Model save
Browse files- README.md +11 -11
- all_results.json +14 -14
- config.json +1 -1
- eval_results.json +11 -11
- model-00001-of-00003.safetensors +1 -1
- model-00002-of-00003.safetensors +1 -1
- model-00003-of-00003.safetensors +1 -1
- train_results.json +3 -3
- trainer_state.json +967 -967
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -17,15 +17,15 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model was trained from scratch on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Logits/chosen: -
|
| 21 |
-
- Logits/rejected:
|
| 22 |
-
- Logps/chosen: -
|
| 23 |
-
- Logps/rejected: -
|
| 24 |
-
- Loss: 0.
|
| 25 |
-
- Rewards/accuracies: 0.
|
| 26 |
-
- Rewards/chosen: -
|
| 27 |
-
- Rewards/margins: 1.
|
| 28 |
-
- Rewards/rejected: -
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
@@ -62,8 +62,8 @@ The following hyperparameters were used during training:
|
|
| 62 |
|
| 63 |
| Training Loss | Epoch | Step | Logits/chosen | Logits/rejected | Logps/chosen | Logps/rejected | Validation Loss | Rewards/accuracies | Rewards/chosen | Rewards/margins | Rewards/rejected |
|
| 64 |
|:-------------:|:------:|:----:|:-------------:|:---------------:|:------------:|:--------------:|:---------------:|:------------------:|:--------------:|:---------------:|:----------------:|
|
| 65 |
-
| 0.
|
| 66 |
-
| 0.
|
| 67 |
|
| 68 |
|
| 69 |
### Framework versions
|
|
|
|
| 17 |
|
| 18 |
This model was trained from scratch on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Logits/chosen: -1.8431
|
| 21 |
+
- Logits/rejected: -1.7970
|
| 22 |
+
- Logps/chosen: -488.8500
|
| 23 |
+
- Logps/rejected: -619.1251
|
| 24 |
+
- Loss: 0.5119
|
| 25 |
+
- Rewards/accuracies: 0.7461
|
| 26 |
+
- Rewards/chosen: -2.1151
|
| 27 |
+
- Rewards/margins: 1.3443
|
| 28 |
+
- Rewards/rejected: -3.4594
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 62 |
|
| 63 |
| Training Loss | Epoch | Step | Logits/chosen | Logits/rejected | Logps/chosen | Logps/rejected | Validation Loss | Rewards/accuracies | Rewards/chosen | Rewards/margins | Rewards/rejected |
|
| 64 |
|:-------------:|:------:|:----:|:-------------:|:---------------:|:------------:|:--------------:|:---------------:|:------------------:|:--------------:|:---------------:|:----------------:|
|
| 65 |
+
| 0.4926 | 0.9984 | 477 | -2.1760 | -2.1455 | -412.9510 | -506.2944 | 0.5082 | 0.7578 | -1.3561 | 0.9750 | -2.3311 |
|
| 66 |
+
| 0.3138 | 1.9969 | 954 | -1.8431 | -1.7970 | -488.8500 | -619.1251 | 0.5119 | 0.7461 | -2.1151 | 1.3443 | -3.4594 |
|
| 67 |
|
| 68 |
|
| 69 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,22 +1,22 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9968602825745683,
|
| 3 |
-
"eval_logits/chosen": -
|
| 4 |
-
"eval_logits/rejected":
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 1.
|
| 8 |
-
"eval_rewards/accuracies": 0.
|
| 9 |
-
"eval_rewards/chosen": -
|
| 10 |
-
"eval_rewards/margins":
|
| 11 |
-
"eval_rewards/rejected": -
|
| 12 |
-
"eval_runtime":
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second": 11.
|
| 15 |
"eval_steps_per_second": 0.177,
|
| 16 |
"total_flos": 0.0,
|
| 17 |
"train_loss": 0.0,
|
| 18 |
-
"train_runtime": 3.
|
| 19 |
"train_samples": 61134,
|
| 20 |
-
"train_samples_per_second":
|
| 21 |
-
"train_steps_per_second":
|
| 22 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9968602825745683,
|
| 3 |
+
"eval_logits/chosen": -1.843076467514038,
|
| 4 |
+
"eval_logits/rejected": -1.7970439195632935,
|
| 5 |
+
"eval_logps/chosen": -488.8499755859375,
|
| 6 |
+
"eval_logps/rejected": -619.1251220703125,
|
| 7 |
+
"eval_loss": 1.217253565788269,
|
| 8 |
+
"eval_rewards/accuracies": 0.74609375,
|
| 9 |
+
"eval_rewards/chosen": -10.575380325317383,
|
| 10 |
+
"eval_rewards/margins": 6.721622467041016,
|
| 11 |
+
"eval_rewards/rejected": -17.2970027923584,
|
| 12 |
+
"eval_runtime": 180.4143,
|
| 13 |
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 11.086,
|
| 15 |
"eval_steps_per_second": 0.177,
|
| 16 |
"total_flos": 0.0,
|
| 17 |
"train_loss": 0.0,
|
| 18 |
+
"train_runtime": 3.478,
|
| 19 |
"train_samples": 61134,
|
| 20 |
+
"train_samples_per_second": 17577.112,
|
| 21 |
+
"train_steps_per_second": 137.146
|
| 22 |
}
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "data/sft/zephyr-7b-sft-1e-every25/checkpoint-
|
| 3 |
"architectures": [
|
| 4 |
"MistralForCausalLM"
|
| 5 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "data/sft/zephyr-7b-sft-1e-every25/checkpoint-200",
|
| 3 |
"architectures": [
|
| 4 |
"MistralForCausalLM"
|
| 5 |
],
|
eval_results.json
CHANGED
|
@@ -1,16 +1,16 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9968602825745683,
|
| 3 |
-
"eval_logits/chosen": -
|
| 4 |
-
"eval_logits/rejected":
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 1.
|
| 8 |
-
"eval_rewards/accuracies": 0.
|
| 9 |
-
"eval_rewards/chosen": -
|
| 10 |
-
"eval_rewards/margins":
|
| 11 |
-
"eval_rewards/rejected": -
|
| 12 |
-
"eval_runtime":
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second": 11.
|
| 15 |
"eval_steps_per_second": 0.177
|
| 16 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9968602825745683,
|
| 3 |
+
"eval_logits/chosen": -1.843076467514038,
|
| 4 |
+
"eval_logits/rejected": -1.7970439195632935,
|
| 5 |
+
"eval_logps/chosen": -488.8499755859375,
|
| 6 |
+
"eval_logps/rejected": -619.1251220703125,
|
| 7 |
+
"eval_loss": 1.217253565788269,
|
| 8 |
+
"eval_rewards/accuracies": 0.74609375,
|
| 9 |
+
"eval_rewards/chosen": -10.575380325317383,
|
| 10 |
+
"eval_rewards/margins": 6.721622467041016,
|
| 11 |
+
"eval_rewards/rejected": -17.2970027923584,
|
| 12 |
+
"eval_runtime": 180.4143,
|
| 13 |
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 11.086,
|
| 15 |
"eval_steps_per_second": 0.177
|
| 16 |
}
|
model-00001-of-00003.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4943162336
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d409451aee7e67f7af769c603992848bb56686f2128725dc784dd7116cf33841
|
| 3 |
size 4943162336
|
model-00002-of-00003.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4999819336
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8587b5070be5d2d68540351a78d713ee93922409df181e41978f6c67a7753634
|
| 3 |
size 4999819336
|
model-00003-of-00003.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4540516344
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:21db7468292c3c37b9fa7faed58f9e587e4299595d99e8b7b0c2873fc4b5238a
|
| 3 |
size 4540516344
|
train_results.json
CHANGED
|
@@ -2,8 +2,8 @@
|
|
| 2 |
"epoch": 1.9968602825745683,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
"train_loss": 0.0,
|
| 5 |
-
"train_runtime": 3.
|
| 6 |
"train_samples": 61134,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second":
|
| 9 |
}
|
|
|
|
| 2 |
"epoch": 1.9968602825745683,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
"train_loss": 0.0,
|
| 5 |
+
"train_runtime": 3.478,
|
| 6 |
"train_samples": 61134,
|
| 7 |
+
"train_samples_per_second": 17577.112,
|
| 8 |
+
"train_steps_per_second": 137.146
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -10,12 +10,12 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0020931449502878076,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 5.208333333333333e-09,
|
| 15 |
-
"logits/chosen": -2.
|
| 16 |
-
"logits/rejected": -2.
|
| 17 |
-
"logps/chosen": -
|
| 18 |
-
"logps/rejected": -
|
| 19 |
"loss": 0.6931,
|
| 20 |
"rewards/accuracies": 0.0,
|
| 21 |
"rewards/chosen": 0.0,
|
|
@@ -25,1459 +25,1459 @@
|
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.020931449502878074,
|
| 28 |
-
"grad_norm":
|
| 29 |
"learning_rate": 5.208333333333333e-08,
|
| 30 |
-
"logits/chosen": -2.
|
| 31 |
-
"logits/rejected": -2.
|
| 32 |
-
"logps/chosen": -
|
| 33 |
-
"logps/rejected": -
|
| 34 |
-
"loss": 0.
|
| 35 |
-
"rewards/accuracies": 0.
|
| 36 |
-
"rewards/chosen":
|
| 37 |
-
"rewards/margins":
|
| 38 |
-
"rewards/rejected": 4.
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.04186289900575615,
|
| 43 |
-
"grad_norm":
|
| 44 |
"learning_rate": 1.0416666666666667e-07,
|
| 45 |
-
"logits/chosen": -2.
|
| 46 |
-
"logits/rejected": -2.
|
| 47 |
-
"logps/chosen": -
|
| 48 |
-
"logps/rejected": -
|
| 49 |
-
"loss": 0.
|
| 50 |
-
"rewards/accuracies": 0.
|
| 51 |
-
"rewards/chosen": 0.
|
| 52 |
-
"rewards/margins":
|
| 53 |
-
"rewards/rejected": 0.
|
| 54 |
"step": 20
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.06279434850863422,
|
| 58 |
-
"grad_norm": 7.
|
| 59 |
"learning_rate": 1.5624999999999999e-07,
|
| 60 |
-
"logits/chosen": -2.
|
| 61 |
-
"logits/rejected": -2.
|
| 62 |
-
"logps/chosen": -
|
| 63 |
-
"logps/rejected": -
|
| 64 |
"loss": 0.6923,
|
| 65 |
-
"rewards/accuracies": 0.
|
| 66 |
-
"rewards/chosen": 0.
|
| 67 |
-
"rewards/margins": 0.
|
| 68 |
-
"rewards/rejected": 0.
|
| 69 |
"step": 30
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.0837257980115123,
|
| 73 |
-
"grad_norm": 7.
|
| 74 |
"learning_rate": 2.0833333333333333e-07,
|
| 75 |
-
"logits/chosen": -2.
|
| 76 |
-
"logits/rejected": -2.
|
| 77 |
-
"logps/chosen": -
|
| 78 |
-
"logps/rejected": -
|
| 79 |
-
"loss": 0.
|
| 80 |
-
"rewards/accuracies": 0.
|
| 81 |
-
"rewards/chosen": 0.
|
| 82 |
-
"rewards/margins": 0.
|
| 83 |
-
"rewards/rejected": 0.
|
| 84 |
"step": 40
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"epoch": 0.10465724751439037,
|
| 88 |
-
"grad_norm":
|
| 89 |
"learning_rate": 2.604166666666667e-07,
|
| 90 |
-
"logits/chosen": -2.
|
| 91 |
-
"logits/rejected": -2.
|
| 92 |
-
"logps/chosen": -
|
| 93 |
-
"logps/rejected": -
|
| 94 |
-
"loss": 0.
|
| 95 |
-
"rewards/accuracies": 0.
|
| 96 |
-
"rewards/chosen": 0.
|
| 97 |
-
"rewards/margins": 0.
|
| 98 |
-
"rewards/rejected": 0.
|
| 99 |
"step": 50
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.12558869701726844,
|
| 103 |
-
"grad_norm":
|
| 104 |
"learning_rate": 3.1249999999999997e-07,
|
| 105 |
-
"logits/chosen": -2.
|
| 106 |
-
"logits/rejected": -2.
|
| 107 |
-
"logps/chosen": -
|
| 108 |
-
"logps/rejected": -
|
| 109 |
-
"loss": 0.
|
| 110 |
-
"rewards/accuracies": 0.
|
| 111 |
-
"rewards/chosen": 0.
|
| 112 |
-
"rewards/margins": 0.
|
| 113 |
-
"rewards/rejected": 0.
|
| 114 |
"step": 60
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.14652014652014653,
|
| 118 |
-
"grad_norm":
|
| 119 |
"learning_rate": 3.645833333333333e-07,
|
| 120 |
-
"logits/chosen": -2.
|
| 121 |
-
"logits/rejected": -2.
|
| 122 |
-
"logps/chosen": -
|
| 123 |
-
"logps/rejected": -
|
| 124 |
-
"loss": 0.
|
| 125 |
-
"rewards/accuracies": 0.
|
| 126 |
-
"rewards/chosen": 0.
|
| 127 |
-
"rewards/margins": 0.
|
| 128 |
-
"rewards/rejected": 0.
|
| 129 |
"step": 70
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.1674515960230246,
|
| 133 |
-
"grad_norm":
|
| 134 |
"learning_rate": 4.1666666666666667e-07,
|
| 135 |
-
"logits/chosen": -2.
|
| 136 |
-
"logits/rejected": -2.
|
| 137 |
-
"logps/chosen": -
|
| 138 |
-
"logps/rejected": -
|
| 139 |
-
"loss": 0.
|
| 140 |
-
"rewards/accuracies": 0.
|
| 141 |
-
"rewards/chosen": 0.
|
| 142 |
-
"rewards/margins": 0.
|
| 143 |
-
"rewards/rejected": -0.
|
| 144 |
"step": 80
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.18838304552590268,
|
| 148 |
-
"grad_norm": 10.
|
| 149 |
"learning_rate": 4.6874999999999996e-07,
|
| 150 |
-
"logits/chosen": -2.
|
| 151 |
-
"logits/rejected": -2.
|
| 152 |
-
"logps/chosen": -
|
| 153 |
-
"logps/rejected": -
|
| 154 |
-
"loss": 0.
|
| 155 |
-
"rewards/accuracies": 0.
|
| 156 |
-
"rewards/chosen": -0.
|
| 157 |
-
"rewards/margins": 0.
|
| 158 |
-
"rewards/rejected": -0.
|
| 159 |
"step": 90
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.20931449502878074,
|
| 163 |
-
"grad_norm":
|
| 164 |
"learning_rate": 4.999731868769026e-07,
|
| 165 |
-
"logits/chosen": -2.
|
| 166 |
-
"logits/rejected": -2.
|
| 167 |
-
"logps/chosen": -
|
| 168 |
-
"logps/rejected": -
|
| 169 |
-
"loss": 0.
|
| 170 |
-
"rewards/accuracies": 0.
|
| 171 |
-
"rewards/chosen": -0.
|
| 172 |
-
"rewards/margins": 0.
|
| 173 |
-
"rewards/rejected": -0.
|
| 174 |
"step": 100
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.2302459445316588,
|
| 178 |
-
"grad_norm":
|
| 179 |
"learning_rate": 4.996716052911017e-07,
|
| 180 |
-
"logits/chosen": -2.
|
| 181 |
-
"logits/rejected": -2.
|
| 182 |
-
"logps/chosen": -
|
| 183 |
-
"logps/rejected": -
|
| 184 |
-
"loss": 0.
|
| 185 |
-
"rewards/accuracies": 0.
|
| 186 |
-
"rewards/chosen": -0.
|
| 187 |
-
"rewards/margins": 0.
|
| 188 |
-
"rewards/rejected": -0.
|
| 189 |
"step": 110
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.25117739403453687,
|
| 193 |
-
"grad_norm":
|
| 194 |
"learning_rate": 4.990353313429303e-07,
|
| 195 |
-
"logits/chosen": -2.
|
| 196 |
-
"logits/rejected": -2.
|
| 197 |
-
"logps/chosen": -
|
| 198 |
-
"logps/rejected": -
|
| 199 |
-
"loss": 0.
|
| 200 |
-
"rewards/accuracies": 0.
|
| 201 |
-
"rewards/chosen": -0.
|
| 202 |
-
"rewards/margins": 0.
|
| 203 |
-
"rewards/rejected": -0.
|
| 204 |
"step": 120
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 0.272108843537415,
|
| 208 |
-
"grad_norm":
|
| 209 |
"learning_rate": 4.980652179769217e-07,
|
| 210 |
-
"logits/chosen": -2.
|
| 211 |
-
"logits/rejected": -2.
|
| 212 |
-
"logps/chosen": -
|
| 213 |
-
"logps/rejected": -
|
| 214 |
-
"loss": 0.
|
| 215 |
-
"rewards/accuracies": 0.
|
| 216 |
-
"rewards/chosen": -0.
|
| 217 |
-
"rewards/margins": 0.
|
| 218 |
-
"rewards/rejected": -0.
|
| 219 |
"step": 130
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.29304029304029305,
|
| 223 |
-
"grad_norm":
|
| 224 |
"learning_rate": 4.967625656594781e-07,
|
| 225 |
-
"logits/chosen": -2.
|
| 226 |
-
"logits/rejected": -2.
|
| 227 |
-
"logps/chosen": -
|
| 228 |
-
"logps/rejected": -
|
| 229 |
-
"loss": 0.
|
| 230 |
-
"rewards/accuracies": 0.
|
| 231 |
-
"rewards/chosen": -0.
|
| 232 |
-
"rewards/margins": 0.
|
| 233 |
-
"rewards/rejected": -0.
|
| 234 |
"step": 140
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.3139717425431711,
|
| 238 |
-
"grad_norm":
|
| 239 |
"learning_rate": 4.951291206355559e-07,
|
| 240 |
-
"logits/chosen": -2.
|
| 241 |
-
"logits/rejected": -2.
|
| 242 |
-
"logps/chosen": -
|
| 243 |
-
"logps/rejected": -
|
| 244 |
-
"loss": 0.
|
| 245 |
-
"rewards/accuracies": 0.
|
| 246 |
-
"rewards/chosen": -0.
|
| 247 |
-
"rewards/margins": 0.
|
| 248 |
-
"rewards/rejected": -0.
|
| 249 |
"step": 150
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.3349031920460492,
|
| 253 |
-
"grad_norm":
|
| 254 |
"learning_rate": 4.93167072587771e-07,
|
| 255 |
-
"logits/chosen": -2.
|
| 256 |
-
"logits/rejected": -2.
|
| 257 |
-
"logps/chosen": -
|
| 258 |
-
"logps/rejected": -
|
| 259 |
-
"loss": 0.
|
| 260 |
-
"rewards/accuracies": 0.
|
| 261 |
-
"rewards/chosen": -0.
|
| 262 |
-
"rewards/margins": 0.
|
| 263 |
-
"rewards/rejected": -0.
|
| 264 |
"step": 160
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 0.35583464154892724,
|
| 268 |
-
"grad_norm":
|
| 269 |
"learning_rate": 4.908790517010636e-07,
|
| 270 |
-
"logits/chosen": -2.
|
| 271 |
-
"logits/rejected": -2.
|
| 272 |
-
"logps/chosen": -
|
| 273 |
-
"logps/rejected": -
|
| 274 |
-
"loss": 0.
