
| 101 |
-
"logps/rejected": -
|
| 102 |
-
"loss": 0.
|
| 103 |
-
"
|
| 104 |
-
"rewards/
|
| 105 |
-
"rewards/
|
| 106 |
-
"rewards/
|
| 107 |
-
"
|
|
|
|
|
|
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07,
|
| 111 |
"learning_rate": 3.645833333333333e-07,
|
| 112 |
-
"logits/chosen": -2.
|
| 113 |
-
"logits/rejected": -2.
|
| 114 |
-
"logps/chosen": -
|
| 115 |
-
"logps/rejected": -
|
| 116 |
-
"loss": 0.
|
| 117 |
-
"
|
| 118 |
-
"rewards/
|
| 119 |
-
"rewards/
|
| 120 |
-
"rewards/
|
| 121 |
-
"
|
|
|
|
|
|
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08,
|
| 125 |
"learning_rate": 4.1666666666666667e-07,
|
| 126 |
-
"logits/chosen": -2.
|
| 127 |
-
"logits/rejected": -2.
|
| 128 |
-
"logps/chosen": -
|
| 129 |
-
"logps/rejected": -
|
| 130 |
-
"loss": 0.
|
| 131 |
-
"
|
| 132 |
-
"rewards/
|
| 133 |
-
"rewards/
|
| 134 |
-
"rewards/
|
| 135 |
-
"
|
|
|
|
|
|
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09,
|
| 139 |
"learning_rate": 4.6874999999999996e-07,
|
| 140 |
-
"logits/chosen": -2.
|
| 141 |
-
"logits/rejected": -2.
|
| 142 |
-
"logps/chosen": -
|
| 143 |
-
"logps/rejected": -
|
| 144 |
-
"loss": 0.
|
| 145 |
-
"
|
| 146 |
-
"rewards/
|
| 147 |
-
"rewards/
|
| 148 |
-
"rewards/
|
| 149 |
-
"
|
|
|
|
|
|
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.1,
|
| 153 |
"learning_rate": 4.976717112922002e-07,
|
| 154 |
-
"logits/chosen": -2.
|
| 155 |
-
"logits/rejected": -2.
|
| 156 |
-
"logps/chosen": -
|
| 157 |
-
"logps/rejected": -
|
| 158 |
-
"loss": 0.
|
| 159 |
-
"
|
| 160 |
-
"rewards/
|
| 161 |
-
"rewards/
|
| 162 |
-
"rewards/
|
| 163 |
-
"
|
|
|
|
|
|
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.12,
|
| 167 |
"learning_rate": 4.918509895227007e-07,
|
| 168 |
-
"logits/chosen": -2.
|
| 169 |
-
"logits/rejected": -2.
|
| 170 |
-
"logps/chosen": -
|
| 171 |
-
"logps/rejected": -
|
| 172 |
-
"loss": 0.
|
| 173 |
-
"
|
| 174 |
-
"rewards/
|
| 175 |
-
"rewards/
|
| 176 |
-
"rewards/
|
| 177 |
-
"
|
|
|
|
|
|
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.13,
|
| 181 |
"learning_rate": 4.860302677532014e-07,
|
| 182 |
-
"logits/chosen": -2.
|
| 183 |
-
"logits/rejected": -2.
|
| 184 |
-
"logps/chosen": -
|
| 185 |
-
"logps/rejected": -
|
| 186 |
-
"loss": 0.
|
| 187 |
-
"
|
| 188 |
-
"rewards/
|
| 189 |
-
"rewards/
|
| 190 |
-
"rewards/
|
| 191 |
-
"
|
|
|
|
|
|
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.14,
|
| 195 |
"learning_rate": 4.802095459837019e-07,
|
| 196 |
-
"logits/chosen": -2.
|
| 197 |
-
"logits/rejected": -2.
|
| 198 |
-
"logps/chosen": -
|
| 199 |
-
"logps/rejected": -
|
| 200 |
-
"loss": 0.
|
| 201 |
-
"
|
| 202 |
-
"rewards/
|
| 203 |
-
"rewards/
|
| 204 |
-
"rewards/
|
| 205 |
-
"
|
|
|
|
|
|
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.15,
|
| 209 |
"learning_rate": 4.743888242142026e-07,
|
| 210 |
-
"logits/chosen": -2.
|
| 211 |
-
"logits/rejected": -2.
|
| 212 |
-
"logps/chosen": -
|
| 213 |
-
"logps/rejected": -
|
| 214 |
-
"loss": 0.
|
| 215 |
-
"
|
| 216 |
-
"rewards/
|
| 217 |
-
"rewards/
|
| 218 |
-
"rewards/
|
| 219 |
-
"
|
|
|
|
|
|
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.16,
|
| 223 |
"learning_rate": 4.685681024447031e-07,
|
| 224 |
-
"logits/chosen": -2.
|
| 225 |
-
"logits/rejected": -2.
|
| 226 |
-
"logps/chosen": -
|
| 227 |
-
"logps/rejected": -260.
|
| 228 |
-
"loss": 0.
|
| 229 |
-
"
|
| 230 |
-
"rewards/
|
| 231 |
-
"rewards/
|
| 232 |
-
"rewards/
|
| 233 |
-
"
|
|
|
|
|
|
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.17,
|
| 237 |
"learning_rate": 4.627473806752037e-07,
|
| 238 |
-
"logits/chosen": -2.
|
| 239 |
-
"logits/rejected": -2.
|
| 240 |
-
"logps/chosen": -
|
| 241 |
-
"logps/rejected": -
|
| 242 |
-
"loss": 0.
|
| 243 |
-
"
|
| 244 |
-
"rewards/
|
| 245 |
-
"rewards/
|
| 246 |
-
"rewards/
|
| 247 |
-
"
|
|
|
|
|
|
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.18,
|
| 251 |
"learning_rate": 4.5692665890570433e-07,
|
| 252 |
-
"logits/chosen": -2.
|
| 253 |
-
"logits/rejected": -2.
|
| 254 |
-
"logps/chosen": -
|
| 255 |
-
"logps/rejected": -
|
| 256 |
-
"loss": 0.
|
| 257 |
-
"
|
| 258 |
-
"rewards/
|
| 259 |
-
"rewards/
|
| 260 |
-
"rewards/
|
| 261 |
-
"
|
|
|
|
|
|
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.19,
|
| 265 |
"learning_rate": 4.5110593713620486e-07,
|
| 266 |
-
"logits/chosen": -2.
|
| 267 |
-
"logits/rejected": -2.
|
| 268 |
-
"logps/chosen": -
|
| 269 |
-
"logps/rejected": -
|
| 270 |
-
"loss": 0.
|
| 271 |
-
"
|
| 272 |
-
"rewards/
|
| 273 |
-
"rewards/
|
| 274 |
-
"rewards/
|
| 275 |
-
"
|
|
|
|
|
|
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2,
|
| 279 |
"learning_rate": 4.4528521536670544e-07,
|
| 280 |
-
"logits/chosen": -2.
|
| 281 |
-
"logits/rejected": -2.
|
| 282 |
-
"logps/chosen": -291.
|
| 283 |
-
"logps/rejected": -
|
| 284 |
-
"loss": 0.
|
| 285 |
-
"
|
| 286 |
-
"rewards/
|
| 287 |
-
"rewards/
|
| 288 |
-
"rewards/
|
| 289 |
-
"
|
|
|
|
|
|
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.21,
|
| 293 |
"learning_rate": 4.3946449359720607e-07,
|
| 294 |
-
"logits/chosen": -2.
|
| 295 |
-
"logits/rejected": -2.
|
| 296 |
-
"logps/chosen": -
|
| 297 |
-
"logps/rejected": -
|
| 298 |
-
"loss": 0.
|
| 299 |
-
"
|
| 300 |
-
"rewards/
|
| 301 |
-
"rewards/
|
| 302 |
-
"rewards/
|
| 303 |
-
"
|
|
|
|
|
|
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.22,
|
| 307 |
"learning_rate": 4.336437718277066e-07,
|
| 308 |
-
"logits/chosen": -2.
|
| 309 |
-
"logits/rejected": -2.
|
| 310 |
-
"logps/chosen": -
|
| 311 |
-
"logps/rejected": -
|
| 312 |
-
"loss": 0.
|
| 313 |
-
"
|
| 314 |
-
"rewards/
|
| 315 |
-
"rewards/
|
| 316 |
-
"rewards/
|
| 317 |
-
"
|
|
|
|
|
|
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.23,
|
| 321 |
"learning_rate": 4.278230500582072e-07,
|
| 322 |
-
"logits/chosen": -2.
|
| 323 |
-
"logits/rejected": -2.
|
| 324 |
-
"logps/chosen": -
|
| 325 |
-
"logps/rejected": -
|
| 326 |
-
"loss": 0.
|
| 327 |
-
"
|
| 328 |
-
"rewards/
|
| 329 |
-
"rewards/
|
| 330 |
-
"rewards/
|
| 331 |
-
"
|
|
|
|
|
|
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.24,
|
| 335 |
"learning_rate": 4.220023282887078e-07,
|
| 336 |
-
"logits/chosen": -2.
|
| 337 |
-
"logits/rejected": -2.
|
| 338 |
-
"logps/chosen": -
|
| 339 |
-
"logps/rejected": -
|
| 340 |
-
"loss": 0.
|
| 341 |
-
"
|
| 342 |
-
"rewards/
|
| 343 |
-
"rewards/
|
| 344 |
-
"rewards/
|
| 345 |
-
"
|
|
|
|
|
|
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.25,
|
| 349 |
"learning_rate": 4.1618160651920834e-07,
|
| 350 |
-
"logits/chosen": -2.
|
| 351 |
-
"logits/rejected": -2.
|
| 352 |
-
"logps/chosen": -
|
| 353 |
-
"logps/rejected": -
|
| 354 |
-
"loss": 0.
|
| 355 |
-
"
|
| 356 |
-
"rewards/
|
| 357 |
-
"rewards/
|
| 358 |
-
"rewards/
|
| 359 |
-
"
|
|
|
|
|
|
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.26,
|
| 363 |
"learning_rate": 4.103608847497089e-07,
|
| 364 |
-
"logits/chosen": -2.
|
| 365 |
-
"logits/rejected": -2.
|
| 366 |
-
"logps/chosen": -
|
| 367 |
-
"logps/rejected": -
|
| 368 |
-
"loss": 0.
|
| 369 |
-
"
|
| 370 |
-
"rewards/
|
| 371 |
-
"rewards/
|
| 372 |
-
"rewards/
|
| 373 |
-
"
|
|
|
|
|
|
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.27,
|
| 377 |
"learning_rate": 4.0454016298020956e-07,
|
| 378 |
-
"logits/chosen": -2.
|
| 379 |
-
"logits/rejected": -2.
|
| 380 |
-
"logps/chosen": -
|
| 381 |
-
"logps/rejected": -
|
| 382 |
-
"loss": 0.
|
| 383 |
-
"
|
| 384 |
-
"rewards/
|
| 385 |
-
"rewards/
|
| 386 |
-
"rewards/
|
| 387 |
-
"
|
|
|
|
|
|
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.28,
|
| 391 |
"learning_rate": 3.987194412107101e-07,
|
| 392 |
-
"logits/chosen": -2.
|
| 393 |
-
"logits/rejected": -2.
|
| 394 |
-
"logps/chosen": -
|
| 395 |
-
"logps/rejected": -
|
| 396 |
"loss": 0.6735,
|
| 397 |
-
"
|
| 398 |
-
"rewards/
|
| 399 |
-
"rewards/
|
| 400 |
-
"rewards/
|
| 401 |
-
"
|
|
|
|
|
|
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.29,
|
| 405 |
"learning_rate": 3.9289871944121066e-07,
|
| 406 |
-
"logits/chosen": -2.
|
| 407 |
-
"logits/rejected": -2.
|
| 408 |
-
"logps/chosen": -
|
| 409 |
-
"logps/rejected": -
|
| 410 |
-
"loss": 0.
|
| 411 |
-
"
|
| 412 |
-
"rewards/
|
| 413 |
-
"rewards/
|
| 414 |
-
"rewards/
|
| 415 |
-
"
|
|
|
|
|
|
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.3,
|
| 419 |
"learning_rate": 3.870779976717113e-07,
|
| 420 |
-
"logits/chosen": -2.
|
| 421 |
-
"logits/rejected": -2.
|
| 422 |
-
"logps/chosen": -
|
| 423 |
-
"logps/rejected": -
|
| 424 |
-
"loss": 0.
|
| 425 |
-
"
|
| 426 |
-
"rewards/
|
| 427 |
-
"rewards/
|
| 428 |
-
"rewards/
|
| 429 |
-
"
|
|
|
|
|
|
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.31,
|
| 433 |
"learning_rate": 3.812572759022118e-07,
|
| 434 |
-
"logits/chosen": -2.
|
| 435 |
-
"logits/rejected": -2.
|
| 436 |
-
"logps/chosen": -
|
| 437 |
-
"logps/rejected": -
|
| 438 |
-
"loss": 0.
|
| 439 |
-
"
|
| 440 |
-
"rewards/
|
| 441 |
-
"rewards/
|
| 442 |
-
"rewards/
|
| 443 |
-
"
|
|
|
|
|
|
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.32,
|
| 447 |
"learning_rate": 3.754365541327124e-07,
|
| 448 |
-
"logits/chosen": -2.
|
| 449 |
-
"logits/rejected": -2.
|
| 450 |
-
"logps/chosen": -
|
| 451 |
-
"logps/rejected": -
|
| 452 |
-
"loss": 0.
|
| 453 |
-
"
|
| 454 |
-
"rewards/
|
| 455 |
-
"rewards/
|
| 456 |
-
"rewards/
|
| 457 |
-
"
|
|
|
|
|
|
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.33,
|
| 461 |
"learning_rate": 3.6961583236321304e-07,
|
| 462 |
-
"logits/chosen": -2.
|
| 463 |
-
"logits/rejected": -2.
|
| 464 |
-
"logps/chosen": -
|
| 465 |
-
"logps/rejected": -
|
| 466 |
-
"loss": 0.
|
| 467 |
-
"
|
| 468 |
-
"rewards/
|
| 469 |
-
"rewards/
|
| 470 |
-
"rewards/
|
| 471 |
-
"
|
|
|
|
|
|
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.35,
|
| 475 |
"learning_rate": 3.637951105937136e-07,
|
| 476 |
-
"logits/chosen": -2.
|
| 477 |
-
"logits/rejected": -2.
|
| 478 |
-
"logps/chosen": -
|
| 479 |
-
"logps/rejected": -
|
| 480 |
-
"loss": 0.
|
| 481 |
-
"
|
| 482 |
-
"rewards/
|
| 483 |
-
"rewards/
|
| 484 |
-
"rewards/
|
| 485 |
-
"
|
|
|
|
|
|
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.36,
|
| 489 |
"learning_rate": 3.579743888242142e-07,
|
| 490 |
-
"logits/chosen": -2.
|
| 491 |
-
"logits/rejected": -2.
|
| 492 |
-
"logps/chosen": -
|
| 493 |
-
"logps/rejected": -
|
| 494 |
-
"loss": 0.
|
| 495 |
-
"
|
| 496 |
-
"rewards/
|
| 497 |
-
"rewards/
|
| 498 |
-
"rewards/
|
| 499 |
-
"
|
|
|
|
|
|
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.37,
|
| 503 |
"learning_rate": 3.521536670547148e-07,
|
| 504 |
-
"logits/chosen": -2.
|
| 505 |
-
"logits/rejected": -2.
|
| 506 |
-
"logps/chosen": -
|
| 507 |
-
"logps/rejected": -
|
| 508 |
-
"loss": 0.
|
| 509 |
-
"
|
| 510 |
-
"rewards/
|
| 511 |
-
"rewards/
|
| 512 |
-
"rewards/
|
| 513 |
-
"
|
|
|
|
|
|
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.38,
|
| 517 |
"learning_rate": 3.4633294528521536e-07,
|
| 518 |
-
"logits/chosen": -2.
|
| 519 |
-
"logits/rejected": -2.
|
| 520 |
-
"logps/chosen": -
|
| 521 |
-
"logps/rejected": -
|
| 522 |
-
"loss": 0.
|
| 523 |
-
"
|
| 524 |
-
"rewards/
|
| 525 |
-
"rewards/
|
| 526 |
-
"rewards/
|
| 527 |
-
"
|
|
|
|
|
|
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.39,
|
| 531 |
"learning_rate": 3.4051222351571594e-07,
|
| 532 |
-
"logits/chosen": -2.
