Model save
Browse files- README.md +19 -23
- all_results.json +15 -15
- 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
- runs/Jun24_18-16-23_facf8d095d45/events.out.tfevents.1719253006.facf8d095d45.359765.0 +3 -0
- runs/Jun24_18-16-23_facf8d095d45/events.out.tfevents.1719273985.facf8d095d45.359765.1 +3 -0
- train_results.json +4 -4
- trainer_state.json +950 -950
- training_args.bin +2 -2
README.md
CHANGED
|
@@ -2,13 +2,9 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: alignment-handbook/zephyr-7b-sft-full
|
| 4 |
tags:
|
| 5 |
-
- alignment-handbook
|
| 6 |
-
- generated_from_trainer
|
| 7 |
- trl
|
| 8 |
- dpo
|
| 9 |
- generated_from_trainer
|
| 10 |
-
datasets:
|
| 11 |
-
- HuggingFaceH4/ultrafeedback_binarized
|
| 12 |
model-index:
|
| 13 |
- name: zephyr-7b-dpo-full
|
| 14 |
results: []
|
|
@@ -19,17 +15,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 19 |
|
| 20 |
# zephyr-7b-dpo-full
|
| 21 |
|
| 22 |
-
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
-
- Loss: 0.
|
| 25 |
-
- Rewards/chosen: -
|
| 26 |
-
- Rewards/rejected: -
|
| 27 |
-
- Rewards/accuracies: 0.
|
| 28 |
-
- Rewards/margins: 1.
|
| 29 |
-
- Logps/rejected: -
|
| 30 |
-
- Logps/chosen: -
|
| 31 |
-
- Logits/rejected:
|
| 32 |
-
- Logits/chosen: 2.
|
| 33 |
|
| 34 |
## Model description
|
| 35 |
|
|
@@ -66,15 +62,15 @@ The following hyperparameters were used during training:
|
|
| 66 |
|
| 67 |
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 68 |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 69 |
-
| 0.
|
| 70 |
-
| 0.
|
| 71 |
-
| 0.
|
| 72 |
-
| 0.
|
| 73 |
-
| 0.
|
| 74 |
-
| 0.
|
| 75 |
-
| 0.
|
| 76 |
-
| 0.
|
| 77 |
-
| 0.
|
| 78 |
|
| 79 |
|
| 80 |
### Framework versions
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: alignment-handbook/zephyr-7b-sft-full
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- trl
|
| 6 |
- dpo
|
| 7 |
- generated_from_trainer
|
|
|
|
|
|
|
| 8 |
model-index:
|
| 9 |
- name: zephyr-7b-dpo-full
|
| 10 |
results: []
|
|
|
|
| 15 |
|
| 16 |
# zephyr-7b-dpo-full
|
| 17 |
|
| 18 |
+
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the None dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.6712
|
| 21 |
+
- Rewards/chosen: -2.0287
|
| 22 |
+
- Rewards/rejected: -3.3245
|
| 23 |
+
- Rewards/accuracies: 0.7639
|
| 24 |
+
- Rewards/margins: 1.2958
|
| 25 |
+
- Logps/rejected: -594.2247
|
| 26 |
+
- Logps/chosen: -486.9804
|
| 27 |
+
- Logits/rejected: 3.7376
|
| 28 |
+
- Logits/chosen: 2.4533
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 62 |
|
| 63 |
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 64 |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 65 |
+
| 0.8957 | 0.1 | 100 | 0.9028 | -0.5210 | -0.6849 | 0.6905 | 0.1639 | -330.2668 | -336.2060 | -2.5107 | -2.5460 |
|
| 66 |
+
| 0.7658 | 0.21 | 200 | 0.7650 | -0.9414 | -1.6932 | 0.7460 | 0.7519 | -431.1015 | -378.2476 | 0.3347 | -0.1529 |
|
| 67 |
+
| 0.7079 | 0.31 | 300 | 0.7289 | -1.3837 | -2.4868 | 0.7560 | 1.1031 | -510.4591 | -422.4754 | 1.8370 | 0.8744 |
|
| 68 |
+
| 0.6806 | 0.42 | 400 | 0.7040 | -1.3285 | -2.4190 | 0.7698 | 1.0904 | -503.6740 | -416.9630 | 1.2713 | 0.0992 |
|
| 69 |
+
| 0.7129 | 0.52 | 500 | 0.6980 | -1.4621 | -2.5268 | 0.7440 | 1.0648 | -514.4609 | -430.3167 | 2.3343 | 1.4091 |
|
| 70 |
+
| 0.6636 | 0.63 | 600 | 0.6877 | -1.3328 | -2.5188 | 0.75 | 1.1861 | -513.6627 | -417.3850 | 2.2082 | 0.7470 |
|
| 71 |
+
| 0.6217 | 0.73 | 700 | 0.6762 | -1.8908 | -3.1786 | 0.7698 | 1.2878 | -579.6354 | -473.1887 | 3.8163 | 2.5932 |
|
| 72 |
+
| 0.6418 | 0.84 | 800 | 0.6712 | -2.0993 | -3.4028 | 0.7679 | 1.3035 | -602.0607 | -494.0422 | 3.8655 | 2.6092 |
|
| 73 |
+
| 0.6678 | 0.94 | 900 | 0.6716 | -2.0307 | -3.3233 | 0.7639 | 1.2926 | -594.1103 | -487.1844 | 3.7332 | 2.4518 |
|
| 74 |
|
| 75 |
|
| 76 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,21 +1,21 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_logits/chosen": 2.
|
| 4 |
-
"eval_logits/rejected":
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_rewards/accuracies": 0.
|
| 9 |
-
"eval_rewards/chosen": -
|
| 10 |
-
"eval_rewards/margins": 1.
|
| 11 |
-
"eval_rewards/rejected": -
|
| 12 |
-
"eval_runtime":
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second": 8.
|
| 15 |
"eval_steps_per_second": 0.258,
|
| 16 |
-
"train_loss": 0.
|
| 17 |
-
"train_runtime":
|
| 18 |
"train_samples": 61135,
|
| 19 |
-
"train_samples_per_second": 2.
|
| 20 |
-
"train_steps_per_second": 0.
|
| 21 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": 2.4533472061157227,
|
| 4 |
+
"eval_logits/rejected": 3.7376418113708496,
|
| 5 |
+
"eval_logps/chosen": -486.9803771972656,
|
| 6 |
+
"eval_logps/rejected": -594.2247314453125,
|
| 7 |
+
"eval_loss": 0.6712061166763306,
|
| 8 |
+
"eval_rewards/accuracies": 0.7638888955116272,
|
| 9 |
+
"eval_rewards/chosen": -2.0287060737609863,
|
| 10 |
+
"eval_rewards/margins": 1.2957550287246704,
|
| 11 |
+
"eval_rewards/rejected": -3.324460983276367,
|
| 12 |
+
"eval_runtime": 244.3063,
|
| 13 |
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 8.186,
|
| 15 |
"eval_steps_per_second": 0.258,
|
| 16 |
+
"train_loss": 0.7263204834224042,
|
| 17 |
+
"train_runtime": 20734.7169,
|
| 18 |
"train_samples": 61135,
|
| 19 |
+
"train_samples_per_second": 2.948,
|
| 20 |
+
"train_steps_per_second": 0.046
|
| 21 |
}
|
config.json
CHANGED
|
@@ -21,6 +21,6 @@
|
|
| 21 |
"tie_word_embeddings": false,
|
| 22 |
"torch_dtype": "bfloat16",
|
| 23 |
"transformers_version": "4.36.2",
|
| 24 |
-
"use_cache":
|
| 25 |
"vocab_size": 32000
|
| 26 |
}
|
|
|
|
| 21 |
"tie_word_embeddings": false,
|
| 22 |
"torch_dtype": "bfloat16",
|
| 23 |
"transformers_version": "4.36.2",
|
| 24 |
+
"use_cache": false,
|
| 25 |
"vocab_size": 32000
|
| 26 |
}
|
eval_results.json
CHANGED
|
@@ -1,16 +1,16 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_logits/chosen": 2.
|
| 4 |
-
"eval_logits/rejected":
|
| 5 |
-
"eval_logps/chosen": -
|
| 6 |
-
"eval_logps/rejected": -
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_rewards/accuracies": 0.
|
| 9 |
-
"eval_rewards/chosen": -
|
| 10 |
-
"eval_rewards/margins": 1.
|
| 11 |
-
"eval_rewards/rejected": -
|
| 12 |
-
"eval_runtime":
|
| 13 |
"eval_samples": 2000,
|
| 14 |
-
"eval_samples_per_second": 8.
|
| 15 |
"eval_steps_per_second": 0.258
|
| 16 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": 2.4533472061157227,
|
| 4 |
+
"eval_logits/rejected": 3.7376418113708496,
|
| 5 |
+
"eval_logps/chosen": -486.9803771972656,
|
| 6 |
+
"eval_logps/rejected": -594.2247314453125,
|
| 7 |
+
"eval_loss": 0.6712061166763306,
|
| 8 |
+
"eval_rewards/accuracies": 0.7638888955116272,
|
| 9 |
+
"eval_rewards/chosen": -2.0287060737609863,
|
| 10 |
+
"eval_rewards/margins": 1.2957550287246704,
|
| 11 |
+
"eval_rewards/rejected": -3.324460983276367,
|
| 12 |
+
"eval_runtime": 244.3063,
|
| 13 |
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 8.186,
|
| 15 |
"eval_steps_per_second": 0.258
|
| 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 sha256:8b1e9e9688ce643625583017cfcba65fa8ff37865db15e78443ca69585dd2607
|
| 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 sha256:3c520ae489cc180f688441279cef32fac2764d4f58b2900b8450d180cbd2d1e3
|
| 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:d7e6172e143dd652d2e901dbedaaa58e54a6eb5247664e6c4669fa1dfb45b15f
|
| 3 |
size 4540516344
|
runs/Jun24_18-16-23_facf8d095d45/events.out.tfevents.1719253006.facf8d095d45.359765.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86428da338c601f6b6affd0dcfe571ba509fb0cd7cee84ad60687dc019f8af40
|
| 3 |
+
size 72293
|
runs/Jun24_18-16-23_facf8d095d45/events.out.tfevents.1719273985.facf8d095d45.359765.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fac01b95a1acb80341bf4c6df1e45b03a848e7e418301c9e1ed5aa85b5589458
|
| 3 |
+
size 828
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 61135,
|
| 6 |
-
"train_samples_per_second": 2.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.7263204834224042,
|
| 4 |
+
"train_runtime": 20734.7169,
|
| 5 |
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 2.948,
|
| 7 |
+
"train_steps_per_second": 0.046
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -15,7 +15,7 @@
|
|
| 15 |
"logits/rejected": -2.686896800994873,
|
| 16 |
"logps/chosen": -229.94229125976562,
|
| 17 |
"logps/rejected": -214.70114135742188,
|
| 18 |
-
"loss": 0.
|
| 19 |
"rewards/accuracies": 0.0,
|
| 20 |
"rewards/chosen": 0.0,
|
| 21 |
"rewards/margins": 0.0,
|
|
@@ -25,1492 +25,1492 @@
|
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
"learning_rate": 5.208333333333333e-08,
|
| 28 |
-
"logits/chosen": -2.
|
| 29 |
-
"logits/rejected": -2.
|
| 30 |
-
"logps/chosen": -295.
|
| 31 |
-
"logps/rejected": -250.
|
| 32 |
-
"loss": 0.
|
| 33 |
-
"rewards/accuracies": 0.
|
| 34 |
-
"rewards/chosen":
|
| 35 |
-
"rewards/margins":
|
| 36 |
-
"rewards/rejected":
|
| 37 |
"step": 10
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02,
|
| 41 |
"learning_rate": 1.0416666666666667e-07,
|
| 42 |
-
"logits/chosen": -2.