|
| 275 |
-
"rewards/accuracies": 0.
|
| 276 |
-
"rewards/chosen": -0.
|
| 277 |
-
"rewards/margins": 0.
|
| 278 |
-
"rewards/rejected": -0.
|
| 279 |
"step": 170
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 0.37676609105180536,
|
| 283 |
-
"grad_norm":
|
| 284 |
"learning_rate": 4.882681251368548e-07,
|
| 285 |
-
"logits/chosen": -2.
|
| 286 |
-
"logits/rejected": -2.
|
| 287 |
-
"logps/chosen": -
|
| 288 |
-
"logps/rejected": -
|
| 289 |
-
"loss": 0.
|
| 290 |
-
"rewards/accuracies": 0.
|
| 291 |
-
"rewards/chosen": -0.
|
| 292 |
-
"rewards/margins": 0.
|
| 293 |
-
"rewards/rejected": -0.
|
| 294 |
"step": 180
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 0.3976975405546834,
|
| 298 |
-
"grad_norm":
|
| 299 |
"learning_rate": 4.853377929214243e-07,
|
| 300 |
-
"logits/chosen": -
|
| 301 |
-
"logits/rejected": -
|
| 302 |
-
"logps/chosen": -
|
| 303 |
-
"logps/rejected": -
|
| 304 |
-
"loss": 0.
|
| 305 |
-
"rewards/accuracies": 0.
|
| 306 |
-
"rewards/chosen": -0.
|
| 307 |
-
"rewards/margins": 0.
|
| 308 |
-
"rewards/rejected": -
|
| 309 |
"step": 190
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 0.4186289900575615,
|
| 313 |
-
"grad_norm":
|
| 314 |
"learning_rate": 4.820919832540181e-07,
|
| 315 |
-
"logits/chosen": -
|
| 316 |
-
"logits/rejected": -
|
| 317 |
-
"logps/chosen": -
|
| 318 |
-
"logps/rejected": -
|
| 319 |
-
"loss": 0.
|
| 320 |
"rewards/accuracies": 0.71875,
|
| 321 |
-
"rewards/chosen": -0.
|
| 322 |
-
"rewards/margins": 0.
|
| 323 |
-
"rewards/rejected": -1.
|
| 324 |
"step": 200
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.43956043956043955,
|
| 328 |
-
"grad_norm":
|
| 329 |
"learning_rate": 4.785350472409791e-07,
|
| 330 |
-
"logits/chosen": -
|
| 331 |
-
"logits/rejected": -
|
| 332 |
-
"logps/chosen": -
|
| 333 |
-
"logps/rejected": -
|
| 334 |
-
"loss": 0.
|
| 335 |
-
"rewards/accuracies": 0.
|
| 336 |
-
"rewards/chosen": -0.
|
| 337 |
-
"rewards/margins": 0.
|
| 338 |
-
"rewards/rejected": -1.
|
| 339 |
"step": 210
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.4604918890633176,
|
| 343 |
-
"grad_norm":
|
| 344 |
"learning_rate": 4.7467175306295647e-07,
|
| 345 |
-
"logits/chosen": -
|
| 346 |
-
"logits/rejected": -
|
| 347 |
-
"logps/chosen": -
|
| 348 |
-
"logps/rejected": -
|
| 349 |
-
"loss": 0.
|
| 350 |
-
"rewards/accuracies": 0.
|
| 351 |
-
"rewards/chosen": -0.
|
| 352 |
-
"rewards/margins": 0.
|
| 353 |
-
"rewards/rejected": -1.
|
| 354 |
"step": 220
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.48142333856619574,
|
| 358 |
-
"grad_norm":
|
| 359 |
"learning_rate": 4.70507279583015e-07,
|
| 360 |
-
"logits/chosen": -
|
| 361 |
-
"logits/rejected": -
|
| 362 |
-
"logps/chosen": -
|
| 363 |
-
"logps/rejected": -
|
| 364 |
-
"loss": 0.
|
| 365 |
-
"rewards/accuracies": 0.
|
| 366 |
-
"rewards/chosen": -0.
|
| 367 |
-
"rewards/margins": 0.
|
| 368 |
-
"rewards/rejected": -1.
|
| 369 |
"step": 230
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 0.5023547880690737,
|
| 373 |
-
"grad_norm":
|
| 374 |
"learning_rate": 4.6604720940421207e-07,
|
| 375 |
-
"logits/chosen": -
|
| 376 |
-
"logits/rejected": -
|
| 377 |
-
"logps/chosen": -
|
| 378 |
-
"logps/rejected": -
|
| 379 |
-
"loss": 0.
|
| 380 |
-
"rewards/accuracies": 0.
|
| 381 |
-
"rewards/chosen": -0.
|
| 382 |
-
"rewards/margins": 0.
|
| 383 |
-
"rewards/rejected": -1.
|
| 384 |
"step": 240
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 0.5232862375719518,
|
| 388 |
-
"grad_norm":
|
| 389 |
"learning_rate": 4.612975213859487e-07,
|
| 390 |
-
"logits/chosen": -
|
| 391 |
-
"logits/rejected": -
|
| 392 |
-
"logps/chosen": -
|
| 393 |
-
"logps/rejected": -
|
| 394 |
-
"loss": 0.
|
| 395 |
-
"rewards/accuracies": 0.
|
| 396 |
-
"rewards/chosen": -0.
|
| 397 |
-
"rewards/margins": 0.
|
| 398 |
-
"rewards/rejected": -1.
|
| 399 |
"step": 250
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 0.54421768707483,
|
| 403 |
-
"grad_norm":
|
| 404 |
"learning_rate": 4.5626458262912735e-07,
|
| 405 |
-
"logits/chosen": -
|
| 406 |
-
"logits/rejected": -
|
| 407 |
-
"logps/chosen": -
|
| 408 |
-
"logps/rejected": -
|
| 409 |
-
"loss": 0.
|
| 410 |
-
"rewards/accuracies": 0.
|
| 411 |
-
"rewards/chosen": -0.
|
| 412 |
-
"rewards/margins": 0.
|
| 413 |
-
"rewards/rejected": -1.
|
| 414 |
"step": 260
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 0.565149136577708,
|
| 418 |
-
"grad_norm":
|
| 419 |
"learning_rate": 4.5095513994085974e-07,
|
| 420 |
-
"logits/chosen": -
|
| 421 |
-
"logits/rejected": -
|
| 422 |
-
"logps/chosen": -
|
| 423 |
-
"logps/rejected": -
|
| 424 |
-
"loss": 0.
|
| 425 |
-
"rewards/accuracies": 0.
|
| 426 |
-
"rewards/chosen": -0.
|
| 427 |
-
"rewards/margins": 0.
|
| 428 |
-
"rewards/rejected": -1.
|
| 429 |
"step": 270
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.5860805860805861,
|
| 433 |
-
"grad_norm": 24.
|
| 434 |
"learning_rate": 4.453763107901675e-07,
|
| 435 |
-
"logits/chosen": -
|
| 436 |
-
"logits/rejected": -
|
| 437 |
-
"logps/chosen": -
|
| 438 |
-
"logps/rejected": -
|
| 439 |
-
"loss": 0.
|
| 440 |
"rewards/accuracies": 0.6937500238418579,
|
| 441 |
-
"rewards/chosen": -0.
|
| 442 |
-
"rewards/margins": 0.
|
| 443 |
-
"rewards/rejected": -1.
|
| 444 |
"step": 280
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.6070120355834642,
|
| 448 |
-
"grad_norm":
|
| 449 |
"learning_rate": 4.395355737667985e-07,
|
| 450 |
-
"logits/chosen": -
|
| 451 |
-
"logits/rejected": -
|
| 452 |
-
"logps/chosen": -
|
| 453 |
-
"logps/rejected": -
|
| 454 |
-
"loss": 0.
|
| 455 |
"rewards/accuracies": 0.7562500238418579,
|
| 456 |
-
"rewards/chosen": -
|
| 457 |
-
"rewards/margins": 0.
|
| 458 |
-
"rewards/rejected": -1.
|
| 459 |
"step": 290
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.6279434850863422,
|
| 463 |
-
"grad_norm":
|
| 464 |
"learning_rate": 4.3344075855595097e-07,
|
| 465 |
-
"logits/chosen": -
|
| 466 |
-
"logits/rejected": -
|
| 467 |
-
"logps/chosen": -
|
| 468 |
-
"logps/rejected": -
|
| 469 |
-
"loss": 0.
|
| 470 |
"rewards/accuracies": 0.731249988079071,
|
| 471 |
-
"rewards/chosen": -0.
|
| 472 |
-
"rewards/margins": 0.
|
| 473 |
-
"rewards/rejected": -1.
|
| 474 |
"step": 300
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.6488749345892203,
|
| 478 |
-
"grad_norm":
|
| 479 |
"learning_rate": 4.271000354423425e-07,
|
| 480 |
-
"logits/chosen": -
|
| 481 |
-
"logits/rejected": -
|
| 482 |
-
"logps/chosen": -
|
| 483 |
-
"logps/rejected": -
|
| 484 |
-
"loss": 0.
|
| 485 |
-
"rewards/accuracies": 0.
|
| 486 |
-
"rewards/chosen": -0.
|
| 487 |
-
"rewards/margins": 0.
|
| 488 |
-
"rewards/rejected": -1.
|
| 489 |
"step": 310
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.6698063840920984,
|
| 493 |
-
"grad_norm":
|
| 494 |
"learning_rate": 4.2052190435769554e-07,
|
| 495 |
-
"logits/chosen": -
|
| 496 |
-
"logits/rejected": -
|
| 497 |
-
"logps/chosen": -
|
| 498 |
-
"logps/rejected": -
|
| 499 |
-
"loss": 0.
|
| 500 |
-
"rewards/accuracies": 0.
|
| 501 |
-
"rewards/chosen": -
|
| 502 |
-
"rewards/margins": 0.
|
| 503 |
-
"rewards/rejected": -
|
| 504 |
"step": 320
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 0.6907378335949764,
|
| 508 |
-
"grad_norm": 25.
|
| 509 |
"learning_rate": 4.137151834863213e-07,
|
| 510 |
-
"logits/chosen": -
|
| 511 |
-
"logits/rejected": -
|
| 512 |
-
"logps/chosen": -
|
| 513 |
-
"logps/rejected": -
|
| 514 |
-
"loss": 0.
|
| 515 |
"rewards/accuracies": 0.7562500238418579,
|
| 516 |
-
"rewards/chosen": -0.
|
| 517 |
-
"rewards/margins": 0.
|
| 518 |
-
"rewards/rejected": -1.
|
| 519 |
"step": 330
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.7116692830978545,
|
| 523 |
-
"grad_norm":
|
| 524 |
"learning_rate": 4.0668899744407567e-07,
|
| 525 |
-
"logits/chosen": -
|
| 526 |
-
"logits/rejected": -
|
| 527 |
-
"logps/chosen": -
|
| 528 |
-
"logps/rejected": -
|
| 529 |
-
"loss": 0.
|
| 530 |
-
"rewards/accuracies": 0.
|
| 531 |
-
"rewards/chosen": -
|
| 532 |
-
"rewards/margins": 0.
|
| 533 |
-
"rewards/rejected": -1.
|
| 534 |
"step": 340
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.7326007326007326,
|
| 538 |
-
"grad_norm":
|
| 539 |
"learning_rate": 3.994527650465352e-07,
|
| 540 |
-
"logits/chosen": -
|
| 541 |
-
"logits/rejected": -
|
| 542 |
-
"logps/chosen": -
|
| 543 |
-
"logps/rejected": -
|
| 544 |
-
"loss": 0.
|
| 545 |
"rewards/accuracies": 0.8187500238418579,
|
| 546 |
-
"rewards/chosen": -
|
| 547 |
-
"rewards/margins": 0.
|
| 548 |
-
"rewards/rejected": -1.
|
| 549 |
"step": 350
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.7535321821036107,
|
| 553 |
-
"grad_norm":
|
| 554 |
"learning_rate": 3.920161866827889e-07,
|
| 555 |
-
"logits/chosen": -
|
| 556 |
-
"logits/rejected": -
|
| 557 |
-
"logps/chosen": -
|
| 558 |
-
"logps/rejected": -
|
| 559 |
-
"loss": 0.
|
| 560 |
-
"rewards/accuracies": 0.
|
| 561 |
-
"rewards/chosen": -
|
| 562 |
-
"rewards/margins": 0.
|
| 563 |
-
"rewards/rejected": -1.
|
| 564 |
"step": 360
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 0.7744636316064888,
|
| 568 |
-
"grad_norm":
|
| 569 |
"learning_rate": 3.8438923131177237e-07,
|
| 570 |
-
"logits/chosen": -
|
| 571 |
-
"logits/rejected": -
|
| 572 |
-
"logps/chosen": -
|
| 573 |
-
"logps/rejected": -
|
| 574 |
-
"loss": 0.
|
| 575 |
-
"rewards/accuracies": 0.
|
| 576 |
-
"rewards/chosen": -
|
| 577 |
-
"rewards/margins": 0.
|
| 578 |
-
"rewards/rejected": -
|
| 579 |
"step": 370
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.7953950811093669,
|
| 583 |
-
"grad_norm":
|
| 584 |
"learning_rate": 3.765821230985757e-07,
|
| 585 |
-
"logits/chosen": -
|
| 586 |
-
"logits/rejected": -
|
| 587 |
-
"logps/chosen": -
|
| 588 |
-
"logps/rejected": -
|
| 589 |
-
"loss": 0.
|
| 590 |
"rewards/accuracies": 0.737500011920929,
|
| 591 |
-
"rewards/chosen": -
|
| 592 |
-
"rewards/margins": 0.
|
| 593 |
-
"rewards/rejected": -
|
| 594 |
"step": 380
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 0.8163265306122449,
|
| 598 |
-
"grad_norm":
|
| 599 |
"learning_rate": 3.6860532770864005e-07,
|
| 600 |
-
"logits/chosen": -
|
| 601 |
-
"logits/rejected": -
|
| 602 |
-
"logps/chosen": -
|
| 603 |
-
"logps/rejected": -
|
| 604 |
-
"loss": 0.
|
| 605 |
-
"rewards/accuracies": 0.
|
| 606 |
-
"rewards/chosen": -
|
| 607 |
-
"rewards/margins": 0.
|
| 608 |
-
"rewards/rejected": -1.
|
| 609 |
"step": 390
|
| 610 |
},
|
| 611 |
{
|
| 612 |
"epoch": 0.837257980115123,
|
| 613 |
-
"grad_norm":
|
| 614 |
"learning_rate": 3.604695382782159e-07,
|
| 615 |
-
"logits/chosen": -
|
| 616 |
-
"logits/rejected": -
|
| 617 |
-
"logps/chosen": -
|
| 618 |
-
"logps/rejected": -
|
| 619 |
-
"loss": 0.
|
| 620 |
-
"rewards/accuracies": 0.
|
| 621 |
-
"rewards/chosen": -
|
| 622 |
-
"rewards/margins": 0.
|
| 623 |
-
"rewards/rejected": -
|
| 624 |
"step": 400
|
| 625 |
},
|
| 626 |
{
|
| 627 |
"epoch": 0.858189429618001,
|
| 628 |
-
"grad_norm":
|
| 629 |
"learning_rate": 3.5218566107988867e-07,
|
| 630 |
-
"logits/chosen": -
|
| 631 |
-
"logits/rejected": -
|
| 632 |
-
"logps/chosen": -
|
| 633 |
-
"logps/rejected": -
|
| 634 |
-
"loss": 0.
|
| 635 |
"rewards/accuracies": 0.7437499761581421,
|
| 636 |
-
"rewards/chosen": -
|
| 637 |
-
"rewards/margins": 0.
|
| 638 |
-
"rewards/rejected": -
|
| 639 |
"step": 410
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.8791208791208791,
|
| 643 |
-
"grad_norm":
|
| 644 |
"learning_rate": 3.4376480090239047e-07,
|
| 645 |
-
"logits/chosen": -
|
| 646 |
-
"logits/rejected": -
|
| 647 |
-
"logps/chosen": -
|
| 648 |
-
"logps/rejected": -
|
| 649 |
-
"loss": 0.
|
| 650 |
-
"rewards/accuracies": 0.
|
| 651 |
-
"rewards/chosen": -
|
| 652 |
-
"rewards/margins": 0.
|
| 653 |
-
"rewards/rejected": -
|
| 654 |
"step": 420
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.9000523286237572,
|
| 658 |
-
"grad_norm":
|
| 659 |
"learning_rate": 3.3521824616429284e-07,
|
| 660 |
-
"logits/chosen": -
|
| 661 |
-
"logits/rejected": -
|
| 662 |
-
"logps/chosen": -
|
| 663 |
-
"logps/rejected": -
|
| 664 |
-
"loss": 0.
|
| 665 |
-
"rewards/accuracies": 0.
|
| 666 |
-
"rewards/chosen": -
|
| 667 |
-
"rewards/margins": 0.
|
| 668 |
-
"rewards/rejected": -
|
| 669 |
"step": 430
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 0.9209837781266352,
|
| 673 |
-
"grad_norm": 27.
|
| 674 |
"learning_rate": 3.265574537815398e-07,
|
| 675 |
-
"logits/chosen": -
|
| 676 |
-
"logits/rejected": -
|
| 677 |
-
"logps/chosen": -
|
| 678 |
-
"logps/rejected": -
|
| 679 |
-
"loss": 0.
|
| 680 |
-
"rewards/accuracies": 0.
|
| 681 |
-
"rewards/chosen": -
|
| 682 |
-
"rewards/margins": 0.
|
| 683 |
-
"rewards/rejected": -
|
| 684 |
"step": 440
|
| 685 |
},
|
| 686 |
{
|
| 687 |
"epoch": 0.9419152276295133,
|
| 688 |
-
"grad_norm":
|
| 689 |
"learning_rate": 3.1779403380910425e-07,
|
| 690 |
-
"logits/chosen": -
|
| 691 |
-
"logits/rejected": -
|
| 692 |
-
"logps/chosen": -
|
| 693 |
-
"logps/rejected": -
|
| 694 |
-
"loss": 0.
|
| 695 |
-
"rewards/accuracies": 0.
|
| 696 |
-
"rewards/chosen": -
|
| 697 |
-
"rewards/margins": 0.
|
| 698 |
-
"rewards/rejected": -
|
| 699 |
"step": 450
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.9628466771323915,
|
| 703 |
-
"grad_norm":
|
| 704 |
"learning_rate": 3.0893973387735683e-07,
|
| 705 |
-
"logits/chosen": -
|
| 706 |
-
"logits/rejected": -
|
| 707 |
-
"logps/chosen": -
|
| 708 |
-
"logps/rejected": -
|
| 709 |
-
"loss": 0.
|
| 710 |
-
"rewards/accuracies": 0.
|
| 711 |
-
"rewards/chosen": -
|
| 712 |
-
"rewards/margins": 0.
|
| 713 |
-
"rewards/rejected": -1.
|
| 714 |
"step": 460
|
| 715 |
},
|
| 716 |
{
|
| 717 |
"epoch": 0.9837781266352695,
|
| 718 |
-
"grad_norm": 23.