|
| 533 |
-
"logits/rejected": -2.
|
| 534 |
-
"logps/chosen": -
|
| 535 |
-
"logps/rejected": -
|
| 536 |
-
"loss": 0.
|
| 537 |
-
"
|
| 538 |
-
"rewards/
|
| 539 |
-
"rewards/
|
| 540 |
-
"rewards/
|
| 541 |
-
"
|
|
|
|
|
|
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.4,
|
| 545 |
"learning_rate": 3.346915017462165e-07,
|
| 546 |
-
"logits/chosen": -2.
|
| 547 |
-
"logits/rejected": -2.
|
| 548 |
-
"logps/chosen": -
|
| 549 |
-
"logps/rejected": -
|
| 550 |
-
"loss": 0.
|
| 551 |
-
"
|
| 552 |
-
"rewards/
|
| 553 |
-
"rewards/
|
| 554 |
-
"rewards/
|
| 555 |
-
"
|
|
|
|
|
|
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.41,
|
| 559 |
"learning_rate": 3.288707799767171e-07,
|
| 560 |
-
"logits/chosen": -2.
|
| 561 |
-
"logits/rejected": -2.
|
| 562 |
-
"logps/chosen": -
|
| 563 |
-
"logps/rejected": -
|
| 564 |
-
"loss": 0.
|
| 565 |
-
"
|
| 566 |
-
"rewards/
|
| 567 |
-
"rewards/
|
| 568 |
-
"rewards/
|
| 569 |
-
"
|
|
|
|
|
|
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.42,
|
| 573 |
"learning_rate": 3.230500582072177e-07,
|
| 574 |
-
"logits/chosen": -2.
|
| 575 |
-
"logits/rejected": -2.
|
| 576 |
-
"logps/chosen": -
|
| 577 |
-
"logps/rejected": -
|
| 578 |
-
"loss": 0.
|
| 579 |
-
"
|
| 580 |
-
"rewards/
|
| 581 |
-
"rewards/
|
| 582 |
-
"rewards/
|
| 583 |
-
"
|
|
|
|
|
|
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.43,
|
| 587 |
"learning_rate": 3.1722933643771827e-07,
|
| 588 |
-
"logits/chosen": -2.
|
| 589 |
-
"logits/rejected": -2.
|
| 590 |
-
"logps/chosen": -
|
| 591 |
-
"logps/rejected": -
|
| 592 |
-
"loss": 0.
|
| 593 |
-
"
|
| 594 |
-
"rewards/
|
| 595 |
-
"rewards/
|
| 596 |
-
"rewards/
|
| 597 |
-
"
|
|
|
|
|
|
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.44,
|
| 601 |
"learning_rate": 3.1140861466821885e-07,
|
| 602 |
-
"logits/chosen": -2.
|
| 603 |
-
"logits/rejected": -2.
|
| 604 |
-
"logps/chosen": -
|
| 605 |
-
"logps/rejected": -243.
|
| 606 |
-
"loss": 0.
|
| 607 |
-
"
|
| 608 |
-
"rewards/
|
| 609 |
-
"rewards/
|
| 610 |
-
"rewards/
|
| 611 |
-
"
|
|
|
|
|
|
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.45,
|
| 615 |
"learning_rate": 3.0558789289871943e-07,
|
| 616 |
-
"logits/chosen": -2.
|
| 617 |
-
"logits/rejected": -2.
|
| 618 |
-
"logps/chosen": -
|
| 619 |
-
"logps/rejected": -
|
| 620 |
-
"loss": 0.
|
| 621 |
-
"
|
| 622 |
-
"rewards/
|
| 623 |
-
"rewards/
|
| 624 |
-
"rewards/
|
| 625 |
-
"
|
|
|
|
|
|
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.46,
|
| 629 |
"learning_rate": 2.9976717112922e-07,
|
| 630 |
-
"logits/chosen": -2.
|
| 631 |
-
"logits/rejected": -2.
|
| 632 |
-
"logps/chosen": -
|
| 633 |
-
"logps/rejected": -
|
| 634 |
-
"loss": 0.
|
| 635 |
-
"
|
| 636 |
-
"rewards/
|
| 637 |
-
"rewards/
|
| 638 |
-
"rewards/
|
| 639 |
-
"
|
|
|
|
|
|
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.47,
|
| 643 |
"learning_rate": 2.939464493597206e-07,
|
| 644 |
-
"logits/chosen": -2.
|
| 645 |
-
"logits/rejected": -2.
|
| 646 |
-
"logps/chosen": -
|
| 647 |
-
"logps/rejected": -
|
| 648 |
-
"loss": 0.
|
| 649 |
-
"
|
| 650 |
-
"rewards/
|
| 651 |
-
"rewards/
|
| 652 |
-
"rewards/
|
| 653 |
-
"
|
|
|
|
|
|
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.48,
|
| 657 |
"learning_rate": 2.8812572759022117e-07,
|
| 658 |
-
"logits/chosen": -2.
|
| 659 |
-
"logits/rejected": -2.
|
| 660 |
-
"logps/chosen": -
|
| 661 |
-
"logps/rejected": -
|
| 662 |
-
"loss": 0.
|
| 663 |
-
"
|
| 664 |
-
"rewards/
|
| 665 |
-
"rewards/
|
| 666 |
-
"rewards/
|
| 667 |
-
"
|
|
|
|
|
|
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.49,
|
| 671 |
"learning_rate": 2.8230500582072175e-07,
|
| 672 |
-
"logits/chosen": -2.
|
| 673 |
-
"logits/rejected": -2.
|
| 674 |
-
"logps/chosen": -
|
| 675 |
-
"logps/rejected": -
|
| 676 |
-
"loss": 0.
|
| 677 |
-
"
|
| 678 |
-
"rewards/
|
| 679 |
-
"rewards/
|
| 680 |
-
"rewards/
|
| 681 |
-
"
|
|
|
|
|
|
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.5,
|
| 685 |
"learning_rate": 2.7648428405122233e-07,
|
| 686 |
-
"logits/chosen": -2.
|
| 687 |
-
"logits/rejected": -2.
|
| 688 |
-
"logps/chosen": -
|
| 689 |
-
"logps/rejected": -
|
| 690 |
-
"loss": 0.
|
| 691 |
-
"
|
| 692 |
-
"rewards/
|
| 693 |
-
"rewards/
|
| 694 |
-
"rewards/
|
| 695 |
-
"
|
|
|
|
|
|
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.51,
|
| 699 |
"learning_rate": 2.706635622817229e-07,
|
| 700 |
-
"logits/chosen": -2.
|
| 701 |
-
"logits/rejected": -2.
|
| 702 |
-
"logps/chosen": -
|
| 703 |
-
"logps/rejected": -
|
| 704 |
-
"loss": 0.
|
| 705 |
-
"
|
| 706 |
-
"rewards/
|
| 707 |
-
"rewards/
|
| 708 |
-
"rewards/
|
| 709 |
-
"
|
|
|
|
|
|
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.52,
|
| 713 |
"learning_rate": 2.648428405122235e-07,
|
| 714 |
-
"logits/chosen": -2.
|
| 715 |
-
"logits/rejected": -2.
|
| 716 |
-
"logps/chosen": -
|
| 717 |
-
"logps/rejected": -
|
| 718 |
-
"loss": 0.
|
| 719 |
-
"
|
| 720 |
-
"rewards/
|
| 721 |
-
"rewards/
|
| 722 |
-
"rewards/
|
| 723 |
-
"
|
|
|
|
|
|
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.53,
|
| 727 |
"learning_rate": 2.590221187427241e-07,
|
| 728 |
-
"logits/chosen": -2.
|
| 729 |
-
"logits/rejected": -2.
|
| 730 |
-
"logps/chosen": -
|
| 731 |
-
"logps/rejected": -
|
| 732 |
-
"loss": 0.
|
| 733 |
-
"
|
| 734 |
-
"rewards/
|
| 735 |
-
"rewards/
|
| 736 |
-
"rewards/
|
| 737 |
-
"
|
|
|
|
|
|
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.54,
|
| 741 |
"learning_rate": 2.5320139697322466e-07,
|
| 742 |
-
"logits/chosen": -2.
|
| 743 |
-
"logits/rejected": -2.
|
| 744 |
-
"logps/chosen": -
|
| 745 |
-
"logps/rejected": -
|
| 746 |
-
"loss": 0.
|
| 747 |
-
"
|
| 748 |
-
"rewards/
|
| 749 |
-
"rewards/
|
| 750 |
-
"rewards/
|
| 751 |
-
"
|
|
|
|
|
|
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.55,
|
| 755 |
"learning_rate": 2.4738067520372524e-07,
|
| 756 |
-
"logits/chosen": -2.
|
| 757 |
-
"logits/rejected": -2.
|
| 758 |
-
"logps/chosen": -
|
| 759 |
-
"logps/rejected": -
|
| 760 |
-
"loss": 0.
|
| 761 |
-
"
|
| 762 |
-
"rewards/
|
| 763 |
-
"rewards/
|
| 764 |
-
"rewards/
|
| 765 |
-
"
|
|
|
|
|
|
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.57,
|
| 769 |
"learning_rate": 2.415599534342258e-07,
|
| 770 |
-
"logits/chosen": -2.
|
| 771 |
-
"logits/rejected": -2.
|
| 772 |
-
"logps/chosen": -
|
| 773 |
-
"logps/rejected": -
|
| 774 |
-
"loss": 0.
|
| 775 |
-
"
|
| 776 |
-
"rewards/
|
| 777 |
-
"rewards/
|
| 778 |
-
"rewards/
|
| 779 |
-
"
|
|
|
|
|
|
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.58,
|
| 783 |
"learning_rate": 2.3573923166472642e-07,
|
| 784 |
-
"logits/chosen": -2.
|
| 785 |
-
"logits/rejected": -2.
|
| 786 |
-
"logps/chosen": -
|
| 787 |
-
"logps/rejected": -
|
| 788 |
-
"loss": 0.
|
| 789 |
-
"
|
| 790 |
-
"rewards/
|
| 791 |
-
"rewards/
|
| 792 |
-
"rewards/
|
| 793 |
-
"
|
|
|
|
|
|
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.59,
|
| 797 |
"learning_rate": 2.2991850989522698e-07,
|
| 798 |
-
"logits/chosen": -2.
|
| 799 |
-
"logits/rejected": -2.
|
| 800 |
-
"logps/chosen": -
|
| 801 |
-
"logps/rejected": -
|
| 802 |
-
"loss": 0.
|
| 803 |
-
"
|
| 804 |
-
"rewards/
|
| 805 |
-
"rewards/
|
| 806 |
-
"rewards/
|
| 807 |
-
"
|
|
|
|
|
|
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.6,
|
| 811 |
"learning_rate": 2.2409778812572759e-07,
|
| 812 |
-
"logits/chosen": -2.
|
| 813 |
-
"logits/rejected": -2.
|
| 814 |
-
"logps/chosen": -
|
| 815 |
-
"logps/rejected": -
|
| 816 |
-
"loss": 0.
|
| 817 |
-
"
|
| 818 |
-
"rewards/
|
| 819 |
-
"rewards/
|
| 820 |
-
"rewards/
|
| 821 |
-
"
|
|
|
|
|
|
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.61,
|
| 825 |
"learning_rate": 2.1827706635622817e-07,
|
| 826 |
-
"logits/chosen": -2.
|
| 827 |
-
"logits/rejected": -2.
|
| 828 |
-
"logps/chosen": -
|
| 829 |
-
"logps/rejected": -
|
| 830 |
-
"loss": 0.
|
| 831 |
-
"
|
| 832 |
-
"rewards/
|
| 833 |
-
"rewards/
|
| 834 |
-
"rewards/
|
| 835 |
-
"
|
|
|
|
|
|
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.62,
|
| 839 |
"learning_rate": 2.1245634458672875e-07,
|
| 840 |
-
"logits/chosen": -2.
|
| 841 |
-
"logits/rejected": -2.
|
| 842 |
-
"logps/chosen": -
|
| 843 |
-
"logps/rejected": -
|
| 844 |
-
"loss": 0.
|
| 845 |
-
"
|
| 846 |
-
"rewards/
|
| 847 |
-
"rewards/
|
| 848 |
-
"rewards/
|
| 849 |
-
"
|
|
|
|
|
|
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.63,
|
| 853 |
"learning_rate": 2.0663562281722933e-07,
|
| 854 |
-
"logits/chosen": -2.
|
| 855 |
-
"logits/rejected": -2.
|
| 856 |
-
"logps/chosen": -
|
| 857 |
-
"logps/rejected": -
|
| 858 |
-
"loss": 0.
|
| 859 |
-
"
|
| 860 |
-
"rewards/
|
| 861 |
-
"rewards/
|
| 862 |
-
"rewards/
|
| 863 |
-
"
|
|
|
|
|
|
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.64,
|
| 867 |
"learning_rate": 2.008149010477299e-07,
|
| 868 |
-
"logits/chosen": -2.
|
| 869 |
-
"logits/rejected": -2.
|
| 870 |
-
"logps/chosen": -
|
| 871 |
-
"logps/rejected": -
|
| 872 |
-
"loss": 0.
|
| 873 |
-
"
|
| 874 |
-
"rewards/
|
| 875 |
-
"rewards/
|
| 876 |
-
"rewards/
|
| 877 |
-
"
|
|
|
|
|
|
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.65,
|
| 881 |
"learning_rate": 1.949941792782305e-07,
|
| 882 |
-
"logits/chosen": -2.
|
| 883 |
-
"logits/rejected": -2.
|
| 884 |
-
"logps/chosen": -
|
| 885 |
-
"logps/rejected": -
|
| 886 |
-
"loss": 0.
|
| 887 |
-
"
|
| 888 |
-
"rewards/
|
| 889 |
-
"rewards/
|
| 890 |
-
"rewards/
|
| 891 |
-
"
|
|
|
|
|
|
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.66,
|
| 895 |
"learning_rate": 1.8917345750873107e-07,
|
| 896 |
-
"logits/chosen": -2.
|
| 897 |
-
"logits/rejected": -2.
|
| 898 |
-
"logps/chosen": -
|
| 899 |
-
"logps/rejected": -
|
| 900 |
-
"loss": 0.
|
| 901 |
-
"
|
| 902 |
-
"rewards/
|
| 903 |
-
"rewards/
|
| 904 |
-
"rewards/
|
| 905 |
-
"
|
|
|
|
|
|
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.67,
|
| 909 |
"learning_rate": 1.8335273573923165e-07,
|
| 910 |
-
"logits/chosen": -2.
|
| 911 |
-
"logits/rejected": -2.
|
| 912 |
-
"logps/chosen": -
|
| 913 |
-
"logps/rejected": -
|
| 914 |
-
"loss": 0.
|
| 915 |
-
"
|
| 916 |
-
"rewards/
|
| 917 |
-
"rewards/
|
| 918 |
-
"rewards/
|
| 919 |
-
"
|
|
|
|
|
|
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.68,
|
| 923 |
"learning_rate": 1.7753201396973226e-07,
|
| 924 |
-
"logits/chosen": -2.
|
| 925 |
-
"logits/rejected": -2.
|
| 926 |
-
"logps/chosen": -
|
| 927 |
-
"logps/rejected": -
|
| 928 |
-
"loss": 0.
|
| 929 |
-
"
|
| 930 |
-
"rewards/
|
| 931 |
-
"rewards/
|
| 932 |
-
"rewards/
|
| 933 |
-
"
|
|
|
|
|
|
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.69,
|
| 937 |
"learning_rate": 1.7171129220023281e-07,
|
| 938 |
-
"logits/chosen": -2.
|
| 939 |
-
"logits/rejected": -2.
|
| 940 |
-
"logps/chosen": -
|
| 941 |
-
"logps/rejected": -
|
| 942 |
-
"loss": 0.
|
| 943 |
-
"
|
| 944 |
-
"rewards/
|
| 945 |
-
"rewards/
|
| 946 |
-
"rewards/
|
| 947 |
-
"
|
|
|
|
|
|
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.7,
|
| 951 |
"learning_rate": 1.658905704307334e-07,
|
| 952 |
-
"logits/chosen": -2.
|
| 953 |
-
"logits/rejected": -2.
|
| 954 |
-
"logps/chosen": -
|
| 955 |
-
"logps/rejected": -
|
| 956 |
-
"loss": 0.
|
| 957 |
-
"
|
| 958 |
-
"rewards/
|
| 959 |
-
"rewards/
|
| 960 |
-
"rewards/
|
| 961 |
-
"
|
|
|
|
|
|
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.71,
|
| 965 |
"learning_rate": 1.60069848661234e-07,
|
| 966 |
-
"logits/chosen": -2.