|
| 43 |
-
"logits/rejected": -2.
|
| 44 |
-
"logps/chosen": -271.
|
| 45 |
-
"logps/rejected": -246.
|
| 46 |
-
"loss": 0.
|
| 47 |
-
"rewards/accuracies": 0.
|
| 48 |
-
"rewards/chosen": 0.
|
| 49 |
-
"rewards/margins": 0.
|
| 50 |
-
"rewards/rejected": -0.
|
| 51 |
"step": 20
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03,
|
| 55 |
"learning_rate": 1.5624999999999999e-07,
|
| 56 |
-
"logits/chosen": -2.
|
| 57 |
-
"logits/rejected": -2.
|
| 58 |
-
"logps/chosen": -278.
|
| 59 |
-
"logps/rejected": -254.
|
| 60 |
-
"loss": 0.
|
| 61 |
-
"rewards/accuracies": 0.
|
| 62 |
-
"rewards/chosen": 0.
|
| 63 |
-
"rewards/margins": 0.
|
| 64 |
-
"rewards/rejected": -0.
|
| 65 |
"step": 30
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.04,
|
| 69 |
"learning_rate": 2.0833333333333333e-07,
|
| 70 |
-
"logits/chosen": -2.
|
| 71 |
-
"logits/rejected": -2.
|
| 72 |
-
"logps/chosen": -273.
|
| 73 |
-
"logps/rejected": -237.
|
| 74 |
-
"loss": 0.
|
| 75 |
-
"rewards/accuracies": 0.
|
| 76 |
-
"rewards/chosen": 0.
|
| 77 |
-
"rewards/margins": 0.
|
| 78 |
-
"rewards/rejected": -0.
|
| 79 |
"step": 40
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05,
|
| 83 |
"learning_rate": 2.604166666666667e-07,
|
| 84 |
-
"logits/chosen": -2.
|
| 85 |
-
"logits/rejected": -2.
|
| 86 |
-
"logps/chosen": -296.
|
| 87 |
-
"logps/rejected": -274.
|
| 88 |
-
"loss": 0.
|
| 89 |
-
"rewards/accuracies": 0.
|
| 90 |
-
"rewards/chosen": 0.
|
| 91 |
-
"rewards/margins": 0.
|
| 92 |
-
"rewards/rejected": -0.
|
| 93 |
"step": 50
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06,
|
| 97 |
"learning_rate": 3.1249999999999997e-07,
|
| 98 |
-
"logits/chosen": -2.
|
| 99 |
-
"logits/rejected": -2.
|
| 100 |
-
"logps/chosen": -285.
|
| 101 |
-
"logps/rejected": -274.
|
| 102 |
-
"loss": 0.
|
| 103 |
-
"rewards/accuracies": 0.
|
| 104 |
-
"rewards/chosen": 0.
|
| 105 |
-
"rewards/margins": 0.
|
| 106 |
-
"rewards/rejected": -0.
|
| 107 |
"step": 60
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07,
|
| 111 |
"learning_rate": 3.645833333333333e-07,
|
| 112 |
-
"logits/chosen": -2.
|
| 113 |
-
"logits/rejected": -2.
|
| 114 |
-
"logps/chosen": -310.
|
| 115 |
-
"logps/rejected": -290.
|
| 116 |
-
"loss": 0.
|
| 117 |
-
"rewards/accuracies": 0.
|
| 118 |
-
"rewards/chosen": 0.
|
| 119 |
-
"rewards/margins": 0.
|
| 120 |
-
"rewards/rejected": -0.
|
| 121 |
"step": 70
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08,
|
| 125 |
"learning_rate": 4.1666666666666667e-07,
|
| 126 |
-
"logits/chosen": -2.
|
| 127 |
-
"logits/rejected": -2.
|
| 128 |
-
"logps/chosen": -304.
|
| 129 |
-
"logps/rejected": -281.
|
| 130 |
-
"loss": 0.
|
| 131 |
-
"rewards/accuracies": 0.
|
| 132 |
-
"rewards/chosen": -0.
|
| 133 |
-
"rewards/margins": 0.
|
| 134 |
-
"rewards/rejected": -0.
|
| 135 |
"step": 80
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09,
|
| 139 |
"learning_rate": 4.6874999999999996e-07,
|
| 140 |
-
"logits/chosen": -2.
|
| 141 |
-
"logits/rejected": -2.
|
| 142 |
-
"logps/chosen": -
|
| 143 |
-
"logps/rejected": -
|
| 144 |
-
"loss": 0.
|
| 145 |
-
"rewards/accuracies": 0.
|
| 146 |
-
"rewards/chosen": 0.
|
| 147 |
-
"rewards/margins": 0.
|
| 148 |
-
"rewards/rejected": -0.
|
| 149 |
"step": 90
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.1,
|
| 153 |
"learning_rate": 4.999732492681437e-07,
|
| 154 |
-
"logits/chosen": -2.
|
| 155 |
-
"logits/rejected": -2.
|
| 156 |
-
"logps/chosen": -
|
| 157 |
-
"logps/rejected": -
|
| 158 |
-
"loss": 0.
|
| 159 |
"rewards/accuracies": 0.7124999761581421,
|
| 160 |
-
"rewards/chosen": -0.
|
| 161 |
-
"rewards/margins": 0.
|
| 162 |
-
"rewards/rejected": -0.
|
| 163 |
"step": 100
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.1,
|
| 167 |
-
"eval_logits/chosen": -2.
|
| 168 |
-
"eval_logits/rejected": -2.
|
| 169 |
-
"eval_logps/chosen": -336.
|
| 170 |
-
"eval_logps/rejected": -
|
| 171 |
-
"eval_loss": 0.
|
| 172 |
-
"eval_rewards/accuracies": 0.
|
| 173 |
-
"eval_rewards/chosen": -0.
|
| 174 |
-
"eval_rewards/margins": 0.
|
| 175 |
-
"eval_rewards/rejected": -0.
|
| 176 |
-
"eval_runtime":
|
| 177 |
-
"eval_samples_per_second": 8.
|
| 178 |
-
"eval_steps_per_second": 0.
|
| 179 |
"step": 100
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.12,
|
| 183 |
"learning_rate": 4.996723692767926e-07,
|
| 184 |
-
"logits/chosen": -2.
|
| 185 |
-
"logits/rejected": -2.
|
| 186 |
-
"logps/chosen": -
|
| 187 |
-
"logps/rejected": -
|
| 188 |
-
"loss": 0.
|
| 189 |
-
"rewards/accuracies": 0.
|
| 190 |
-
"rewards/chosen": -0.
|
| 191 |
-
"rewards/margins": 0.
|
| 192 |
-
"rewards/rejected": -0.
|
| 193 |
"step": 110
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.13,
|
| 197 |
"learning_rate": 4.990375746213598e-07,
|
| 198 |
-
"logits/chosen": -
|
| 199 |
-
"logits/rejected": -
|
| 200 |
-
"logps/chosen": -
|
| 201 |
-
"logps/rejected": -
|
| 202 |
-
"loss": 0.
|
| 203 |
-
"rewards/accuracies": 0.
|
| 204 |
-
"rewards/chosen": -0.
|
| 205 |
-
"rewards/margins": 0.
|
| 206 |
-
"rewards/rejected": -
|
| 207 |
"step": 120
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.14,
|
| 211 |
"learning_rate": 4.980697142834314e-07,
|
| 212 |
-
"logits/chosen": -
|
| 213 |
-
"logits/rejected": -
|
| 214 |
-
"logps/chosen": -
|
| 215 |
-
"logps/rejected": -
|
| 216 |
-
"loss": 0.
|
| 217 |
-
"rewards/accuracies": 0.
|
| 218 |
-
"rewards/chosen": -0.
|
| 219 |
-
"rewards/margins": 0.
|
| 220 |
-
"rewards/rejected": -
|
| 221 |
"step": 130
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 0.15,
|
| 225 |
"learning_rate": 4.967700826904229e-07,
|
| 226 |
-
"logits/chosen":
|
| 227 |
-
"logits/rejected":
|
| 228 |
-
"logps/chosen": -
|
| 229 |
-
"logps/rejected": -
|
| 230 |
-
"loss": 0.
|
| 231 |
-
"rewards/accuracies": 0.
|
| 232 |
-
"rewards/chosen": -0.
|
| 233 |
-
"rewards/margins": 0.
|
| 234 |
-
"rewards/rejected": -1.
|
| 235 |
"step": 140
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 0.16,
|
| 239 |
"learning_rate": 4.951404179843962e-07,
|
| 240 |
-
"logits/chosen":
|
| 241 |
-
"logits/rejected":
|
| 242 |
-
"logps/chosen": -
|
| 243 |
-
"logps/rejected": -
|
| 244 |
-
"loss": 0.
|
| 245 |
-
"rewards/accuracies": 0.
|
| 246 |
-
"rewards/chosen": -0.
|
| 247 |
-
"rewards/margins": 0.
|
| 248 |
-
"rewards/rejected": -1.
|
| 249 |
"step": 150
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.17,
|
| 253 |
"learning_rate": 4.931828996974498e-07,
|
| 254 |
-
"logits/chosen":
|
| 255 |
-
"logits/rejected":
|
| 256 |
-
"logps/chosen": -
|
| 257 |
-
"logps/rejected": -
|
| 258 |
-
"loss": 0.
|
| 259 |
"rewards/accuracies": 0.6937500238418579,
|
| 260 |
-
"rewards/chosen": -
|
| 261 |
-
"rewards/margins": 0.
|
| 262 |
-
"rewards/rejected": -1.
|
| 263 |
"step": 160
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.18,
|
| 267 |
"learning_rate": 4.909001458367866e-07,
|
| 268 |
-
"logits/chosen": 0.
|
| 269 |
-
"logits/rejected":
|
| 270 |
-
"logps/chosen": -
|
| 271 |
-
"logps/rejected": -
|
| 272 |
-
"loss": 0.
|
| 273 |
-
"rewards/accuracies": 0.
|
| 274 |
-
"rewards/chosen": -1.
|
| 275 |
-
"rewards/margins": 0.
|
| 276 |
-
"rewards/rejected": -
|
| 277 |
"step": 170
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 0.19,
|
| 281 |
"learning_rate": 4.882952093833627e-07,
|
| 282 |
-
"logits/chosen": 0.
|
| 283 |
-
"logits/rejected": 0.
|
| 284 |
-
"logps/chosen": -
|
| 285 |
-
"logps/rejected": -
|
| 286 |
-
"loss": 0.
|
| 287 |
-
"rewards/accuracies": 0.
|
| 288 |
-
"rewards/chosen": -1.
|
| 289 |
-
"rewards/margins": 0.
|
| 290 |
-
"rewards/rejected": -1.
|
| 291 |
"step": 180
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 0.2,
|
| 295 |
"learning_rate": 4.853715742087946e-07,
|
| 296 |
-
"logits/chosen": 0.
|
| 297 |
-
"logits/rejected": 0.
|
| 298 |
-
"logps/chosen": -
|
| 299 |
-
"logps/rejected": -
|
| 300 |
-
"loss": 0.
|
| 301 |
-
"rewards/accuracies": 0.
|
| 302 |
-
"rewards/chosen": -1.
|
| 303 |
-
"rewards/margins": 0.
|
| 304 |
-
"rewards/rejected": -
|
| 305 |
"step": 190
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 0.21,
|
| 309 |
"learning_rate": 4.821331504159906e-07,
|
| 310 |
-
"logits/chosen": 0.
|
| 311 |
-
"logits/rejected":
|
| 312 |
-
"logps/chosen": -
|
| 313 |
-
"logps/rejected": -
|
| 314 |
-
"loss": 0.
|
| 315 |
-
"rewards/accuracies": 0.