|
| 719 |
"learning_rate": 3.000064234440111e-07,
|
| 720 |
-
"logits/chosen": -
|
| 721 |
-
"logits/rejected": -
|
| 722 |
-
"logps/chosen": -
|
| 723 |
-
"logps/rejected": -
|
| 724 |
-
"loss": 0.
|
| 725 |
-
"rewards/accuracies": 0.
|
| 726 |
-
"rewards/chosen": -
|
| 727 |
-
"rewards/margins": 0.
|
| 728 |
-
"rewards/rejected": -1.
|
| 729 |
"step": 470
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.9984301412872841,
|
| 733 |
-
"eval_logits/chosen": -
|
| 734 |
-
"eval_logits/rejected": -
|
| 735 |
-
"eval_logps/chosen": -
|
| 736 |
-
"eval_logps/rejected": -
|
| 737 |
-
"eval_loss": 0.
|
| 738 |
-
"eval_rewards/accuracies": 0.
|
| 739 |
-
"eval_rewards/chosen": -
|
| 740 |
-
"eval_rewards/margins": 0.
|
| 741 |
-
"eval_rewards/rejected": -
|
| 742 |
-
"eval_runtime":
|
| 743 |
-
"eval_samples_per_second": 11.
|
| 744 |
-
"eval_steps_per_second": 0.
|
| 745 |
"step": 477
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 1.0047095761381475,
|
| 749 |
-
"grad_norm":
|
| 750 |
"learning_rate": 2.910060778827554e-07,
|
| 751 |
-
"logits/chosen": -
|
| 752 |
-
"logits/rejected": -
|
| 753 |
-
"logps/chosen": -
|
| 754 |
-
"logps/rejected": -
|
| 755 |
-
"loss": 0.
|
| 756 |
-
"rewards/accuracies": 0.
|
| 757 |
-
"rewards/chosen": -
|
| 758 |
-
"rewards/margins": 0.
|
| 759 |
-
"rewards/rejected": -
|
| 760 |
"step": 480
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 1.0256410256410255,
|
| 764 |
-
"grad_norm":
|
| 765 |
"learning_rate": 2.8195076242990116e-07,
|
| 766 |
-
"logits/chosen": -
|
| 767 |
-
"logits/rejected": -
|
| 768 |
-
"logps/chosen": -
|
| 769 |
-
"logps/rejected": -
|
| 770 |
-
"loss": 0.
|
| 771 |
-
"rewards/accuracies": 0.
|
| 772 |
-
"rewards/chosen": -
|
| 773 |
-
"rewards/margins": 1.
|
| 774 |
-
"rewards/rejected": -
|
| 775 |
"step": 490
|
| 776 |
},
|
| 777 |
{
|
| 778 |
"epoch": 1.0465724751439036,
|
| 779 |
-
"grad_norm":
|
| 780 |
"learning_rate": 2.7285261601056697e-07,
|
| 781 |
-
"logits/chosen": -
|
| 782 |
-
"logits/rejected": -
|
| 783 |
-
"logps/chosen": -
|
| 784 |
-
"logps/rejected": -
|
| 785 |
-
"loss": 0.
|
| 786 |
-
"rewards/accuracies": 0.
|
| 787 |
-
"rewards/chosen": -
|
| 788 |
-
"rewards/margins": 1.
|
| 789 |
-
"rewards/rejected": -
|
| 790 |
"step": 500
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 1.0675039246467817,
|
| 794 |
-
"grad_norm":
|
| 795 |
"learning_rate": 2.6372383496608186e-07,
|
| 796 |
-
"logits/chosen": -
|
| 797 |
-
"logits/rejected":
|
| 798 |
-
"logps/chosen": -
|
| 799 |
-
"logps/rejected": -
|
| 800 |
-
"loss": 0.
|
| 801 |
-
"rewards/accuracies": 0.
|
| 802 |
-
"rewards/chosen": -
|
| 803 |
-
"rewards/margins": 1.
|
| 804 |
-
"rewards/rejected": -
|
| 805 |
"step": 510
|
| 806 |
},
|
| 807 |
{
|
| 808 |
"epoch": 1.08843537414966,
|
| 809 |
-
"grad_norm":
|
| 810 |
"learning_rate": 2.5457665670441937e-07,
|
| 811 |
-
"logits/chosen": -
|
| 812 |
-
"logits/rejected":
|
| 813 |
-
"logps/chosen": -
|
| 814 |
-
"logps/rejected": -
|
| 815 |
-
"loss": 0.
|
| 816 |
-
"rewards/accuracies": 0.
|
| 817 |
-
"rewards/chosen": -
|
| 818 |
-
"rewards/margins": 1.
|
| 819 |
-
"rewards/rejected": -
|
| 820 |
"step": 520
|
| 821 |
},
|
| 822 |
{
|
| 823 |
"epoch": 1.109366823652538,
|
| 824 |
-
"grad_norm":
|
| 825 |
"learning_rate": 2.454233432955807e-07,
|
| 826 |
-
"logits/chosen": -
|
| 827 |
-
"logits/rejected":
|
| 828 |
-
"logps/chosen": -
|
| 829 |
-
"logps/rejected": -
|
| 830 |
-
"loss": 0.
|
| 831 |
-
"rewards/accuracies": 0.
|
| 832 |
-
"rewards/chosen": -1.
|
| 833 |
-
"rewards/margins": 1.
|
| 834 |
-
"rewards/rejected": -
|
| 835 |
"step": 530
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 1.130298273155416,
|
| 839 |
-
"grad_norm":
|
| 840 |
"learning_rate": 2.3627616503391812e-07,
|
| 841 |
-
"logits/chosen": -
|
| 842 |
-
"logits/rejected":
|
| 843 |
-
"logps/chosen": -
|
| 844 |
-
"logps/rejected": -
|
| 845 |
-
"loss": 0.
|
| 846 |
"rewards/accuracies": 0.8812500238418579,
|
| 847 |
-
"rewards/chosen": -1.
|
| 848 |
-
"rewards/margins": 1.
|
| 849 |
-
"rewards/rejected": -
|
| 850 |
"step": 540
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 1.1512297226582942,
|
| 854 |
-
"grad_norm":
|
| 855 |
"learning_rate": 2.2714738398943308e-07,
|
| 856 |
-
"logits/chosen": -
|
| 857 |
-
"logits/rejected": -
|
| 858 |
-
"logps/chosen": -
|
| 859 |
-
"logps/rejected": -
|
| 860 |
-
"loss": 0.
|
| 861 |
-
"rewards/accuracies": 0.
|
| 862 |
-
"rewards/chosen": -1.
|
| 863 |
-
"rewards/margins": 1.
|
| 864 |
-
"rewards/rejected": -
|
| 865 |
"step": 550
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 1.1721611721611722,
|
| 869 |
-
"grad_norm":
|
| 870 |
"learning_rate": 2.1804923757009882e-07,
|
| 871 |
-
"logits/chosen": -
|
| 872 |
-
"logits/rejected": -
|
| 873 |
-
"logps/chosen": -
|
| 874 |
-
"logps/rejected": -
|
| 875 |
-
"loss": 0.
|
| 876 |
-
"rewards/accuracies": 0.
|
| 877 |
-
"rewards/chosen": -
|
| 878 |
-
"rewards/margins": 1.
|
| 879 |
-
"rewards/rejected": -
|
| 880 |
"step": 560
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 1.1930926216640503,
|
| 884 |
-
"grad_norm":
|
| 885 |
"learning_rate": 2.089939221172446e-07,
|
| 886 |
-
"logits/chosen": -
|
| 887 |
-
"logits/rejected": -
|
| 888 |
-
"logps/chosen": -
|
| 889 |
-
"logps/rejected": -
|
| 890 |
-
"loss": 0.
|
| 891 |
-
"rewards/accuracies": 0.
|
| 892 |
-
"rewards/chosen": -1.
|
| 893 |
-
"rewards/margins": 1.
|
| 894 |
-
"rewards/rejected": -
|
| 895 |
"step": 570
|
| 896 |
},
|
| 897 |
{
|
| 898 |
"epoch": 1.2140240711669283,
|
| 899 |
-
"grad_norm":
|
| 900 |
"learning_rate": 1.9999357655598891e-07,
|
| 901 |
-
"logits/chosen": -
|
| 902 |
-
"logits/rejected": -
|
| 903 |
-
"logps/chosen": -
|
| 904 |
-
"logps/rejected": -
|
| 905 |
-
"loss": 0.
|
| 906 |
"rewards/accuracies": 0.875,
|
| 907 |
-
"rewards/chosen": -1.
|
| 908 |
-
"rewards/margins": 1.
|
| 909 |
-
"rewards/rejected": -
|
| 910 |
"step": 580
|
| 911 |
},
|
| 912 |
{
|
| 913 |
"epoch": 1.2349555206698064,
|
| 914 |
-
"grad_norm":
|
| 915 |
"learning_rate": 1.9106026612264315e-07,
|
| 916 |
-
"logits/chosen": -
|
| 917 |
-
"logits/rejected": -
|
| 918 |
-
"logps/chosen": -
|
| 919 |
-
"logps/rejected": -
|
| 920 |
-
"loss": 0.
|
| 921 |
-
"rewards/accuracies": 0.
|
| 922 |
-
"rewards/chosen": -
|
| 923 |
-
"rewards/margins": 1.
|
| 924 |
-
"rewards/rejected": -
|
| 925 |
"step": 590
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 1.2558869701726845,
|
| 929 |
-
"grad_norm":
|
| 930 |
"learning_rate": 1.8220596619089573e-07,
|
| 931 |
-
"logits/chosen": -
|
| 932 |
-
"logits/rejected": -
|
| 933 |
-
"logps/chosen": -
|
| 934 |
-
"logps/rejected": -
|
| 935 |
-
"loss": 0.
|
| 936 |
-
"rewards/accuracies": 0.
|
| 937 |
-
"rewards/chosen": -1.
|
| 938 |
-
"rewards/margins": 1.
|
| 939 |
-
"rewards/rejected": -
|
| 940 |
"step": 600
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 1.2768184196755625,
|
| 944 |
-
"grad_norm":
|
| 945 |
"learning_rate": 1.7344254621846017e-07,
|
| 946 |
-
"logits/chosen": -
|
| 947 |
-
"logits/rejected":
|
| 948 |
-
"logps/chosen": -
|
| 949 |
-
"logps/rejected": -
|
| 950 |
-
"loss": 0.
|
| 951 |
-
"rewards/accuracies": 0.
|
| 952 |
-
"rewards/chosen": -
|
| 953 |
-
"rewards/margins": 1.
|
| 954 |
-
"rewards/rejected": -
|
| 955 |
"step": 610
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 1.2977498691784406,
|
| 959 |
-
"grad_norm":
|
| 960 |
"learning_rate": 1.647817538357072e-07,
|
| 961 |
-
"logits/chosen": -
|
| 962 |
-
"logits/rejected": -
|
| 963 |
-
"logps/chosen": -
|
| 964 |
-
"logps/rejected": -
|
| 965 |
-
"loss": 0.
|
| 966 |
"rewards/accuracies": 0.824999988079071,
|
| 967 |
-
"rewards/chosen": -1.
|
| 968 |
-
"rewards/margins": 1.
|
| 969 |
-
"rewards/rejected": -
|
| 970 |
"step": 620
|
| 971 |
},
|
| 972 |
{
|
| 973 |
"epoch": 1.3186813186813187,
|
| 974 |
-
"grad_norm":
|
| 975 |
"learning_rate": 1.562351990976095e-07,
|
| 976 |
-
"logits/chosen": -
|
| 977 |
-
"logits/rejected":
|
| 978 |
-
"logps/chosen": -
|
| 979 |
-
"logps/rejected": -
|
| 980 |
-
"loss": 0.
|
| 981 |
-
"rewards/accuracies": 0.
|
| 982 |
-
"rewards/chosen": -1.
|
| 983 |
-
"rewards/margins": 1.
|
| 984 |
-
"rewards/rejected": -
|
| 985 |
"step": 630
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 1.3396127681841967,
|
| 989 |
-
"grad_norm":
|
| 990 |
"learning_rate": 1.478143389201113e-07,
|
| 991 |
-
"logits/chosen": -
|
| 992 |
-
"logits/rejected":
|
| 993 |
-
"logps/chosen": -
|
| 994 |
-
"logps/rejected": -
|
| 995 |
-
"loss": 0.
|
| 996 |
-
"rewards/accuracies": 0.
|
| 997 |
-
"rewards/chosen": -1.
|
| 998 |
-
"rewards/margins": 1.
|
| 999 |
-
"rewards/rejected": -
|
| 1000 |
"step": 640
|
| 1001 |
},
|
| 1002 |
{
|
| 1003 |
"epoch": 1.3605442176870748,
|
| 1004 |
-
"grad_norm":
|
| 1005 |
"learning_rate": 1.3953046172178413e-07,
|
| 1006 |
-
"logits/chosen": -
|
| 1007 |
-
"logits/rejected":
|
| 1008 |
-
"logps/chosen": -
|
| 1009 |
-
"logps/rejected": -
|
| 1010 |
-
"loss": 0.
|
| 1011 |
-
"rewards/accuracies": 0.
|
| 1012 |
-
"rewards/chosen": -1.
|
| 1013 |
-
"rewards/margins": 1.
|
| 1014 |
-
"rewards/rejected": -
|
| 1015 |
"step": 650
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
"epoch": 1.3814756671899528,
|
| 1019 |
-
"grad_norm": 34.
|
| 1020 |
"learning_rate": 1.3139467229135998e-07,
|
| 1021 |
-
"logits/chosen": -
|
| 1022 |
-
"logits/rejected":
|
| 1023 |
-
"logps/chosen": -
|
| 1024 |
-
"logps/rejected": -
|
| 1025 |
-
"loss": 0.
|
| 1026 |
-
"rewards/accuracies": 0.
|
| 1027 |
-
"rewards/chosen": -1.
|
| 1028 |
-
"rewards/margins": 1.
|
| 1029 |
-
"rewards/rejected": -
|
| 1030 |
"step": 660
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 1.402407116692831,
|
| 1034 |
-
"grad_norm":
|
| 1035 |
"learning_rate": 1.2341787690142435e-07,
|
| 1036 |
-
"logits/chosen":
|
| 1037 |
-
"logits/rejected":
|
| 1038 |
-
"logps/chosen": -
|
| 1039 |
-
"logps/rejected": -
|
| 1040 |
-
"loss": 0.
|
| 1041 |
-
"rewards/accuracies": 0.
|
| 1042 |
-
"rewards/chosen": -
|
| 1043 |
-
"rewards/margins": 1.
|
| 1044 |
-
"rewards/rejected": -
|
| 1045 |
"step": 670
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 1.423338566195709,
|
| 1049 |
-
"grad_norm":
|
| 1050 |
"learning_rate": 1.1561076868822755e-07,
|
| 1051 |
-
"logits/chosen": -
|
| 1052 |
-
"logits/rejected":
|
| 1053 |
-
"logps/chosen": -
|
| 1054 |
-
"logps/rejected": -
|
| 1055 |
-
"loss": 0.
|
| 1056 |
-
"rewards/accuracies": 0.
|
| 1057 |
-
"rewards/chosen": -1.
|
| 1058 |
-
"rewards/margins": 1.
|
| 1059 |
-
"rewards/rejected": -
|
| 1060 |
"step": 680
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 1.4442700156985873,
|
| 1064 |
-
"grad_norm":
|
| 1065 |
"learning_rate": 1.0798381331721107e-07,
|
| 1066 |
-
"logits/chosen":
|
| 1067 |
-
"logits/rejected":
|
| 1068 |
-
"logps/chosen": -
|
| 1069 |
-
"logps/rejected": -
|
| 1070 |
-
"loss": 0.
|
| 1071 |
-
"rewards/accuracies": 0.
|
| 1072 |
-
"rewards/chosen": -1.
|
| 1073 |
-
"rewards/margins": 1.
|
| 1074 |
-
"rewards/rejected": -
|
| 1075 |
"step": 690
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
"epoch": 1.4652014652014653,
|
| 1079 |
-
"grad_norm":
|
| 1080 |
"learning_rate": 1.0054723495346482e-07,
|
| 1081 |
-
"logits/chosen": -
|
| 1082 |
-
"logits/rejected":
|
| 1083 |
-
"logps/chosen": -
|
| 1084 |
-
"logps/rejected": -
|
| 1085 |
-
"loss": 0.
|
| 1086 |
-
"rewards/accuracies": 0.
|
| 1087 |
-
"rewards/chosen": -1.
|
| 1088 |
-
"rewards/margins": 1.
|
| 1089 |
-
"rewards/rejected": -3.
|
| 1090 |
"step": 700
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 1.4861329147043434,
|
| 1094 |
-
"grad_norm":
|
| 1095 |
"learning_rate": 9.331100255592436e-08,
|
| 1096 |
-
"logits/chosen": -
|
| 1097 |
-
"logits/rejected":
|
| 1098 |
-
"logps/chosen": -
|
| 1099 |
-
"logps/rejected": -
|
| 1100 |
-
"loss": 0.
|
| 1101 |
-
"rewards/accuracies": 0.
|
| 1102 |
-
"rewards/chosen": -1.
|
| 1103 |
-
"rewards/margins": 1.
|
| 1104 |
-
"rewards/rejected": -
|
| 1105 |
"step": 710
|
| 1106 |
},
|
| 1107 |
{
|
| 1108 |
"epoch": 1.5070643642072215,
|
| 1109 |
-
"grad_norm":
|
| 1110 |
"learning_rate": 8.628481651367875e-08,
|
| 1111 |
-
"logits/chosen": -
|
| 1112 |
-
"logits/rejected":
|
| 1113 |
-
"logps/chosen": -
|
| 1114 |
-
"logps/rejected": -
|
| 1115 |
-
"loss": 0.
|
| 1116 |
-
"rewards/accuracies": 0.
|
| 1117 |
-
"rewards/chosen": -
|
| 1118 |
-
"rewards/margins": 1.
|
| 1119 |
-
"rewards/rejected": -3.
|
| 1120 |
"step": 720
|
| 1121 |
},
|
| 1122 |
{
|
| 1123 |
"epoch": 1.5279958137100995,
|
| 1124 |
-
"grad_norm":
|
| 1125 |
"learning_rate": 7.947809564230445e-08,
|
| 1126 |
-
"logits/chosen": -
|
| 1127 |
-
"logits/rejected": -
|
| 1128 |
-
"logps/chosen": -
|
| 1129 |
-
"logps/rejected": -
|
| 1130 |
-
"loss": 0.
|
| 1131 |
-
"rewards/accuracies": 0.
|
| 1132 |
-
"rewards/chosen": -1.
|
| 1133 |
-
"rewards/margins": 1.
|
| 1134 |
-
"rewards/rejected": -
|
| 1135 |
"step": 730
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
"epoch": 1.5489272632129776,
|
| 1139 |
-
"grad_norm":
|
| 1140 |
"learning_rate": 7.289996455765748e-08,
|
| 1141 |
-
"logits/chosen": -
|
| 1142 |
-
"logits/rejected":
|
| 1143 |
-
"logps/chosen": -
|
| 1144 |
-
"logps/rejected": -
|
| 1145 |
-
"loss": 0.
|
| 1146 |
-
"rewards/accuracies": 0.
|
| 1147 |
-
"rewards/chosen": -1.
|
| 1148 |
-
"rewards/margins": 1.