|
| 967 |
-
"logits/rejected": -2.
|
| 968 |
-
"logps/chosen": -
|
| 969 |
-
"logps/rejected": -
|
| 970 |
-
"loss": 0.
|
| 971 |
-
"
|
| 972 |
-
"rewards/
|
| 973 |
-
"rewards/
|
| 974 |
-
"rewards/
|
| 975 |
-
"
|
|
|
|
|
|
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.72,
|
| 979 |
"learning_rate": 1.5424912689173456e-07,
|
| 980 |
-
"logits/chosen": -2.
|
| 981 |
-
"logits/rejected": -2.
|
| 982 |
-
"logps/chosen": -
|
| 983 |
-
"logps/rejected": -
|
| 984 |
-
"loss": 0.
|
| 985 |
-
"
|
| 986 |
-
"rewards/
|
| 987 |
-
"rewards/
|
| 988 |
-
"rewards/
|
| 989 |
-
"
|
|
|
|
|
|
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.73,
|
| 993 |
"learning_rate": 1.4842840512223514e-07,
|
| 994 |
-
"logits/chosen": -2.
|
| 995 |
-
"logits/rejected": -2.
|
| 996 |
-
"logps/chosen": -
|
| 997 |
-
"logps/rejected": -232.
|
| 998 |
-
"loss": 0.
|
| 999 |
-
"
|
| 1000 |
-
"rewards/
|
| 1001 |
-
"rewards/
|
| 1002 |
-
"rewards/
|
| 1003 |
-
"
|
|
|
|
|
|
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.74,
|
| 1007 |
"learning_rate": 1.4260768335273574e-07,
|
| 1008 |
-
"logits/chosen": -2.
|
| 1009 |
-
"logits/rejected": -2.
|
| 1010 |
-
"logps/chosen": -
|
| 1011 |
-
"logps/rejected": -
|
| 1012 |
-
"loss": 0.
|
| 1013 |
-
"
|
| 1014 |
-
"rewards/
|
| 1015 |
-
"rewards/
|
| 1016 |
-
"rewards/
|
| 1017 |
-
"
|
|
|
|
|
|
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.75,
|
| 1021 |
"learning_rate": 1.3678696158323632e-07,
|
| 1022 |
-
"logits/chosen": -2.
|
| 1023 |
-
"logits/rejected": -2.
|
| 1024 |
-
"logps/chosen": -
|
| 1025 |
-
"logps/rejected": -
|
| 1026 |
-
"loss": 0.
|
| 1027 |
-
"
|
| 1028 |
-
"rewards/
|
| 1029 |
-
"rewards/
|
| 1030 |
-
"rewards/
|
| 1031 |
-
"
|
|
|
|
|
|
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.76,
|
| 1035 |
"learning_rate": 1.3096623981373688e-07,
|
| 1036 |
-
"logits/chosen": -2.
|
| 1037 |
-
"logits/rejected": -2.
|
| 1038 |
-
"logps/chosen": -
|
| 1039 |
-
"logps/rejected": -
|
| 1040 |
-
"loss": 0.
|
| 1041 |
-
"
|
| 1042 |
-
"rewards/
|
| 1043 |
-
"rewards/
|
| 1044 |
-
"rewards/
|
| 1045 |
-
"
|
|
|
|
|
|
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.77,
|
| 1049 |
"learning_rate": 1.2514551804423749e-07,
|
| 1050 |
-
"logits/chosen": -2.
|
| 1051 |
-
"logits/rejected": -2.
|
| 1052 |
-
"logps/chosen": -
|
| 1053 |
-
"logps/rejected": -
|
| 1054 |
-
"loss": 0.
|
| 1055 |
-
"
|
| 1056 |
-
"rewards/
|
| 1057 |
-
"rewards/
|
| 1058 |
-
"rewards/
|
| 1059 |
-
"
|
|
|
|
|
|
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.79,
|
| 1063 |
"learning_rate": 1.1932479627473807e-07,
|
| 1064 |
-
"logits/chosen": -2.
|
| 1065 |
-
"logits/rejected": -2.
|
| 1066 |
-
"logps/chosen": -
|
| 1067 |
-
"logps/rejected": -
|
| 1068 |
-
"loss": 0.
|
| 1069 |
-
"
|
| 1070 |
-
"rewards/
|
| 1071 |
-
"rewards/
|
| 1072 |
-
"rewards/
|
| 1073 |
-
"
|
|
|
|
|
|
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.8,
|
| 1077 |
"learning_rate": 1.1350407450523865e-07,
|
| 1078 |
-
"logits/chosen": -2.
|
| 1079 |
-
"logits/rejected": -2.
|
| 1080 |
-
"logps/chosen": -
|
| 1081 |
-
"logps/rejected": -
|
| 1082 |
-
"loss": 0.
|
| 1083 |
-
"
|
| 1084 |
-
"rewards/
|
| 1085 |
-
"rewards/
|
| 1086 |
-
"rewards/
|
| 1087 |
-
"
|
|
|
|
|
|
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.81,
|
| 1091 |
"learning_rate": 1.0768335273573923e-07,
|
| 1092 |
-
"logits/chosen": -2.
|
| 1093 |
-
"logits/rejected": -2.
|
| 1094 |
-
"logps/chosen": -
|
| 1095 |
-
"logps/rejected": -276.
|
| 1096 |
-
"loss": 0.
|
| 1097 |
-
"
|
| 1098 |
-
"rewards/
|
| 1099 |
-
"rewards/
|
| 1100 |
-
"rewards/
|
| 1101 |
-
"
|
|
|
|
|
|
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.82,
|
| 1105 |
"learning_rate": 1.0186263096623981e-07,
|
| 1106 |
-
"logits/chosen": -2.
|
| 1107 |
-
"logits/rejected": -2.
|
| 1108 |
-
"logps/chosen": -
|
| 1109 |
-
"logps/rejected": -
|
| 1110 |
-
"loss": 0.
|
| 1111 |
-
"
|
| 1112 |
-
"rewards/
|
| 1113 |
-
"rewards/
|
| 1114 |
-
"rewards/
|
| 1115 |
-
"
|
|
|
|
|
|
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.83,
|
| 1119 |
"learning_rate": 9.604190919674039e-08,
|
| 1120 |
-
"logits/chosen": -2.
|
| 1121 |
-
"logits/rejected": -2.
|
| 1122 |
-
"logps/chosen": -
|
| 1123 |
-
"logps/rejected": -
|
| 1124 |
-
"loss": 0.
|
| 1125 |
-
"
|
| 1126 |
-
"rewards/
|
| 1127 |
-
"rewards/
|
| 1128 |
-
"rewards/
|
| 1129 |
-
"
|
|
|
|
|
|
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.84,
|
| 1133 |
"learning_rate": 9.022118742724097e-08,
|
| 1134 |
-
"logits/chosen": -2.
|
| 1135 |
-
"logits/rejected": -2.
|
| 1136 |
-
"logps/chosen": -
|
| 1137 |
-
"logps/rejected": -
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"
|
| 1140 |
-
"rewards/
|
| 1141 |
-
"rewards/
|
| 1142 |
-
"rewards/
|
| 1143 |
-
"
|
|
|
|
|
|
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.85,
|
| 1147 |
"learning_rate": 8.440046565774157e-08,
|
| 1148 |
-
"logits/chosen": -2.
|
| 1149 |
-
"logits/rejected": -2.
|
| 1150 |
-
"logps/chosen": -272.
|
| 1151 |
-
"logps/rejected": -
|
| 1152 |
-
"loss": 0.
|
| 1153 |
-
"
|
| 1154 |
-
"rewards/
|
| 1155 |
-
"rewards/
|
| 1156 |
-
"rewards/
|
| 1157 |
-
"
|
|
|
|
|
|
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.86,
|
| 1161 |
"learning_rate": 7.857974388824213e-08,
|
| 1162 |
-
"logits/chosen": -2.
|
| 1163 |
-
"logits/rejected": -2.
|
| 1164 |
-
"logps/chosen": -
|
| 1165 |
-
"logps/rejected": -
|
| 1166 |
-
"loss": 0.
|
| 1167 |
-
"
|
| 1168 |
-
"rewards/
|
| 1169 |
-
"rewards/
|
| 1170 |
-
"rewards/
|
| 1171 |
-
"
|
|
|
|
|
|
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.87,
|
| 1175 |
"learning_rate": 7.275902211874273e-08,
|
| 1176 |
-
"logits/chosen": -2.
|
| 1177 |
-
"logits/rejected": -2.
|
| 1178 |
-
"logps/chosen": -
|
| 1179 |
-
"logps/rejected": -
|
| 1180 |
-
"loss": 0.
|
| 1181 |
-
"
|
| 1182 |
-
"rewards/
|
| 1183 |
-
"rewards/
|
| 1184 |
-
"rewards/
|
| 1185 |
-
"
|
|
|
|
|
|
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.88,
|
| 1189 |
"learning_rate": 6.693830034924331e-08,
|
| 1190 |
-
"logits/chosen": -2.
|
| 1191 |
-
"logits/rejected": -2.
|
| 1192 |
-
"logps/chosen": -
|
| 1193 |
-
"logps/rejected": -
|
| 1194 |
-
"loss": 0.
|
| 1195 |
-
"
|
| 1196 |
-
"rewards/
|
| 1197 |
-
"rewards/
|
| 1198 |
-
"rewards/
|
| 1199 |
-
"
|
|
|
|
|
|
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.89,
|
| 1203 |
"learning_rate": 6.111757857974389e-08,
|
| 1204 |
-
"logits/chosen": -2.
|
| 1205 |
-
"logits/rejected": -2.
|
| 1206 |
-
"logps/chosen": -
|
| 1207 |
-
"logps/rejected": -
|
| 1208 |
-
"loss": 0.
|
| 1209 |
-
"
|
| 1210 |
-
"rewards/
|
| 1211 |
-
"rewards/
|
| 1212 |
-
"rewards/
|
| 1213 |
-
"
|
|
|
|
|
|
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.9,
|
| 1217 |
"learning_rate": 5.529685681024446e-08,
|
| 1218 |
-
"logits/chosen": -2.
|
| 1219 |
-
"logits/rejected": -2.
|
| 1220 |
-
"logps/chosen": -
|
| 1221 |
-
"logps/rejected": -
|
| 1222 |
-
"loss": 0.
|
| 1223 |
-
"
|
| 1224 |
-
"rewards/
|
| 1225 |
-
"rewards/
|
| 1226 |
-
"rewards/
|
| 1227 |
-
"
|
|
|
|
|
|
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.91,
|
| 1231 |
"learning_rate": 4.947613504074505e-08,
|
| 1232 |
-
"logits/chosen": -2.
|
| 1233 |
-
"logits/rejected": -2.
|
| 1234 |
-
"logps/chosen": -
|
| 1235 |
-
"logps/rejected": -
|
| 1236 |
-
"loss": 0.
|
| 1237 |
-
"
|
| 1238 |
-
"rewards/
|
| 1239 |
-
"rewards/
|
| 1240 |
-
"rewards/
|
| 1241 |
-
"
|
|
|
|
|
|
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.92,
|
| 1245 |
"learning_rate": 4.365541327124563e-08,
|
| 1246 |
-
"logits/chosen": -2.
|
| 1247 |
-
"logits/rejected": -2.
|
| 1248 |
-
"logps/chosen": -
|
| 1249 |
-
"logps/rejected": -
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"
|
| 1252 |
-
"rewards/
|
| 1253 |
-
"rewards/
|
| 1254 |
-
"rewards/
|
| 1255 |
-
"
|
|
|
|
|
|
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.93,
|
| 1259 |
"learning_rate": 3.783469150174622e-08,
|
| 1260 |
-
"logits/chosen": -2.
|
| 1261 |
-
"logits/rejected": -2.
|
| 1262 |
-
"logps/chosen": -
|
| 1263 |
-
"logps/rejected": -
|
| 1264 |
-
"loss": 0.
|
| 1265 |
-
"
|
| 1266 |
-
"rewards/
|
| 1267 |
-
"rewards/
|
| 1268 |
-
"rewards/
|
| 1269 |
-
"
|
|
|
|
|
|
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.94,
|
| 1273 |
"learning_rate": 3.20139697322468e-08,
|
| 1274 |
-
"logits/chosen": -2.
|
| 1275 |
-
"logits/rejected": -2.
|
| 1276 |
-
"logps/chosen": -264.
|
| 1277 |
-
"logps/rejected": -
|
| 1278 |
-
"loss": 0.
|
| 1279 |
-
"
|
| 1280 |
-
"rewards/
|
| 1281 |
-
"rewards/
|
| 1282 |
-
"rewards/
|
| 1283 |
-
"
|
|
|
|
|
|
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.95,
|
| 1287 |
"learning_rate": 2.619324796274738e-08,
|
| 1288 |
-
"logits/chosen": -2.
|
| 1289 |
-
"logits/rejected": -2.
|
| 1290 |
-
"logps/chosen": -
|
| 1291 |
-
"logps/rejected": -
|
| 1292 |
-
"loss": 0.
|
| 1293 |
-
"
|
| 1294 |
-
"rewards/
|
| 1295 |
-
"rewards/
|
| 1296 |
-
"rewards/
|
| 1297 |
-
"
|
|
|
|
|
|
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.96,
|
| 1301 |
"learning_rate": 2.037252619324796e-08,
|
| 1302 |
-
"logits/chosen": -2.
|
| 1303 |
-
"logits/rejected": -2.
|
| 1304 |
-
"logps/chosen": -
|
| 1305 |
-
"logps/rejected": -
|
| 1306 |
-
"loss": 0.
|
| 1307 |
-
"
|
| 1308 |
-
"rewards/
|
| 1309 |
-
"rewards/
|
| 1310 |
-
"rewards/
|
| 1311 |
-
"
|
|
|
|
|
|
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.97,
|
| 1315 |
"learning_rate": 1.4551804423748545e-08,
|
| 1316 |
-
"logits/chosen": -2.
|
| 1317 |
-
"logits/rejected": -2.
|
| 1318 |
-
"logps/chosen": -
|
| 1319 |
-
"logps/rejected": -
|
| 1320 |
-
"loss": 0.
|
| 1321 |
-
"
|
| 1322 |
-
"rewards/
|
| 1323 |
-
"rewards/
|
| 1324 |
-
"rewards/
|
| 1325 |
-
"
|
|
|
|
|
|
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.98,
|
| 1329 |
"learning_rate": 8.731082654249125e-09,
|
| 1330 |
-
"logits/chosen": -2.
|
| 1331 |
-
"logits/rejected": -2.
|
| 1332 |
-
"logps/chosen": -
|
| 1333 |
-
"logps/rejected": -
|
| 1334 |
-
"loss": 0.
|
| 1335 |
-
"
|
| 1336 |
-
"rewards/
|
| 1337 |
-
"rewards/
|
| 1338 |
-
"rewards/
|
| 1339 |
-
"
|
|
|
|
|
|
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 0.99,
|
| 1343 |
"learning_rate": 2.910360884749709e-09,
|
| 1344 |
-
"logits/chosen": -2.
|
| 1345 |
-
"logits/rejected": -2.
|
| 1346 |
-
"logps/chosen": -
|
| 1347 |
-
"logps/rejected": -
|
| 1348 |
-
"loss": 0.
|
| 1349 |
-
"
|
| 1350 |
-
"rewards/
|
| 1351 |
-
"rewards/
|
| 1352 |
-
"rewards/
|
| 1353 |
-
"
|
|
|
|
|
|
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 1.0,
|
| 1357 |
-
"eval_logits/chosen": -2.
|
| 1358 |
-
"eval_logits/rejected": -2.
|
| 1359 |
-
"eval_logps/chosen": -
|
| 1360 |
-
"eval_logps/rejected": -
|
| 1361 |
-
"eval_loss": 0.
|
| 1362 |
-
"
|
| 1363 |
-
"eval_rewards/
|
| 1364 |
-
"eval_rewards/
|
| 1365 |
-
"eval_rewards/
|
| 1366 |
-
"
|
| 1367 |
-
"
|
| 1368 |
-
"
|
|
|
|
|
|
|
| 1369 |
"step": 955
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 1.0,
|
| 1373 |
"step": 955,
|
| 1374 |
"total_flos": 0.0,
|
| 1375 |
-
"train_loss": 0.
|
| 1376 |
-
"train_runtime":
|
| 1377 |
-
"train_samples_per_second":
|
| 1378 |
-
"train_steps_per_second": 0.