|
| 316 |
-
"rewards/chosen": -1.
|
| 317 |
-
"rewards/margins": 0.
|
| 318 |
-
"rewards/rejected": -1.
|
| 319 |
"step": 200
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 0.21,
|
| 323 |
-
"eval_logits/chosen": -0.
|
| 324 |
-
"eval_logits/rejected": 0.
|
| 325 |
-
"eval_logps/chosen": -
|
| 326 |
-
"eval_logps/rejected": -
|
| 327 |
-
"eval_loss": 0.
|
| 328 |
-
"eval_rewards/accuracies": 0.
|
| 329 |
-
"eval_rewards/chosen": -
|
| 330 |
-
"eval_rewards/margins": 0.
|
| 331 |
-
"eval_rewards/rejected": -1.
|
| 332 |
-
"eval_runtime":
|
| 333 |
-
"eval_samples_per_second": 8.
|
| 334 |
-
"eval_steps_per_second": 0.
|
| 335 |
"step": 200
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 0.22,
|
| 339 |
"learning_rate": 4.785842691097342e-07,
|
| 340 |
-
"logits/chosen": -0.
|
| 341 |
-
"logits/rejected": 0.
|
| 342 |
-
"logps/chosen": -
|
| 343 |
-
"logps/rejected": -
|
| 344 |
-
"loss": 0.
|
| 345 |
"rewards/accuracies": 0.7250000238418579,
|
| 346 |
-
"rewards/chosen": -0.
|
| 347 |
-
"rewards/margins": 0.
|
| 348 |
-
"rewards/rejected": -1.
|
| 349 |
"step": 210
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 0.23,
|
| 353 |
"learning_rate": 4.7472967660421603e-07,
|
| 354 |
-
"logits/chosen":
|
| 355 |
-
"logits/rejected": 0.
|
| 356 |
-
"logps/chosen": -
|
| 357 |
-
"logps/rejected": -
|
| 358 |
-
"loss": 0.
|
| 359 |
-
"rewards/accuracies": 0.
|
| 360 |
-
"rewards/chosen": -0.
|
| 361 |
-
"rewards/margins": 0.
|
| 362 |
-
"rewards/rejected": -1.
|
| 363 |
"step": 220
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 0.24,
|
| 367 |
"learning_rate": 4.705745280752585e-07,
|
| 368 |
-
"logits/chosen":
|
| 369 |
-
"logits/rejected":
|
| 370 |
-
"logps/chosen": -
|
| 371 |
-
"logps/rejected": -
|
| 372 |
-
"loss": 0.
|
| 373 |
-
"rewards/accuracies": 0.
|
| 374 |
-
"rewards/chosen": -1.
|
| 375 |
-
"rewards/margins": 0.
|
| 376 |
-
"rewards/rejected": -
|
| 377 |
"step": 230
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.25,
|
| 381 |
"learning_rate": 4.6612438066572555e-07,
|
| 382 |
-
"logits/chosen":
|
| 383 |
-
"logits/rejected":
|
| 384 |
-
"logps/chosen": -
|
| 385 |
-
"logps/rejected": -
|
| 386 |
-
"loss": 0.
|
| 387 |
-
"rewards/accuracies": 0.
|
| 388 |
-
"rewards/chosen": -
|
| 389 |
-
"rewards/margins": 0.
|
| 390 |
-
"rewards/rejected": -
|
| 391 |
"step": 240
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 0.26,
|
| 395 |
"learning_rate": 4.6138518605333664e-07,
|
| 396 |
-
"logits/chosen": 0.
|
| 397 |
-
"logits/rejected": 0.
|
| 398 |
-
"logps/chosen": -
|
| 399 |
-
"logps/rejected": -
|
| 400 |
-
"loss": 0.
|
| 401 |
-
"rewards/accuracies": 0.
|
| 402 |
-
"rewards/chosen": -
|
| 403 |
-
"rewards/margins": 0.
|
| 404 |
-
"rewards/rejected": -1.
|
| 405 |
"step": 250
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 0.27,
|
| 409 |
"learning_rate": 4.5636328249082514e-07,
|
| 410 |
-
"logits/chosen": -
|
| 411 |
-
"logits/rejected": -0.
|
| 412 |
-
"logps/chosen": -
|
| 413 |
-
"logps/rejected": -
|
| 414 |
-
"loss": 0.
|
| 415 |
-
"rewards/accuracies": 0.
|
| 416 |
-
"rewards/chosen": -0.
|
| 417 |
-
"rewards/margins": 0.
|
| 418 |
-
"rewards/rejected": -1.
|
| 419 |
"step": 260
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 0.28,
|
| 423 |
"learning_rate": 4.510653863290871e-07,
|
| 424 |
-
"logits/chosen": -0.
|
| 425 |
-
"logits/rejected":
|
| 426 |
-
"logps/chosen": -
|
| 427 |
-
"logps/rejected": -
|
| 428 |
-
"loss": 0.
|
| 429 |
-
"rewards/accuracies": 0.
|
| 430 |
-
"rewards/chosen": -1.
|
| 431 |
-
"rewards/margins":
|
| 432 |
-
"rewards/rejected": -
|
| 433 |
"step": 270
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 0.29,
|
| 437 |
"learning_rate": 4.4549858303465737e-07,
|
| 438 |
-
"logits/chosen":
|
| 439 |
-
"logits/rejected": 0.
|
| 440 |
-
"logps/chosen": -
|
| 441 |
-
"logps/rejected": -
|
| 442 |
-
"loss": 0.
|
| 443 |
-
"rewards/accuracies": 0.
|
| 444 |
-
"rewards/chosen": -1.
|
| 445 |
-
"rewards/margins": 0.
|
| 446 |
-
"rewards/rejected": -
|
| 447 |
"step": 280
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 0.3,
|
| 451 |
"learning_rate": 4.396703177135261e-07,
|
| 452 |
-
"logits/chosen":
|
| 453 |
-
"logits/rejected": 0.
|
| 454 |
-
"logps/chosen": -
|
| 455 |
-
"logps/rejected": -
|
| 456 |
-
"loss": 0.
|
| 457 |
-
"rewards/accuracies": 0.
|
| 458 |
-
"rewards/chosen": -1.
|
| 459 |
-
"rewards/margins": 0.
|
| 460 |
-
"rewards/rejected": -
|
| 461 |
"step": 290
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 0.31,
|
| 465 |
"learning_rate": 4.335883851539693e-07,
|
| 466 |
-
"logits/chosen": 0.
|
| 467 |
-
"logits/rejected":
|
| 468 |
-
"logps/chosen": -
|
| 469 |
-
"logps/rejected": -
|
| 470 |
-
"loss": 0.
|
| 471 |
-
"rewards/accuracies": 0.
|
| 472 |
-
"rewards/chosen": -1.
|
| 473 |
-
"rewards/margins":
|
| 474 |
-
"rewards/rejected": -
|
| 475 |
"step": 300
|
| 476 |
},
|
| 477 |
{
|
| 478 |
"epoch": 0.31,
|
| 479 |
-
"eval_logits/chosen": 0.
|
| 480 |
-
"eval_logits/rejected":
|
| 481 |
-
"eval_logps/chosen": -
|
| 482 |
-
"eval_logps/rejected": -
|
| 483 |
-
"eval_loss": 0.
|
| 484 |
-
"eval_rewards/accuracies": 0.
|
| 485 |
-
"eval_rewards/chosen": -1.
|
| 486 |
-
"eval_rewards/margins":
|
| 487 |
-
"eval_rewards/rejected": -
|
| 488 |
-
"eval_runtime": 242.
|
| 489 |
-
"eval_samples_per_second": 8.
|
| 490 |
"eval_steps_per_second": 0.26,
|
| 491 |
"step": 300
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 0.32,
|
| 495 |
"learning_rate": 4.272609194017105e-07,
|
| 496 |
-
"logits/chosen": 0.
|
| 497 |
-
"logits/rejected": 1.
|
| 498 |
-
"logps/chosen": -
|
| 499 |
-
"logps/rejected": -
|
| 500 |
-
"loss": 0.
|
| 501 |
-
"rewards/accuracies": 0.
|
| 502 |
-
"rewards/chosen": -
|
| 503 |
-
"rewards/margins": 1.
|
| 504 |
-
"rewards/rejected": -2.
|
| 505 |
"step": 310
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.33,
|
| 509 |
"learning_rate": 4.2069638288135547e-07,
|
| 510 |
-
"logits/chosen":
|
| 511 |
-
"logits/rejected": 1.
|
| 512 |
-
"logps/chosen": -
|
| 513 |
-
"logps/rejected": -
|
| 514 |
-
"loss": 0.
|
| 515 |
-
"rewards/accuracies": 0.
|
| 516 |
-
"rewards/chosen": -1.
|
| 517 |
-
"rewards/margins": 0.
|
| 518 |
-
"rewards/rejected": -2.
|
| 519 |
"step": 320
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.35,
|
| 523 |
"learning_rate": 4.139035550786494e-07,
|
| 524 |
-
"logits/chosen": 0.
|
| 525 |
-
"logits/rejected":
|
| 526 |
-
"logps/chosen": -
|
| 527 |
-
"logps/rejected": -
|
| 528 |
-
"loss": 0.
|
| 529 |
"rewards/accuracies": 0.7437499761581421,
|
| 530 |
-
"rewards/chosen": -1.
|
| 531 |
-
"rewards/margins": 0.
|
| 532 |
-
"rewards/rejected": -1.
|
| 533 |
"step": 330
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 0.36,
|
| 537 |
"learning_rate": 4.0689152079869306e-07,
|
| 538 |
-
"logits/chosen": 0.
|
| 539 |
-
"logits/rejected": 0.
|
| 540 |
-
"logps/chosen": -
|
| 541 |
-
"logps/rejected": -
|
| 542 |
-
"loss": 0.
|
| 543 |
-
"rewards/accuracies": 0.
|
| 544 |
-
"rewards/chosen": -0.
|
| 545 |
-
"rewards/margins": 0.
|
| 546 |
-
"rewards/rejected": -1.
|
| 547 |
"step": 340
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 0.37,
|
| 551 |
"learning_rate": 3.99669658015821e-07,
|
| 552 |
-
"logits/chosen": 0.
|
| 553 |
-
"logits/rejected": 0.
|
| 554 |
-
"logps/chosen": -
|
| 555 |
-
"logps/rejected": -
|
| 556 |
-
"loss": 0.
|
| 557 |
-
"rewards/accuracies": 0.
|
| 558 |
-
"rewards/chosen": -1.
|
| 559 |
-
"rewards/margins": 0.
|
| 560 |
-
"rewards/rejected": -
|
| 561 |
"step": 350
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 0.38,
|
| 565 |
"learning_rate": 3.92247625331392e-07,
|
| 566 |
-
"logits/chosen":
|
| 567 |
-
"logits/rejected": 1.
|
| 568 |
-
"logps/chosen": -
|
| 569 |
-
"logps/rejected": -
|
| 570 |
-
"loss": 0.
|
| 571 |
-
"rewards/accuracies": 0.
|
| 572 |
-
"rewards/chosen": -1.
|
| 573 |
-
"rewards/margins": 0.
|
| 574 |
-
"rewards/rejected": -2.
|
| 575 |
"step": 360
|
| 576 |
},
|
| 577 |
{
|
| 578 |
"epoch": 0.39,
|
| 579 |
"learning_rate": 3.846353490562664e-07,
|
| 580 |
-
"logits/chosen":
|
| 581 |
-
"logits/rejected":
|
| 582 |
-
"logps/chosen": -
|
| 583 |
-
"logps/rejected": -
|
| 584 |
-
"loss": 0.
|
| 585 |
"rewards/accuracies": 0.7749999761581421,
|
| 586 |
-
"rewards/chosen": -1.