|
| 1149 |
-
"rewards/rejected": -
|
| 1150 |
"step": 740
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 1.5698587127158556,
|
| 1154 |
-
"grad_norm":
|
| 1155 |
"learning_rate": 6.655924144404906e-08,
|
| 1156 |
-
"logits/chosen": -
|
| 1157 |
-
"logits/rejected":
|
| 1158 |
-
"logps/chosen": -
|
| 1159 |
-
"logps/rejected": -
|
| 1160 |
-
"loss": 0.
|
| 1161 |
-
"rewards/accuracies": 0.
|
| 1162 |
-
"rewards/chosen": -1.
|
| 1163 |
-
"rewards/margins": 1.
|
| 1164 |
-
"rewards/rejected": -
|
| 1165 |
"step": 750
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 1.5907901622187337,
|
| 1169 |
-
"grad_norm": 33.
|
| 1170 |
"learning_rate": 6.046442623320145e-08,
|
| 1171 |
-
"logits/chosen": -
|
| 1172 |
-
"logits/rejected":
|
| 1173 |
-
"logps/chosen": -
|
| 1174 |
-
"logps/rejected": -
|
| 1175 |
-
"loss": 0.
|
| 1176 |
-
"rewards/accuracies": 0.
|
| 1177 |
-
"rewards/chosen": -
|
| 1178 |
-
"rewards/margins": 1.
|
| 1179 |
-
"rewards/rejected": -3.
|
| 1180 |
"step": 760
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
"epoch": 1.6117216117216118,
|
| 1184 |
-
"grad_norm":
|
| 1185 |
"learning_rate": 5.4623689209832484e-08,
|
| 1186 |
-
"logits/chosen": -
|
| 1187 |
-
"logits/rejected":
|
| 1188 |
-
"logps/chosen": -
|
| 1189 |
-
"logps/rejected": -
|
| 1190 |
-
"loss": 0.
|
| 1191 |
-
"rewards/accuracies": 0.
|
| 1192 |
-
"rewards/chosen": -
|
| 1193 |
-
"rewards/margins": 1.
|
| 1194 |
-
"rewards/rejected": -
|
| 1195 |
"step": 770
|
| 1196 |
},
|
| 1197 |
{
|
| 1198 |
"epoch": 1.6326530612244898,
|
| 1199 |
-
"grad_norm":
|
| 1200 |
"learning_rate": 4.904486005914027e-08,
|
| 1201 |
-
"logits/chosen": -
|
| 1202 |
-
"logits/rejected":
|
| 1203 |
-
"logps/chosen": -
|
| 1204 |
-
"logps/rejected": -
|
| 1205 |
-
"loss": 0.
|
| 1206 |
-
"rewards/accuracies": 0.
|
| 1207 |
-
"rewards/chosen": -1.
|
| 1208 |
-
"rewards/margins": 1.
|
| 1209 |
-
"rewards/rejected": -
|
| 1210 |
"step": 780
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 1.653584510727368,
|
| 1214 |
-
"grad_norm":
|
| 1215 |
"learning_rate": 4.373541737087263e-08,
|
| 1216 |
-
"logits/chosen": -
|
| 1217 |
-
"logits/rejected":
|
| 1218 |
-
"logps/chosen": -
|
| 1219 |
-
"logps/rejected": -
|
| 1220 |
-
"loss": 0.
|
| 1221 |
-
"rewards/accuracies": 0.
|
| 1222 |
-
"rewards/chosen": -1.
|
| 1223 |
-
"rewards/margins": 1.
|
| 1224 |
-
"rewards/rejected": -
|
| 1225 |
"step": 790
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"epoch": 1.674515960230246,
|
| 1229 |
-
"grad_norm":
|
| 1230 |
"learning_rate": 3.8702478614051345e-08,
|
| 1231 |
-
"logits/chosen": -
|
| 1232 |
-
"logits/rejected":
|
| 1233 |
-
"logps/chosen": -
|
| 1234 |
-
"logps/rejected": -
|
| 1235 |
-
"loss": 0.
|
| 1236 |
"rewards/accuracies": 0.90625,
|
| 1237 |
-
"rewards/chosen": -1.
|
| 1238 |
-
"rewards/margins": 1.
|
| 1239 |
-
"rewards/rejected": -
|
| 1240 |
"step": 800
|
| 1241 |
},
|
| 1242 |
{
|
| 1243 |
"epoch": 1.695447409733124,
|
| 1244 |
-
"grad_norm":
|
| 1245 |
"learning_rate": 3.3952790595787986e-08,
|
| 1246 |
-
"logits/chosen": -
|
| 1247 |
-
"logits/rejected":
|
| 1248 |
-
"logps/chosen": -
|
| 1249 |
-
"logps/rejected": -
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"rewards/accuracies": 0.
|
| 1252 |
-
"rewards/chosen": -1.
|
| 1253 |
-
"rewards/margins":
|
| 1254 |
-
"rewards/rejected": -3.
|
| 1255 |
"step": 810
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 1.716378859236002,
|
| 1259 |
-
"grad_norm":
|
| 1260 |
"learning_rate": 2.9492720416985e-08,
|
| 1261 |
-
"logits/chosen": -
|
| 1262 |
-
"logits/rejected": -
|
| 1263 |
-
"logps/chosen": -
|
| 1264 |
-
"logps/rejected": -
|
| 1265 |
-
"loss": 0.
|
| 1266 |
"rewards/accuracies": 0.8687499761581421,
|
| 1267 |
-
"rewards/chosen": -1.
|
| 1268 |
-
"rewards/margins": 1.
|
| 1269 |
-
"rewards/rejected": -
|
| 1270 |
"step": 820
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 1.7373103087388801,
|
| 1274 |
-
"grad_norm":
|
| 1275 |
"learning_rate": 2.5328246937043525e-08,
|
| 1276 |
-
"logits/chosen": -
|
| 1277 |
-
"logits/rejected": -
|
| 1278 |
-
"logps/chosen": -
|
| 1279 |
-
"logps/rejected": -
|
| 1280 |
-
"loss": 0.
|
| 1281 |
-
"rewards/accuracies": 0.
|
| 1282 |
-
"rewards/chosen": -1.
|
| 1283 |
-
"rewards/margins": 1.
|
| 1284 |
-
"rewards/rejected": -
|
| 1285 |
"step": 830
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 1.7582417582417582,
|
| 1289 |
-
"grad_norm":
|
| 1290 |
"learning_rate": 2.1464952759020856e-08,
|
| 1291 |
-
"logits/chosen": -
|
| 1292 |
-
"logits/rejected":
|
| 1293 |
-
"logps/chosen": -
|
| 1294 |
-
"logps/rejected": -
|
| 1295 |
-
"loss": 0.
|
| 1296 |
-
"rewards/accuracies": 0.
|
| 1297 |
-
"rewards/chosen": -
|
| 1298 |
-
"rewards/margins": 1.
|
| 1299 |
-
"rewards/rejected": -3.
|
| 1300 |
"step": 840
|
| 1301 |
},
|
| 1302 |
{
|
| 1303 |
"epoch": 1.7791732077446363,
|
| 1304 |
-
"grad_norm":
|
| 1305 |
"learning_rate": 1.7908016745981856e-08,
|
| 1306 |
-
"logits/chosen": -
|
| 1307 |
-
"logits/rejected":
|
| 1308 |
-
"logps/chosen": -
|
| 1309 |
-
"logps/rejected": -
|
| 1310 |
-
"loss": 0.
|
| 1311 |
-
"rewards/accuracies": 0.
|
| 1312 |
-
"rewards/chosen": -1.
|
| 1313 |
-
"rewards/margins": 1.
|
| 1314 |
-
"rewards/rejected": -
|
| 1315 |
"step": 850
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 1.8001046572475143,
|
| 1319 |
-
"grad_norm":
|
| 1320 |
"learning_rate": 1.4662207078575684e-08,
|
| 1321 |
-
"logits/chosen": -
|
| 1322 |
-
"logits/rejected": -
|
| 1323 |
-
"logps/chosen": -
|
| 1324 |
-
"logps/rejected": -
|
| 1325 |
-
"loss": 0.
|
| 1326 |
-
"rewards/accuracies": 0.
|
| 1327 |
-
"rewards/chosen": -1.
|
| 1328 |
-
"rewards/margins": 1.
|
| 1329 |
-
"rewards/rejected": -
|
| 1330 |
"step": 860
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 1.8210361067503924,
|
| 1334 |
-
"grad_norm":
|
| 1335 |
"learning_rate": 1.1731874863145142e-08,
|
| 1336 |
-
"logits/chosen": -
|
| 1337 |
-
"logits/rejected":
|
| 1338 |
-
"logps/chosen": -
|
| 1339 |
-
"logps/rejected": -
|
| 1340 |
-
"loss": 0.
|
| 1341 |
-
"rewards/accuracies": 0.
|
| 1342 |
-
"rewards/chosen": -1.
|
| 1343 |
-
"rewards/margins": 1.
|
| 1344 |
-
"rewards/rejected": -
|
| 1345 |
"step": 870
|
| 1346 |
},
|
| 1347 |
{
|
| 1348 |
"epoch": 1.8419675562532705,
|
| 1349 |
-
"grad_norm":
|
| 1350 |
"learning_rate": 9.12094829893642e-09,
|
| 1351 |
-
"logits/chosen": -
|
| 1352 |
-
"logits/rejected":
|
| 1353 |
-
"logps/chosen": -
|
| 1354 |
-
"logps/rejected": -
|
| 1355 |
-
"loss": 0.
|
| 1356 |
-
"rewards/accuracies": 0.
|
| 1357 |
-
"rewards/chosen": -
|
| 1358 |
-
"rewards/margins": 1.
|
| 1359 |
-
"rewards/rejected": -
|
| 1360 |
"step": 880
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
"epoch": 1.8628990057561485,
|
| 1364 |
-
"grad_norm":
|
| 1365 |
"learning_rate": 6.832927412229017e-09,
|
| 1366 |
-
"logits/chosen": -
|
| 1367 |
-
"logits/rejected": -
|
| 1368 |
-
"logps/chosen": -
|
| 1369 |
-
"logps/rejected": -
|
| 1370 |
-
"loss": 0.
|
| 1371 |
-
"rewards/accuracies": 0.
|
| 1372 |
-
"rewards/chosen": -1.
|
| 1373 |
-
"rewards/margins":
|
| 1374 |
-
"rewards/rejected": -
|
| 1375 |
"step": 890
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 1.8838304552590266,
|
| 1379 |
-
"grad_norm":
|
| 1380 |
"learning_rate": 4.8708793644441086e-09,
|
| 1381 |
-
"logits/chosen": -
|
| 1382 |
-
"logits/rejected":
|
| 1383 |
-
"logps/chosen": -
|
| 1384 |
-
"logps/rejected": -
|
| 1385 |
-
"loss": 0.
|
| 1386 |
-
"rewards/accuracies": 0.
|
| 1387 |
-
"rewards/chosen": -1.
|
| 1388 |
-
"rewards/margins": 1.
|
| 1389 |
-
"rewards/rejected": -
|
| 1390 |
"step": 900
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 1.9047619047619047,
|
| 1394 |
-
"grad_norm":
|
| 1395 |
"learning_rate": 3.2374343405217884e-09,
|
| 1396 |
-
"logits/chosen": -
|
| 1397 |
-
"logits/rejected":
|
| 1398 |
-
"logps/chosen": -
|
| 1399 |
-
"logps/rejected": -
|
| 1400 |
-
"loss": 0.
|
| 1401 |
-
"rewards/accuracies": 0.
|
| 1402 |
-
"rewards/chosen": -1.
|
| 1403 |
-
"rewards/margins": 1.
|
| 1404 |
-
"rewards/rejected": -
|
| 1405 |
"step": 910
|
| 1406 |
},
|
| 1407 |
{
|
| 1408 |
"epoch": 1.9256933542647827,
|
| 1409 |
-
"grad_norm":
|
| 1410 |
"learning_rate": 1.9347820230782295e-09,
|
| 1411 |
-
"logits/chosen": -
|
| 1412 |
-
"logits/rejected": -
|
| 1413 |
-
"logps/chosen": -
|
| 1414 |
-
"logps/rejected": -
|
| 1415 |
-
"loss": 0.
|
| 1416 |
"rewards/accuracies": 0.8999999761581421,
|
| 1417 |
-
"rewards/chosen": -1.
|
| 1418 |
-
"rewards/margins": 1.
|
| 1419 |
-
"rewards/rejected": -
|
| 1420 |
"step": 920
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 1.9466248037676608,
|
| 1424 |
-
"grad_norm":
|
| 1425 |
"learning_rate": 9.64668657069706e-10,
|
| 1426 |
-
"logits/chosen": -
|
| 1427 |
-
"logits/rejected": -
|
| 1428 |
-
"logps/chosen": -
|
| 1429 |
-
"logps/rejected": -
|
| 1430 |
-
"loss": 0.
|
| 1431 |
-
"rewards/accuracies": 0.
|
| 1432 |
-
"rewards/chosen": -1.
|
| 1433 |
-
"rewards/margins": 1.
|
| 1434 |
-
"rewards/rejected": -
|
| 1435 |
"step": 930
|
| 1436 |
},
|
| 1437 |
{
|
| 1438 |
"epoch": 1.9675562532705388,
|
| 1439 |
-
"grad_norm":
|
| 1440 |
"learning_rate": 3.2839470889836627e-10,
|
| 1441 |
-
"logits/chosen": -
|
| 1442 |
-
"logits/rejected":
|
| 1443 |
-
"logps/chosen": -
|
| 1444 |
-
"logps/rejected": -
|
| 1445 |
-
"loss": 0.
|
| 1446 |
-
"rewards/accuracies": 0.
|
| 1447 |
-
"rewards/chosen": -1.
|
| 1448 |
-
"rewards/margins": 1.
|
| 1449 |
-
"rewards/rejected": -
|
| 1450 |
"step": 940
|
| 1451 |
},
|
| 1452 |
{
|
| 1453 |
"epoch": 1.988487702773417,
|
| 1454 |
-
"grad_norm":
|
| 1455 |
"learning_rate": 2.6813123097352287e-11,
|
| 1456 |
-
"logits/chosen": -
|
| 1457 |
-
"logits/rejected":
|
| 1458 |
-
"logps/chosen": -
|
| 1459 |
-
"logps/rejected": -
|
| 1460 |
-
"loss": 0.
|
| 1461 |
-
"rewards/accuracies": 0.
|
| 1462 |
-
"rewards/chosen": -
|
| 1463 |
-
"rewards/margins": 1.
|
| 1464 |
-
"rewards/rejected": -3.
|
| 1465 |
"step": 950
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 1.9968602825745683,
|
| 1469 |
-
"eval_logits/chosen": -
|
| 1470 |
-
"eval_logits/rejected":
|
| 1471 |
-
"eval_logps/chosen": -
|
| 1472 |
-
"eval_logps/rejected": -
|
| 1473 |
-
"eval_loss": 0.
|
| 1474 |
-
"eval_rewards/accuracies": 0.
|
| 1475 |
-
"eval_rewards/chosen": -
|
| 1476 |
-
"eval_rewards/margins": 1.
|
| 1477 |
-
"eval_rewards/rejected": -
|
| 1478 |
-
"eval_runtime": 169.
|
| 1479 |
-
"eval_samples_per_second": 11.
|
| 1480 |
-
"eval_steps_per_second": 0.
|
| 1481 |
"step": 954
|
| 1482 |
},
|
| 1483 |
{
|
|
@@ -1485,9 +1485,9 @@
|
|
| 1485 |
"step": 954,
|
| 1486 |
"total_flos": 0.0,
|
| 1487 |
"train_loss": 0.0,
|
| 1488 |
-
"train_runtime": 3.