|
| 1379 |
}
|
| 1380 |
],
|
| 1381 |
"logging_steps": 10,
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
"learning_rate": 5.208333333333333e-09,
|
| 14 |
+
"logits/chosen": -2.676934003829956,
|
| 15 |
+
"logits/rejected": -2.509021043777466,
|
| 16 |
+
"logps/chosen": -304.709228515625,
|
| 17 |
+
"logps/rejected": -229.49505615234375,
|
| 18 |
"loss": 0.6931,
|
| 19 |
+
"pred_label": 0.0,
|
| 20 |
"rewards/accuracies": 0.0,
|
| 21 |
"rewards/chosen": 0.0,
|
| 22 |
"rewards/margins": 0.0,
|
| 23 |
"rewards/rejected": 0.0,
|
| 24 |
+
"step": 1,
|
| 25 |
+
"use_label": 10.0
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.01,
|
| 29 |
"learning_rate": 5.208333333333333e-08,
|
| 30 |
+
"logits/chosen": -2.5309348106384277,
|
| 31 |
+
"logits/rejected": -2.5985612869262695,
|
| 32 |
+
"logps/chosen": -313.4403381347656,
|
| 33 |
+
"logps/rejected": -277.2276611328125,
|
| 34 |
+
"loss": 0.6812,
|
| 35 |
+
"pred_label": 18.55555534362793,
|
| 36 |
+
"rewards/accuracies": 0.4861111044883728,
|
| 37 |
+
"rewards/chosen": 0.0022232607007026672,
|
| 38 |
+
"rewards/margins": -7.403641939163208e-05,
|
| 39 |
+
"rewards/rejected": 0.0022972968872636557,
|
| 40 |
+
"step": 10,
|
| 41 |
+
"use_label": 71.44444274902344
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 0.02,
|
| 45 |
"learning_rate": 1.0416666666666667e-07,
|
| 46 |
+
"logits/chosen": -2.390097141265869,
|
| 47 |
+
"logits/rejected": -2.4532761573791504,
|
| 48 |
+
"logps/chosen": -229.01559448242188,
|
| 49 |
+
"logps/rejected": -232.5874786376953,
|
| 50 |
+
"loss": 0.6791,
|
| 51 |
+
"pred_label": 61.54999923706055,
|
| 52 |
+
"rewards/accuracies": 0.53125,
|
| 53 |
+
"rewards/chosen": -0.001973286736756563,
|
| 54 |
+
"rewards/margins": -0.0003373834188096225,
|
| 55 |
+
"rewards/rejected": -0.0016359034925699234,
|
| 56 |
+
"step": 20,
|
| 57 |
+
"use_label": 180.4499969482422
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.03,
|
| 61 |
"learning_rate": 1.5624999999999999e-07,
|
| 62 |
+
"logits/chosen": -2.529320001602173,
|
| 63 |
+
"logits/rejected": -2.4789929389953613,
|
| 64 |
+
"logps/chosen": -270.6199951171875,
|
| 65 |
+
"logps/rejected": -244.80044555664062,
|
| 66 |
+
"loss": 0.6782,
|
| 67 |
+
"pred_label": 106.80000305175781,
|
| 68 |
+
"rewards/accuracies": 0.48124998807907104,
|
| 69 |
+
"rewards/chosen": -0.006186917424201965,
|
| 70 |
+
"rewards/margins": -0.005202265456318855,
|
| 71 |
+
"rewards/rejected": -0.0009846522007137537,
|
| 72 |
+
"step": 30,
|
| 73 |
+
"use_label": 295.20001220703125
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.04,
|
| 77 |
"learning_rate": 2.0833333333333333e-07,
|
| 78 |
+
"logits/chosen": -2.4986624717712402,
|
| 79 |
+
"logits/rejected": -2.5215377807617188,
|
| 80 |
+
"logps/chosen": -270.62274169921875,
|
| 81 |
+
"logps/rejected": -263.65087890625,
|
| 82 |
+
"loss": 0.6789,
|
| 83 |
+
"pred_label": 156.22500610351562,
|
| 84 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 85 |
+
"rewards/chosen": -0.0010104719549417496,
|
| 86 |
+
"rewards/margins": 0.0030316715128719807,
|
| 87 |
+
"rewards/rejected": -0.004042143002152443,
|
| 88 |
+
"step": 40,
|
| 89 |
+
"use_label": 405.7749938964844
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 0.05,
|
| 93 |
"learning_rate": 2.604166666666667e-07,
|
| 94 |
+
"logits/chosen": -2.531646490097046,
|
| 95 |
+
"logits/rejected": -2.5199639797210693,
|
| 96 |
+
"logps/chosen": -255.382080078125,
|
| 97 |
+
"logps/rejected": -249.44775390625,
|
| 98 |
+
"loss": 0.6778,
|
| 99 |
+
"pred_label": 213.4250030517578,
|
| 100 |
+
"rewards/accuracies": 0.4937500059604645,
|
| 101 |
+
"rewards/chosen": -0.0030100971926003695,
|
| 102 |
+
"rewards/margins": -0.00096442288486287,
|
| 103 |
+
"rewards/rejected": -0.0020456742495298386,
|
| 104 |
+
"step": 50,
|
| 105 |
+
"use_label": 508.57501220703125
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.06,
|
| 109 |
"learning_rate": 3.1249999999999997e-07,
|
| 110 |
+
"logits/chosen": -2.4867427349090576,
|
| 111 |
+
"logits/rejected": -2.473141670227051,
|
| 112 |
+
"logps/chosen": -278.27655029296875,
|
| 113 |
+
"logps/rejected": -257.5716247558594,
|
| 114 |
+
"loss": 0.6784,
|
| 115 |
+
"pred_label": 266.17498779296875,
|
| 116 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 117 |
+
"rewards/chosen": -0.0048277489840984344,
|
| 118 |
+
"rewards/margins": -0.004098966252058744,
|
| 119 |
+
"rewards/rejected": -0.0007287820335477591,
|
| 120 |
+
"step": 60,
|
| 121 |
+
"use_label": 615.8250122070312
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07,
|
| 125 |
"learning_rate": 3.645833333333333e-07,
|
| 126 |
+
"logits/chosen": -2.4988815784454346,
|
| 127 |
+
"logits/rejected": -2.492649793624878,
|
| 128 |
+
"logps/chosen": -285.4491271972656,
|
| 129 |
+
"logps/rejected": -263.70635986328125,
|
| 130 |
+
"loss": 0.6793,
|
| 131 |
+
"pred_label": 313.125,
|
| 132 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 133 |
+
"rewards/chosen": -0.004624291323125362,
|
| 134 |
+
"rewards/margins": -0.003086260985583067,
|
| 135 |
+
"rewards/rejected": -0.0015380297554656863,
|
| 136 |
+
"step": 70,
|
| 137 |
+
"use_label": 728.875
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 0.08,
|
| 141 |
"learning_rate": 4.1666666666666667e-07,
|
| 142 |
+
"logits/chosen": -2.447936534881592,
|
| 143 |
+
"logits/rejected": -2.440683126449585,
|
| 144 |
+
"logps/chosen": -292.83795166015625,
|
| 145 |
+
"logps/rejected": -262.8486328125,
|
| 146 |
+
"loss": 0.6787,
|
| 147 |
+
"pred_label": 365.5,
|
| 148 |
+
"rewards/accuracies": 0.53125,
|
| 149 |
+
"rewards/chosen": -0.00346105033531785,
|
| 150 |
+
"rewards/margins": -0.00289200060069561,
|
| 151 |
+
"rewards/rejected": -0.0005690503749065101,
|
| 152 |
+
"step": 80,
|
| 153 |
+
"use_label": 836.5
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.09,
|
| 157 |
"learning_rate": 4.6874999999999996e-07,
|
| 158 |
+
"logits/chosen": -2.410609006881714,
|
| 159 |
+
"logits/rejected": -2.4672443866729736,
|
| 160 |
+
"logps/chosen": -281.00421142578125,
|
| 161 |
+
"logps/rejected": -256.7078552246094,
|
| 162 |
+
"loss": 0.6794,
|
| 163 |
+
"pred_label": 423.875,
|
| 164 |
+
"rewards/accuracies": 0.512499988079071,
|
| 165 |
+
"rewards/chosen": 0.0006400573183782399,
|
| 166 |
+
"rewards/margins": -0.0020913761109113693,
|
| 167 |
+
"rewards/rejected": 0.002731433603912592,
|
| 168 |
+
"step": 90,
|
| 169 |
+
"use_label": 938.125
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 0.1,
|
| 173 |
"learning_rate": 4.976717112922002e-07,
|
| 174 |
+
"logits/chosen": -2.4385340213775635,
|
| 175 |
+
"logits/rejected": -2.401470899581909,
|
| 176 |
+
"logps/chosen": -263.50323486328125,
|
| 177 |
+
"logps/rejected": -282.2510070800781,
|
| 178 |
+
"loss": 0.6781,
|
| 179 |
+
"pred_label": 479.6000061035156,
|
| 180 |
+
"rewards/accuracies": 0.550000011920929,
|
| 181 |
+
"rewards/chosen": 0.007567421533167362,
|
| 182 |
+
"rewards/margins": 0.010446270927786827,
|
| 183 |
+
"rewards/rejected": -0.002878849394619465,
|
| 184 |
+
"step": 100,
|
| 185 |
+
"use_label": 1042.4000244140625
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.12,
|
| 189 |
"learning_rate": 4.918509895227007e-07,
|
| 190 |
+
"logits/chosen": -2.4747567176818848,
|
| 191 |
+
"logits/rejected": -2.4313316345214844,
|
| 192 |
+
"logps/chosen": -255.5030975341797,
|
| 193 |
+
"logps/rejected": -241.1230926513672,
|
| 194 |
+
"loss": 0.6776,
|
| 195 |
+
"pred_label": 530.625,
|
| 196 |
+
"rewards/accuracies": 0.5,
|
| 197 |
+
"rewards/chosen": 0.0008927445742301643,
|
| 198 |
+
"rewards/margins": 0.001805878826417029,
|
| 199 |
+
"rewards/rejected": -0.0009131338447332382,
|
| 200 |
+
"step": 110,
|
| 201 |
+
"use_label": 1151.375
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 0.13,
|
| 205 |
"learning_rate": 4.860302677532014e-07,
|
| 206 |
+
"logits/chosen": -2.432814836502075,
|
| 207 |
+
"logits/rejected": -2.4882383346557617,
|
| 208 |
+
"logps/chosen": -266.65386962890625,
|
| 209 |
+
"logps/rejected": -237.3206787109375,
|
| 210 |
+
"loss": 0.6794,
|
| 211 |
+
"pred_label": 576.875,
|
| 212 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 213 |
+
"rewards/chosen": 0.008614806458353996,
|
| 214 |
+
"rewards/margins": 0.01159285381436348,
|
| 215 |
+
"rewards/rejected": -0.0029780478216707706,
|
| 216 |
+
"step": 120,
|
| 217 |
+
"use_label": 1265.125
|
| 218 |
},
|
| 219 |
{
|
| 220 |
"epoch": 0.14,
|
| 221 |
"learning_rate": 4.802095459837019e-07,
|
| 222 |
+
"logits/chosen": -2.498955249786377,
|
| 223 |
+
"logits/rejected": -2.49072003364563,
|
| 224 |
+
"logps/chosen": -295.6458435058594,
|
| 225 |
+
"logps/rejected": -263.4668273925781,
|
| 226 |
+
"loss": 0.6774,
|
| 227 |
+
"pred_label": 627.4749755859375,
|
| 228 |
+
"rewards/accuracies": 0.5062500238418579,
|
| 229 |
+
"rewards/chosen": -0.00027961478917859495,
|
| 230 |
+
"rewards/margins": 0.0010693027870729566,
|
| 231 |
+
"rewards/rejected": -0.001348917605355382,
|
| 232 |
+
"step": 130,
|
| 233 |
+
"use_label": 1374.5250244140625
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.15,
|
| 237 |
"learning_rate": 4.743888242142026e-07,
|
| 238 |
+
"logits/chosen": -2.50268292427063,
|
| 239 |
+
"logits/rejected": -2.506229877471924,
|
| 240 |
+
"logps/chosen": -249.8160400390625,
|
| 241 |
+
"logps/rejected": -228.35537719726562,
|
| 242 |
+
"loss": 0.6792,
|
| 243 |
+
"pred_label": 681.7999877929688,
|
| 244 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 245 |
+
"rewards/chosen": 0.0006030676886439323,
|
| 246 |
+
"rewards/margins": -0.00023779459297657013,
|
| 247 |
+
"rewards/rejected": 0.0008408633293583989,
|
| 248 |
+
"step": 140,
|
| 249 |
+
"use_label": 1480.199951171875
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.16,
|
| 253 |
"learning_rate": 4.685681024447031e-07,
|
| 254 |
+
"logits/chosen": -2.430985689163208,
|
| 255 |
+
"logits/rejected": -2.438312292098999,
|
| 256 |
+
"logps/chosen": -282.99285888671875,
|
| 257 |
+
"logps/rejected": -260.1983947753906,
|
| 258 |
+
"loss": 0.6764,
|
| 259 |
+
"pred_label": 730.6749877929688,
|
| 260 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 261 |
+
"rewards/chosen": 0.007840326987206936,
|
| 262 |
+
"rewards/margins": 0.008648511953651905,
|
| 263 |
+
"rewards/rejected": -0.0008081849664449692,
|
| 264 |
+
"step": 150,
|
| 265 |
+
"use_label": 1591.324951171875
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"epoch": 0.17,
|
| 269 |
"learning_rate": 4.627473806752037e-07,
|
| 270 |
+
"logits/chosen": -2.513338565826416,
|
| 271 |
+
"logits/rejected": -2.4252638816833496,
|
| 272 |
+
"logps/chosen": -251.98593139648438,
|
| 273 |
+
"logps/rejected": -258.3878479003906,
|
| 274 |
+
"loss": 0.6779,
|
| 275 |
+
"pred_label": 785.4500122070312,
|
| 276 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 277 |
+
"rewards/chosen": 0.009579241275787354,
|
| 278 |
+
"rewards/margins": 0.013932722620666027,
|
| 279 |
+
"rewards/rejected": -0.004353481810539961,
|
| 280 |
+
"step": 160,
|
| 281 |
+
"use_label": 1696.550048828125
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 0.18,
|
| 285 |
"learning_rate": 4.5692665890570433e-07,
|
| 286 |
+
"logits/chosen": -2.51324725151062,
|
| 287 |
+
"logits/rejected": -2.5183651447296143,
|
| 288 |
+
"logps/chosen": -286.78240966796875,
|
| 289 |
+
"logps/rejected": -242.28451538085938,
|
| 290 |
+
"loss": 0.6781,
|
| 291 |
+
"pred_label": 837.375,
|
| 292 |
+
"rewards/accuracies": 0.625,
|
| 293 |
+
"rewards/chosen": 0.010844933800399303,
|
| 294 |
+
"rewards/margins": 0.014444952830672264,
|
| 295 |
+
"rewards/rejected": -0.003600016701966524,
|
| 296 |
+
"step": 170,
|
| 297 |
+
"use_label": 1804.625
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.19,
|
| 301 |
"learning_rate": 4.5110593713620486e-07,
|
| 302 |
+
"logits/chosen": -2.4405276775360107,
|
| 303 |
+
"logits/rejected": -2.418701410293579,
|
| 304 |
+
"logps/chosen": -245.29843139648438,
|
| 305 |
+
"logps/rejected": -261.94500732421875,
|
| 306 |
+
"loss": 0.6779,
|
| 307 |
+
"pred_label": 900.5250244140625,
|
| 308 |
+
"rewards/accuracies": 0.512499988079071,
|
| 309 |
+
"rewards/chosen": -9.365100413560867e-05,
|
| 310 |
+
"rewards/margins": 0.0037569026462733746,
|
| 311 |
+
"rewards/rejected": -0.003850553184747696,
|
| 312 |
+
"step": 180,
|
| 313 |
+
"use_label": 1901.4749755859375
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"epoch": 0.2,
|
| 317 |
"learning_rate": 4.4528521536670544e-07,
|
| 318 |
+
"logits/chosen": -2.549407482147217,
|
| 319 |
+
"logits/rejected": -2.5117835998535156,
|
| 320 |
+
"logps/chosen": -291.4512939453125,
|
| 321 |
+
"logps/rejected": -250.8589324951172,
|
| 322 |
+
"loss": 0.6749,
|
| 323 |
+
"pred_label": 960.4500122070312,
|
| 324 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 325 |
+
"rewards/chosen": 0.0050302534364163876,
|
| 326 |
+
"rewards/margins": 0.01830260455608368,