|
| 587 |
-
"rewards/margins":
|
| 588 |
-
"rewards/rejected": -2.
|
| 589 |
"step": 370
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 0.4,
|
| 593 |
"learning_rate": 3.768430099352445e-07,
|
| 594 |
-
"logits/chosen": 0.
|
| 595 |
-
"logits/rejected": 1.
|
| 596 |
-
"logps/chosen": -
|
| 597 |
-
"logps/rejected": -
|
| 598 |
-
"loss": 0.
|
| 599 |
-
"rewards/accuracies": 0.
|
| 600 |
-
"rewards/chosen": -1.
|
| 601 |
-
"rewards/margins":
|
| 602 |
-
"rewards/rejected": -2.
|
| 603 |
"step": 380
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 0.41,
|
| 607 |
"learning_rate": 3.6888102953122304e-07,
|
| 608 |
-
"logits/chosen":
|
| 609 |
-
"logits/rejected":
|
| 610 |
-
"logps/chosen": -
|
| 611 |
-
"logps/rejected": -
|
| 612 |
-
"loss": 0.
|
| 613 |
-
"rewards/accuracies": 0.
|
| 614 |
-
"rewards/chosen": -1.
|
| 615 |
-
"rewards/margins":
|
| 616 |
-
"rewards/rejected": -2.
|
| 617 |
"step": 390
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 0.42,
|
| 621 |
"learning_rate": 3.607600562872785e-07,
|
| 622 |
-
"logits/chosen":
|
| 623 |
-
"logits/rejected":
|
| 624 |
-
"logps/chosen": -
|
| 625 |
-
"logps/rejected": -
|
| 626 |
-
"loss": 0.
|
| 627 |
-
"rewards/accuracies": 0.
|
| 628 |
-
"rewards/chosen": -1.
|
| 629 |
-
"rewards/margins": 0.
|
| 630 |
-
"rewards/rejected": -2.
|
| 631 |
"step": 400
|
| 632 |
},
|
| 633 |
{
|
| 634 |
"epoch": 0.42,
|
| 635 |
-
"eval_logits/chosen":
|
| 636 |
-
"eval_logits/rejected":
|
| 637 |
-
"eval_logps/chosen": -
|
| 638 |
-
"eval_logps/rejected": -
|
| 639 |
-
"eval_loss": 0.
|
| 640 |
-
"eval_rewards/accuracies": 0.
|
| 641 |
-
"eval_rewards/chosen": -1.
|
| 642 |
-
"eval_rewards/margins":
|
| 643 |
-
"eval_rewards/rejected": -2.
|
| 644 |
-
"eval_runtime":
|
| 645 |
-
"eval_samples_per_second": 8.
|
| 646 |
-
"eval_steps_per_second": 0.
|
| 647 |
"step": 400
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.43,
|
| 651 |
"learning_rate": 3.5249095128531856e-07,
|
| 652 |
-
"logits/chosen": 0.
|
| 653 |
-
"logits/rejected": 1.
|
| 654 |
-
"logps/chosen": -
|
| 655 |
-
"logps/rejected": -
|
| 656 |
-
"loss": 0.
|
| 657 |
-
"rewards/accuracies": 0.
|
| 658 |
-
"rewards/chosen": -1.
|
| 659 |
-
"rewards/margins":
|
| 660 |
-
"rewards/rejected": -
|
| 661 |
"step": 410
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.44,
|
| 665 |
"learning_rate": 3.4408477372034736e-07,
|
| 666 |
-
"logits/chosen":
|
| 667 |
-
"logits/rejected":
|
| 668 |
-
"logps/chosen": -
|
| 669 |
-
"logps/rejected": -
|
| 670 |
-
"loss": 0.
|
| 671 |
-
"rewards/accuracies": 0.
|
| 672 |
-
"rewards/chosen": -1.
|
| 673 |
-
"rewards/margins": 0.
|
| 674 |
-
"rewards/rejected": -
|
| 675 |
"step": 420
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.45,
|
| 679 |
"learning_rate": 3.3555276610977276e-07,
|
| 680 |
-
"logits/chosen":
|
| 681 |
-
"logits/rejected":
|
| 682 |
-
"logps/chosen": -
|
| 683 |
-
"logps/rejected": -
|
| 684 |
-
"loss": 0.
|
| 685 |
-
"rewards/accuracies": 0.
|
| 686 |
-
"rewards/chosen": -1.
|
| 687 |
-
"rewards/margins": 0.
|
| 688 |
-
"rewards/rejected": -1.
|
| 689 |
"step": 430
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.46,
|
| 693 |
"learning_rate": 3.269063392575352e-07,
|
| 694 |
-
"logits/chosen":
|
| 695 |
-
"logits/rejected":
|
| 696 |
-
"logps/chosen": -
|
| 697 |
-
"logps/rejected": -
|
| 698 |
-
"loss": 0.
|
| 699 |
-
"rewards/accuracies": 0.
|
| 700 |
-
"rewards/chosen": -1.
|
| 701 |
-
"rewards/margins": 0.
|
| 702 |
-
"rewards/rejected": -
|
| 703 |
"step": 440
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.47,
|
| 707 |
"learning_rate": 3.1815705699316964e-07,
|
| 708 |
-
"logits/chosen":
|
| 709 |
-
"logits/rejected":
|
| 710 |
-
"logps/chosen": -
|
| 711 |
-
"logps/rejected": -
|
| 712 |
-
"loss": 0.
|
| 713 |
"rewards/accuracies": 0.762499988079071,
|
| 714 |
-
"rewards/chosen": -1.
|
| 715 |
-
"rewards/margins":
|
| 716 |
-
"rewards/rejected": -2.
|
| 717 |
"step": 450
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.48,
|
| 721 |
"learning_rate": 3.0931662070620794e-07,
|
| 722 |
-
"logits/chosen":
|
| 723 |
-
"logits/rejected": 2.
|
| 724 |
-
"logps/chosen": -
|
| 725 |
-
"logps/rejected": -
|
| 726 |
-
"loss": 0.
|
| 727 |
-
"rewards/accuracies": 0.
|
| 728 |
-
"rewards/chosen": -1.
|
| 729 |
-
"rewards/margins":
|
| 730 |
-
"rewards/rejected": -
|
| 731 |
"step": 460
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.49,
|
| 735 |
"learning_rate": 3.003968536966078e-07,
|
| 736 |
-
"logits/chosen": 1.
|
| 737 |
-
"logits/rejected": 2.
|
| 738 |
-
"logps/chosen": -
|
| 739 |
-
"logps/rejected": -
|
| 740 |
-
"loss": 0.
|
| 741 |
-
"rewards/accuracies": 0.
|
| 742 |
-
"rewards/chosen": -1.
|
| 743 |
-
"rewards/margins":
|
| 744 |
-
"rewards/rejected": -2.
|
| 745 |
"step": 470
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.5,
|
| 749 |
"learning_rate": 2.9140968536213693e-07,
|
| 750 |
-
"logits/chosen":
|
| 751 |
-
"logits/rejected":
|
| 752 |
-
"logps/chosen": -372.
|
| 753 |
-
"logps/rejected": -
|
| 754 |
-
"loss": 0.
|
| 755 |
-
"rewards/accuracies": 0.
|
| 756 |
-
"rewards/chosen": -1.
|
| 757 |
-
"rewards/margins": 0.
|
| 758 |
-
"rewards/rejected": -2.
|
| 759 |
"step": 480
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.51,
|
| 763 |
"learning_rate": 2.823671352438608e-07,
|
| 764 |
-
"logits/chosen":
|
| 765 |
-
"logits/rejected": 2.
|
| 766 |
-
"logps/chosen": -
|
| 767 |
-
"logps/rejected": -
|
| 768 |
-
"loss": 0.
|
| 769 |
-
"rewards/accuracies": 0.
|
| 770 |
-
"rewards/chosen": -1.
|
| 771 |
-
"rewards/margins": 0.
|
| 772 |
-
"rewards/rejected": -2.
|
| 773 |
"step": 490
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.52,
|
| 777 |
"learning_rate": 2.73281296951072e-07,
|
| 778 |
-
"logits/chosen":
|
| 779 |
-
"logits/rejected": 2.
|
| 780 |
-
"logps/chosen": -
|
| 781 |
-
"logps/rejected": -
|
| 782 |
-
"loss": 0.
|
| 783 |
-
"rewards/accuracies": 0.
|
| 784 |
-
"rewards/chosen": -1.
|
| 785 |
-
"rewards/margins":
|
| 786 |
-
"rewards/rejected": -2.
|
| 787 |
"step": 500
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.52,
|
| 791 |
-
"eval_logits/chosen":
|
| 792 |
-
"eval_logits/rejected": 2.
|
| 793 |
-
"eval_logps/chosen": -
|
| 794 |
-
"eval_logps/rejected": -
|
| 795 |
-
"eval_loss": 0.
|
| 796 |
-
"eval_rewards/accuracies": 0.
|
| 797 |
-
"eval_rewards/chosen": -1.
|
| 798 |
-
"eval_rewards/margins":
|
| 799 |
-
"eval_rewards/rejected": -2.
|
| 800 |
-
"eval_runtime": 243.
|
| 801 |
-
"eval_samples_per_second": 8.
|
| 802 |
"eval_steps_per_second": 0.259,
|
| 803 |
"step": 500
|
| 804 |
},
|
| 805 |
{
|
| 806 |
"epoch": 0.53,
|
| 807 |
"learning_rate": 2.641643219871597e-07,
|
| 808 |
-
"logits/chosen":
|
| 809 |
-
"logits/rejected":
|
| 810 |
-
"logps/chosen": -
|
| 811 |
-
"logps/rejected": -
|
| 812 |
-
"loss": 0.
|
| 813 |
-
"rewards/accuracies": 0.
|
| 814 |
-
"rewards/chosen": -1.
|
| 815 |
-
"rewards/margins":
|
| 816 |
-
"rewards/rejected": -2.
|
| 817 |
"step": 510
|
| 818 |
},
|
| 819 |
{
|
| 820 |
"epoch": 0.54,
|
| 821 |
"learning_rate": 2.550284034980507e-07,
|
| 822 |
-
"logits/chosen":
|
| 823 |
-
"logits/rejected":
|
| 824 |
-
"logps/chosen": -
|
| 825 |
-
"logps/rejected": -
|
| 826 |
-
"loss": 0.
|
| 827 |
-
"rewards/accuracies": 0.
|
| 828 |
-
"rewards/chosen": -1.
|
| 829 |
-
"rewards/margins":
|
| 830 |
-
"rewards/rejected": -2.
|
| 831 |
"step": 520
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 0.55,
|
| 835 |
"learning_rate": 2.4588575996495794e-07,
|
| 836 |
-
"logits/chosen":
|
| 837 |
-
"logits/rejected":
|
| 838 |
-
"logps/chosen": -
|
| 839 |
-
"logps/rejected": -
|
| 840 |
-
"loss": 0.
|
| 841 |
-
"rewards/accuracies": 0.
|
| 842 |
-
"rewards/chosen": -1.
|
| 843 |
-
"rewards/margins":
|
| 844 |
-
"rewards/rejected": -2.
|
| 845 |
"step": 530
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 0.57,
|
| 849 |
"learning_rate": 2.367486188632446e-07,
|
| 850 |
-
"logits/chosen":
|
| 851 |
-
"logits/rejected": 2.
|
| 852 |
-
"logps/chosen": -
|
| 853 |
-
"logps/rejected": -
|
| 854 |
-
"loss": 0.
|
| 855 |
-
"rewards/accuracies": 0.
|
| 856 |
-
"rewards/chosen": -1.
|
| 857 |
-
"rewards/margins": 1.
|
| 858 |
-
"rewards/rejected": -2.