|
| 1489 |
-
"train_samples_per_second":
|
| 1490 |
-
"train_steps_per_second":
|
| 1491 |
}
|
| 1492 |
],
|
| 1493 |
"logging_steps": 10,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0020931449502878076,
|
| 13 |
+
"grad_norm": 8.648654168519636,
|
| 14 |
"learning_rate": 5.208333333333333e-09,
|
| 15 |
+
"logits/chosen": -2.911508798599243,
|
| 16 |
+
"logits/rejected": -2.8825056552886963,
|
| 17 |
+
"logps/chosen": -328.824462890625,
|
| 18 |
+
"logps/rejected": -375.2520751953125,
|
| 19 |
"loss": 0.6931,
|
| 20 |
"rewards/accuracies": 0.0,
|
| 21 |
"rewards/chosen": 0.0,
|
|
|
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.020931449502878074,
|
| 28 |
+
"grad_norm": 8.529629402090206,
|
| 29 |
"learning_rate": 5.208333333333333e-08,
|
| 30 |
+
"logits/chosen": -2.738391876220703,
|
| 31 |
+
"logits/rejected": -2.715987205505371,
|
| 32 |
+
"logps/chosen": -320.3747863769531,
|
| 33 |
+
"logps/rejected": -286.63323974609375,
|
| 34 |
+
"loss": 0.6932,
|
| 35 |
+
"rewards/accuracies": 0.4375,
|
| 36 |
+
"rewards/chosen": -1.6565963960601948e-05,
|
| 37 |
+
"rewards/margins": -6.0656500863842666e-05,
|
| 38 |
+
"rewards/rejected": 4.4090484152548015e-05,
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.04186289900575615,
|
| 43 |
+
"grad_norm": 8.121160296339491,
|
| 44 |
"learning_rate": 1.0416666666666667e-07,
|
| 45 |
+
"logits/chosen": -2.812523365020752,
|
| 46 |
+
"logits/rejected": -2.765321731567383,
|
| 47 |
+
"logps/chosen": -326.799072265625,
|
| 48 |
+
"logps/rejected": -291.5594177246094,
|
| 49 |
+
"loss": 0.693,
|
| 50 |
+
"rewards/accuracies": 0.543749988079071,
|
| 51 |
+
"rewards/chosen": 0.0013127574929967523,
|
| 52 |
+
"rewards/margins": 0.0006053571123629808,
|
| 53 |
+
"rewards/rejected": 0.0007074003806337714,
|
| 54 |
"step": 20
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.06279434850863422,
|
| 58 |
+
"grad_norm": 7.841014680458238,
|
| 59 |
"learning_rate": 1.5624999999999999e-07,
|
| 60 |
+
"logits/chosen": -2.791351318359375,
|
| 61 |
+
"logits/rejected": -2.737368583679199,
|
| 62 |
+
"logps/chosen": -304.5863952636719,
|
| 63 |
+
"logps/rejected": -255.7802734375,
|
| 64 |
"loss": 0.6923,
|
| 65 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 66 |
+
"rewards/chosen": 0.005904044955968857,
|
| 67 |
+
"rewards/margins": 0.0021464950405061245,
|
| 68 |
+
"rewards/rejected": 0.003757549449801445,
|
| 69 |
"step": 30
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.0837257980115123,
|
| 73 |
+
"grad_norm": 7.780689883117447,
|
| 74 |
"learning_rate": 2.0833333333333333e-07,
|
| 75 |
+
"logits/chosen": -2.7076315879821777,
|
| 76 |
+
"logits/rejected": -2.685647487640381,
|
| 77 |
+
"logps/chosen": -276.40045166015625,
|
| 78 |
+
"logps/rejected": -266.44720458984375,
|
| 79 |
+
"loss": 0.69,
|
| 80 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 81 |
+
"rewards/chosen": 0.015561707317829132,
|
| 82 |
+
"rewards/margins": 0.00809282623231411,
|
| 83 |
+
"rewards/rejected": 0.007468880619853735,
|
| 84 |
"step": 40
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"epoch": 0.10465724751439037,
|
| 88 |
+
"grad_norm": 7.837366185969384,
|
| 89 |
"learning_rate": 2.604166666666667e-07,
|
| 90 |
+
"logits/chosen": -2.7266910076141357,
|
| 91 |
+
"logits/rejected": -2.665973424911499,
|
| 92 |
+
"logps/chosen": -287.90960693359375,
|
| 93 |
+
"logps/rejected": -280.13787841796875,
|
| 94 |
+
"loss": 0.6871,
|
| 95 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 96 |
+
"rewards/chosen": 0.026710817590355873,
|
| 97 |
+
"rewards/margins": 0.01235121302306652,
|
| 98 |
+
"rewards/rejected": 0.014359605498611927,
|
| 99 |
"step": 50
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.12558869701726844,
|
| 103 |
+
"grad_norm": 8.68562739775634,
|
| 104 |
"learning_rate": 3.1249999999999997e-07,
|
| 105 |
+
"logits/chosen": -2.738069534301758,
|
| 106 |
+
"logits/rejected": -2.694655656814575,
|
| 107 |
+
"logps/chosen": -261.654296875,
|
| 108 |
+
"logps/rejected": -249.44241333007812,
|
| 109 |
+
"loss": 0.6811,
|
| 110 |
+
"rewards/accuracies": 0.71875,
|
| 111 |
+
"rewards/chosen": 0.039315465837717056,
|
| 112 |
+
"rewards/margins": 0.02459501288831234,
|
| 113 |
+
"rewards/rejected": 0.014720454812049866,
|
| 114 |
"step": 60
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.14652014652014653,
|
| 118 |
+
"grad_norm": 8.17805459442737,
|
| 119 |
"learning_rate": 3.645833333333333e-07,
|
| 120 |
+
"logits/chosen": -2.6594691276550293,
|
| 121 |
+
"logits/rejected": -2.6348719596862793,
|
| 122 |
+
"logps/chosen": -304.16754150390625,
|
| 123 |
+
"logps/rejected": -292.7433166503906,
|
| 124 |
+
"loss": 0.6713,
|
| 125 |
+
"rewards/accuracies": 0.699999988079071,
|
| 126 |
+
"rewards/chosen": 0.03804076835513115,
|
| 127 |
+
"rewards/margins": 0.0464973971247673,
|
| 128 |
+
"rewards/rejected": -0.008456626906991005,
|
| 129 |
"step": 70
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.1674515960230246,
|
| 133 |
+
"grad_norm": 9.10750617456122,
|
| 134 |
"learning_rate": 4.1666666666666667e-07,
|
| 135 |
+
"logits/chosen": -2.713014602661133,
|
| 136 |
+
"logits/rejected": -2.63926362991333,
|
| 137 |
+
"logps/chosen": -343.3531494140625,
|
| 138 |
+
"logps/rejected": -272.9491271972656,
|
| 139 |
+
"loss": 0.6603,
|
| 140 |
+
"rewards/accuracies": 0.737500011920929,
|
| 141 |
+
"rewards/chosen": 0.01621154509484768,
|
| 142 |
+
"rewards/margins": 0.09821438044309616,
|
| 143 |
+
"rewards/rejected": -0.08200283348560333,
|
| 144 |
"step": 80
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.18838304552590268,
|
| 148 |
+
"grad_norm": 10.344063716824344,
|
| 149 |
"learning_rate": 4.6874999999999996e-07,
|
| 150 |
+
"logits/chosen": -2.6566672325134277,
|
| 151 |
+
"logits/rejected": -2.616140127182007,
|
| 152 |
+
"logps/chosen": -273.27459716796875,
|
| 153 |
+
"logps/rejected": -285.70361328125,
|
| 154 |
+
"loss": 0.6362,
|
| 155 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 156 |
+
"rewards/chosen": -0.09348011016845703,
|
| 157 |
+
"rewards/margins": 0.12727037072181702,
|
| 158 |
+
"rewards/rejected": -0.22075048089027405,
|
| 159 |
"step": 90
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.20931449502878074,
|
| 163 |
+
"grad_norm": 18.46434787727296,
|
| 164 |
"learning_rate": 4.999731868769026e-07,
|
| 165 |
+
"logits/chosen": -2.62548828125,
|
| 166 |
+
"logits/rejected": -2.604613780975342,
|
| 167 |
+
"logps/chosen": -326.0329895019531,
|
| 168 |
+
"logps/rejected": -328.9161071777344,
|
| 169 |
+
"loss": 0.6209,
|
| 170 |
+
"rewards/accuracies": 0.706250011920929,
|
| 171 |
+
"rewards/chosen": -0.15337222814559937,
|
| 172 |
+
"rewards/margins": 0.27119967341423035,
|
| 173 |
+
"rewards/rejected": -0.4245719313621521,
|
| 174 |
"step": 100
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.2302459445316588,
|
| 178 |
+
"grad_norm": 12.853014931259976,
|
| 179 |
"learning_rate": 4.996716052911017e-07,
|
| 180 |
+
"logits/chosen": -2.6524977684020996,
|
| 181 |
+
"logits/rejected": -2.613800287246704,
|
| 182 |
+
"logps/chosen": -333.86810302734375,
|
| 183 |
+
"logps/rejected": -339.0367431640625,
|
| 184 |
+
"loss": 0.6155,
|
| 185 |
+
"rewards/accuracies": 0.65625,
|
| 186 |
+
"rewards/chosen": -0.2943059504032135,
|
| 187 |
+
"rewards/margins": 0.2476470172405243,
|
| 188 |
+
"rewards/rejected": -0.5419529676437378,
|
| 189 |
"step": 110
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.25117739403453687,
|
| 193 |
+
"grad_norm": 13.434978559765035,
|
| 194 |
"learning_rate": 4.990353313429303e-07,
|
| 195 |
+
"logits/chosen": -2.5527138710021973,
|
| 196 |
+
"logits/rejected": -2.5590224266052246,
|
| 197 |
+
"logps/chosen": -309.571044921875,
|
| 198 |
+
"logps/rejected": -332.9461364746094,
|
| 199 |
+
"loss": 0.598,
|
| 200 |
+
"rewards/accuracies": 0.737500011920929,
|
| 201 |
+
"rewards/chosen": -0.2963547110557556,
|
| 202 |
+
"rewards/margins": 0.33014291524887085,
|
| 203 |
+
"rewards/rejected": -0.6264976263046265,
|
| 204 |
"step": 120
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 0.272108843537415,
|
| 208 |
+
"grad_norm": 13.77722186805988,
|
| 209 |
"learning_rate": 4.980652179769217e-07,
|
| 210 |
+
"logits/chosen": -2.5915229320526123,
|
| 211 |
+
"logits/rejected": -2.5797364711761475,
|
| 212 |
+
"logps/chosen": -358.6594543457031,
|
| 213 |
+
"logps/rejected": -371.49652099609375,
|
| 214 |
+
"loss": 0.588,
|
| 215 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 216 |
+
"rewards/chosen": -0.6360726356506348,
|
| 217 |
+
"rewards/margins": 0.286227822303772,
|
| 218 |
+
"rewards/rejected": -0.9223003387451172,
|
| 219 |
"step": 130
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.29304029304029305,
|
| 223 |
+
"grad_norm": 19.98956871690929,
|
| 224 |
"learning_rate": 4.967625656594781e-07,
|
| 225 |
+
"logits/chosen": -2.6256630420684814,
|
| 226 |
+
"logits/rejected": -2.6000983715057373,
|
| 227 |
+
"logps/chosen": -368.37274169921875,
|
| 228 |
+
"logps/rejected": -357.54522705078125,
|
| 229 |
+
"loss": 0.5846,
|
| 230 |
+
"rewards/accuracies": 0.75,
|
| 231 |
+
"rewards/chosen": -0.3657643496990204,
|
| 232 |
+
"rewards/margins": 0.411146879196167,
|
| 233 |
+
"rewards/rejected": -0.776911199092865,
|
| 234 |
"step": 140
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.3139717425431711,
|
| 238 |
+
"grad_norm": 18.241862609597362,
|
| 239 |
"learning_rate": 4.951291206355559e-07,
|
| 240 |
+
"logits/chosen": -2.5957682132720947,
|
| 241 |
+
"logits/rejected": -2.570385456085205,
|
| 242 |
+
"logps/chosen": -348.45361328125,
|
| 243 |
+
"logps/rejected": -350.3853759765625,
|
| 244 |
+
"loss": 0.5898,
|
| 245 |
+
"rewards/accuracies": 0.706250011920929,
|
| 246 |
+
"rewards/chosen": -0.5664342641830444,
|
| 247 |
+
"rewards/margins": 0.37359124422073364,
|
| 248 |
+
"rewards/rejected": -0.9400255084037781,
|
| 249 |
"step": 150
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.3349031920460492,
|
| 253 |
+
"grad_norm": 16.695399210826476,
|
| 254 |
"learning_rate": 4.93167072587771e-07,
|
| 255 |
+
"logits/chosen": -2.6461105346679688,
|
| 256 |
+
"logits/rejected": -2.6081748008728027,
|
| 257 |
+
"logps/chosen": -334.31024169921875,
|
| 258 |
+
"logps/rejected": -353.1805419921875,
|
| 259 |
+
"loss": 0.5763,
|
| 260 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 261 |
+
"rewards/chosen": -0.45854535698890686,
|
| 262 |
+
"rewards/margins": 0.46823421120643616,
|
| 263 |
+
"rewards/rejected": -0.926779568195343,
|
| 264 |
"step": 160
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 0.35583464154892724,
|
| 268 |
+
"grad_norm": 20.770623175746024,
|
| 269 |
"learning_rate": 4.908790517010636e-07,
|
| 270 |
+
"logits/chosen": -2.5199551582336426,
|
| 271 |
+
"logits/rejected": -2.5082385540008545,
|
| 272 |
+
"logps/chosen": -325.3274230957031,
|
| 273 |
+
"logps/rejected": -337.1798400878906,
|
| 274 |
+
"loss": 0.5817,
|
| 275 |
+
"rewards/accuracies": 0.71875,
|
| 276 |
+
"rewards/chosen": -0.5404187440872192,
|
| 277 |
+
"rewards/margins": 0.3684747815132141,
|
| 278 |
+
"rewards/rejected": -0.9088934659957886,
|
| 279 |
"step": 170
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 0.37676609105180536,
|
| 283 |
+
"grad_norm": 17.458454310533863,
|
| 284 |
"learning_rate": 4.882681251368548e-07,
|
| 285 |
+
"logits/chosen": -2.5944693088531494,
|
| 286 |
+
"logits/rejected": -2.57043719291687,
|
| 287 |
+
"logps/chosen": -339.6058044433594,
|
| 288 |
+
"logps/rejected": -346.19708251953125,
|
| 289 |
+
"loss": 0.5499,
|
| 290 |
+
"rewards/accuracies": 0.71875,
|
| 291 |
+
"rewards/chosen": -0.5662221312522888,
|
| 292 |
+
"rewards/margins": 0.3738660216331482,
|
| 293 |
+
"rewards/rejected": -0.9400882720947266,
|
| 294 |
"step": 180
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 0.3976975405546834,
|
| 298 |
+
"grad_norm": 21.162626385312407,
|
| 299 |
"learning_rate": 4.853377929214243e-07,
|
| 300 |
+
"logits/chosen": -2.5164248943328857,
|
| 301 |
+
"logits/rejected": -2.490086793899536,
|
| 302 |
+
"logps/chosen": -326.11981201171875,
|
| 303 |
+
"logps/rejected": -356.9112548828125,
|
| 304 |
+
"loss": 0.567,
|
| 305 |
+
"rewards/accuracies": 0.699999988079071,
|
| 306 |
+
"rewards/chosen": -0.5691932439804077,
|
| 307 |
+
"rewards/margins": 0.3675684630870819,
|
| 308 |
+
"rewards/rejected": -0.9367616772651672,
|
| 309 |
"step": 190
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 0.4186289900575615,
|
| 313 |
+
"grad_norm": 81.07404869193526,
|
| 314 |
"learning_rate": 4.820919832540181e-07,
|
| 315 |
+
"logits/chosen": -2.495347499847412,
|
| 316 |
+
"logits/rejected": -2.5034618377685547,
|
| 317 |
+
"logps/chosen": -325.27166748046875,
|
| 318 |
+
"logps/rejected": -383.38934326171875,
|
| 319 |
+
"loss": 0.5841,
|
| 320 |
"rewards/accuracies": 0.71875,
|
| 321 |
+
"rewards/chosen": -0.6413096189498901,
|
| 322 |
+
"rewards/margins": 0.3838884234428406,
|
| 323 |
+
"rewards/rejected": -1.0251981019973755,
|
| 324 |
"step": 200
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.43956043956043955,
|
| 328 |
+
"grad_norm": 20.383161365084906,
|
| 329 |
"learning_rate": 4.785350472409791e-07,
|
| 330 |
+
"logits/chosen": -2.5574824810028076,
|
| 331 |
+
"logits/rejected": -2.5132758617401123,
|
| 332 |
+
"logps/chosen": -382.79547119140625,
|
| 333 |
+
"logps/rejected": -362.0451354980469,
|
| 334 |
+
"loss": 0.5673,
|
| 335 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 336 |
+
"rewards/chosen": -0.5895588397979736,
|
| 337 |
+
"rewards/margins": 0.43525028228759766,
|
| 338 |
+
"rewards/rejected": -1.0248091220855713,
|
| 339 |
"step": 210
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.4604918890633176,
|
| 343 |
+
"grad_norm": 19.259994881877645,
|
| 344 |
"learning_rate": 4.7467175306295647e-07,
|
| 345 |
+
"logits/chosen": -2.493474245071411,
|
| 346 |
+
"logits/rejected": -2.5184781551361084,
|
| 347 |
+
"logps/chosen": -318.1018981933594,
|
| 348 |
+
"logps/rejected": -364.78167724609375,
|
| 349 |
+
"loss": 0.561,
|
| 350 |
+
"rewards/accuracies": 0.768750011920929,
|
| 351 |
+
"rewards/chosen": -0.6697767376899719,
|
| 352 |
+
"rewards/margins": 0.4676804542541504,
|
| 353 |
+
"rewards/rejected": -1.137457251548767,
|
| 354 |
"step": 220
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.48142333856619574,
|
| 358 |
+
"grad_norm": 19.775692922777765,
|
| 359 |
"learning_rate": 4.70507279583015e-07,
|
| 360 |
+
"logits/chosen": -2.6063382625579834,
|
| 361 |
+
"logits/rejected": -2.5750064849853516,
|
| 362 |
+
"logps/chosen": -334.16009521484375,
|
| 363 |
+
"logps/rejected": -381.5287780761719,
|
| 364 |
+
"loss": 0.5634,
|
| 365 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 366 |
+
"rewards/chosen": -0.6287840008735657,
|
| 367 |
+
"rewards/margins": 0.7023510932922363,
|
| 368 |
+
"rewards/rejected": -1.3311350345611572,
|
| 369 |
"step": 230
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 0.5023547880690737,
|
| 373 |
+
"grad_norm": 26.289584097515366,
|
| 374 |
"learning_rate": 4.6604720940421207e-07,
|
| 375 |
+
"logits/chosen": -2.5913660526275635,
|
| 376 |
+
"logits/rejected": -2.5510830879211426,
|
| 377 |
+
"logps/chosen": -363.0348205566406,
|
| 378 |
+
"logps/rejected": -385.59173583984375,
|
| 379 |
+
"loss": 0.5304,
|
| 380 |
+
"rewards/accuracies": 0.699999988079071,
|
| 381 |
+
"rewards/chosen": -0.599415123462677,
|
| 382 |
+
"rewards/margins": 0.555901050567627,
|
| 383 |
+
"rewards/rejected": -1.1553161144256592,
|
| 384 |