|
| 327 |
+
"rewards/rejected": -0.013272350654006004,
|
| 328 |
+
"step": 190,
|
| 329 |
+
"use_label": 2001.550048828125
|
| 330 |
},
|
| 331 |
{
|
| 332 |
"epoch": 0.21,
|
| 333 |
"learning_rate": 4.3946449359720607e-07,
|
| 334 |
+
"logits/chosen": -2.521691083908081,
|
| 335 |
+
"logits/rejected": -2.522428035736084,
|
| 336 |
+
"logps/chosen": -251.04867553710938,
|
| 337 |
+
"logps/rejected": -259.01055908203125,
|
| 338 |
+
"loss": 0.6754,
|
| 339 |
+
"pred_label": 1024.925048828125,
|
| 340 |
+
"rewards/accuracies": 0.512499988079071,
|
| 341 |
+
"rewards/chosen": 0.004073199350386858,
|
| 342 |
+
"rewards/margins": 0.008589169010519981,
|
| 343 |
+
"rewards/rejected": -0.004515970591455698,
|
| 344 |
+
"step": 200,
|
| 345 |
+
"use_label": 2097.074951171875
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.22,
|
| 349 |
"learning_rate": 4.336437718277066e-07,
|
| 350 |
+
"logits/chosen": -2.5071074962615967,
|
| 351 |
+
"logits/rejected": -2.5461485385894775,
|
| 352 |
+
"logps/chosen": -288.81842041015625,
|
| 353 |
+
"logps/rejected": -246.823486328125,
|
| 354 |
+
"loss": 0.6752,
|
| 355 |
+
"pred_label": 1087.875,
|
| 356 |
+
"rewards/accuracies": 0.53125,
|
| 357 |
+
"rewards/chosen": 0.011537188664078712,
|
| 358 |
+
"rewards/margins": 0.013686036691069603,
|
| 359 |
+
"rewards/rejected": -0.0021488501224666834,
|
| 360 |
+
"step": 210,
|
| 361 |
+
"use_label": 2194.125
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.23,
|
| 365 |
"learning_rate": 4.278230500582072e-07,
|
| 366 |
+
"logits/chosen": -2.4844157695770264,
|
| 367 |
+
"logits/rejected": -2.4883625507354736,
|
| 368 |
+
"logps/chosen": -263.35015869140625,
|
| 369 |
+
"logps/rejected": -261.3338623046875,
|
| 370 |
+
"loss": 0.6738,
|
| 371 |
+
"pred_label": 1153.574951171875,
|
| 372 |
+
"rewards/accuracies": 0.612500011920929,
|
| 373 |
+
"rewards/chosen": 0.0034189000725746155,
|
| 374 |
+
"rewards/margins": 0.01800468936562538,
|
| 375 |
+
"rewards/rejected": -0.014585788361728191,
|
| 376 |
+
"step": 220,
|
| 377 |
+
"use_label": 2288.425048828125
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.24,
|
| 381 |
"learning_rate": 4.220023282887078e-07,
|
| 382 |
+
"logits/chosen": -2.4569809436798096,
|
| 383 |
+
"logits/rejected": -2.4612298011779785,
|
| 384 |
+
"logps/chosen": -294.7366943359375,
|
| 385 |
+
"logps/rejected": -254.261962890625,
|
| 386 |
+
"loss": 0.672,
|
| 387 |
+
"pred_label": 1221.175048828125,
|
| 388 |
+
"rewards/accuracies": 0.581250011920929,
|
| 389 |
+
"rewards/chosen": 0.01110165473073721,
|
| 390 |
+
"rewards/margins": 0.014376315288245678,
|
| 391 |
+
"rewards/rejected": -0.0032746598590165377,
|
| 392 |
+
"step": 230,
|
| 393 |
+
"use_label": 2380.824951171875
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 0.25,
|
| 397 |
"learning_rate": 4.1618160651920834e-07,
|
| 398 |
+
"logits/chosen": -2.440027952194214,
|
| 399 |
+
"logits/rejected": -2.415015935897827,
|
| 400 |
+
"logps/chosen": -257.82342529296875,
|
| 401 |
+
"logps/rejected": -215.18408203125,
|
| 402 |
+
"loss": 0.6712,
|
| 403 |
+
"pred_label": 1294.2750244140625,
|
| 404 |
+
"rewards/accuracies": 0.5625,
|
| 405 |
+
"rewards/chosen": 0.011741789057850838,
|
| 406 |
+
"rewards/margins": 0.014540466479957104,
|
| 407 |
+
"rewards/rejected": -0.0027986769564449787,
|
| 408 |
+
"step": 240,
|
| 409 |
+
"use_label": 2467.72509765625
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.26,
|
| 413 |
"learning_rate": 4.103608847497089e-07,
|
| 414 |
+
"logits/chosen": -2.4385263919830322,
|
| 415 |
+
"logits/rejected": -2.4420063495635986,
|
| 416 |
+
"logps/chosen": -277.59747314453125,
|
| 417 |
+
"logps/rejected": -258.9275817871094,
|
| 418 |
+
"loss": 0.672,
|
| 419 |
+
"pred_label": 1364.699951171875,
|
| 420 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 421 |
+
"rewards/chosen": 0.01612050086259842,
|
| 422 |
+
"rewards/margins": 0.02269991859793663,
|
| 423 |
+
"rewards/rejected": -0.0065794168040156364,
|
| 424 |
+
"step": 250,
|
| 425 |
+
"use_label": 2557.300048828125
|
| 426 |
},
|
| 427 |
{
|
| 428 |
"epoch": 0.27,
|
| 429 |
"learning_rate": 4.0454016298020956e-07,
|
| 430 |
+
"logits/chosen": -2.42997407913208,
|
| 431 |
+
"logits/rejected": -2.424861431121826,
|
| 432 |
+
"logps/chosen": -239.51095581054688,
|
| 433 |
+
"logps/rejected": -240.63623046875,
|
| 434 |
+
"loss": 0.6707,
|
| 435 |
+
"pred_label": 1427.824951171875,
|
| 436 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 437 |
+
"rewards/chosen": 0.000991971348412335,
|
| 438 |
+
"rewards/margins": 0.004964248277246952,
|
| 439 |
+
"rewards/rejected": -0.003972277976572514,
|
| 440 |
+
"step": 260,
|
| 441 |
+
"use_label": 2654.175048828125
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 0.28,
|
| 445 |
"learning_rate": 3.987194412107101e-07,
|
| 446 |
+
"logits/chosen": -2.4381518363952637,
|
| 447 |
+
"logits/rejected": -2.46921706199646,
|
| 448 |
+
"logps/chosen": -280.505859375,
|
| 449 |
+
"logps/rejected": -266.14080810546875,
|
| 450 |
"loss": 0.6735,
|
| 451 |
+
"pred_label": 1494.574951171875,
|
| 452 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 453 |
+
"rewards/chosen": 0.01380440779030323,
|
| 454 |
+
"rewards/margins": 0.023902520537376404,
|
| 455 |
+
"rewards/rejected": -0.010098112747073174,
|
| 456 |
+
"step": 270,
|
| 457 |
+
"use_label": 2747.425048828125
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.29,
|
| 461 |
"learning_rate": 3.9289871944121066e-07,
|
| 462 |
+
"logits/chosen": -2.527064800262451,
|
| 463 |
+
"logits/rejected": -2.552640438079834,
|
| 464 |
+
"logps/chosen": -284.7080078125,
|
| 465 |
+
"logps/rejected": -255.3748016357422,
|
| 466 |
+
"loss": 0.6706,
|
| 467 |
+
"pred_label": 1554.8499755859375,
|
| 468 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 469 |
+
"rewards/chosen": 0.0216833408921957,
|
| 470 |
+
"rewards/margins": 0.026800300925970078,
|
| 471 |
+
"rewards/rejected": -0.005116959102451801,
|
| 472 |
+
"step": 280,
|
| 473 |
+
"use_label": 2847.14990234375
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 0.3,
|
| 477 |
"learning_rate": 3.870779976717113e-07,
|
| 478 |
+
"logits/chosen": -2.5222768783569336,
|
| 479 |
+
"logits/rejected": -2.508826732635498,
|
| 480 |
+
"logps/chosen": -278.7129821777344,
|
| 481 |
+
"logps/rejected": -220.00210571289062,
|
| 482 |
+
"loss": 0.669,
|
| 483 |
+
"pred_label": 1625.199951171875,
|
| 484 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 485 |
+
"rewards/chosen": 0.004253073129802942,
|
| 486 |
+
"rewards/margins": 0.014592866413295269,
|
| 487 |
+
"rewards/rejected": -0.010339794680476189,
|
| 488 |
+
"step": 290,
|
| 489 |
+
"use_label": 2936.800048828125
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.31,
|
| 493 |
"learning_rate": 3.812572759022118e-07,
|
| 494 |
+
"logits/chosen": -2.4862565994262695,
|
| 495 |
+
"logits/rejected": -2.4585397243499756,
|
| 496 |
+
"logps/chosen": -254.63589477539062,
|
| 497 |
+
"logps/rejected": -230.22238159179688,
|
| 498 |
+
"loss": 0.669,
|
| 499 |
+
"pred_label": 1695.0250244140625,
|
| 500 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 501 |
+
"rewards/chosen": 0.013738395646214485,
|
| 502 |
+
"rewards/margins": 0.030270254239439964,
|
| 503 |
+
"rewards/rejected": -0.01653185673058033,
|
| 504 |
+
"step": 300,
|
| 505 |
+
"use_label": 3026.97509765625
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.32,
|
| 509 |
"learning_rate": 3.754365541327124e-07,
|
| 510 |
+
"logits/chosen": -2.490318775177002,
|
| 511 |
+
"logits/rejected": -2.411966323852539,
|
| 512 |
+
"logps/chosen": -255.95016479492188,
|
| 513 |
+
"logps/rejected": -278.394775390625,
|
| 514 |
+
"loss": 0.6677,
|
| 515 |
+
"pred_label": 1769.550048828125,
|
| 516 |
+
"rewards/accuracies": 0.612500011920929,
|
| 517 |
+
"rewards/chosen": 0.011918948963284492,
|
| 518 |
+
"rewards/margins": 0.024249419569969177,
|
| 519 |
+
"rewards/rejected": -0.012330473400652409,
|
| 520 |
+
"step": 310,
|
| 521 |
+
"use_label": 3112.449951171875
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.33,
|
| 525 |
"learning_rate": 3.6961583236321304e-07,
|
| 526 |
+
"logits/chosen": -2.4655749797821045,
|
| 527 |
+
"logits/rejected": -2.429624080657959,
|
| 528 |
+
"logps/chosen": -279.8106384277344,
|
| 529 |
+
"logps/rejected": -272.11993408203125,
|
| 530 |
+
"loss": 0.6648,
|
| 531 |
+
"pred_label": 1836.0,
|
| 532 |
+
"rewards/accuracies": 0.606249988079071,
|
| 533 |
+
"rewards/chosen": 0.012938129715621471,
|
| 534 |
+
"rewards/margins": 0.027768433094024658,
|
| 535 |
+
"rewards/rejected": -0.014830301515758038,
|
| 536 |
+
"step": 320,
|
| 537 |
+
"use_label": 3206.0
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.35,
|
| 541 |
"learning_rate": 3.637951105937136e-07,
|
| 542 |
+
"logits/chosen": -2.494778633117676,
|
| 543 |
+
"logits/rejected": -2.445328712463379,
|
| 544 |
+
"logps/chosen": -258.9544982910156,
|
| 545 |
+
"logps/rejected": -226.5720977783203,
|
| 546 |
+
"loss": 0.6674,
|
| 547 |
+
"pred_label": 1910.1500244140625,
|
| 548 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 549 |
+
"rewards/chosen": 0.00947931781411171,
|
| 550 |
+
"rewards/margins": 0.01889108493924141,
|
| 551 |
+
"rewards/rejected": -0.009411768987774849,
|
| 552 |
+
"step": 330,
|
| 553 |
+
"use_label": 3291.85009765625
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"epoch": 0.36,
|
| 557 |
"learning_rate": 3.579743888242142e-07,
|
| 558 |
+
"logits/chosen": -2.44804048538208,
|
| 559 |
+
"logits/rejected": -2.4573702812194824,
|
| 560 |
+
"logps/chosen": -267.03192138671875,
|
| 561 |
+
"logps/rejected": -228.086181640625,
|
| 562 |
+
"loss": 0.6657,
|
| 563 |
+
"pred_label": 1974.324951171875,
|
| 564 |
+
"rewards/accuracies": 0.581250011920929,
|
| 565 |
+
"rewards/chosen": 0.018211424350738525,
|
| 566 |
+
"rewards/margins": 0.021834325045347214,
|
| 567 |
+
"rewards/rejected": -0.0036229020915925503,
|
| 568 |
+
"step": 340,
|
| 569 |
+
"use_label": 3387.675048828125
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.37,
|
| 573 |
"learning_rate": 3.521536670547148e-07,
|
| 574 |
+
"logits/chosen": -2.4245553016662598,
|
| 575 |
+
"logits/rejected": -2.4514193534851074,
|
| 576 |
+
"logps/chosen": -284.3450927734375,
|
| 577 |
+
"logps/rejected": -268.92486572265625,
|
| 578 |
+
"loss": 0.6631,
|
| 579 |
+
"pred_label": 2047.8499755859375,
|
| 580 |
+
"rewards/accuracies": 0.71875,
|
| 581 |
+
"rewards/chosen": 0.02513638138771057,
|
| 582 |
+
"rewards/margins": 0.04776451736688614,
|
| 583 |
+
"rewards/rejected": -0.02262813411653042,
|
| 584 |
+
"step": 350,
|
| 585 |
+
"use_label": 3474.14990234375
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.38,
|
| 589 |
"learning_rate": 3.4633294528521536e-07,
|
| 590 |
+
"logits/chosen": -2.5096325874328613,
|
| 591 |
+
"logits/rejected": -2.5362212657928467,
|
| 592 |
+
"logps/chosen": -262.33953857421875,
|
| 593 |
+
"logps/rejected": -256.80279541015625,
|
| 594 |
+
"loss": 0.6611,
|
| 595 |
+
"pred_label": 2130.89990234375,
|
| 596 |
+
"rewards/accuracies": 0.612500011920929,
|
| 597 |
+
"rewards/chosen": 0.008055051788687706,
|
| 598 |
+
"rewards/margins": 0.03220166265964508,
|
| 599 |
+
"rewards/rejected": -0.024146610870957375,
|
| 600 |
+
"step": 360,
|
| 601 |
+
"use_label": 3551.10009765625
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"epoch": 0.39,
|
| 605 |
"learning_rate": 3.4051222351571594e-07,
|
| 606 |
+
"logits/chosen": -2.4874494075775146,
|
| 607 |
+
"logits/rejected": -2.4494900703430176,
|
| 608 |
+
"logps/chosen": -263.681884765625,
|
| 609 |
+
"logps/rejected": -261.24871826171875,
|
| 610 |
+
"loss": 0.6645,
|
| 611 |
+
"pred_label": 2212.10009765625,
|
| 612 |
+
"rewards/accuracies": 0.637499988079071,
|
| 613 |
+
"rewards/chosen": 0.01168682612478733,
|
| 614 |
+
"rewards/margins": 0.035164039582014084,
|
| 615 |
+
"rewards/rejected": -0.023477211594581604,
|
| 616 |
+
"step": 370,
|
| 617 |
+
"use_label": 3629.89990234375
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 0.4,
|
| 621 |
"learning_rate": 3.346915017462165e-07,
|
| 622 |
+
"logits/chosen": -2.5527141094207764,
|
| 623 |
+
"logits/rejected": -2.500330686569214,
|
| 624 |
+
"logps/chosen": -270.70123291015625,
|
| 625 |
+
"logps/rejected": -251.94638061523438,
|
| 626 |
+
"loss": 0.6568,
|
| 627 |
+
"pred_label": 2298.27490234375,
|
| 628 |
+
"rewards/accuracies": 0.65625,
|
| 629 |
+
"rewards/chosen": 0.015008668415248394,
|
| 630 |
+
"rewards/margins": 0.03496234118938446,
|
| 631 |
+
"rewards/rejected": -0.019953671842813492,
|
| 632 |
+
"step": 380,
|
| 633 |
+
"use_label": 3703.72509765625
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.41,
|
| 637 |
"learning_rate": 3.288707799767171e-07,
|
| 638 |
+
"logits/chosen": -2.4154019355773926,
|
| 639 |
+
"logits/rejected": -2.410980224609375,
|
| 640 |
+
"logps/chosen": -267.69012451171875,
|
| 641 |
+