|
| 859 |
"step": 540
|
| 860 |
},
|
| 861 |
{
|
| 862 |
"epoch": 0.58,
|
| 863 |
"learning_rate": 2.276292003092593e-07,
|
| 864 |
-
"logits/chosen":
|
| 865 |
-
"logits/rejected": 2.
|
| 866 |
-
"logps/chosen": -
|
| 867 |
-
"logps/rejected": -
|
| 868 |
-
"loss": 0.
|
| 869 |
"rewards/accuracies": 0.75,
|
| 870 |
-
"rewards/chosen": -1.
|
| 871 |
-
"rewards/margins":
|
| 872 |
-
"rewards/rejected": -2.
|
| 873 |
"step": 550
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.59,
|
| 877 |
"learning_rate": 2.185397007170141e-07,
|
| 878 |
-
"logits/chosen":
|
| 879 |
-
"logits/rejected":
|
| 880 |
-
"logps/chosen": -
|
| 881 |
-
"logps/rejected": -
|
| 882 |
-
"loss": 0.
|
| 883 |
"rewards/accuracies": 0.731249988079071,
|
| 884 |
-
"rewards/chosen": -1.
|
| 885 |
-
"rewards/margins": 0.
|
| 886 |
-
"rewards/rejected": -
|
| 887 |
"step": 560
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 0.6,
|
| 891 |
"learning_rate": 2.094922764865619e-07,
|
| 892 |
-
"logits/chosen":
|
| 893 |
-
"logits/rejected":
|
| 894 |
-
"logps/chosen": -
|
| 895 |
-
"logps/rejected": -
|
| 896 |
-
"loss": 0.
|
| 897 |
-
"rewards/accuracies": 0.
|
| 898 |
-
"rewards/chosen": -1.
|
| 899 |
-
"rewards/margins": 0.
|
| 900 |
-
"rewards/rejected": -2.
|
| 901 |
"step": 570
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 0.61,
|
| 905 |
"learning_rate": 2.0049902774588797e-07,
|
| 906 |
-
"logits/chosen":
|
| 907 |
-
"logits/rejected":
|
| 908 |
-
"logps/chosen": -
|
| 909 |
-
"logps/rejected": -
|
| 910 |
-
"loss": 0.
|
| 911 |
-
"rewards/accuracies": 0.
|
| 912 |
-
"rewards/chosen": -1.
|
| 913 |
-
"rewards/margins":
|
| 914 |
-
"rewards/rejected": -2.
|
| 915 |
"step": 580
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 0.62,
|
| 919 |
"learning_rate": 1.9157198216806238e-07,
|
| 920 |
-
"logits/chosen":
|
| 921 |
-
"logits/rejected":
|
| 922 |
-
"logps/chosen": -
|
| 923 |
-
"logps/rejected": -
|
| 924 |
-
"loss": 0.
|
| 925 |
-
"rewards/accuracies": 0.
|
| 926 |
-
"rewards/chosen": -1.
|
| 927 |
-
"rewards/margins": 0.
|
| 928 |
-
"rewards/rejected": -
|
| 929 |
"step": 590
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 0.63,
|
| 933 |
"learning_rate": 1.8272307888529274e-07,
|
| 934 |
-
"logits/chosen":
|
| 935 |
-
"logits/rejected":
|
| 936 |
-
"logps/chosen": -
|
| 937 |
-
"logps/rejected": -
|
| 938 |
-
"loss": 0.
|
| 939 |
-
"rewards/accuracies": 0.
|
| 940 |
-
"rewards/chosen": -1.
|
| 941 |
-
"rewards/margins":
|
| 942 |
-
"rewards/rejected": -2.
|
| 943 |
"step": 600
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 0.63,
|
| 947 |
-
"eval_logits/chosen":
|
| 948 |
-
"eval_logits/rejected":
|
| 949 |
-
"eval_logps/chosen": -
|
| 950 |
-
"eval_logps/rejected": -
|
| 951 |
-
"eval_loss": 0.
|
| 952 |
-
"eval_rewards/accuracies": 0.
|
| 953 |
-
"eval_rewards/chosen": -1.
|
| 954 |
-
"eval_rewards/margins":
|
| 955 |
-
"eval_rewards/rejected": -2.
|
| 956 |
-
"eval_runtime":
|
| 957 |
-
"eval_samples_per_second": 8.
|
| 958 |
"eval_steps_per_second": 0.258,
|
| 959 |
"step": 600
|
| 960 |
},
|
| 961 |
{
|
| 962 |
"epoch": 0.64,
|
| 963 |
"learning_rate": 1.7396415252139288e-07,
|
| 964 |
-
"logits/chosen":
|
| 965 |
-
"logits/rejected":
|
| 966 |
-
"logps/chosen": -
|
| 967 |
-
"logps/rejected": -
|
| 968 |
-
"loss": 0.
|
| 969 |
-
"rewards/accuracies": 0.
|
| 970 |
-
"rewards/chosen": -1.
|
| 971 |
-
"rewards/margins":
|
| 972 |
-
"rewards/rejected": -2.
|
| 973 |
"step": 610
|
| 974 |
},
|
| 975 |
{
|
| 976 |
"epoch": 0.65,
|
| 977 |
"learning_rate": 1.6530691736402316e-07,
|
| 978 |
-
"logits/chosen": 2.
|
| 979 |
-
"logits/rejected": 3.
|
| 980 |
-
"logps/chosen": -
|
| 981 |
-
"logps/rejected": -
|
| 982 |
-
"loss": 0.
|
| 983 |
-
"rewards/accuracies": 0.
|
| 984 |
-
"rewards/chosen": -1.
|
| 985 |
-
"rewards/margins":
|
| 986 |
-
"rewards/rejected": -2.
|
| 987 |
"step": 620
|
| 988 |
},
|
| 989 |
{
|
| 990 |
"epoch": 0.66,
|
| 991 |
"learning_rate": 1.5676295169786864e-07,
|
| 992 |
-
"logits/chosen":
|
| 993 |
-
"logits/rejected": 3.
|
| 994 |
-
"logps/chosen": -
|
| 995 |
-
"logps/rejected": -
|
| 996 |
-
"loss": 0.
|
| 997 |
-
"rewards/accuracies": 0.
|
| 998 |
-
"rewards/chosen": -1.
|
| 999 |
-
"rewards/margins": 1.
|
| 1000 |
-
"rewards/rejected": -
|
| 1001 |
"step": 630
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"epoch": 0.67,
|
| 1005 |
"learning_rate": 1.483436823197092e-07,
|
| 1006 |
-
"logits/chosen":
|
| 1007 |
-
"logits/rejected":
|
| 1008 |
-
"logps/chosen": -
|
| 1009 |
-
"logps/rejected": -
|
| 1010 |
-
"loss": 0.
|
| 1011 |
"rewards/accuracies": 0.800000011920929,
|
| 1012 |
-
"rewards/chosen": -1.
|
| 1013 |
-
"rewards/margins":
|
| 1014 |
-
"rewards/rejected": -2.
|
| 1015 |
"step": 640
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
"epoch": 0.68,
|
| 1019 |
"learning_rate": 1.4006036925609243e-07,
|
| 1020 |
-
"logits/chosen":
|
| 1021 |
-
"logits/rejected": 2.
|
| 1022 |
-
"logps/chosen": -
|
| 1023 |
-
"logps/rejected": -
|
| 1024 |
-
"loss": 0.
|
| 1025 |
-
"rewards/accuracies": 0.
|
| 1026 |
-
"rewards/chosen": -1.
|
| 1027 |
-
"rewards/margins": 0.
|
| 1028 |
-
"rewards/rejected": -2.
|
| 1029 |
"step": 650
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
"epoch": 0.69,
|
| 1033 |
"learning_rate": 1.319240907040458e-07,
|
| 1034 |
-
"logits/chosen":
|
| 1035 |
-
"logits/rejected": 2.
|
| 1036 |
-
"logps/chosen": -
|
| 1037 |
-
"logps/rejected": -
|
| 1038 |
-
"loss": 0.
|
| 1039 |
-
"rewards/accuracies": 0.
|
| 1040 |
-
"rewards/chosen": -1.
|
| 1041 |
-
"rewards/margins":
|
| 1042 |
-
"rewards/rejected": -2.
|
| 1043 |
"step": 660
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
"epoch": 0.7,
|
| 1047 |
"learning_rate": 1.239457282149695e-07,
|
| 1048 |
-
"logits/chosen":
|
| 1049 |
-
"logits/rejected": 3.
|
| 1050 |
-
"logps/chosen": -
|
| 1051 |
-
"logps/rejected": -
|
| 1052 |
-
"loss": 0.
|
| 1053 |
-
"rewards/accuracies": 0.
|
| 1054 |
-
"rewards/chosen": -1.
|
| 1055 |
-
"rewards/margins":
|
| 1056 |
-
"rewards/rejected": -2.
|
| 1057 |
"step": 670
|
| 1058 |
},
|
| 1059 |
{
|
| 1060 |
"epoch": 0.71,
|
| 1061 |
"learning_rate": 1.1613595214152711e-07,
|
| 1062 |
-
"logits/chosen": 2.
|
| 1063 |
-
"logits/rejected": 2.
|
| 1064 |
-
"logps/chosen": -
|
| 1065 |
-
"logps/rejected": -
|
| 1066 |
-
"loss": 0.
|
| 1067 |
-
"rewards/accuracies": 0.
|
| 1068 |
-
"rewards/chosen": -1.
|
| 1069 |
-
"rewards/margins": 0.
|
| 1070 |
-
"rewards/rejected": -
|
| 1071 |
"step": 680
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
"epoch": 0.72,
|
| 1075 |
"learning_rate": 1.0850520736699362e-07,
|
| 1076 |
-
"logits/chosen":
|
| 1077 |
-
"logits/rejected":
|
| 1078 |
-
"logps/chosen": -
|
| 1079 |
-
"logps/rejected": -
|
| 1080 |
-
"loss": 0.
|
| 1081 |
-
"rewards/accuracies": 0.
|
| 1082 |
-
"rewards/chosen": -1.
|
| 1083 |
-
"rewards/margins":
|
| 1084 |
-
"rewards/rejected": -2.
|
| 1085 |
"step": 690
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
"epoch": 0.73,
|
| 1089 |
"learning_rate": 1.0106369933615042e-07,
|
| 1090 |
-
"logits/chosen": 2.
|
| 1091 |
-
"logits/rejected":
|
| 1092 |
-
"logps/chosen": -
|
| 1093 |
-
"logps/rejected": -
|
| 1094 |
-
"loss": 0.
|
| 1095 |
-
"rewards/accuracies": 0.
|
| 1096 |
-
"rewards/chosen": -1.
|
| 1097 |
-
"rewards/margins": 1.
|
| 1098 |
-
"rewards/rejected": -
|
| 1099 |
"step": 700
|
| 1100 |
},
|
| 1101 |
{
|
| 1102 |
"epoch": 0.73,
|
| 1103 |
-
"eval_logits/chosen": 2.
|
| 1104 |
-
"eval_logits/rejected":
|
| 1105 |
-
"eval_logps/chosen": -
|
| 1106 |
-
"eval_logps/rejected": -
|
| 1107 |
-
"eval_loss": 0.
|
| 1108 |
-
"eval_rewards/accuracies": 0.
|
| 1109 |
-
"eval_rewards/chosen": -1.
|
| 1110 |
-
"eval_rewards/margins":
|
| 1111 |
-
"eval_rewards/rejected": -
|
| 1112 |
-
"eval_runtime":
|
| 1113 |
-
"eval_samples_per_second": 8.
|
| 1114 |
-
"eval_steps_per_second": 0.
|
| 1115 |
"step": 700
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.74,
|
| 1119 |
"learning_rate": 9.382138040640714e-08,
|
| 1120 |
-
"logits/chosen": 2.