"step": 240
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 0.5232862375719518,
|
| 388 |
+
"grad_norm": 20.96954894885707,
|
| 389 |
"learning_rate": 4.612975213859487e-07,
|
| 390 |
+
"logits/chosen": -2.5373878479003906,
|
| 391 |
+
"logits/rejected": -2.4917821884155273,
|
| 392 |
+
"logps/chosen": -379.81591796875,
|
| 393 |
+
"logps/rejected": -413.7845764160156,
|
| 394 |
+
"loss": 0.5326,
|
| 395 |
+
"rewards/accuracies": 0.8062499761581421,
|
| 396 |
+
"rewards/chosen": -0.8831035494804382,
|
| 397 |
+
"rewards/margins": 0.783009946346283,
|
| 398 |
+
"rewards/rejected": -1.6661134958267212,
|
| 399 |
"step": 250
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 0.54421768707483,
|
| 403 |
+
"grad_norm": 20.35055226575624,
|
| 404 |
"learning_rate": 4.5626458262912735e-07,
|
| 405 |
+
"logits/chosen": -2.5308146476745605,
|
| 406 |
+
"logits/rejected": -2.5432188510894775,
|
| 407 |
+
"logps/chosen": -362.18731689453125,
|
| 408 |
+
"logps/rejected": -401.601318359375,
|
| 409 |
+
"loss": 0.527,
|
| 410 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 411 |
+
"rewards/chosen": -0.8506525754928589,
|
| 412 |
+
"rewards/margins": 0.6240705847740173,
|
| 413 |
+
"rewards/rejected": -1.474723219871521,
|
| 414 |
"step": 260
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 0.565149136577708,
|
| 418 |
+
"grad_norm": 21.42532206002976,
|
| 419 |
"learning_rate": 4.5095513994085974e-07,
|
| 420 |
+
"logits/chosen": -2.396043300628662,
|
| 421 |
+
"logits/rejected": -2.3887431621551514,
|
| 422 |
+
"logps/chosen": -395.9293212890625,
|
| 423 |
+
"logps/rejected": -445.9908142089844,
|
| 424 |
+
"loss": 0.5364,
|
| 425 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 426 |
+
"rewards/chosen": -0.8952659368515015,
|
| 427 |
+
"rewards/margins": 0.8105502128601074,
|
| 428 |
+
"rewards/rejected": -1.7058160305023193,
|
| 429 |
"step": 270
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.5860805860805861,
|
| 433 |
+
"grad_norm": 24.442795762265213,
|
| 434 |
"learning_rate": 4.453763107901675e-07,
|
| 435 |
+
"logits/chosen": -2.251249074935913,
|
| 436 |
+
"logits/rejected": -2.2444510459899902,
|
| 437 |
+
"logps/chosen": -361.8575744628906,
|
| 438 |
+
"logps/rejected": -387.69476318359375,
|
| 439 |
+
"loss": 0.5385,
|
| 440 |
"rewards/accuracies": 0.6937500238418579,
|
| 441 |
+
"rewards/chosen": -0.8830081820487976,
|
| 442 |
+
"rewards/margins": 0.6050563454627991,
|
| 443 |
+
"rewards/rejected": -1.4880646467208862,
|
| 444 |
"step": 280
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.6070120355834642,
|
| 448 |
+
"grad_norm": 21.25843011088378,
|
| 449 |
"learning_rate": 4.395355737667985e-07,
|
| 450 |
+
"logits/chosen": -2.3289453983306885,
|
| 451 |
+
"logits/rejected": -2.2574009895324707,
|
| 452 |
+
"logps/chosen": -364.3363342285156,
|
| 453 |
+
"logps/rejected": -418.8705139160156,
|
| 454 |
+
"loss": 0.5286,
|
| 455 |
"rewards/accuracies": 0.7562500238418579,
|
| 456 |
+
"rewards/chosen": -1.0685086250305176,
|
| 457 |
+
"rewards/margins": 0.6174401044845581,
|
| 458 |
+
"rewards/rejected": -1.6859489679336548,
|
| 459 |
"step": 290
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.6279434850863422,
|
| 463 |
+
"grad_norm": 21.179507487381557,
|
| 464 |
"learning_rate": 4.3344075855595097e-07,
|
| 465 |
+
"logits/chosen": -2.2611992359161377,
|
| 466 |
+
"logits/rejected": -2.218019962310791,
|
| 467 |
+
"logps/chosen": -428.0699768066406,
|
| 468 |
+
"logps/rejected": -442.8394470214844,
|
| 469 |
+
"loss": 0.5118,
|
| 470 |
"rewards/accuracies": 0.731249988079071,
|
| 471 |
+
"rewards/chosen": -0.9603120684623718,
|
| 472 |
+
"rewards/margins": 0.6408740878105164,
|
| 473 |
+
"rewards/rejected": -1.6011863946914673,
|
| 474 |
"step": 300
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.6488749345892203,
|
| 478 |
+
"grad_norm": 26.830608117395943,
|
| 479 |
"learning_rate": 4.271000354423425e-07,
|
| 480 |
+
"logits/chosen": -2.3035359382629395,
|
| 481 |
+
"logits/rejected": -2.2721774578094482,
|
| 482 |
+
"logps/chosen": -399.39703369140625,
|
| 483 |
+
"logps/rejected": -419.53594970703125,
|
| 484 |
+
"loss": 0.5142,
|
| 485 |
+
"rewards/accuracies": 0.762499988079071,
|
| 486 |
+
"rewards/chosen": -0.8477271199226379,
|
| 487 |
+
"rewards/margins": 0.7570197582244873,
|
| 488 |
+
"rewards/rejected": -1.6047470569610596,
|
| 489 |
"step": 310
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.6698063840920984,
|
| 493 |
+
"grad_norm": 25.48809105017836,
|
| 494 |
"learning_rate": 4.2052190435769554e-07,
|
| 495 |
+
"logits/chosen": -2.2605085372924805,
|
| 496 |
+
"logits/rejected": -2.199134349822998,
|
| 497 |
+
"logps/chosen": -384.046142578125,
|
| 498 |
+
"logps/rejected": -435.5370178222656,
|
| 499 |
+
"loss": 0.5029,
|
| 500 |
+
"rewards/accuracies": 0.831250011920929,
|
| 501 |
+
"rewards/chosen": -1.1322637796401978,
|
| 502 |
+
"rewards/margins": 0.8748885989189148,
|
| 503 |
+
"rewards/rejected": -2.007152557373047,
|
| 504 |
"step": 320
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 0.6907378335949764,
|
| 508 |
+
"grad_norm": 25.806873877392455,
|
| 509 |
"learning_rate": 4.137151834863213e-07,
|
| 510 |
+
"logits/chosen": -2.2592532634735107,
|
| 511 |
+
"logits/rejected": -2.2413222789764404,
|
| 512 |
+
"logps/chosen": -370.25860595703125,
|
| 513 |
+
"logps/rejected": -438.299072265625,
|
| 514 |
+
"loss": 0.5145,
|
| 515 |
"rewards/accuracies": 0.7562500238418579,
|
| 516 |
+
"rewards/chosen": -0.8921791911125183,
|
| 517 |
+
"rewards/margins": 0.7555935382843018,
|
| 518 |
+
"rewards/rejected": -1.6477725505828857,
|
| 519 |
"step": 330
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.7116692830978545,
|
| 523 |
+
"grad_norm": 31.495534978222874,
|
| 524 |
"learning_rate": 4.0668899744407567e-07,
|
| 525 |
+
"logits/chosen": -2.151144504547119,
|
| 526 |
+
"logits/rejected": -2.1227824687957764,
|
| 527 |
+
"logps/chosen": -416.178955078125,
|
| 528 |
+
"logps/rejected": -469.052490234375,
|
| 529 |
+
"loss": 0.5129,
|
| 530 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 531 |
+
"rewards/chosen": -1.136067271232605,
|
| 532 |
+
"rewards/margins": 0.7855342626571655,
|
| 533 |
+
"rewards/rejected": -1.9216015338897705,
|
| 534 |
"step": 340
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.7326007326007326,
|
| 538 |
+
"grad_norm": 22.383341966309303,
|
| 539 |
"learning_rate": 3.994527650465352e-07,
|
| 540 |
+
"logits/chosen": -2.245666027069092,
|
| 541 |
+
"logits/rejected": -2.1971018314361572,
|
| 542 |
+
"logps/chosen": -376.11187744140625,
|
| 543 |
+
"logps/rejected": -433.7845764160156,
|
| 544 |
+
"loss": 0.4804,
|
| 545 |
"rewards/accuracies": 0.8187500238418579,
|
| 546 |
+
"rewards/chosen": -1.0256567001342773,
|
| 547 |
+
"rewards/margins": 0.8312705159187317,
|
| 548 |
+
"rewards/rejected": -1.856927514076233,
|
| 549 |
"step": 350
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.7535321821036107,
|
| 553 |
+
"grad_norm": 22.50493532567032,
|
| 554 |
"learning_rate": 3.920161866827889e-07,
|
| 555 |
+
"logits/chosen": -2.2463080883026123,
|
| 556 |
+
"logits/rejected": -2.190325975418091,
|
| 557 |
+
"logps/chosen": -426.4788513183594,
|
| 558 |
+
"logps/rejected": -468.1583557128906,
|
| 559 |
+
"loss": 0.5312,
|
| 560 |
+
"rewards/accuracies": 0.768750011920929,
|
| 561 |
+
"rewards/chosen": -1.0429120063781738,
|
| 562 |
+
"rewards/margins": 0.8944908976554871,
|
| 563 |
+
"rewards/rejected": -1.9374029636383057,
|
| 564 |
"step": 360
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 0.7744636316064888,
|
| 568 |
+
"grad_norm": 31.530089384956717,
|
| 569 |
"learning_rate": 3.8438923131177237e-07,
|
| 570 |
+
"logits/chosen": -2.1957874298095703,
|
| 571 |
+
"logits/rejected": -2.1412596702575684,
|
| 572 |
+
"logps/chosen": -406.10504150390625,
|
| 573 |
+
"logps/rejected": -453.65252685546875,
|
| 574 |
+
"loss": 0.5133,
|
| 575 |
+
"rewards/accuracies": 0.762499988079071,
|
| 576 |
+
"rewards/chosen": -1.265554666519165,
|
| 577 |
+
"rewards/margins": 0.8195264935493469,
|
| 578 |
+
"rewards/rejected": -2.085081100463867,
|
| 579 |
"step": 370
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.7953950811093669,
|
| 583 |
+
"grad_norm": 22.98180976919898,
|
| 584 |
"learning_rate": 3.765821230985757e-07,
|
| 585 |
+
"logits/chosen": -2.0363900661468506,
|
| 586 |
+
"logits/rejected": -2.024770736694336,
|
| 587 |
+
"logps/chosen": -385.59698486328125,
|
| 588 |
+
"logps/rejected": -494.7015075683594,
|
| 589 |
+
"loss": 0.5061,
|
| 590 |
"rewards/accuracies": 0.737500011920929,
|
| 591 |
+
"rewards/chosen": -1.2817010879516602,
|
| 592 |
+
"rewards/margins": 0.9372233152389526,
|
| 593 |
+
"rewards/rejected": -2.2189245223999023,
|
| 594 |
"step": 380
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 0.8163265306122449,
|
| 598 |
+
"grad_norm": 25.12613071409362,
|
| 599 |
"learning_rate": 3.6860532770864005e-07,
|
| 600 |
+
"logits/chosen": -2.099245548248291,
|
| 601 |
+
"logits/rejected": -2.04612135887146,
|
| 602 |
+
"logps/chosen": -457.05682373046875,
|
| 603 |
+
"logps/rejected": -488.36944580078125,
|
| 604 |
+
"loss": 0.5015,
|
| 605 |
+
"rewards/accuracies": 0.731249988079071,
|
| 606 |
+
"rewards/chosen": -1.019736647605896,
|
| 607 |
+
"rewards/margins": 0.7141124606132507,
|
| 608 |
+
"rewards/rejected": -1.7338489294052124,
|
| 609 |
"step": 390
|
| 610 |
},
|
| 611 |
{
|
| 612 |
"epoch": 0.837257980115123,
|
| 613 |
+
"grad_norm": 23.959018744941538,
|
| 614 |
"learning_rate": 3.604695382782159e-07,
|
| 615 |
+
"logits/chosen": -1.991755485534668,
|
| 616 |
+
"logits/rejected": -1.9470754861831665,
|
| 617 |
+
"logps/chosen": -408.3104248046875,
|
| 618 |
+
"logps/rejected": -473.34637451171875,
|
| 619 |
+
"loss": 0.5099,
|
| 620 |
+
"rewards/accuracies": 0.78125,
|
| 621 |
+
"rewards/chosen": -1.4101927280426025,
|
| 622 |
+
"rewards/margins": 0.8306114077568054,
|
| 623 |
+
"rewards/rejected": -2.2408041954040527,
|
| 624 |
"step": 400
|
| 625 |
},
|
| 626 |
{
|
| 627 |
"epoch": 0.858189429618001,
|
| 628 |
+
"grad_norm": 21.1631977830508,
|
| 629 |
"learning_rate": 3.5218566107988867e-07,
|
| 630 |
+
"logits/chosen": -2.091287136077881,
|
| 631 |
+
"logits/rejected": -2.0397820472717285,
|
| 632 |
+
"logps/chosen": -457.28131103515625,
|
| 633 |
+
"logps/rejected": -507.02667236328125,
|
| 634 |
+
"loss": 0.5122,
|
| 635 |
"rewards/accuracies": 0.7437499761581421,
|
| 636 |
+
"rewards/chosen": -1.4175649881362915,
|
| 637 |
+
"rewards/margins": 0.8866469264030457,
|
| 638 |
+
"rewards/rejected": -2.3042116165161133,
|
| 639 |
"step": 410
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.8791208791208791,
|
| 643 |
+
"grad_norm": 27.242164262823913,
|
| 644 |
"learning_rate": 3.4376480090239047e-07,
|
| 645 |
+
"logits/chosen": -2.0190649032592773,
|
| 646 |
+
"logits/rejected": -2.000903844833374,
|
| 647 |
+
"logps/chosen": -379.5609436035156,
|
| 648 |
+
"logps/rejected": -469.435302734375,
|
| 649 |
+
"loss": 0.5039,
|
| 650 |
+
"rewards/accuracies": 0.793749988079071,
|
| 651 |
+
"rewards/chosen": -1.0910499095916748,
|
| 652 |
+
"rewards/margins": 0.9324447512626648,
|
| 653 |
+
"rewards/rejected": -2.0234947204589844,
|
| 654 |
"step": 420
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.9000523286237572,
|
| 658 |
+
"grad_norm": 30.81437560389435,
|
| 659 |
"learning_rate": 3.3521824616429284e-07,
|
| 660 |
+
"logits/chosen": -2.035524606704712,
|
| 661 |
+
"logits/rejected": -1.9534003734588623,
|
| 662 |
+
"logps/chosen": -431.333740234375,
|
| 663 |
+
"logps/rejected": -497.05035400390625,
|
| 664 |
+
"loss": 0.4921,
|
| 665 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 666 |
+
"rewards/chosen": -1.2844098806381226,
|
| 667 |
+
"rewards/margins": 0.9379785656929016,
|
| 668 |
+
"rewards/rejected": -2.222388505935669,
|
| 669 |
"step": 430
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 0.9209837781266352,
|
| 673 |
+
"grad_norm": 27.873966573342,
|
| 674 |
"learning_rate": 3.265574537815398e-07,
|
| 675 |
+
"logits/chosen": -2.1473166942596436,
|
| 676 |
+
"logits/rejected": -2.0772416591644287,
|
| 677 |
+
"logps/chosen": -405.3023986816406,
|
| 678 |
+
"logps/rejected": -452.0018615722656,
|
| 679 |
+
"loss": 0.4995,
|
| 680 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 681 |
+
"rewards/chosen": -1.3173587322235107,
|
| 682 |
+
"rewards/margins": 0.8507854342460632,
|
| 683 |
+
"rewards/rejected": -2.1681439876556396,
|
| 684 |
"step": 440
|
| 685 |
},
|
| 686 |
{
|
| 687 |
"epoch": 0.9419152276295133,
|
| 688 |
+
"grad_norm": 27.057353974975,
|
| 689 |
"learning_rate": 3.1779403380910425e-07,
|
| 690 |
+
"logits/chosen": -1.9614715576171875,
|
| 691 |
+
"logits/rejected": -1.947115182876587,
|
| 692 |
+
"logps/chosen": -397.4813232421875,
|
| 693 |
+
"logps/rejected": -475.9383850097656,
|
| 694 |
+
"loss": 0.5059,
|
| 695 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 696 |
+
"rewards/chosen": -1.1171411275863647,
|
| 697 |
+
"rewards/margins": 0.9366900324821472,
|
| 698 |
+
"rewards/rejected": -2.053831100463867,
|
| 699 |
"step": 450
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.9628466771323915,
|
| 703 |
+
"grad_norm": 23.906392544335443,
|
| 704 |
"learning_rate": 3.0893973387735683e-07,
|
| 705 |
+
"logits/chosen": -2.1360177993774414,
|
| 706 |
+
"logits/rejected": -2.067448139190674,
|
| 707 |
+
"logps/chosen": -395.38140869140625,
|
| 708 |
+
"logps/rejected": -425.06060791015625,
|
| 709 |
+
"loss": 0.5181,
|
| 710 |
+
"rewards/accuracies": 0.762499988079071,
|
| 711 |
+
"rewards/chosen": -1.0835450887680054,
|
| 712 |
+
"rewards/margins": 0.7977073192596436,
|
| 713 |
+
"rewards/rejected": -1.8812522888183594,
|
| 714 |
"step": 460
|
| 715 |
},
|
| 716 |
{
|
| 717 |
"epoch": 0.9837781266352695,
|
| 718 |
+
"grad_norm": 23.415439117206944,
|
| 719 |
"learning_rate": 3.000064234440111e-07,
|
| 720 |
+
"logits/chosen": -2.1340832710266113,
|
| 721 |
+
"logits/rejected": -2.099747657775879,
|
| 722 |
+
"logps/chosen": -433.472900390625,
|
| 723 |
+
"logps/rejected": -478.5433654785156,
|
| 724 |
+
"loss": 0.4926,
|
| 725 |
+
"rewards/accuracies": 0.706250011920929,
|
| 726 |
+
"rewards/chosen": -1.1599546670913696,
|
| 727 |
+
"rewards/margins": 0.8136328458786011,
|
| 728 |
+
"rewards/rejected": -1.9735876321792603,
|
| 729 |
"step": 470
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.9984301412872841,
|
| 733 |
+
"eval_logits/chosen": -2.1760072708129883,
|
| 734 |
+
"eval_logits/rejected": -2.14554762840271,
|
| 735 |
+
"eval_logps/chosen": -412.9510192871094,
|
| 736 |
+
"eval_logps/rejected": -506.2944030761719,
|
| 737 |
+
"eval_loss": 0.5081629753112793,
|
| 738 |
+
"eval_rewards/accuracies": 0.7578125,
|
| 739 |
+
"eval_rewards/chosen": -1.3560869693756104,
|
| 740 |
+
"eval_rewards/margins": 0.9750058054924011,
|
| 741 |
+
"eval_rewards/rejected": -2.331092596054077,
|
| 742 |
+
"eval_runtime": 169.9844,
|
| 743 |
+
"eval_samples_per_second": 11.766,
|
| 744 |
+
"eval_steps_per_second": 0.188,
|
| 745 |
"step": 477
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 1.0047095761381475,
|
| 749 |
+
"grad_norm": 23.808566042067955,
|
| 750 |
"learning_rate": 2.910060778827554e-07,
|
| 751 |
+
"logits/chosen": -2.1567416191101074,
|
| 752 |
+
"logits/rejected": -2.1457903385162354,
|
| 753 |
+
"logps/chosen": -434.89910888671875,
|
| 754 |
+
"logps/rejected": -504.3150329589844,
|
| 755 |
+
"loss": 0.4859,
|
| 756 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 757 |