"logps/rejected": -234.7194061279297,
|
| 642 |
+
"loss": 0.6612,
|
| 643 |
+
"pred_label": 2388.699951171875,
|
| 644 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 645 |
+
"rewards/chosen": 0.016574550420045853,
|
| 646 |
+
"rewards/margins": 0.03597176447510719,
|
| 647 |
+
"rewards/rejected": -0.01939721405506134,
|
| 648 |
+
"step": 390,
|
| 649 |
+
"use_label": 3773.300048828125
|
| 650 |
},
|
| 651 |
{
|
| 652 |
"epoch": 0.42,
|
| 653 |
"learning_rate": 3.230500582072177e-07,
|
| 654 |
+
"logits/chosen": -2.573418378829956,
|
| 655 |
+
"logits/rejected": -2.5670266151428223,
|
| 656 |
+
"logps/chosen": -296.6780090332031,
|
| 657 |
+
"logps/rejected": -251.03109741210938,
|
| 658 |
+
"loss": 0.6603,
|
| 659 |
+
"pred_label": 2472.550048828125,
|
| 660 |
+
"rewards/accuracies": 0.625,
|
| 661 |
+
"rewards/chosen": 0.019582100212574005,
|
| 662 |
+
"rewards/margins": 0.03700990229845047,
|
| 663 |
+
"rewards/rejected": -0.017427802085876465,
|
| 664 |
+
"step": 400,
|
| 665 |
+
"use_label": 3849.449951171875
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"epoch": 0.43,
|
| 669 |
"learning_rate": 3.1722933643771827e-07,
|
| 670 |
+
"logits/chosen": -2.5456345081329346,
|
| 671 |
+
"logits/rejected": -2.5357089042663574,
|
| 672 |
+
"logps/chosen": -287.56024169921875,
|
| 673 |
+
"logps/rejected": -250.7117462158203,
|
| 674 |
+
"loss": 0.6565,
|
| 675 |
+
"pred_label": 2563.89990234375,
|
| 676 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 677 |
+
"rewards/chosen": 0.02417120710015297,
|
| 678 |
+
"rewards/margins": 0.03269309923052788,
|
| 679 |
+
"rewards/rejected": -0.008521895855665207,
|
| 680 |
+
"step": 410,
|
| 681 |
+
"use_label": 3918.10009765625
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.44,
|
| 685 |
"learning_rate": 3.1140861466821885e-07,
|
| 686 |
+
"logits/chosen": -2.425220251083374,
|
| 687 |
+
"logits/rejected": -2.403921127319336,
|
| 688 |
+
"logps/chosen": -231.62960815429688,
|
| 689 |
+
"logps/rejected": -243.0712432861328,
|
| 690 |
+
"loss": 0.6565,
|
| 691 |
+
"pred_label": 2649.25,
|
| 692 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 693 |
+
"rewards/chosen": 0.014951197430491447,
|
| 694 |
+
"rewards/margins": 0.026637399569153786,
|
| 695 |
+
"rewards/rejected": -0.011686199344694614,
|
| 696 |
+
"step": 420,
|
| 697 |
+
"use_label": 3992.75
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.45,
|
| 701 |
"learning_rate": 3.0558789289871943e-07,
|
| 702 |
+
"logits/chosen": -2.4625051021575928,
|
| 703 |
+
"logits/rejected": -2.4735419750213623,
|
| 704 |
+
"logps/chosen": -266.99554443359375,
|
| 705 |
+
"logps/rejected": -251.20022583007812,
|
| 706 |
+
"loss": 0.6572,
|
| 707 |
+
"pred_label": 2728.300048828125,
|
| 708 |
+
"rewards/accuracies": 0.606249988079071,
|
| 709 |
+
"rewards/chosen": 0.02481374330818653,
|
| 710 |
+
"rewards/margins": 0.031031513586640358,
|
| 711 |
+
"rewards/rejected": -0.006217771675437689,
|
| 712 |
+
"step": 430,
|
| 713 |
+
"use_label": 4073.699951171875
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.46,
|
| 717 |
"learning_rate": 2.9976717112922e-07,
|
| 718 |
+
"logits/chosen": -2.51550555229187,
|
| 719 |
+
"logits/rejected": -2.5493228435516357,
|
| 720 |
+
"logps/chosen": -274.2099914550781,
|
| 721 |
+
"logps/rejected": -245.31491088867188,
|
| 722 |
+
"loss": 0.6567,
|
| 723 |
+
"pred_label": 2806.375,
|
| 724 |
+
"rewards/accuracies": 0.5625,
|
| 725 |
+
"rewards/chosen": 0.01486388873308897,
|
| 726 |
+
"rewards/margins": 0.032243579626083374,
|
| 727 |
+
"rewards/rejected": -0.017379695549607277,
|
| 728 |
+
"step": 440,
|
| 729 |
+
"use_label": 4155.625
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.47,
|
| 733 |
"learning_rate": 2.939464493597206e-07,
|
| 734 |
+
"logits/chosen": -2.3210535049438477,
|
| 735 |
+
"logits/rejected": -2.3065693378448486,
|
| 736 |
+
"logps/chosen": -217.0265655517578,
|
| 737 |
+
"logps/rejected": -216.3203582763672,
|
| 738 |
+
"loss": 0.6561,
|
| 739 |
+
"pred_label": 2881.75,
|
| 740 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 741 |
+
"rewards/chosen": 0.009125987999141216,
|
| 742 |
+
"rewards/margins": 0.017552336677908897,
|
| 743 |
+
"rewards/rejected": -0.008426347747445107,
|
| 744 |
+
"step": 450,
|
| 745 |
+
"use_label": 4240.25
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.48,
|
| 749 |
"learning_rate": 2.8812572759022117e-07,
|
| 750 |
+
"logits/chosen": -2.528326988220215,
|
| 751 |
+
"logits/rejected": -2.513597011566162,
|
| 752 |
+
"logps/chosen": -290.8050231933594,
|
| 753 |
+
"logps/rejected": -260.4567565917969,
|
| 754 |
+
"loss": 0.6526,
|
| 755 |
+
"pred_label": 2965.02490234375,
|
| 756 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 757 |
+
"rewards/chosen": 0.024029741063714027,
|
| 758 |
+
"rewards/margins": 0.03901149705052376,
|
| 759 |
+
"rewards/rejected": -0.014981756918132305,
|
| 760 |
+
"step": 460,
|
| 761 |
+
"use_label": 4316.97509765625
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"epoch": 0.49,
|
| 765 |
"learning_rate": 2.8230500582072175e-07,
|
| 766 |
+
"logits/chosen": -2.424431085586548,
|
| 767 |
+
"logits/rejected": -2.41347336769104,
|
| 768 |
+
"logps/chosen": -254.66506958007812,
|
| 769 |
+
"logps/rejected": -234.6154022216797,
|
| 770 |
+
"loss": 0.6498,
|
| 771 |
+
"pred_label": 3055.699951171875,
|
| 772 |
+
"rewards/accuracies": 0.606249988079071,
|
| 773 |
+
"rewards/chosen": 0.001125166891142726,
|
| 774 |
+
"rewards/margins": 0.027449512854218483,
|
| 775 |
+
"rewards/rejected": -0.02632434293627739,
|
| 776 |
+
"step": 470,
|
| 777 |
+
"use_label": 4386.2998046875
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.5,
|
| 781 |
"learning_rate": 2.7648428405122233e-07,
|
| 782 |
+
"logits/chosen": -2.400242567062378,
|
| 783 |
+
"logits/rejected": -2.3543686866760254,
|
| 784 |
+
"logps/chosen": -235.65719604492188,
|
| 785 |
+
"logps/rejected": -239.40701293945312,
|
| 786 |
+
"loss": 0.6533,
|
| 787 |
+
"pred_label": 3150.449951171875,
|
| 788 |
+
"rewards/accuracies": 0.59375,
|
| 789 |
+
"rewards/chosen": 0.011093830689787865,
|
| 790 |
+
"rewards/margins": 0.0209193117916584,
|
| 791 |
+
"rewards/rejected": -0.009825478307902813,
|
| 792 |
+
"step": 480,
|
| 793 |
+
"use_label": 4451.5498046875
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.51,
|
| 797 |
"learning_rate": 2.706635622817229e-07,
|
| 798 |
+
"logits/chosen": -2.4891812801361084,
|
| 799 |
+
"logits/rejected": -2.4687983989715576,
|
| 800 |
+
"logps/chosen": -257.9729919433594,
|
| 801 |
+
"logps/rejected": -231.1624298095703,
|
| 802 |
+
"loss": 0.6504,
|
| 803 |
+
"pred_label": 3233.625,
|
| 804 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 805 |
+
"rewards/chosen": 0.018697496503591537,
|
| 806 |
+
"rewards/margins": 0.042968858033418655,
|
| 807 |
+
"rewards/rejected": -0.02427135966718197,
|
| 808 |
+
"step": 490,
|
| 809 |
+
"use_label": 4528.375
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 0.52,
|
| 813 |
"learning_rate": 2.648428405122235e-07,
|
| 814 |
+
"logits/chosen": -2.4956674575805664,
|
| 815 |
+
"logits/rejected": -2.444516658782959,
|
| 816 |
+
"logps/chosen": -274.7785949707031,
|
| 817 |
+
"logps/rejected": -247.5068359375,
|
| 818 |
+
"loss": 0.6497,
|
| 819 |
+
"pred_label": 3323.5,
|
| 820 |
+
"rewards/accuracies": 0.706250011920929,
|
| 821 |
+
"rewards/chosen": 0.020702462643384933,
|
| 822 |
+
"rewards/margins": 0.048901624977588654,
|
| 823 |
+
"rewards/rejected": -0.02819916605949402,
|
| 824 |
+
"step": 500,
|
| 825 |
+
"use_label": 4598.5
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 0.53,
|
| 829 |
"learning_rate": 2.590221187427241e-07,
|
| 830 |
+
"logits/chosen": -2.490265369415283,
|
| 831 |
+
"logits/rejected": -2.4258570671081543,
|
| 832 |
+
"logps/chosen": -262.3305969238281,
|
| 833 |
+
"logps/rejected": -225.7815704345703,
|
| 834 |
+
"loss": 0.6488,
|
| 835 |
+
"pred_label": 3414.699951171875,
|
| 836 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 837 |
+
"rewards/chosen": 0.022988121956586838,
|
| 838 |
+
"rewards/margins": 0.04825048893690109,
|
| 839 |
+
"rewards/rejected": -0.025262365117669106,
|
| 840 |
+
"step": 510,
|
| 841 |
+
"use_label": 4667.2998046875
|
| 842 |
},
|
| 843 |
{
|
| 844 |
"epoch": 0.54,
|
| 845 |
"learning_rate": 2.5320139697322466e-07,
|
| 846 |
+
"logits/chosen": -2.533203125,
|
| 847 |
+
"logits/rejected": -2.481980800628662,
|
| 848 |
+
"logps/chosen": -277.6612243652344,
|
| 849 |
+
"logps/rejected": -259.1783752441406,
|
| 850 |
+
"loss": 0.6496,
|
| 851 |
+
"pred_label": 3500.675048828125,
|
| 852 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 853 |
+
"rewards/chosen": 0.03759980946779251,
|
| 854 |
+
"rewards/margins": 0.07107619941234589,
|
| 855 |
+
"rewards/rejected": -0.033476393669843674,
|
| 856 |
+
"step": 520,
|
| 857 |
+
"use_label": 4741.3251953125
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.55,
|
| 861 |
"learning_rate": 2.4738067520372524e-07,
|
| 862 |
+
"logits/chosen": -2.478482484817505,
|
| 863 |
+
"logits/rejected": -2.4806153774261475,
|
| 864 |
+
"logps/chosen": -288.8346252441406,
|
| 865 |
+
"logps/rejected": -279.86871337890625,
|
| 866 |
+
"loss": 0.6453,
|
| 867 |
+
"pred_label": 3598.425048828125,
|
| 868 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 869 |
+
"rewards/chosen": 0.029218871146440506,
|
| 870 |
+
"rewards/margins": 0.05338066816329956,
|
| 871 |
+
"rewards/rejected": -0.024161797016859055,
|
| 872 |
+
"step": 530,
|
| 873 |
+
"use_label": 4803.5751953125
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.57,
|
| 877 |
"learning_rate": 2.415599534342258e-07,
|
| 878 |
+
"logits/chosen": -2.528951644897461,
|
| 879 |
+
"logits/rejected": -2.4700913429260254,
|
| 880 |
+
"logps/chosen": -278.3326416015625,
|
| 881 |
+
"logps/rejected": -256.6160888671875,
|
| 882 |
+
"loss": 0.6441,
|
| 883 |
+
"pred_label": 3706.47509765625,
|
| 884 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 885 |
+
"rewards/chosen": 0.03666474670171738,
|
| 886 |
+
"rewards/margins": 0.06262056529521942,
|
| 887 |
+
"rewards/rejected": -0.025955811142921448,
|
| 888 |
+
"step": 540,
|
| 889 |
+
"use_label": 4855.52490234375
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.58,
|
| 893 |
"learning_rate": 2.3573923166472642e-07,
|
| 894 |
+
"logits/chosen": -2.514728546142578,
|
| 895 |
+
"logits/rejected": -2.4884872436523438,
|
| 896 |
+
"logps/chosen": -278.8384094238281,
|
| 897 |
+
"logps/rejected": -251.11495971679688,
|
| 898 |
+
"loss": 0.6419,
|
| 899 |
+
"pred_label": 3806.125,
|
| 900 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 901 |
+
"rewards/chosen": 0.028542999178171158,
|
| 902 |
+
"rewards/margins": 0.05743313580751419,
|
| 903 |
+
"rewards/rejected": -0.028890132904052734,
|
| 904 |
+
"step": 550,
|
| 905 |
+
"use_label": 4915.875
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.59,
|
| 909 |
"learning_rate": 2.2991850989522698e-07,
|
| 910 |
+
"logits/chosen": -2.4250597953796387,
|
| 911 |
+
"logits/rejected": -2.4664273262023926,
|
| 912 |
+
"logps/chosen": -263.1305847167969,
|
| 913 |
+
"logps/rejected": -209.12753295898438,
|
| 914 |
+
"loss": 0.6473,
|
| 915 |
+
"pred_label": 3901.925048828125,
|
| 916 |
+
"rewards/accuracies": 0.612500011920929,
|
| 917 |
+
"rewards/chosen": 0.01411795150488615,
|
| 918 |
+
"rewards/margins": 0.043625928461551666,
|
| 919 |
+
"rewards/rejected": -0.02950797602534294,
|
| 920 |
+
"step": 560,
|
| 921 |
+
"use_label": 4980.0751953125
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"epoch": 0.6,
|
| 925 |
"learning_rate": 2.2409778812572759e-07,
|
| 926 |
+
"logits/chosen": -2.481841802597046,
|
| 927 |
+
"logits/rejected": -2.4556803703308105,
|
| 928 |
+
"logps/chosen": -256.8991394042969,
|
| 929 |
+
"logps/rejected": -253.22006225585938,
|
| 930 |
+
"loss": 0.6455,
|
| 931 |
+
"pred_label": 3991.375,
|
| 932 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 933 |
+
"rewards/chosen": 0.021176273003220558,
|
| 934 |
+
"rewards/margins": 0.04561372101306915,
|
| 935 |
+
"rewards/rejected": -0.024437451735138893,
|
| 936 |
+
"step": 570,
|
| 937 |
+
"use_label": 5050.625
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 0.61,
|
| 941 |
"learning_rate": 2.1827706635622817e-07,
|
| 942 |
+
"logits/chosen": -2.513284683227539,
|
| 943 |
+
"logits/rejected": -2.4957690238952637,
|
| 944 |
+
"logps/chosen": -265.62261962890625,
|
| 945 |
+
"logps/rejected": -255.05160522460938,
|
| 946 |
+
"loss": 0.6459,
|
| 947 |
+
"pred_label": 4092.97509765625,
|
| 948 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 949 |
+
"rewards/chosen": 0.023949166759848595,
|
| 950 |
+
"rewards/margins": 0.05676042288541794,
|
| 951 |
+
"rewards/rejected": -0.0328112468123436,
|
| 952 |
+
"step": 580,
|
| 953 |
+
"use_label": 5109.02490234375
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.62,
|
| 957 |
"learning_rate": 2.1245634458672875e-07,