|
| 1121 |
-
"logits/rejected":
|
| 1122 |
-
"logps/chosen": -
|
| 1123 |
-
"logps/rejected": -
|
| 1124 |
-
"loss": 0.
|
| 1125 |
-
"rewards/accuracies": 0.
|
| 1126 |
-
"rewards/chosen": -
|
| 1127 |
-
"rewards/margins":
|
| 1128 |
-
"rewards/rejected": -
|
| 1129 |
"step": 710
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.75,
|
| 1133 |
"learning_rate": 8.678793653740632e-08,
|
| 1134 |
-
"logits/chosen":
|
| 1135 |
-
"logits/rejected":
|
| 1136 |
-
"logps/chosen": -
|
| 1137 |
-
"logps/rejected": -
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"rewards/accuracies": 0.
|
| 1140 |
-
"rewards/chosen": -
|
| 1141 |
-
"rewards/margins":
|
| 1142 |
-
"rewards/rejected": -
|
| 1143 |
"step": 720
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.76,
|
| 1147 |
"learning_rate": 7.997277433690983e-08,
|
| 1148 |
-
"logits/chosen":
|
| 1149 |
-
"logits/rejected":
|
| 1150 |
-
"logps/chosen": -
|
| 1151 |
-
"logps/rejected": -
|
| 1152 |
-
"loss": 0.
|
| 1153 |
-
"rewards/accuracies": 0.
|
| 1154 |
-
"rewards/chosen": -
|
| 1155 |
-
"rewards/margins":
|
| 1156 |
-
"rewards/rejected": -
|
| 1157 |
"step": 730
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.77,
|
| 1161 |
"learning_rate": 7.338500848029602e-08,
|
| 1162 |
-
"logits/chosen":
|
| 1163 |
-
"logits/rejected":
|
| 1164 |
-
"logps/chosen": -
|
| 1165 |
-
"logps/rejected": -
|
| 1166 |
-
"loss": 0.
|
| 1167 |
-
"rewards/accuracies": 0.
|
| 1168 |
-
"rewards/chosen": -
|
| 1169 |
-
"rewards/margins": 1.
|
| 1170 |
-
"rewards/rejected": -
|
| 1171 |
"step": 740
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.78,
|
| 1175 |
"learning_rate": 6.70334495204884e-08,
|
| 1176 |
-
"logits/chosen": 2.
|
| 1177 |
-
"logits/rejected":
|
| 1178 |
-
"logps/chosen": -
|
| 1179 |
-
"logps/rejected": -
|
| 1180 |
-
"loss": 0.
|
| 1181 |
-
"rewards/accuracies": 0.
|
| 1182 |
-
"rewards/chosen": -
|
| 1183 |
-
"rewards/margins":
|
| 1184 |
-
"rewards/rejected": -
|
| 1185 |
"step": 750
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.8,
|
| 1189 |
"learning_rate": 6.092659210462231e-08,
|
| 1190 |
-
"logits/chosen":
|
| 1191 |
-
"logits/rejected":
|
| 1192 |
-
"logps/chosen": -
|
| 1193 |
-
"logps/rejected": -
|
| 1194 |
-
"loss": 0.
|
| 1195 |
-
"rewards/accuracies": 0.
|
| 1196 |
-
"rewards/chosen": -
|
| 1197 |
-
"rewards/margins":
|
| 1198 |
-
"rewards/rejected": -
|
| 1199 |
"step": 760
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.81,
|
| 1203 |
"learning_rate": 5.507260361320737e-08,
|
| 1204 |
-
"logits/chosen":
|
| 1205 |
-
"logits/rejected":
|
| 1206 |
-
"logps/chosen": -
|
| 1207 |
-
"logps/rejected": -
|
| 1208 |
-
"loss": 0.
|
| 1209 |
-
"rewards/accuracies": 0.
|
| 1210 |
-
"rewards/chosen": -
|
| 1211 |
-
"rewards/margins":
|
| 1212 |
-
"rewards/rejected": -
|
| 1213 |
"step": 770
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.82,
|
| 1217 |
"learning_rate": 4.947931323697982e-08,
|
| 1218 |
-
"logits/chosen": 3.
|
| 1219 |
-
"logits/rejected":
|
| 1220 |
-
"logps/chosen": -
|
| 1221 |
-
"logps/rejected": -
|
| 1222 |
-
"loss": 0.
|
| 1223 |
"rewards/accuracies": 0.706250011920929,
|
| 1224 |
-
"rewards/chosen": -
|
| 1225 |
-
"rewards/margins":
|
| 1226 |
-
"rewards/rejected": -
|
| 1227 |
"step": 780
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.83,
|
| 1231 |
"learning_rate": 4.415420150605398e-08,
|
| 1232 |
-
"logits/chosen": 2.
|
| 1233 |
-
"logits/rejected": 3.
|
| 1234 |
-
"logps/chosen": -
|
| 1235 |
-
"logps/rejected": -
|
| 1236 |
-
"loss": 0.
|
| 1237 |
-
"rewards/accuracies": 0.
|
| 1238 |
-
"rewards/chosen": -
|
| 1239 |
-
"rewards/margins": 1.
|
| 1240 |
-
"rewards/rejected": -
|
| 1241 |
"step": 790
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.84,
|
| 1245 |
"learning_rate": 3.9104390285376374e-08,
|
| 1246 |
-
"logits/chosen": 2.
|
| 1247 |
-
"logits/rejected":
|
| 1248 |
-
"logps/chosen": -
|
| 1249 |
-
"logps/rejected": -
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"rewards/accuracies": 0.
|
| 1252 |
-
"rewards/chosen": -
|
| 1253 |
-
"rewards/margins": 1.
|
| 1254 |
-
"rewards/rejected": -
|
| 1255 |
"step": 800
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.84,
|
| 1259 |
-
"eval_logits/chosen": 2.
|
| 1260 |
-
"eval_logits/rejected": 3.
|
| 1261 |
-
"eval_logps/chosen": -
|
| 1262 |
-
"eval_logps/rejected": -
|
| 1263 |
-
"eval_loss": 0.
|
| 1264 |
"eval_rewards/accuracies": 0.7678571343421936,
|
| 1265 |
-
"eval_rewards/chosen": -
|
| 1266 |
-
"eval_rewards/margins": 1.
|
| 1267 |
-
"eval_rewards/rejected": -
|
| 1268 |
-
"eval_runtime": 243.
|
| 1269 |
-
"eval_samples_per_second": 8.
|
| 1270 |
"eval_steps_per_second": 0.259,
|
| 1271 |
"step": 800
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
"epoch": 0.85,
|
| 1275 |
"learning_rate": 3.433663324986208e-08,
|
| 1276 |
-
"logits/chosen":
|
| 1277 |
-
"logits/rejected":
|
| 1278 |
-
"logps/chosen": -
|
| 1279 |
-
"logps/rejected": -
|
| 1280 |
-
"loss": 0.
|
| 1281 |
-
"rewards/accuracies": 0.
|
| 1282 |
-
"rewards/chosen": -
|
| 1283 |
-
"rewards/margins": 0.
|
| 1284 |
-
"rewards/rejected": -
|
| 1285 |
"step": 810
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 0.86,
|
| 1289 |
"learning_rate": 2.9857306851953897e-08,
|
| 1290 |
-
"logits/chosen": 3.
|
| 1291 |
-
"logits/rejected": 3.
|
| 1292 |
-
"logps/chosen": -
|
| 1293 |
-
"logps/rejected": -
|
| 1294 |
-
"loss": 0.
|
| 1295 |
"rewards/accuracies": 0.762499988079071,
|
| 1296 |
-
"rewards/chosen": -
|
| 1297 |
-
"rewards/margins":
|
| 1298 |
-
"rewards/rejected": -
|
| 1299 |
"step": 820
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
"epoch": 0.87,
|
| 1303 |
"learning_rate": 2.567240179368185e-08,
|
| 1304 |
-
"logits/chosen": 2.
|
| 1305 |
-
"logits/rejected": 3.
|
| 1306 |
-
"logps/chosen": -
|
| 1307 |
-
"logps/rejected": -
|
| 1308 |
-
"loss": 0.
|
| 1309 |
-
"rewards/accuracies": 0.
|
| 1310 |
-
"rewards/chosen": -
|
| 1311 |
-
"rewards/margins":
|
| 1312 |
-
"rewards/rejected": -
|
| 1313 |
"step": 830
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
"epoch": 0.88,
|
| 1317 |
"learning_rate": 2.1787515014630357e-08,
|
| 1318 |
-
"logits/chosen":
|
| 1319 |
-
"logits/rejected":
|
| 1320 |
-
"logps/chosen": -
|
| 1321 |
-
"logps/rejected": -
|
| 1322 |
-
"loss": 0.
|
| 1323 |
-
"rewards/accuracies": 0.
|
| 1324 |
-
"rewards/chosen": -
|
| 1325 |
-
"rewards/margins":
|
| 1326 |
-
"rewards/rejected": -
|
| 1327 |
"step": 840
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
"epoch": 0.89,
|
| 1331 |
"learning_rate": 1.820784220652766e-08,
|
| 1332 |
-
"logits/chosen": 2.
|
| 1333 |
-
"logits/rejected": 3.
|
| 1334 |
-
"logps/chosen": -
|
| 1335 |
-
"logps/rejected": -
|
| 1336 |
-
"loss": 0.
|
| 1337 |
"rewards/accuracies": 0.7437499761581421,
|
| 1338 |
-
"rewards/chosen": -1.
|
| 1339 |
-
"rewards/margins":
|
| 1340 |
-
"rewards/rejected": -
|
| 1341 |
"step": 850
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
"epoch": 0.9,
|
| 1345 |
"learning_rate": 1.4938170864468636e-08,
|
| 1346 |
-
"logits/chosen": 2.
|
| 1347 |
-
"logits/rejected":
|
| 1348 |
-
"logps/chosen": -
|
| 1349 |
-
"logps/rejected": -
|
| 1350 |
-
"loss": 0.
|
| 1351 |
-
"rewards/accuracies": 0.
|
| 1352 |
-
"rewards/chosen": -
|
| 1353 |
-
"rewards/margins":
|
| 1354 |
-
"rewards/rejected": -
|
| 1355 |
"step": 860
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 0.91,
|
| 1359 |
"learning_rate": 1.1982873884064465e-08,
|
| 1360 |
-
"logits/chosen": 2.
|
| 1361 |
-
"logits/rejected":
|
| 1362 |
-
"logps/chosen": -
|
| 1363 |
-
"logps/rejected": -
|
| 1364 |
-
"loss": 0.
|
| 1365 |
-
"rewards/accuracies": 0.
|
| 1366 |
-
"rewards/chosen": -1.
|
| 1367 |
-
"rewards/margins": 1.
|
| 1368 |
-
"rewards/rejected": -
|
| 1369 |
"step": 870
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.92,
|
| 1373 |
"learning_rate": 9.345903713082304e-09,
|
| 1374 |
-
"logits/chosen": 2.
|
| 1375 |
-
"logits/rejected":
|
| 1376 |
-
"logps/chosen": -
|
| 1377 |
-
"logps/rejected": -
|
| 1378 |
-
"loss": 0.
|
| 1379 |
-
"rewards/accuracies": 0.
|
| 1380 |
-
"rewards/chosen": -
|
| 1381 |
-
"rewards/margins":
|
| 1382 |
-
"rewards/rejected": -
|
| 1383 |
"step": 880
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 0.93,
|
| 1387 |
"learning_rate": 7.030787065396865e-09,
|
| 1388 |
-
"logits/chosen": 2.
|
| 1389 |
-
"logits/rejected": 3.
|
| 1390 |
-
"logps/chosen": -
|
| 1391 |
-
"logps/rejected": -
|
| 1392 |
-
"loss": 0.