+
"rewards/chosen": -1.2960957288742065,
|
| 758 |
+
"rewards/margins": 0.8546605110168457,
|
| 759 |
+
"rewards/rejected": -2.1507561206817627,
|
| 760 |
"step": 480
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 1.0256410256410255,
|
| 764 |
+
"grad_norm": 21.238305952154406,
|
| 765 |
"learning_rate": 2.8195076242990116e-07,
|
| 766 |
+
"logits/chosen": -2.1224186420440674,
|
| 767 |
+
"logits/rejected": -2.067232131958008,
|
| 768 |
+
"logps/chosen": -452.2091369628906,
|
| 769 |
+
"logps/rejected": -532.13623046875,
|
| 770 |
+
"loss": 0.3595,
|
| 771 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 772 |
+
"rewards/chosen": -1.5007661581039429,
|
| 773 |
+
"rewards/margins": 1.2642806768417358,
|
| 774 |
+
"rewards/rejected": -2.7650468349456787,
|
| 775 |
"step": 490
|
| 776 |
},
|
| 777 |
{
|
| 778 |
"epoch": 1.0465724751439036,
|
| 779 |
+
"grad_norm": 29.689018536225188,
|
| 780 |
"learning_rate": 2.7285261601056697e-07,
|
| 781 |
+
"logits/chosen": -2.0156354904174805,
|
| 782 |
+
"logits/rejected": -1.971776008605957,
|
| 783 |
+
"logps/chosen": -468.71783447265625,
|
| 784 |
+
"logps/rejected": -602.7720947265625,
|
| 785 |
+
"loss": 0.339,
|
| 786 |
+
"rewards/accuracies": 0.862500011920929,
|
| 787 |
+
"rewards/chosen": -1.8369992971420288,
|
| 788 |
+
"rewards/margins": 1.511438012123108,
|
| 789 |
+
"rewards/rejected": -3.348437547683716,
|
| 790 |
"step": 500
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 1.0675039246467817,
|
| 794 |
+
"grad_norm": 29.243062750459888,
|
| 795 |
"learning_rate": 2.6372383496608186e-07,
|
| 796 |
+
"logits/chosen": -1.928108811378479,
|
| 797 |
+
"logits/rejected": -1.8696941137313843,
|
| 798 |
+
"logps/chosen": -471.1832580566406,
|
| 799 |
+
"logps/rejected": -599.2332763671875,
|
| 800 |
+
"loss": 0.3513,
|
| 801 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 802 |
+
"rewards/chosen": -1.8683710098266602,
|
| 803 |
+
"rewards/margins": 1.7321943044662476,
|
| 804 |
+
"rewards/rejected": -3.6005654335021973,
|
| 805 |
"step": 510
|
| 806 |
},
|
| 807 |
{
|
| 808 |
"epoch": 1.08843537414966,
|
| 809 |
+
"grad_norm": 25.479631786694963,
|
| 810 |
"learning_rate": 2.5457665670441937e-07,
|
| 811 |
+
"logits/chosen": -1.9516363143920898,
|
| 812 |
+
"logits/rejected": -1.8393253087997437,
|
| 813 |
+
"logps/chosen": -469.7832946777344,
|
| 814 |
+
"logps/rejected": -591.2280883789062,
|
| 815 |
+
"loss": 0.3432,
|
| 816 |
+
"rewards/accuracies": 0.875,
|
| 817 |
+
"rewards/chosen": -1.8596889972686768,
|
| 818 |
+
"rewards/margins": 1.560598373413086,
|
| 819 |
+
"rewards/rejected": -3.4202873706817627,
|
| 820 |
"step": 520
|
| 821 |
},
|
| 822 |
{
|
| 823 |
"epoch": 1.109366823652538,
|
| 824 |
+
"grad_norm": 41.52483663106515,
|
| 825 |
"learning_rate": 2.454233432955807e-07,
|
| 826 |
+
"logits/chosen": -1.983980417251587,
|
| 827 |
+
"logits/rejected": -1.8693246841430664,
|
| 828 |
+
"logps/chosen": -452.64208984375,
|
| 829 |
+
"logps/rejected": -564.1638793945312,
|
| 830 |
+
"loss": 0.3472,
|
| 831 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 832 |
+
"rewards/chosen": -1.847670316696167,
|
| 833 |
+
"rewards/margins": 1.559227705001831,
|
| 834 |
+
"rewards/rejected": -3.406898021697998,
|
| 835 |
"step": 530
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 1.130298273155416,
|
| 839 |
+
"grad_norm": 26.257090591710913,
|
| 840 |
"learning_rate": 2.3627616503391812e-07,
|
| 841 |
+
"logits/chosen": -1.9792553186416626,
|
| 842 |
+
"logits/rejected": -1.9432990550994873,
|
| 843 |
+
"logps/chosen": -432.3143615722656,
|
| 844 |
+
"logps/rejected": -584.3980102539062,
|
| 845 |
+
"loss": 0.3457,
|
| 846 |
"rewards/accuracies": 0.8812500238418579,
|
| 847 |
+
"rewards/chosen": -1.673628568649292,
|
| 848 |
+
"rewards/margins": 1.6501662731170654,
|
| 849 |
+
"rewards/rejected": -3.3237948417663574,
|
| 850 |
"step": 540
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 1.1512297226582942,
|
| 854 |
+
"grad_norm": 25.747556328035525,
|
| 855 |
"learning_rate": 2.2714738398943308e-07,
|
| 856 |
+
"logits/chosen": -1.9460952281951904,
|
| 857 |
+
"logits/rejected": -1.875335693359375,
|
| 858 |
+
"logps/chosen": -468.60198974609375,
|
| 859 |
+
"logps/rejected": -596.0556640625,
|
| 860 |
+
"loss": 0.3378,
|
| 861 |
+
"rewards/accuracies": 0.856249988079071,
|
| 862 |
+
"rewards/chosen": -1.7640159130096436,
|
| 863 |
+
"rewards/margins": 1.6453392505645752,
|
| 864 |
+
"rewards/rejected": -3.409355640411377,
|
| 865 |
"step": 550
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 1.1721611721611722,
|
| 869 |
+
"grad_norm": 41.43400202902694,
|
| 870 |
"learning_rate": 2.1804923757009882e-07,
|
| 871 |
+
"logits/chosen": -1.9045295715332031,
|
| 872 |
+
"logits/rejected": -1.8370736837387085,
|
| 873 |
+
"logps/chosen": -432.0585021972656,
|
| 874 |
+
"logps/rejected": -571.2145385742188,
|
| 875 |
+
"loss": 0.3533,
|
| 876 |
+
"rewards/accuracies": 0.925000011920929,
|
| 877 |
+
"rewards/chosen": -1.6177046298980713,
|
| 878 |
+
"rewards/margins": 1.5667366981506348,
|
| 879 |
+
"rewards/rejected": -3.184441328048706,
|
| 880 |
"step": 560
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 1.1930926216640503,
|
| 884 |
+
"grad_norm": 34.75832087075438,
|
| 885 |
"learning_rate": 2.089939221172446e-07,
|
| 886 |
+
"logits/chosen": -1.939172387123108,
|
| 887 |
+
"logits/rejected": -1.8915830850601196,
|
| 888 |
+
"logps/chosen": -472.9189453125,
|
| 889 |
+
"logps/rejected": -638.5263061523438,
|
| 890 |
+
"loss": 0.3399,
|
| 891 |
+
"rewards/accuracies": 0.918749988079071,
|
| 892 |
+
"rewards/chosen": -1.6979265213012695,
|
| 893 |
+
"rewards/margins": 1.673963189125061,
|
| 894 |
+
"rewards/rejected": -3.371889591217041,
|
| 895 |
"step": 570
|
| 896 |
},
|
| 897 |
{
|
| 898 |
"epoch": 1.2140240711669283,
|
| 899 |
+
"grad_norm": 30.420080915117918,
|
| 900 |
"learning_rate": 1.9999357655598891e-07,
|
| 901 |
+
"logits/chosen": -1.8901792764663696,
|
| 902 |
+
"logits/rejected": -1.8162235021591187,
|
| 903 |
+
"logps/chosen": -475.49859619140625,
|
| 904 |
+
"logps/rejected": -622.9483032226562,
|
| 905 |
+
"loss": 0.3291,
|
| 906 |
"rewards/accuracies": 0.875,
|
| 907 |
+
"rewards/chosen": -1.903842568397522,
|
| 908 |
+
"rewards/margins": 1.5753123760223389,
|
| 909 |
+
"rewards/rejected": -3.4791553020477295,
|
| 910 |
"step": 580
|
| 911 |
},
|
| 912 |
{
|
| 913 |
"epoch": 1.2349555206698064,
|
| 914 |
+
"grad_norm": 22.335542729275428,
|
| 915 |
"learning_rate": 1.9106026612264315e-07,
|
| 916 |
+
"logits/chosen": -1.867323637008667,
|
| 917 |
+
"logits/rejected": -1.8158063888549805,
|
| 918 |
+
"logps/chosen": -455.6680603027344,
|
| 919 |
+
"logps/rejected": -624.3533935546875,
|
| 920 |
+
"loss": 0.3306,
|
| 921 |
+
"rewards/accuracies": 0.887499988079071,
|
| 922 |
+
"rewards/chosen": -1.6538959741592407,
|
| 923 |
+
"rewards/margins": 1.7916101217269897,
|
| 924 |
+
"rewards/rejected": -3.4455063343048096,
|
| 925 |
"step": 590
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 1.2558869701726845,
|
| 929 |
+
"grad_norm": 36.90747690001929,
|
| 930 |
"learning_rate": 1.8220596619089573e-07,
|
| 931 |
+
"logits/chosen": -1.8775126934051514,
|
| 932 |
+
"logits/rejected": -1.8093585968017578,
|
| 933 |
+
"logps/chosen": -513.2244873046875,
|
| 934 |
+
"logps/rejected": -664.3628540039062,
|
| 935 |
+
"loss": 0.3203,
|
| 936 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 937 |
+
"rewards/chosen": -1.8860762119293213,
|
| 938 |
+
"rewards/margins": 1.6223599910736084,
|
| 939 |
+
"rewards/rejected": -3.5084362030029297,
|
| 940 |
"step": 600
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 1.2768184196755625,
|
| 944 |
+
"grad_norm": 33.78718364461237,
|
| 945 |
"learning_rate": 1.7344254621846017e-07,
|
| 946 |
+
"logits/chosen": -1.8397912979125977,
|
| 947 |
+
"logits/rejected": -1.718785047531128,
|
| 948 |
+
"logps/chosen": -515.6451416015625,
|
| 949 |
+
"logps/rejected": -648.025390625,
|
| 950 |
+
"loss": 0.3529,
|
| 951 |
+
"rewards/accuracies": 0.8812500238418579,
|
| 952 |
+
"rewards/chosen": -2.0395894050598145,
|
| 953 |
+
"rewards/margins": 1.8018648624420166,
|
| 954 |
+
"rewards/rejected": -3.841454267501831,
|
| 955 |
"step": 610
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 1.2977498691784406,
|
| 959 |
+
"grad_norm": 31.52253218044683,
|
| 960 |
"learning_rate": 1.647817538357072e-07,
|
| 961 |
+
"logits/chosen": -1.8304967880249023,
|
| 962 |
+
"logits/rejected": -1.7391868829727173,
|
| 963 |
+
"logps/chosen": -446.7728576660156,
|
| 964 |
+
"logps/rejected": -589.5924682617188,
|
| 965 |
+
"loss": 0.3282,
|
| 966 |
"rewards/accuracies": 0.824999988079071,
|
| 967 |
+
"rewards/chosen": -1.8400465250015259,
|
| 968 |
+
"rewards/margins": 1.6443830728530884,
|
| 969 |
+
"rewards/rejected": -3.4844298362731934,
|
| 970 |
"step": 620
|
| 971 |
},
|
| 972 |
{
|
| 973 |
"epoch": 1.3186813186813187,
|
| 974 |
+
"grad_norm": 27.590908392591455,
|
| 975 |
"learning_rate": 1.562351990976095e-07,
|
| 976 |
+
"logits/chosen": -1.8083171844482422,
|
| 977 |
+
"logits/rejected": -1.7326056957244873,
|
| 978 |
+
"logps/chosen": -467.4588928222656,
|
| 979 |
+
"logps/rejected": -620.0399169921875,
|
| 980 |
+
"loss": 0.3287,
|
| 981 |
+
"rewards/accuracies": 0.893750011920929,
|
| 982 |
+
"rewards/chosen": -1.8651363849639893,
|
| 983 |
+
"rewards/margins": 1.7772023677825928,
|
| 984 |
+
"rewards/rejected": -3.642338514328003,
|
| 985 |
"step": 630
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 1.3396127681841967,
|
| 989 |
+
"grad_norm": 32.562595361242664,
|
| 990 |
"learning_rate": 1.478143389201113e-07,
|
| 991 |
+
"logits/chosen": -1.8179775476455688,
|
| 992 |
+
"logits/rejected": -1.7229331731796265,
|
| 993 |
+
"logps/chosen": -464.788818359375,
|
| 994 |
+
"logps/rejected": -596.3846435546875,
|
| 995 |
+
"loss": 0.3404,
|
| 996 |
+
"rewards/accuracies": 0.84375,
|
| 997 |
+
"rewards/chosen": -1.8788200616836548,
|
| 998 |
+
"rewards/margins": 1.667773962020874,
|
| 999 |
+
"rewards/rejected": -3.5465941429138184,
|
| 1000 |
"step": 640
|
| 1001 |
},
|
| 1002 |
{
|
| 1003 |
"epoch": 1.3605442176870748,
|
| 1004 |
+
"grad_norm": 25.048068766086473,
|
| 1005 |
"learning_rate": 1.3953046172178413e-07,
|
| 1006 |
+
"logits/chosen": -1.831610083580017,
|
| 1007 |
+
"logits/rejected": -1.7702150344848633,
|
| 1008 |
+
"logps/chosen": -464.4568786621094,
|
| 1009 |
+
"logps/rejected": -618.4747924804688,
|
| 1010 |
+
"loss": 0.3413,
|
| 1011 |
+
"rewards/accuracies": 0.90625,
|
| 1012 |
+
"rewards/chosen": -1.7352968454360962,
|
| 1013 |
+
"rewards/margins": 1.7301437854766846,
|
| 1014 |
+
"rewards/rejected": -3.4654407501220703,
|
| 1015 |
"step": 650
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
"epoch": 1.3814756671899528,
|
| 1019 |
+
"grad_norm": 34.73968044504535,
|
| 1020 |
"learning_rate": 1.3139467229135998e-07,
|
| 1021 |
+
"logits/chosen": -1.850510835647583,
|
| 1022 |
+
"logits/rejected": -1.768938660621643,
|
| 1023 |
+
"logps/chosen": -475.8479919433594,
|
| 1024 |
+
"logps/rejected": -614.0346069335938,
|
| 1025 |
+
"loss": 0.3299,
|
| 1026 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 1027 |
+
"rewards/chosen": -1.8521826267242432,
|
| 1028 |
+
"rewards/margins": 1.6953092813491821,
|
| 1029 |
+
"rewards/rejected": -3.5474917888641357,
|
| 1030 |
"step": 660
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 1.402407116692831,
|
| 1034 |
+
"grad_norm": 32.80513573876802,
|
| 1035 |
"learning_rate": 1.2341787690142435e-07,
|
| 1036 |
+
"logits/chosen": -1.828988790512085,
|
| 1037 |
+
"logits/rejected": -1.7655560970306396,
|
| 1038 |
+
"logps/chosen": -496.58599853515625,
|
| 1039 |
+
"logps/rejected": -668.3373413085938,
|
| 1040 |
+
"loss": 0.3336,
|
| 1041 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1042 |
+
"rewards/chosen": -2.0432398319244385,
|
| 1043 |
+
"rewards/margins": 1.6564356088638306,
|
| 1044 |
+
"rewards/rejected": -3.6996757984161377,
|
| 1045 |
"step": 670
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 1.423338566195709,
|
| 1049 |
+
"grad_norm": 23.481598085104388,
|
| 1050 |
"learning_rate": 1.1561076868822755e-07,
|
| 1051 |
+
"logits/chosen": -1.82489812374115,
|
| 1052 |
+
"logits/rejected": -1.736617088317871,
|
| 1053 |
+
"logps/chosen": -481.0087890625,
|
| 1054 |
+
"logps/rejected": -625.2978515625,
|
| 1055 |
+
"loss": 0.3322,
|
| 1056 |
+
"rewards/accuracies": 0.856249988079071,
|
| 1057 |
+
"rewards/chosen": -1.9523290395736694,
|
| 1058 |
+
"rewards/margins": 1.745566725730896,
|
| 1059 |
+
"rewards/rejected": -3.6978955268859863,
|
| 1060 |
"step": 680
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 1.4442700156985873,
|
| 1064 |
+
"grad_norm": 25.58609752457987,
|
| 1065 |
"learning_rate": 1.0798381331721107e-07,
|
| 1066 |
+
"logits/chosen": -1.7810049057006836,
|
| 1067 |
+
"logits/rejected": -1.7480462789535522,
|
| 1068 |
+
"logps/chosen": -467.09161376953125,
|
| 1069 |
+
"logps/rejected": -561.844482421875,
|
| 1070 |
+
"loss": 0.3345,
|
| 1071 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 1072 |
+
"rewards/chosen": -1.8833472728729248,
|
| 1073 |
+
"rewards/margins": 1.3580461740493774,
|
| 1074 |
+
"rewards/rejected": -3.241393566131592,
|
| 1075 |
"step": 690
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
"epoch": 1.4652014652014653,
|
| 1079 |
+
"grad_norm": 24.893301883943476,
|
| 1080 |
"learning_rate": 1.0054723495346482e-07,
|
| 1081 |
+
"logits/chosen": -1.8719736337661743,
|
| 1082 |
+
"logits/rejected": -1.7901394367218018,
|
| 1083 |
+
"logps/chosen": -495.5224609375,
|
| 1084 |
+
"logps/rejected": -657.6851806640625,
|
| 1085 |
+
"loss": 0.325,
|
| 1086 |
+
"rewards/accuracies": 0.856249988079071,
|
| 1087 |
+
"rewards/chosen": -1.8440284729003906,
|
| 1088 |
+
"rewards/margins": 1.9203050136566162,
|
| 1089 |
+
"rewards/rejected": -3.7643332481384277,
|
| 1090 |
"step": 700
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 1.4861329147043434,
|
| 1094 |
+
"grad_norm": 31.496542283071072,
|
| 1095 |
"learning_rate": 9.331100255592436e-08,
|
| 1096 |
+
"logits/chosen": -1.8443610668182373,
|
| 1097 |
+
"logits/rejected": -1.8108571767807007,
|
| 1098 |
+
"logps/chosen": -477.08660888671875,
|
| 1099 |
+
"logps/rejected": -641.9728393554688,
|
| 1100 |
+
"loss": 0.3221,
|
| 1101 |
+
"rewards/accuracies": 0.875,
|
| 1102 |
+
"rewards/chosen": -1.9340569972991943,
|
| 1103 |
+
"rewards/margins": 1.6980880498886108,
|
| 1104 |
+
"rewards/rejected": -3.6321449279785156,
|
| 1105 |
"step": 710
|
| 1106 |
},
|
| 1107 |
{
|
| 1108 |
"epoch": 1.5070643642072215,
|
| 1109 |
+
"grad_norm": 32.42007210036511,
|
| 1110 |
"learning_rate": 8.628481651367875e-08,
|
| 1111 |
+
"logits/chosen": -1.8817088603973389,
|
| 1112 |
+
"logits/rejected": -1.7917859554290771,
|
| 1113 |
+
"logps/chosen": -490.68963623046875,
|
| 1114 |
+
"logps/rejected": -666.5413818359375,
|
| 1115 |
+
"loss": 0.3147,
|
| 1116 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1117 |
+
"rewards/chosen": -2.0003697872161865,
|
| 1118 |
+
"rewards/margins": 1.905077338218689,
|
| 1119 |
+
"rewards/rejected": -3.905447483062744,
|
| 1120 |
"step": 720
|
| 1121 |
},
|
| 1122 |
{
|
| 1123 |
"epoch": 1.5279958137100995,
|
| 1124 |
+
"grad_norm": 31.861142046155326,
|
| 1125 |
"learning_rate": 7.947809564230445e-08,
|
| 1126 |
+
"logits/chosen": -1.943153738975525,
|
| 1127 |
+
"logits/rejected": -1.8634049892425537,
|
| 1128 |
+