|
| 958 |
+
"logits/chosen": -2.4279496669769287,
|
| 959 |
+
"logits/rejected": -2.3903157711029053,
|
| 960 |
+
"logps/chosen": -251.9611358642578,
|
| 961 |
+
"logps/rejected": -243.91909790039062,
|
| 962 |
+
"loss": 0.6421,
|
| 963 |
+
"pred_label": 4190.25,
|
| 964 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 965 |
+
"rewards/chosen": 0.02141127921640873,
|
| 966 |
+
"rewards/margins": 0.04126313328742981,
|
| 967 |
+
"rewards/rejected": -0.01985185407102108,
|
| 968 |
+
"step": 590,
|
| 969 |
+
"use_label": 5171.75
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.63,
|
| 973 |
"learning_rate": 2.0663562281722933e-07,
|
| 974 |
+
"logits/chosen": -2.4996018409729004,
|
| 975 |
+
"logits/rejected": -2.505694627761841,
|
| 976 |
+
"logps/chosen": -294.9464111328125,
|
| 977 |
+
"logps/rejected": -294.47930908203125,
|
| 978 |
+
"loss": 0.6445,
|
| 979 |
+
"pred_label": 4285.97509765625,
|
| 980 |
+
"rewards/accuracies": 0.675000011920929,
|
| 981 |
+
"rewards/chosen": 0.03435974568128586,
|
| 982 |
+
"rewards/margins": 0.053214918822050095,
|
| 983 |
+
"rewards/rejected": -0.018855175003409386,
|
| 984 |
+
"step": 600,
|
| 985 |
+
"use_label": 5236.02490234375
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 0.64,
|
| 989 |
"learning_rate": 2.008149010477299e-07,
|
| 990 |
+
"logits/chosen": -2.500229597091675,
|
| 991 |
+
"logits/rejected": -2.44234037399292,
|
| 992 |
+
"logps/chosen": -288.0011291503906,
|
| 993 |
+
"logps/rejected": -231.5128936767578,
|
| 994 |
+
"loss": 0.6418,
|
| 995 |
+
"pred_label": 4384.60009765625,
|
| 996 |
+
"rewards/accuracies": 0.643750011920929,
|
| 997 |
+
"rewards/chosen": 0.03441072255373001,
|
| 998 |
+
"rewards/margins": 0.06906420737504959,
|
| 999 |
+
"rewards/rejected": -0.03465348482131958,
|
| 1000 |
+
"step": 610,
|
| 1001 |
+
"use_label": 5297.39990234375
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"epoch": 0.65,
|
| 1005 |
"learning_rate": 1.949941792782305e-07,
|
| 1006 |
+
"logits/chosen": -2.4949872493743896,
|
| 1007 |
+
"logits/rejected": -2.4376912117004395,
|
| 1008 |
+
"logps/chosen": -271.5177917480469,
|
| 1009 |
+
"logps/rejected": -241.4095458984375,
|
| 1010 |
+
"loss": 0.6436,
|
| 1011 |
+
"pred_label": 4473.35009765625,
|
| 1012 |
+
"rewards/accuracies": 0.643750011920929,
|
| 1013 |
+
"rewards/chosen": 0.02876521274447441,
|
| 1014 |
+
"rewards/margins": 0.04532798379659653,
|
| 1015 |
+
"rewards/rejected": -0.016562769189476967,
|
| 1016 |
+
"step": 620,
|
| 1017 |
+
"use_label": 5368.64990234375
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.66,
|
| 1021 |
"learning_rate": 1.8917345750873107e-07,
|
| 1022 |
+
"logits/chosen": -2.4714274406433105,
|
| 1023 |
+
"logits/rejected": -2.4965994358062744,
|
| 1024 |
+
"logps/chosen": -272.4942321777344,
|
| 1025 |
+
"logps/rejected": -245.6349334716797,
|
| 1026 |
+
"loss": 0.6423,
|
| 1027 |
+
"pred_label": 4572.5,
|
| 1028 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1029 |
+
"rewards/chosen": 0.033678844571113586,
|
| 1030 |
+
"rewards/margins": 0.053819943219423294,
|
| 1031 |
+
"rewards/rejected": -0.02014109678566456,
|
| 1032 |
+
"step": 630,
|
| 1033 |
+
"use_label": 5429.5
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 0.67,
|
| 1037 |
"learning_rate": 1.8335273573923165e-07,
|
| 1038 |
+
"logits/chosen": -2.4581923484802246,
|
| 1039 |
+
"logits/rejected": -2.474356174468994,
|
| 1040 |
+
"logps/chosen": -241.07748413085938,
|
| 1041 |
+
"logps/rejected": -239.5830535888672,
|
| 1042 |
+
"loss": 0.6398,
|
| 1043 |
+
"pred_label": 4672.5751953125,
|
| 1044 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1045 |
+
"rewards/chosen": 0.036277130246162415,
|
| 1046 |
+
"rewards/margins": 0.06419126689434052,
|
| 1047 |
+
"rewards/rejected": -0.027914145961403847,
|
| 1048 |
+
"step": 640,
|
| 1049 |
+
"use_label": 5489.4248046875
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
"epoch": 0.68,
|
| 1053 |
"learning_rate": 1.7753201396973226e-07,
|
| 1054 |
+
"logits/chosen": -2.4635560512542725,
|
| 1055 |
+
"logits/rejected": -2.4887423515319824,
|
| 1056 |
+
"logps/chosen": -286.7969665527344,
|
| 1057 |
+
"logps/rejected": -265.29522705078125,
|
| 1058 |
+
"loss": 0.6357,
|
| 1059 |
+
"pred_label": 4776.0751953125,
|
| 1060 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 1061 |
+
"rewards/chosen": 0.028953587636351585,
|
| 1062 |
+
"rewards/margins": 0.07534638047218323,
|
| 1063 |
+
"rewards/rejected": -0.04639279097318649,
|
| 1064 |
+
"step": 650,
|
| 1065 |
+
"use_label": 5545.9248046875
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 0.69,
|
| 1069 |
"learning_rate": 1.7171129220023281e-07,
|
| 1070 |
+
"logits/chosen": -2.5433027744293213,
|
| 1071 |
+
"logits/rejected": -2.4936447143554688,
|
| 1072 |
+
"logps/chosen": -285.5106506347656,
|
| 1073 |
+
"logps/rejected": -242.0889434814453,
|
| 1074 |
+
"loss": 0.6451,
|
| 1075 |
+
"pred_label": 4884.875,
|
| 1076 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1077 |
+
"rewards/chosen": 0.024147575721144676,
|
| 1078 |
+
"rewards/margins": 0.03708204999566078,
|
| 1079 |
+
"rewards/rejected": -0.012934470549225807,
|
| 1080 |
+
"step": 660,
|
| 1081 |
+
"use_label": 5597.125
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.7,
|
| 1085 |
"learning_rate": 1.658905704307334e-07,
|
| 1086 |
+
"logits/chosen": -2.461125612258911,
|
| 1087 |
+
"logits/rejected": -2.419694185256958,
|
| 1088 |
+
"logps/chosen": -269.296142578125,
|
| 1089 |
+
"logps/rejected": -234.6914520263672,
|
| 1090 |
+
"loss": 0.6392,
|
| 1091 |
+
"pred_label": 4982.97509765625,
|
| 1092 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 1093 |
+
"rewards/chosen": 0.032385144382715225,
|
| 1094 |
+
"rewards/margins": 0.07203620672225952,
|
| 1095 |
+
"rewards/rejected": -0.039651062339544296,
|
| 1096 |
+
"step": 670,
|
| 1097 |
+
"use_label": 5659.02490234375
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 0.71,
|
| 1101 |
"learning_rate": 1.60069848661234e-07,
|
| 1102 |
+
"logits/chosen": -2.4804444313049316,
|
| 1103 |
+
"logits/rejected": -2.5111165046691895,
|
| 1104 |
+
"logps/chosen": -274.47027587890625,
|
| 1105 |
+
"logps/rejected": -250.04782104492188,
|
| 1106 |
+
"loss": 0.6424,
|
| 1107 |
+
"pred_label": 5080.27490234375,
|
| 1108 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 1109 |
+
"rewards/chosen": 0.020998146384954453,
|
| 1110 |
+
"rewards/margins": 0.029987860471010208,
|
| 1111 |
+
"rewards/rejected": -0.008989715948700905,
|
| 1112 |
+
"step": 680,
|
| 1113 |
+
"use_label": 5721.72509765625
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"epoch": 0.72,
|
| 1117 |
"learning_rate": 1.5424912689173456e-07,
|
| 1118 |
+
"logits/chosen": -2.475935220718384,
|
| 1119 |
+
"logits/rejected": -2.4738192558288574,
|
| 1120 |
+
"logps/chosen": -239.99905395507812,
|
| 1121 |
+
"logps/rejected": -201.9076690673828,
|
| 1122 |
+
"loss": 0.6373,
|
| 1123 |
+
"pred_label": 5173.0751953125,
|
| 1124 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1125 |
+
"rewards/chosen": 0.02569044753909111,
|
| 1126 |
+
"rewards/margins": 0.05879662558436394,
|
| 1127 |
+
"rewards/rejected": -0.03310617804527283,
|
| 1128 |
+
"step": 690,
|
| 1129 |
+
"use_label": 5788.9248046875
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.73,
|
| 1133 |
"learning_rate": 1.4842840512223514e-07,
|
| 1134 |
+
"logits/chosen": -2.5004830360412598,
|
| 1135 |
+
"logits/rejected": -2.4815783500671387,
|
| 1136 |
+
"logps/chosen": -262.7047424316406,
|
| 1137 |
+
"logps/rejected": -232.9453887939453,
|
| 1138 |
+
"loss": 0.6416,
|
| 1139 |
+
"pred_label": 5271.14990234375,
|
| 1140 |
+
"rewards/accuracies": 0.59375,
|
| 1141 |
+
"rewards/chosen": 0.021065320819616318,
|
| 1142 |
+
"rewards/margins": 0.05442025512456894,
|
| 1143 |
+
"rewards/rejected": -0.03335493057966232,
|
| 1144 |
+
"step": 700,
|
| 1145 |
+
"use_label": 5850.85009765625
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"epoch": 0.74,
|
| 1149 |
"learning_rate": 1.4260768335273574e-07,
|
| 1150 |
+
"logits/chosen": -2.5241992473602295,
|
| 1151 |
+
"logits/rejected": -2.4814987182617188,
|
| 1152 |
+
"logps/chosen": -279.4823303222656,
|
| 1153 |
+
"logps/rejected": -233.07373046875,
|
| 1154 |
+
"loss": 0.6346,
|
| 1155 |
+
"pred_label": 5374.75,
|
| 1156 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1157 |
+
"rewards/chosen": 0.02996956743299961,
|
| 1158 |
+
"rewards/margins": 0.0639830082654953,
|
| 1159 |
+
"rewards/rejected": -0.03401344642043114,
|
| 1160 |
+
"step": 710,
|
| 1161 |
+
"use_label": 5907.25
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.75,
|
| 1165 |
"learning_rate": 1.3678696158323632e-07,
|
| 1166 |
+
"logits/chosen": -2.4806971549987793,
|
| 1167 |
+
"logits/rejected": -2.4919962882995605,
|
| 1168 |
+
"logps/chosen": -261.57806396484375,
|
| 1169 |
+
"logps/rejected": -246.28042602539062,
|
| 1170 |
+
"loss": 0.6397,
|
| 1171 |
+
"pred_label": 5479.1748046875,
|
| 1172 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1173 |
+
"rewards/chosen": 0.034241896122694016,
|
| 1174 |
+
"rewards/margins": 0.06610169261693954,
|
| 1175 |
+
"rewards/rejected": -0.03185979649424553,
|
| 1176 |
+
"step": 720,
|
| 1177 |
+
"use_label": 5962.8251953125
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"epoch": 0.76,
|
| 1181 |
"learning_rate": 1.3096623981373688e-07,
|
| 1182 |
+
"logits/chosen": -2.4852230548858643,
|
| 1183 |
+
"logits/rejected": -2.473806142807007,
|
| 1184 |
+
"logps/chosen": -287.5245056152344,
|
| 1185 |
+
"logps/rejected": -234.40615844726562,
|
| 1186 |
+
"loss": 0.6392,
|
| 1187 |
+
"pred_label": 5584.5,
|
| 1188 |
+
"rewards/accuracies": 0.643750011920929,
|
| 1189 |
+
"rewards/chosen": 0.04039750620722771,
|
| 1190 |
+
"rewards/margins": 0.07070399820804596,
|
| 1191 |
+
"rewards/rejected": -0.0303064975887537,
|
| 1192 |
+
"step": 730,
|
| 1193 |
+
"use_label": 6017.5
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.77,
|
| 1197 |
"learning_rate": 1.2514551804423749e-07,
|
| 1198 |
+
"logits/chosen": -2.466087818145752,
|
| 1199 |
+
"logits/rejected": -2.5063462257385254,
|
| 1200 |
+
"logps/chosen": -295.0579528808594,
|
| 1201 |
+
"logps/rejected": -279.02740478515625,
|
| 1202 |
+
"loss": 0.6358,
|
| 1203 |
+
"pred_label": 5692.77490234375,
|
| 1204 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 1205 |
+
"rewards/chosen": 0.040842343121767044,
|
| 1206 |
+
"rewards/margins": 0.07114710658788681,
|
| 1207 |
+
"rewards/rejected": -0.030304765328764915,
|
| 1208 |
+
"step": 740,
|
| 1209 |
+
"use_label": 6069.22509765625
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 0.79,
|
| 1213 |
"learning_rate": 1.1932479627473807e-07,
|
| 1214 |
+
"logits/chosen": -2.4788904190063477,
|
| 1215 |
+
"logits/rejected": -2.460747718811035,
|
| 1216 |
+
"logps/chosen": -250.11758422851562,
|
| 1217 |
+
"logps/rejected": -250.706787109375,
|
| 1218 |
+
"loss": 0.6351,
|
| 1219 |
+
"pred_label": 5804.22509765625,
|
| 1220 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1221 |
+
"rewards/chosen": 0.02599816396832466,
|
| 1222 |
+
"rewards/margins": 0.05924210697412491,
|
| 1223 |
+
"rewards/rejected": -0.03324393928050995,
|
| 1224 |
+
"step": 750,
|
| 1225 |
+
"use_label": 6117.77490234375
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"epoch": 0.8,
|
| 1229 |
"learning_rate": 1.1350407450523865e-07,
|
| 1230 |
+
"logits/chosen": -2.4908533096313477,
|
| 1231 |
+
"logits/rejected": -2.5631959438323975,
|
| 1232 |
+
"logps/chosen": -280.65643310546875,
|
| 1233 |
+
"logps/rejected": -252.4939422607422,
|
| 1234 |
+
"loss": 0.6382,
|
| 1235 |
+
"pred_label": 5910.4248046875,
|
| 1236 |
+
"rewards/accuracies": 0.59375,
|
| 1237 |
+
"rewards/chosen": 0.035374678671360016,
|
| 1238 |
+
"rewards/margins": 0.06185116618871689,
|
| 1239 |
+
"rewards/rejected": -0.026476481929421425,
|
| 1240 |
+
"step": 760,
|
| 1241 |
+
"use_label": 6171.5751953125
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.81,
|
| 1245 |
"learning_rate": 1.0768335273573923e-07,
|
| 1246 |
+
"logits/chosen": -2.5305213928222656,
|
| 1247 |
+
"logits/rejected": -2.548992156982422,
|
| 1248 |
+
"logps/chosen": -291.19085693359375,
|
| 1249 |
+
"logps/rejected": -276.0303649902344,
|
| 1250 |
+
"loss": 0.6357,
|
| 1251 |
+
"pred_label": 6023.6748046875,
|
| 1252 |
+
"rewards/accuracies": 0.612500011920929,
|
| 1253 |
+
"rewards/chosen": 0.038612816482782364,
|
| 1254 |
+
"rewards/margins": 0.0703565925359726,
|
| 1255 |
+
"rewards/rejected": -0.03174378350377083,
|
| 1256 |
+
"step": 770,
|
| 1257 |
+
"use_label": 6218.3251953125
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"epoch": 0.82,
|
| 1261 |
"learning_rate": 1.0186263096623981e-07,
|
| 1262 |
+
"logits/chosen": -2.500197172164917,
|
| 1263 |
+
"logits/rejected": -2.5004265308380127,
|
| 1264 |
+
"logps/chosen": -269.6267395019531,
|
| 1265 |
+