|
| 1393 |
-
"rewards/accuracies": 0.
|
| 1394 |
-
"rewards/chosen": -
|
| 1395 |
-
"rewards/margins":
|
| 1396 |
-
"rewards/rejected": -
|
| 1397 |
"step": 890
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 0.94,
|
| 1401 |
"learning_rate": 5.04062020432286e-09,
|
| 1402 |
-
"logits/chosen": 2.
|
| 1403 |
-
"logits/rejected": 3.
|
| 1404 |
-
"logps/chosen": -
|
| 1405 |
-
"logps/rejected": -
|
| 1406 |
-
"loss": 0.
|
| 1407 |
"rewards/accuracies": 0.71875,
|
| 1408 |
-
"rewards/chosen": -
|
| 1409 |
-
"rewards/margins":
|
| 1410 |
-
"rewards/rejected": -
|
| 1411 |
"step": 900
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 0.94,
|
| 1415 |
-
"eval_logits/chosen": 2.
|
| 1416 |
-
"eval_logits/rejected":
|
| 1417 |
-
"eval_logps/chosen": -
|
| 1418 |
-
"eval_logps/rejected": -
|
| 1419 |
-
"eval_loss": 0.
|
| 1420 |
-
"eval_rewards/accuracies": 0.
|
| 1421 |
-
"eval_rewards/chosen": -
|
| 1422 |
-
"eval_rewards/margins":
|
| 1423 |
-
"eval_rewards/rejected": -
|
| 1424 |
-
"eval_runtime":
|
| 1425 |
-
"eval_samples_per_second": 8.
|
| 1426 |
-
"eval_steps_per_second": 0.
|
| 1427 |
"step": 900
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 0.95,
|
| 1431 |
"learning_rate": 3.3780648016376866e-09,
|
| 1432 |
-
"logits/chosen": 2.
|
| 1433 |
-
"logits/rejected":
|
| 1434 |
-
"logps/chosen": -
|
| 1435 |
-
"logps/rejected": -
|
| 1436 |
-
"loss": 0.
|
| 1437 |
-
"rewards/accuracies": 0.
|
| 1438 |
-
"rewards/chosen": -
|
| 1439 |
-
"rewards/margins":
|
| 1440 |
-
"rewards/rejected": -
|
| 1441 |
"step": 910
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 0.96,
|
| 1445 |
"learning_rate": 2.0453443778310766e-09,
|
| 1446 |
-
"logits/chosen": 2.
|
| 1447 |
-
"logits/rejected":
|
| 1448 |
-
"logps/chosen": -
|
| 1449 |
-
"logps/rejected": -
|
| 1450 |
-
"loss": 0.
|
| 1451 |
"rewards/accuracies": 0.8187500238418579,
|
| 1452 |
-
"rewards/chosen": -
|
| 1453 |
-
"rewards/margins": 1.
|
| 1454 |
-
"rewards/rejected": -
|
| 1455 |
"step": 920
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
"epoch": 0.97,
|
| 1459 |
"learning_rate": 1.0442413283435758e-09,
|
| 1460 |
-
"logits/chosen": 2.
|
| 1461 |
-
"logits/rejected":
|
| 1462 |
-
"logps/chosen": -
|
| 1463 |
-
"logps/rejected": -
|
| 1464 |
-
"loss": 0.
|
| 1465 |
-
"rewards/accuracies": 0.
|
| 1466 |
-
"rewards/chosen": -
|
| 1467 |
-
"rewards/margins": 1.
|
| 1468 |
-
"rewards/rejected": -
|
| 1469 |
"step": 930
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
"epoch": 0.98,
|
| 1473 |
"learning_rate": 3.760945397705828e-10,
|
| 1474 |
-
"logits/chosen": 2.
|
| 1475 |
-
"logits/rejected":
|
| 1476 |
-
"logps/chosen": -
|
| 1477 |
-
"logps/rejected": -
|
| 1478 |
-
"loss": 0.
|
| 1479 |
-
"rewards/accuracies": 0.
|
| 1480 |
-
"rewards/chosen": -
|
| 1481 |
-
"rewards/margins":
|
| 1482 |
-
"rewards/rejected": -
|
| 1483 |
"step": 940
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
"epoch": 0.99,
|
| 1487 |
"learning_rate": 4.17975992204056e-11,
|
| 1488 |
-
"logits/chosen": 2.
|
| 1489 |
-
"logits/rejected": 3.
|
| 1490 |
-
"logps/chosen": -
|
| 1491 |
-
"logps/rejected": -
|
| 1492 |
-
"loss": 0.
|
| 1493 |
-
"rewards/accuracies": 0.
|
| 1494 |
-
"rewards/chosen": -
|
| 1495 |
-
"rewards/margins": 1.
|
| 1496 |
-
"rewards/rejected": -
|
| 1497 |
"step": 950
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 1.0,
|
| 1501 |
"step": 955,
|
| 1502 |
"total_flos": 0.0,
|
| 1503 |
-
"train_loss": 0.
|
| 1504 |
-
"train_runtime":
|
| 1505 |
-
"train_samples_per_second": 2.
|
| 1506 |
-
"train_steps_per_second": 0.
|
| 1507 |
}
|
| 1508 |
],
|
| 1509 |
"logging_steps": 10,
|
| 1510 |
"max_steps": 955,
|
| 1511 |
"num_input_tokens_seen": 0,
|
| 1512 |
"num_train_epochs": 1,
|
| 1513 |
-
"save_steps":
|
| 1514 |
"total_flos": 0.0,
|
| 1515 |
"train_batch_size": 8,
|
| 1516 |
"trial_name": null,
|
|
|
|
| 15 |
"logits/rejected": -2.686896800994873,
|
| 16 |
"logps/chosen": -229.94229125976562,
|
| 17 |
"logps/rejected": -214.70114135742188,
|
| 18 |
+
"loss": 0.9741,
|
| 19 |
"rewards/accuracies": 0.0,
|
| 20 |
"rewards/chosen": 0.0,
|
| 21 |
"rewards/margins": 0.0,
|
|
|
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
"learning_rate": 5.208333333333333e-08,
|
| 28 |
+
"logits/chosen": -2.6810548305511475,
|
| 29 |
+
"logits/rejected": -2.709120035171509,
|
| 30 |
+
"logps/chosen": -295.81451416015625,
|
| 31 |
+
"logps/rejected": -250.5977325439453,
|
| 32 |
+
"loss": 0.9742,
|
| 33 |
+
"rewards/accuracies": 0.4375,
|
| 34 |
+
"rewards/chosen": 0.0003499284212011844,
|
| 35 |
+
"rewards/margins": 0.0001925795222632587,
|
| 36 |
+
"rewards/rejected": 0.00015734886983409524,
|
| 37 |
"step": 10
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02,
|
| 41 |
"learning_rate": 1.0416666666666667e-07,
|
| 42 |
+
"logits/chosen": -2.6195099353790283,
|
| 43 |
+
"logits/rejected": -2.625662088394165,
|
| 44 |
+
"logps/chosen": -271.3158264160156,
|
| 45 |
+
"logps/rejected": -246.94711303710938,
|
| 46 |
+
"loss": 0.9738,
|
| 47 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 48 |
+
"rewards/chosen": 0.00036363088293001056,
|
| 49 |
+
"rewards/margins": 0.0009133815765380859,
|
| 50 |
+
"rewards/rejected": -0.0005497508682310581,
|
| 51 |
"step": 20
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03,
|
| 55 |
"learning_rate": 1.5624999999999999e-07,
|
| 56 |
+
"logits/chosen": -2.7032079696655273,
|
| 57 |
+
"logits/rejected": -2.666191577911377,
|
| 58 |
+
"logps/chosen": -278.3299865722656,
|
| 59 |
+
"logps/rejected": -254.5498809814453,
|
| 60 |
+
"loss": 0.9731,
|
| 61 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 62 |
+
"rewards/chosen": 0.0010723542654886842,
|
| 63 |
+
"rewards/margins": 0.0018442096188664436,
|
| 64 |
+
"rewards/rejected": -0.0007718555280007422,
|
| 65 |
"step": 30
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.04,
|
| 69 |
"learning_rate": 2.0833333333333333e-07,
|
| 70 |
+
"logits/chosen": -2.6499533653259277,
|
| 71 |
+
"logits/rejected": -2.6374642848968506,
|
| 72 |
+
"logps/chosen": -273.9149475097656,
|
| 73 |
+
"logps/rejected": -237.7373504638672,
|
| 74 |
+
"loss": 0.9713,
|
| 75 |
+
"rewards/accuracies": 0.675000011920929,
|
| 76 |
+
"rewards/chosen": 0.003091720398515463,
|
| 77 |
+
"rewards/margins": 0.005813647527247667,
|
| 78 |
+
"rewards/rejected": -0.002721927361562848,
|
| 79 |
"step": 40
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05,
|
| 83 |
"learning_rate": 2.604166666666667e-07,
|
| 84 |
+
"logits/chosen": -2.6743884086608887,
|
| 85 |
+
"logits/rejected": -2.6382362842559814,
|
| 86 |
+
"logps/chosen": -296.06744384765625,
|
| 87 |
+
"logps/rejected": -274.7203063964844,
|
| 88 |
+
"loss": 0.9657,
|
| 89 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 90 |
+
"rewards/chosen": 0.009085027500987053,
|
| 91 |
+
"rewards/margins": 0.016131814569234848,
|
| 92 |
+
"rewards/rejected": -0.007046787533909082,
|
| 93 |
"step": 50
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06,
|
| 97 |
"learning_rate": 3.1249999999999997e-07,
|
| 98 |
+
"logits/chosen": -2.630288600921631,
|
| 99 |
+
"logits/rejected": -2.6329100131988525,
|
| 100 |
+
"logps/chosen": -285.3112487792969,
|
| 101 |
+
"logps/rejected": -274.3540954589844,
|
| 102 |
+
"loss": 0.9577,
|
| 103 |
+
"rewards/accuracies": 0.668749988079071,
|
| 104 |
+
"rewards/chosen": 0.02476242184638977,
|
| 105 |
+
"rewards/margins": 0.0284078661352396,
|
| 106 |
+
"rewards/rejected": -0.0036454431246966124,
|
| 107 |
"step": 60
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07,
|
| 111 |
"learning_rate": 3.645833333333333e-07,
|
| 112 |
+
"logits/chosen": -2.6492714881896973,
|
| 113 |
+
"logits/rejected": -2.6846468448638916,
|
| 114 |
+
"logps/chosen": -310.99761962890625,
|
| 115 |
+
"logps/rejected": -290.5529479980469,
|
| 116 |
+
"loss": 0.9406,
|
| 117 |
+
"rewards/accuracies": 0.6875,
|
| 118 |
+
"rewards/chosen": 0.03506559133529663,
|
| 119 |
+
"rewards/margins": 0.05608881637454033,
|
| 120 |
+
"rewards/rejected": -0.021023228764533997,
|
| 121 |
"step": 70
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08,
|
| 125 |
"learning_rate": 4.1666666666666667e-07,
|
| 126 |
+
"logits/chosen": -2.523487091064453,
|
| 127 |
+
"logits/rejected": -2.464901924133301,
|
| 128 |
+
"logps/chosen": -304.66845703125,
|
| 129 |
+
"logps/rejected": -281.81732177734375,
|
| 130 |
+
"loss": 0.9095,
|
| 131 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 132 |
+
"rewards/chosen": -0.04107608273625374,
|
| 133 |
+
"rewards/margins": 0.11836276948451996,
|
| 134 |
+
"rewards/rejected": -0.1594388782978058,
|
| 135 |
"step": 80
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09,
|
| 139 |
"learning_rate": 4.6874999999999996e-07,
|
| 140 |
+
"logits/chosen": -2.5307559967041016,
|
| 141 |
+
"logits/rejected": -2.5012693405151367,
|
| 142 |
+
"logps/chosen": -292.33392333984375,
|
| 143 |
+
"logps/rejected": -284.6061096191406,
|
| 144 |
+
"loss": 0.8948,
|
| 145 |
+
"rewards/accuracies": 0.706250011920929,
|