"logps/chosen": -467.35162353515625,
|
| 1129 |
+
"logps/rejected": -635.6701049804688,
|
| 1130 |
+
"loss": 0.311,
|
| 1131 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1132 |
+
"rewards/chosen": -1.7533117532730103,
|
| 1133 |
+
"rewards/margins": 1.6374591588974,
|
| 1134 |
+
"rewards/rejected": -3.3907711505889893,
|
| 1135 |
"step": 730
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
"epoch": 1.5489272632129776,
|
| 1139 |
+
"grad_norm": 31.38323843580912,
|
| 1140 |
"learning_rate": 7.289996455765748e-08,
|
| 1141 |
+
"logits/chosen": -1.9120324850082397,
|
| 1142 |
+
"logits/rejected": -1.8584543466567993,
|
| 1143 |
+
"logps/chosen": -483.58868408203125,
|
| 1144 |
+
"logps/rejected": -613.0440673828125,
|
| 1145 |
+
"loss": 0.3397,
|
| 1146 |
+
"rewards/accuracies": 0.90625,
|
| 1147 |
+
"rewards/chosen": -1.846814513206482,
|
| 1148 |
+
"rewards/margins": 1.7059099674224854,
|
| 1149 |
+
"rewards/rejected": -3.5527243614196777,
|
| 1150 |
"step": 740
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 1.5698587127158556,
|
| 1154 |
+
"grad_norm": 23.167192529713482,
|
| 1155 |
"learning_rate": 6.655924144404906e-08,
|
| 1156 |
+
"logits/chosen": -1.7680591344833374,
|
| 1157 |
+
"logits/rejected": -1.7400696277618408,
|
| 1158 |
+
"logps/chosen": -473.86248779296875,
|
| 1159 |
+
"logps/rejected": -644.2675170898438,
|
| 1160 |
+
"loss": 0.313,
|
| 1161 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1162 |
+
"rewards/chosen": -1.8474416732788086,
|
| 1163 |
+
"rewards/margins": 1.7949087619781494,
|
| 1164 |
+
"rewards/rejected": -3.642350673675537,
|
| 1165 |
"step": 750
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 1.5907901622187337,
|
| 1169 |
+
"grad_norm": 33.669973383827035,
|
| 1170 |
"learning_rate": 6.046442623320145e-08,
|
| 1171 |
+
"logits/chosen": -1.8686587810516357,
|
| 1172 |
+
"logits/rejected": -1.7762336730957031,
|
| 1173 |
+
"logps/chosen": -554.06591796875,
|
| 1174 |
+
"logps/rejected": -700.1908569335938,
|
| 1175 |
+
"loss": 0.3261,
|
| 1176 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 1177 |
+
"rewards/chosen": -2.0161635875701904,
|
| 1178 |
+
"rewards/margins": 1.855608582496643,
|
| 1179 |
+
"rewards/rejected": -3.871772050857544,
|
| 1180 |
"step": 760
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
"epoch": 1.6117216117216118,
|
| 1184 |
+
"grad_norm": 30.7976860953366,
|
| 1185 |
"learning_rate": 5.4623689209832484e-08,
|
| 1186 |
+
"logits/chosen": -1.9193071126937866,
|
| 1187 |
+
"logits/rejected": -1.7936471700668335,
|
| 1188 |
+
"logps/chosen": -525.6559448242188,
|
| 1189 |
+
"logps/rejected": -623.5465087890625,
|
| 1190 |
+
"loss": 0.3316,
|
| 1191 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 1192 |
+
"rewards/chosen": -2.0721254348754883,
|
| 1193 |
+
"rewards/margins": 1.649618148803711,
|
| 1194 |
+
"rewards/rejected": -3.72174334526062,
|
| 1195 |
"step": 770
|
| 1196 |
},
|
| 1197 |
{
|
| 1198 |
"epoch": 1.6326530612244898,
|
| 1199 |
+
"grad_norm": 25.816934052591822,
|
| 1200 |
"learning_rate": 4.904486005914027e-08,
|
| 1201 |
+
"logits/chosen": -1.8367938995361328,
|
| 1202 |
+
"logits/rejected": -1.7610937356948853,
|
| 1203 |
+
"logps/chosen": -509.4143981933594,
|
| 1204 |
+
"logps/rejected": -686.8128662109375,
|
| 1205 |
+
"loss": 0.3352,
|
| 1206 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1207 |
+
"rewards/chosen": -1.8448041677474976,
|
| 1208 |
+
"rewards/margins": 1.9604244232177734,
|
| 1209 |
+
"rewards/rejected": -3.8052284717559814,
|
| 1210 |
"step": 780
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 1.653584510727368,
|
| 1214 |
+
"grad_norm": 22.36666573974123,
|
| 1215 |
"learning_rate": 4.373541737087263e-08,
|
| 1216 |
+
"logits/chosen": -1.8902270793914795,
|
| 1217 |
+
"logits/rejected": -1.7814185619354248,
|
| 1218 |
+
"logps/chosen": -464.0545959472656,
|
| 1219 |
+
"logps/rejected": -618.5113525390625,
|
| 1220 |
+
"loss": 0.2998,
|
| 1221 |
+
"rewards/accuracies": 0.875,
|
| 1222 |
+
"rewards/chosen": -1.7587671279907227,
|
| 1223 |
+
"rewards/margins": 1.7177417278289795,
|
| 1224 |
+
"rewards/rejected": -3.4765090942382812,
|
| 1225 |
"step": 790
|
| 1226 |
},
|
| 1227 |
{
|
| 1228 |
"epoch": 1.674515960230246,
|
| 1229 |
+
"grad_norm": 40.0165337984474,
|
| 1230 |
"learning_rate": 3.8702478614051345e-08,
|
| 1231 |
+
"logits/chosen": -1.8753862380981445,
|
| 1232 |
+
"logits/rejected": -1.803836464881897,
|
| 1233 |
+
"logps/chosen": -516.0680541992188,
|
| 1234 |
+
"logps/rejected": -638.5137329101562,
|
| 1235 |
+
"loss": 0.3013,
|
| 1236 |
"rewards/accuracies": 0.90625,
|
| 1237 |
+
"rewards/chosen": -1.8721412420272827,
|
| 1238 |
+
"rewards/margins": 1.763323187828064,
|
| 1239 |
+
"rewards/rejected": -3.635464906692505,
|
| 1240 |
"step": 800
|
| 1241 |
},
|
| 1242 |
{
|
| 1243 |
"epoch": 1.695447409733124,
|
| 1244 |
+
"grad_norm": 27.862986136186375,
|
| 1245 |
"learning_rate": 3.3952790595787986e-08,
|
| 1246 |
+
"logits/chosen": -1.8482224941253662,
|
| 1247 |
+
"logits/rejected": -1.7838103771209717,
|
| 1248 |
+
"logps/chosen": -477.73760986328125,
|
| 1249 |
+
"logps/rejected": -661.8113403320312,
|
| 1250 |
+
"loss": 0.3248,
|
| 1251 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1252 |
+
"rewards/chosen": -1.789380431175232,
|
| 1253 |
+
"rewards/margins": 2.0652589797973633,
|
| 1254 |
+
"rewards/rejected": -3.8546395301818848,
|
| 1255 |
"step": 810
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 1.716378859236002,
|
| 1259 |
+
"grad_norm": 34.28604882984395,
|
| 1260 |
"learning_rate": 2.9492720416985e-08,
|
| 1261 |
+
"logits/chosen": -1.898856520652771,
|
| 1262 |
+
"logits/rejected": -1.8373451232910156,
|
| 1263 |
+
"logps/chosen": -471.712890625,
|
| 1264 |
+
"logps/rejected": -652.3070068359375,
|
| 1265 |
+
"loss": 0.322,
|
| 1266 |
"rewards/accuracies": 0.8687499761581421,
|
| 1267 |
+
"rewards/chosen": -1.9994008541107178,
|
| 1268 |
+
"rewards/margins": 1.8078997135162354,
|
| 1269 |
+
"rewards/rejected": -3.807300567626953,
|
| 1270 |
"step": 820
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 1.7373103087388801,
|
| 1274 |
+
"grad_norm": 25.18301550597887,
|
| 1275 |
"learning_rate": 2.5328246937043525e-08,
|
| 1276 |
+
"logits/chosen": -1.8851368427276611,
|
| 1277 |
+
"logits/rejected": -1.8264014720916748,
|
| 1278 |
+
"logps/chosen": -510.86798095703125,
|
| 1279 |
+
"logps/rejected": -643.9739379882812,
|
| 1280 |
+
"loss": 0.3037,
|
| 1281 |
+
"rewards/accuracies": 0.8812500238418579,
|
| 1282 |
+
"rewards/chosen": -1.9238808155059814,
|
| 1283 |
+
"rewards/margins": 1.7123737335205078,
|
| 1284 |
+
"rewards/rejected": -3.636254072189331,
|
| 1285 |
"step": 830
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 1.7582417582417582,
|
| 1289 |
+
"grad_norm": 39.715305232221894,
|
| 1290 |
"learning_rate": 2.1464952759020856e-08,
|
| 1291 |
+
"logits/chosen": -1.7454639673233032,
|
| 1292 |
+
"logits/rejected": -1.719659447669983,
|
| 1293 |
+
"logps/chosen": -476.7300720214844,
|
| 1294 |
+
"logps/rejected": -662.8555908203125,
|
| 1295 |
+
"loss": 0.3174,
|
| 1296 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1297 |
+
"rewards/chosen": -2.1026253700256348,
|
| 1298 |
+
"rewards/margins": 1.8342987298965454,
|
| 1299 |
+
"rewards/rejected": -3.9369239807128906,
|
| 1300 |
"step": 840
|
| 1301 |
},
|
| 1302 |
{
|
| 1303 |
"epoch": 1.7791732077446363,
|
| 1304 |
+
"grad_norm": 26.712746983305596,
|
| 1305 |
"learning_rate": 1.7908016745981856e-08,
|
| 1306 |
+
"logits/chosen": -1.8332322835922241,
|
| 1307 |
+
"logits/rejected": -1.7802534103393555,
|
| 1308 |
+
"logps/chosen": -459.216064453125,
|
| 1309 |
+
"logps/rejected": -635.7386474609375,
|
| 1310 |
+
"loss": 0.3118,
|
| 1311 |
+
"rewards/accuracies": 0.856249988079071,
|
| 1312 |
+
"rewards/chosen": -1.981248140335083,
|
| 1313 |
+
"rewards/margins": 1.7714271545410156,
|
| 1314 |
+
"rewards/rejected": -3.7526748180389404,
|
| 1315 |
"step": 850
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 1.8001046572475143,
|
| 1319 |
+
"grad_norm": 37.61857911792585,
|
| 1320 |
"learning_rate": 1.4662207078575684e-08,
|
| 1321 |
+
"logits/chosen": -1.8996174335479736,
|
| 1322 |
+
"logits/rejected": -1.843186616897583,
|
| 1323 |
+
"logps/chosen": -505.267822265625,
|
| 1324 |
+
"logps/rejected": -616.5883178710938,
|
| 1325 |
+
"loss": 0.3276,
|
| 1326 |
+
"rewards/accuracies": 0.84375,
|
| 1327 |
+
"rewards/chosen": -1.9595191478729248,
|
| 1328 |
+
"rewards/margins": 1.6286653280258179,
|
| 1329 |
+
"rewards/rejected": -3.588184356689453,
|
| 1330 |
"step": 860
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 1.8210361067503924,
|
| 1334 |
+
"grad_norm": 36.465175683499524,
|
| 1335 |
"learning_rate": 1.1731874863145142e-08,
|
| 1336 |
+
"logits/chosen": -1.8656623363494873,
|
| 1337 |
+
"logits/rejected": -1.7836997509002686,
|
| 1338 |
+
"logps/chosen": -524.4869995117188,
|
| 1339 |
+
"logps/rejected": -655.26025390625,
|
| 1340 |
+
"loss": 0.3114,
|
| 1341 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1342 |
+
"rewards/chosen": -1.984437346458435,
|
| 1343 |
+
"rewards/margins": 1.840402603149414,
|
| 1344 |
+
"rewards/rejected": -3.8248400688171387,
|
| 1345 |
"step": 870
|
| 1346 |
},
|
| 1347 |
{
|
| 1348 |
"epoch": 1.8419675562532705,
|
| 1349 |
+
"grad_norm": 31.317239416373415,
|
| 1350 |
"learning_rate": 9.12094829893642e-09,
|
| 1351 |
+
"logits/chosen": -1.8914934396743774,
|
| 1352 |
+
"logits/rejected": -1.8087772130966187,
|
| 1353 |
+
"logps/chosen": -499.74639892578125,
|
| 1354 |
+
"logps/rejected": -632.9876708984375,
|
| 1355 |
+
"loss": 0.3185,
|
| 1356 |
+
"rewards/accuracies": 0.8812500238418579,
|
| 1357 |
+
"rewards/chosen": -2.0976243019104004,
|
| 1358 |
+
"rewards/margins": 1.6914640665054321,
|
| 1359 |
+
"rewards/rejected": -3.789088010787964,
|
| 1360 |
"step": 880
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
"epoch": 1.8628990057561485,
|
| 1364 |
+
"grad_norm": 25.15346806696213,
|
| 1365 |
"learning_rate": 6.832927412229017e-09,
|
| 1366 |
+
"logits/chosen": -1.954000473022461,
|
| 1367 |
+
"logits/rejected": -1.8538668155670166,
|
| 1368 |
+
"logps/chosen": -519.0631713867188,
|
| 1369 |
+
"logps/rejected": -677.8694458007812,
|
| 1370 |
+
"loss": 0.2969,
|
| 1371 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1372 |
+
"rewards/chosen": -1.8931020498275757,
|
| 1373 |
+
"rewards/margins": 2.0227460861206055,
|
| 1374 |
+
"rewards/rejected": -3.9158482551574707,
|
| 1375 |
"step": 890
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 1.8838304552590266,
|
| 1379 |
+
"grad_norm": 47.22034878743445,
|
| 1380 |
"learning_rate": 4.8708793644441086e-09,
|
| 1381 |
+
"logits/chosen": -1.7922518253326416,
|
| 1382 |
+
"logits/rejected": -1.7178666591644287,
|
| 1383 |
+
"logps/chosen": -470.116455078125,
|
| 1384 |
+
"logps/rejected": -621.2272338867188,
|
| 1385 |
+
"loss": 0.3031,
|
| 1386 |
+
"rewards/accuracies": 0.90625,
|
| 1387 |
+
"rewards/chosen": -1.822932481765747,
|
| 1388 |
+
"rewards/margins": 1.9056673049926758,
|
| 1389 |
+
"rewards/rejected": -3.7285995483398438,
|
| 1390 |
"step": 900
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 1.9047619047619047,
|
| 1394 |
+
"grad_norm": 33.30456524388558,
|
| 1395 |
"learning_rate": 3.2374343405217884e-09,
|
| 1396 |
+
"logits/chosen": -1.8453195095062256,
|
| 1397 |
+
"logits/rejected": -1.783247947692871,
|
| 1398 |
+
"logps/chosen": -464.78131103515625,
|
| 1399 |
+
"logps/rejected": -612.1450805664062,
|
| 1400 |
+
"loss": 0.3202,
|
| 1401 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1402 |
+
"rewards/chosen": -1.9610507488250732,
|
| 1403 |
+
"rewards/margins": 1.8240289688110352,
|
| 1404 |
+
"rewards/rejected": -3.7850799560546875,
|
| 1405 |
"step": 910
|
| 1406 |
},
|
| 1407 |
{
|
| 1408 |
"epoch": 1.9256933542647827,
|
| 1409 |
+
"grad_norm": 34.393381899703904,
|
| 1410 |
"learning_rate": 1.9347820230782295e-09,
|
| 1411 |
+
"logits/chosen": -1.8662745952606201,
|
| 1412 |
+
"logits/rejected": -1.8287159204483032,
|
| 1413 |
+
"logps/chosen": -490.3973083496094,
|
| 1414 |
+
"logps/rejected": -676.7599487304688,
|
| 1415 |
+
"loss": 0.3221,
|
| 1416 |
"rewards/accuracies": 0.8999999761581421,
|
| 1417 |
+
"rewards/chosen": -1.9537601470947266,
|
| 1418 |
+
"rewards/margins": 1.8785762786865234,
|
| 1419 |
+
"rewards/rejected": -3.832336902618408,
|
| 1420 |
"step": 920
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 1.9466248037676608,
|
| 1424 |
+
"grad_norm": 48.439087658815986,
|
| 1425 |
"learning_rate": 9.64668657069706e-10,
|
| 1426 |
+
"logits/chosen": -1.8716627359390259,
|
| 1427 |
+
"logits/rejected": -1.8195035457611084,
|
| 1428 |
+
"logps/chosen": -469.0355529785156,
|
| 1429 |
+
"logps/rejected": -609.7899780273438,
|
| 1430 |
+
"loss": 0.3118,
|
| 1431 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1432 |
+
"rewards/chosen": -1.8933093547821045,
|
| 1433 |
+
"rewards/margins": 1.6561877727508545,
|
| 1434 |
+
"rewards/rejected": -3.549497127532959,
|
| 1435 |
"step": 930
|
| 1436 |
},
|
| 1437 |
{
|
| 1438 |
"epoch": 1.9675562532705388,
|
| 1439 |
+
"grad_norm": 33.514570074363135,
|
| 1440 |
"learning_rate": 3.2839470889836627e-10,
|
| 1441 |
+
"logits/chosen": -1.8711134195327759,
|
| 1442 |
+
"logits/rejected": -1.7610361576080322,
|
| 1443 |
+
"logps/chosen": -504.5443420410156,
|
| 1444 |
+
"logps/rejected": -611.5713500976562,
|
| 1445 |
+
"loss": 0.3216,
|
| 1446 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1447 |
+
"rewards/chosen": -1.9376933574676514,
|
| 1448 |
+
"rewards/margins": 1.6980724334716797,
|
| 1449 |
+
"rewards/rejected": -3.635765790939331,
|
| 1450 |
"step": 940
|
| 1451 |
},
|
| 1452 |
{
|
| 1453 |
"epoch": 1.988487702773417,
|
| 1454 |
+
"grad_norm": 33.13760490600229,
|
| 1455 |
"learning_rate": 2.6813123097352287e-11,
|
| 1456 |
+
"logits/chosen": -1.8425813913345337,
|
| 1457 |
+
"logits/rejected": -1.7561390399932861,
|
| 1458 |
+
"logps/chosen": -512.4994506835938,
|
| 1459 |
+
"logps/rejected": -666.7471313476562,
|
| 1460 |
+
"loss": 0.3138,
|
| 1461 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1462 |
+
"rewards/chosen": -2.029484748840332,
|
| 1463 |
+
"rewards/margins": 1.9451849460601807,
|
| 1464 |
+
"rewards/rejected": -3.9746696949005127,
|
| 1465 |
"step": 950
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 1.9968602825745683,
|
| 1469 |
+
"eval_logits/chosen": -1.843076467514038,
|
| 1470 |
+
"eval_logits/rejected": -1.7970439195632935,
|
| 1471 |
+
"eval_logps/chosen": -488.8499755859375,
|
| 1472 |
+
"eval_logps/rejected": -619.1251220703125,
|
| 1473 |
+
"eval_loss": 0.5119243860244751,
|
| 1474 |
+
"eval_rewards/accuracies": 0.74609375,
|
| 1475 |
+
"eval_rewards/chosen": -2.1150760650634766,
|
| 1476 |
+
"eval_rewards/margins": 1.3443243503570557,
|
| 1477 |
+
"eval_rewards/rejected": -3.4594006538391113,
|
| 1478 |
+
"eval_runtime": 169.1222,
|
| 1479 |
+
"eval_samples_per_second": 11.826,
|
| 1480 |
+
"eval_steps_per_second": 0.189,
|
| 1481 |
"step": 954
|
| 1482 |
},
|
| 1483 |
{
|
|
|
|
| 1485 |
"step": 954,
|
| 1486 |
"total_flos": 0.0,
|
| 1487 |
"train_loss": 0.0,
|
| 1488 |
+
"train_runtime": 3.478,
|
| 1489 |
+
"train_samples_per_second": 17577.112,
|
| 1490 |
+
"train_steps_per_second": 137.146
|
| 1491 |
}
|
| 1492 |
],
|
| 1493 |
"logging_steps": 10,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c6d2d89a16be19cf14d1d8c8ab9be0c7839e54c975f498671e2e6605832100e
|
| 3 |
size 7672
|