"logps/rejected": -237.23226928710938,
|
| 1266 |
+
"loss": 0.6382,
|
| 1267 |
+
"pred_label": 6124.77490234375,
|
| 1268 |
+
"rewards/accuracies": 0.574999988079071,
|
| 1269 |
+
"rewards/chosen": 0.03419669717550278,
|
| 1270 |
+
"rewards/margins": 0.05804433301091194,
|
| 1271 |
+
"rewards/rejected": -0.023847635835409164,
|
| 1272 |
+
"step": 780,
|
| 1273 |
+
"use_label": 6277.22509765625
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 0.83,
|
| 1277 |
"learning_rate": 9.604190919674039e-08,
|
| 1278 |
+
"logits/chosen": -2.5122509002685547,
|
| 1279 |
+
"logits/rejected": -2.527552843093872,
|
| 1280 |
+
"logps/chosen": -258.3880920410156,
|
| 1281 |
+
"logps/rejected": -250.8657684326172,
|
| 1282 |
+
"loss": 0.6408,
|
| 1283 |
+
"pred_label": 6225.125,
|
| 1284 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 1285 |
+
"rewards/chosen": 0.024235274642705917,
|
| 1286 |
+
"rewards/margins": 0.03934397175908089,
|
| 1287 |
+
"rewards/rejected": -0.015108692459762096,
|
| 1288 |
+
"step": 790,
|
| 1289 |
+
"use_label": 6336.875
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
"epoch": 0.84,
|
| 1293 |
"learning_rate": 9.022118742724097e-08,
|
| 1294 |
+
"logits/chosen": -2.409346103668213,
|
| 1295 |
+
"logits/rejected": -2.458906650543213,
|
| 1296 |
+
"logps/chosen": -291.98406982421875,
|
| 1297 |
+
"logps/rejected": -263.28204345703125,
|
| 1298 |
+
"loss": 0.632,
|
| 1299 |
+
"pred_label": 6330.6748046875,
|
| 1300 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1301 |
+
"rewards/chosen": 0.03406571224331856,
|
| 1302 |
+
"rewards/margins": 0.06302747130393982,
|
| 1303 |
+
"rewards/rejected": -0.028961753472685814,
|
| 1304 |
+
"step": 800,
|
| 1305 |
+
"use_label": 6391.3251953125
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.85,
|
| 1309 |
"learning_rate": 8.440046565774157e-08,
|
| 1310 |
+
"logits/chosen": -2.4541163444519043,
|
| 1311 |
+
"logits/rejected": -2.426970958709717,
|
| 1312 |
+
"logps/chosen": -272.098388671875,
|
| 1313 |
+
"logps/rejected": -219.48318481445312,
|
| 1314 |
+
"loss": 0.6397,
|
| 1315 |
+
"pred_label": 6430.9248046875,
|
| 1316 |
+
"rewards/accuracies": 0.643750011920929,
|
| 1317 |
+
"rewards/chosen": 0.0267607681453228,
|
| 1318 |
+
"rewards/margins": 0.049745358526706696,
|
| 1319 |
+
"rewards/rejected": -0.022984590381383896,
|
| 1320 |
+
"step": 810,
|
| 1321 |
+
"use_label": 6451.0751953125
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"epoch": 0.86,
|
| 1325 |
"learning_rate": 7.857974388824213e-08,
|
| 1326 |
+
"logits/chosen": -2.443247079849243,
|
| 1327 |
+
"logits/rejected": -2.4559197425842285,
|
| 1328 |
+
"logps/chosen": -278.1483459472656,
|
| 1329 |
+
"logps/rejected": -233.55221557617188,
|
| 1330 |
+
"loss": 0.6401,
|
| 1331 |
+
"pred_label": 6533.77490234375,
|
| 1332 |
+
"rewards/accuracies": 0.65625,
|
| 1333 |
+
"rewards/chosen": 0.03652986139059067,
|
| 1334 |
+
"rewards/margins": 0.0706639438867569,
|
| 1335 |
+
"rewards/rejected": -0.03413407504558563,
|
| 1336 |
+
"step": 820,
|
| 1337 |
+
"use_label": 6508.22509765625
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"epoch": 0.87,
|
| 1341 |
"learning_rate": 7.275902211874273e-08,
|
| 1342 |
+
"logits/chosen": -2.492734432220459,
|
| 1343 |
+
"logits/rejected": -2.5021793842315674,
|
| 1344 |
+
"logps/chosen": -252.682373046875,
|
| 1345 |
+
"logps/rejected": -271.76885986328125,
|
| 1346 |
+
"loss": 0.6402,
|
| 1347 |
+
"pred_label": 6633.2001953125,
|
| 1348 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1349 |
+
"rewards/chosen": 0.02864421345293522,
|
| 1350 |
+
"rewards/margins": 0.06612807512283325,
|
| 1351 |
+
"rewards/rejected": -0.03748386353254318,
|
| 1352 |
+
"step": 830,
|
| 1353 |
+
"use_label": 6568.7998046875
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.88,
|
| 1357 |
"learning_rate": 6.693830034924331e-08,
|
| 1358 |
+
"logits/chosen": -2.431647300720215,
|
| 1359 |
+
"logits/rejected": -2.4801371097564697,
|
| 1360 |
+
"logps/chosen": -254.59475708007812,
|
| 1361 |
+
"logps/rejected": -252.9589080810547,
|
| 1362 |
+
"loss": 0.6335,
|
| 1363 |
+
"pred_label": 6738.25,
|
| 1364 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1365 |
+
"rewards/chosen": 0.02285713143646717,
|
| 1366 |
+
"rewards/margins": 0.05092828720808029,
|
| 1367 |
+
"rewards/rejected": -0.02807115949690342,
|
| 1368 |
+
"step": 840,
|
| 1369 |
+
"use_label": 6623.75
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.89,
|
| 1373 |
"learning_rate": 6.111757857974389e-08,
|
| 1374 |
+
"logits/chosen": -2.4767704010009766,
|
| 1375 |
+
"logits/rejected": -2.5524322986602783,
|
| 1376 |
+
"logps/chosen": -284.7981872558594,
|
| 1377 |
+
"logps/rejected": -236.7716522216797,
|
| 1378 |
+
"loss": 0.6388,
|
| 1379 |
+
"pred_label": 6841.77490234375,
|
| 1380 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1381 |
+
"rewards/chosen": 0.05144144967198372,
|
| 1382 |
+
"rewards/margins": 0.09114910662174225,
|
| 1383 |
+
"rewards/rejected": -0.03970765322446823,
|
| 1384 |
+
"step": 850,
|
| 1385 |
+
"use_label": 6680.22509765625
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 0.9,
|
| 1389 |
"learning_rate": 5.529685681024446e-08,
|
| 1390 |
+
"logits/chosen": -2.440410614013672,
|
| 1391 |
+
"logits/rejected": -2.4388670921325684,
|
| 1392 |
+
"logps/chosen": -274.1076354980469,
|
| 1393 |
+
"logps/rejected": -265.62774658203125,
|
| 1394 |
+
"loss": 0.639,
|
| 1395 |
+
"pred_label": 6939.10009765625,
|
| 1396 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1397 |
+
"rewards/chosen": 0.021466780453920364,
|
| 1398 |
+
"rewards/margins": 0.053748417645692825,
|
| 1399 |
+
"rewards/rejected": -0.032281629741191864,
|
| 1400 |
+
"step": 860,
|
| 1401 |
+
"use_label": 6742.89990234375
|
| 1402 |
},
|
| 1403 |
{
|
| 1404 |
"epoch": 0.91,
|
| 1405 |
"learning_rate": 4.947613504074505e-08,
|
| 1406 |
+
"logits/chosen": -2.4695324897766113,
|
| 1407 |
+
"logits/rejected": -2.5007336139678955,
|
| 1408 |
+
"logps/chosen": -234.1849822998047,
|
| 1409 |
+
"logps/rejected": -233.623046875,
|
| 1410 |
+
"loss": 0.6369,
|
| 1411 |
+
"pred_label": 7033.625,
|
| 1412 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1413 |
+
"rewards/chosen": 0.03171985596418381,
|
| 1414 |
+
"rewards/margins": 0.0530589334666729,
|
| 1415 |
+
"rewards/rejected": -0.02133907750248909,
|
| 1416 |
+
"step": 870,
|
| 1417 |
+
"use_label": 6808.375
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 0.92,
|
| 1421 |
"learning_rate": 4.365541327124563e-08,
|
| 1422 |
+
"logits/chosen": -2.5348362922668457,
|
| 1423 |
+
"logits/rejected": -2.48889422416687,
|
| 1424 |
+
"logps/chosen": -275.351806640625,
|
| 1425 |
+
"logps/rejected": -252.8624267578125,
|
| 1426 |
+
"loss": 0.6378,
|
| 1427 |
+
"pred_label": 7132.27490234375,
|
| 1428 |
+
"rewards/accuracies": 0.706250011920929,
|
| 1429 |
+
"rewards/chosen": 0.04395347461104393,
|
| 1430 |
+
"rewards/margins": 0.07273541390895844,
|
| 1431 |
+
"rewards/rejected": -0.028781946748495102,
|
| 1432 |
+
"step": 880,
|
| 1433 |
+
"use_label": 6869.72509765625
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
"epoch": 0.93,
|
| 1437 |
"learning_rate": 3.783469150174622e-08,
|
| 1438 |
+
"logits/chosen": -2.5349538326263428,
|
| 1439 |
+
"logits/rejected": -2.5645627975463867,
|
| 1440 |
+
"logps/chosen": -269.9933166503906,
|
| 1441 |
+
"logps/rejected": -262.0923156738281,
|
| 1442 |
+
"loss": 0.6362,
|
| 1443 |
+
"pred_label": 7239.27490234375,
|
| 1444 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 1445 |
+
"rewards/chosen": 0.024751801043748856,
|
| 1446 |
+
"rewards/margins": 0.05075649172067642,
|
| 1447 |
+
"rewards/rejected": -0.026004692539572716,
|
| 1448 |
+
"step": 890,
|
| 1449 |
+
"use_label": 6922.72509765625
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.94,
|
| 1453 |
"learning_rate": 3.20139697322468e-08,
|
| 1454 |
+
"logits/chosen": -2.4200382232666016,
|
| 1455 |
+
"logits/rejected": -2.4404149055480957,
|
| 1456 |
+
"logps/chosen": -264.64501953125,
|
| 1457 |
+
"logps/rejected": -250.6757354736328,
|
| 1458 |
+
"loss": 0.6397,
|
| 1459 |
+
"pred_label": 7349.375,
|
| 1460 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 1461 |
+
"rewards/chosen": 0.02098809741437435,
|
| 1462 |
+
"rewards/margins": 0.051287733018398285,
|
| 1463 |
+
"rewards/rejected": -0.030299633741378784,
|
| 1464 |
+
"step": 900,
|
| 1465 |
+
"use_label": 6972.625
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 0.95,
|
| 1469 |
"learning_rate": 2.619324796274738e-08,
|
| 1470 |
+
"logits/chosen": -2.485106945037842,
|
| 1471 |
+
"logits/rejected": -2.509174346923828,
|
| 1472 |
+
"logps/chosen": -282.1513671875,
|
| 1473 |
+
"logps/rejected": -253.21719360351562,
|
| 1474 |
+
"loss": 0.6356,
|
| 1475 |
+
"pred_label": 7449.4501953125,
|
| 1476 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1477 |
+
"rewards/chosen": 0.0359305813908577,
|
| 1478 |
+
"rewards/margins": 0.05107826739549637,
|
| 1479 |
+
"rewards/rejected": -0.015147687867283821,
|
| 1480 |
+
"step": 910,
|
| 1481 |
+
"use_label": 7032.5498046875
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.96,
|
| 1485 |
"learning_rate": 2.037252619324796e-08,
|
| 1486 |
+
"logits/chosen": -2.4887633323669434,
|
| 1487 |
+
"logits/rejected": -2.4791512489318848,
|
| 1488 |
+
"logps/chosen": -279.5574951171875,
|
| 1489 |
+
"logps/rejected": -249.22793579101562,
|
| 1490 |
+
"loss": 0.6371,
|
| 1491 |
+
"pred_label": 7561.125,
|
| 1492 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1493 |
+
"rewards/chosen": 0.034208498895168304,
|
| 1494 |
+
"rewards/margins": 0.06242678314447403,
|
| 1495 |
+
"rewards/rejected": -0.028218284249305725,
|
| 1496 |
+
"step": 920,
|
| 1497 |
+
"use_label": 7080.875
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 0.97,
|
| 1501 |
"learning_rate": 1.4551804423748545e-08,
|
| 1502 |
+
"logits/chosen": -2.499810218811035,
|
| 1503 |
+
"logits/rejected": -2.465851306915283,
|
| 1504 |
+
"logps/chosen": -258.83099365234375,
|
| 1505 |
+
"logps/rejected": -234.7783203125,
|
| 1506 |
+
"loss": 0.6375,
|
| 1507 |
+
"pred_label": 7669.27490234375,
|
| 1508 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1509 |
+
"rewards/chosen": 0.04888115078210831,
|
| 1510 |
+
"rewards/margins": 0.0783516988158226,
|
| 1511 |
+
"rewards/rejected": -0.02947053872048855,
|
| 1512 |
+
"step": 930,
|
| 1513 |
+
"use_label": 7132.72509765625
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
"epoch": 0.98,
|
| 1517 |
"learning_rate": 8.731082654249125e-09,
|
| 1518 |
+
"logits/chosen": -2.530332565307617,
|
| 1519 |
+
"logits/rejected": -2.512815237045288,
|
| 1520 |
+
"logps/chosen": -277.4713439941406,
|
| 1521 |
+
"logps/rejected": -260.1994934082031,
|
| 1522 |
+
"loss": 0.6359,
|
| 1523 |
+
"pred_label": 7771.25,
|
| 1524 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 1525 |
+
"rewards/chosen": 0.009665247052907944,
|
| 1526 |
+
"rewards/margins": 0.03832637146115303,
|
| 1527 |
+
"rewards/rejected": -0.028661120682954788,
|
| 1528 |
+
"step": 940,
|
| 1529 |
+
"use_label": 7190.75
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 0.99,
|
| 1533 |
"learning_rate": 2.910360884749709e-09,
|
| 1534 |
+
"logits/chosen": -2.4885025024414062,
|
| 1535 |
+
"logits/rejected": -2.4381051063537598,
|
| 1536 |
+
"logps/chosen": -255.72409057617188,
|
| 1537 |
+
"logps/rejected": -244.0724639892578,
|
| 1538 |
+
"loss": 0.6403,
|
| 1539 |
+
"pred_label": 7870.75,
|
| 1540 |
+
"rewards/accuracies": 0.625,
|
| 1541 |
+
"rewards/chosen": 0.0019513871520757675,
|
| 1542 |
+
"rewards/margins": 0.03540351241827011,
|
| 1543 |
+
"rewards/rejected": -0.03345213085412979,
|
| 1544 |
+
"step": 950,
|
| 1545 |
+
"use_label": 7251.25
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 1.0,
|
| 1549 |
+
"eval_logits/chosen": -2.5062618255615234,
|
| 1550 |
+
"eval_logits/rejected": -2.5199859142303467,
|
| 1551 |
+
"eval_logps/chosen": -272.93438720703125,
|
| 1552 |
+
"eval_logps/rejected": -250.28643798828125,
|
| 1553 |
+
"eval_loss": 0.6350826025009155,
|
| 1554 |
+
"eval_pred_label": 8131.07177734375,
|
| 1555 |
+
"eval_rewards/accuracies": 0.6200000047683716,
|
| 1556 |
+
"eval_rewards/chosen": 0.029984984546899796,
|
| 1557 |
+
"eval_rewards/margins": 0.06345725804567337,
|
| 1558 |
+
"eval_rewards/rejected": -0.033472273498773575,
|
| 1559 |
+
"eval_runtime": 444.1068,
|
| 1560 |
+
"eval_samples_per_second": 4.503,
|
| 1561 |
+
"eval_steps_per_second": 0.281,
|
| 1562 |
+
"eval_use_label": 7400.92822265625,
|
| 1563 |
"step": 955
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 1.0,
|
| 1567 |
"step": 955,
|
| 1568 |
"total_flos": 0.0,
|
| 1569 |
+
"train_loss": 0.6552608817035616,
|
| 1570 |
+
"train_runtime": 24261.6882,
|
| 1571 |
+
"train_samples_per_second": 2.52,
|
| 1572 |
+
"train_steps_per_second": 0.039
|
| 1573 |
}
|
| 1574 |
],
|
| 1575 |
"logging_steps": 10,
|