| 146 |
+
"rewards/chosen": -0.00477323355153203,
|
| 147 |
+
"rewards/margins": 0.17751149833202362,
|
| 148 |
+
"rewards/rejected": -0.18228471279144287,
|
| 149 |
"step": 90
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.1,
|
| 153 |
"learning_rate": 4.999732492681437e-07,
|
| 154 |
+
"logits/chosen": -2.501075267791748,
|
| 155 |
+
"logits/rejected": -2.491670846939087,
|
| 156 |
+
"logps/chosen": -340.92401123046875,
|
| 157 |
+
"logps/rejected": -345.025390625,
|
| 158 |
+
"loss": 0.8957,
|
| 159 |
"rewards/accuracies": 0.7124999761581421,
|
| 160 |
+
"rewards/chosen": -0.35498708486557007,
|
| 161 |
+
"rewards/margins": 0.17348773777484894,
|
| 162 |
+
"rewards/rejected": -0.5284748077392578,
|
| 163 |
"step": 100
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.1,
|
| 167 |
+
"eval_logits/chosen": -2.54604434967041,
|
| 168 |
+
"eval_logits/rejected": -2.5107295513153076,
|
| 169 |
+
"eval_logps/chosen": -336.2060241699219,
|
| 170 |
+
"eval_logps/rejected": -330.2667541503906,
|
| 171 |
+
"eval_loss": 0.9028440117835999,
|
| 172 |
+
"eval_rewards/accuracies": 0.6904761791229248,
|
| 173 |
+
"eval_rewards/chosen": -0.5209627151489258,
|
| 174 |
+
"eval_rewards/margins": 0.1639193296432495,
|
| 175 |
+
"eval_rewards/rejected": -0.6848820447921753,
|
| 176 |
+
"eval_runtime": 245.63,
|
| 177 |
+
"eval_samples_per_second": 8.142,
|
| 178 |
+
"eval_steps_per_second": 0.256,
|
| 179 |
"step": 100
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.12,
|
| 183 |
"learning_rate": 4.996723692767926e-07,
|
| 184 |
+
"logits/chosen": -2.258904218673706,
|
| 185 |
+
"logits/rejected": -2.185375452041626,
|
| 186 |
+
"logps/chosen": -319.14654541015625,
|
| 187 |
+
"logps/rejected": -304.89739990234375,
|
| 188 |
+
"loss": 0.8703,
|
| 189 |
+
"rewards/accuracies": 0.668749988079071,
|
| 190 |
+
"rewards/chosen": -0.5882319211959839,
|
| 191 |
+
"rewards/margins": 0.2905711531639099,
|
| 192 |
+
"rewards/rejected": -0.8788030743598938,
|
| 193 |
"step": 110
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.13,
|
| 197 |
"learning_rate": 4.990375746213598e-07,
|
| 198 |
+
"logits/chosen": -1.46225106716156,
|
| 199 |
+
"logits/rejected": -1.342179775238037,
|
| 200 |
+
"logps/chosen": -348.0282897949219,
|
| 201 |
+
"logps/rejected": -337.330078125,
|
| 202 |
+
"loss": 0.8274,
|
| 203 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 204 |
+
"rewards/chosen": -0.42451825737953186,
|
| 205 |
+
"rewards/margins": 0.41508832573890686,
|
| 206 |
+
"rewards/rejected": -0.8396065831184387,
|
| 207 |
"step": 120
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.14,
|
| 211 |
"learning_rate": 4.980697142834314e-07,
|
| 212 |
+
"logits/chosen": -1.1638177633285522,
|
| 213 |
+
"logits/rejected": -0.918566107749939,
|
| 214 |
+
"logps/chosen": -371.7701721191406,
|
| 215 |
+
"logps/rejected": -366.410400390625,
|
| 216 |
+
"loss": 0.8256,
|
| 217 |
+
"rewards/accuracies": 0.706250011920929,
|
| 218 |
+
"rewards/chosen": -0.5147355198860168,
|
| 219 |
+
"rewards/margins": 0.4547084867954254,
|
| 220 |
+
"rewards/rejected": -0.9694439172744751,
|
| 221 |
"step": 130
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 0.15,
|
| 225 |
"learning_rate": 4.967700826904229e-07,
|
| 226 |
+
"logits/chosen": 0.09236583858728409,
|
| 227 |
+
"logits/rejected": 0.07060788571834564,
|
| 228 |
+
"logps/chosen": -294.2372741699219,
|
| 229 |
+
"logps/rejected": -336.0912170410156,
|
| 230 |
+
"loss": 0.771,
|
| 231 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 232 |
+
"rewards/chosen": -0.5445362329483032,
|
| 233 |
+
"rewards/margins": 0.5613424181938171,
|
| 234 |
+
"rewards/rejected": -1.1058785915374756,
|
| 235 |
"step": 140
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 0.16,
|
| 239 |
"learning_rate": 4.951404179843962e-07,
|
| 240 |
+
"logits/chosen": 0.7083513140678406,
|
| 241 |
+
"logits/rejected": 0.5464950203895569,
|
| 242 |
+
"logps/chosen": -364.0424499511719,
|
| 243 |
+
"logps/rejected": -411.08209228515625,
|
| 244 |
+
"loss": 0.8494,
|
| 245 |
+
"rewards/accuracies": 0.637499988079071,
|
| 246 |
+
"rewards/chosen": -0.8699826002120972,
|
| 247 |
+
"rewards/margins": 0.39457255601882935,
|
| 248 |
+
"rewards/rejected": -1.2645552158355713,
|
| 249 |
"step": 150
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.17,
|
| 253 |
"learning_rate": 4.931828996974498e-07,
|
| 254 |
+
"logits/chosen": 0.4938809871673584,
|
| 255 |
+
"logits/rejected": 0.9410598874092102,
|
| 256 |
+
"logps/chosen": -424.72430419921875,
|
| 257 |
+
"logps/rejected": -468.349609375,
|
| 258 |
+
"loss": 0.7631,
|
| 259 |
"rewards/accuracies": 0.6937500238418579,
|
| 260 |
+
"rewards/chosen": -1.2891677618026733,
|
| 261 |
+
"rewards/margins": 0.488609254360199,
|
| 262 |
+
"rewards/rejected": -1.777777075767517,
|
| 263 |
"step": 160
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.18,
|
| 267 |
"learning_rate": 4.909001458367866e-07,
|
| 268 |
+
"logits/chosen": 0.14566074311733246,
|
| 269 |
+
"logits/rejected": 0.20485401153564453,
|
| 270 |
+
"logps/chosen": -388.87158203125,
|
| 271 |
+
"logps/rejected": -427.6263732910156,
|
| 272 |
+
"loss": 0.7772,
|
| 273 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 274 |
+
"rewards/chosen": -1.1580675840377808,
|
| 275 |
+
"rewards/margins": 0.6463179588317871,
|
| 276 |
+
"rewards/rejected": -1.8043855428695679,
|
| 277 |
"step": 170
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 0.19,
|
| 281 |
"learning_rate": 4.882952093833627e-07,
|
| 282 |
+
"logits/chosen": 0.1968054324388504,
|
| 283 |
+
"logits/rejected": 0.5141702890396118,
|
| 284 |
+
"logps/chosen": -351.22821044921875,
|
| 285 |
+
"logps/rejected": -409.0249938964844,
|
| 286 |
+
"loss": 0.7395,
|
| 287 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 288 |
+
"rewards/chosen": -1.02028489112854,
|
| 289 |
+
"rewards/margins": 0.7496393322944641,
|
| 290 |
+
"rewards/rejected": -1.7699241638183594,
|
| 291 |
"step": 180
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 0.2,
|
| 295 |
"learning_rate": 4.853715742087946e-07,
|
| 296 |
+
"logits/chosen": 0.310161292552948,
|
| 297 |
+
"logits/rejected": 0.9174222946166992,
|
| 298 |
+
"logps/chosen": -406.7144775390625,
|
| 299 |
+
"logps/rejected": -464.5382385253906,
|
| 300 |
+
"loss": 0.7373,
|
| 301 |
+
"rewards/accuracies": 0.762499988079071,
|
| 302 |
+
"rewards/chosen": -1.2787965536117554,
|
| 303 |
+
"rewards/margins": 0.8055577278137207,
|
| 304 |
+
"rewards/rejected": -2.0843544006347656,
|
| 305 |
"step": 190
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 0.21,
|
| 309 |
"learning_rate": 4.821331504159906e-07,
|
| 310 |
+
"logits/chosen": 0.49894601106643677,
|
| 311 |
+
"logits/rejected": 1.0194227695465088,
|
| 312 |
+
"logps/chosen": -405.57330322265625,
|
| 313 |
+
"logps/rejected": -467.26641845703125,
|
| 314 |
+
"loss": 0.7658,
|
| 315 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 316 |
+
"rewards/chosen": -1.1890487670898438,
|
| 317 |
+
"rewards/margins": 0.7286871671676636,
|
| 318 |
+
"rewards/rejected": -1.9177358150482178,
|
| 319 |
"step": 200
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 0.21,
|
| 323 |
+
"eval_logits/chosen": -0.152938574552536,
|
| 324 |
+
"eval_logits/rejected": 0.33469492197036743,
|
| 325 |
+
"eval_logps/chosen": -378.24761962890625,
|
| 326 |
+
"eval_logps/rejected": -431.1014709472656,
|
| 327 |
+
"eval_loss": 0.7649896144866943,
|
| 328 |
+
"eval_rewards/accuracies": 0.7460317611694336,
|
| 329 |
+
"eval_rewards/chosen": -0.941378653049469,
|
| 330 |
+
"eval_rewards/margins": 0.7518512010574341,
|
| 331 |
+
"eval_rewards/rejected": -1.6932299137115479,
|
| 332 |
+
"eval_runtime": 243.5152,
|
| 333 |
+
"eval_samples_per_second": 8.213,
|
| 334 |
+
"eval_steps_per_second": 0.259,
|
| 335 |
"step": 200
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 0.22,
|
| 339 |
"learning_rate": 4.785842691097342e-07,
|
| 340 |
+
"logits/chosen": -0.3292608857154846,
|
| 341 |
+
"logits/rejected": 0.1720762550830841,
|
| 342 |
+
"logps/chosen": -389.8594665527344,
|
| 343 |
+
"logps/rejected": -401.64581298828125,
|
| 344 |
+
"loss": 0.7634,
|
| 345 |
"rewards/accuracies": 0.7250000238418579,
|
| 346 |
+
"rewards/chosen": -0.8458383679389954,
|
| 347 |
+
"rewards/margins": 0.6316131353378296,
|
| 348 |
+
"rewards/rejected": -1.4774516820907593,
|
| 349 |
"step": 210
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 0.23,
|
| 353 |
"learning_rate": 4.7472967660421603e-07,
|
| 354 |
+
"logits/chosen": 0.5375509858131409,
|
| 355 |
+
"logits/rejected": 0.9775497317314148,
|
| 356 |
+
"logps/chosen": -387.09014892578125,
|
| 357 |
+
"logps/rejected": -446.6785583496094,
|
| 358 |
+
"loss": 0.7559,
|
| 359 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 360 |
+
"rewards/chosen": -0.9755613207817078,
|
| 361 |
+
"rewards/margins": 0.7280157208442688,
|
| 362 |
+
"rewards/rejected": -1.7035770416259766,
|
| 363 |
"step": 220
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 0.24,
|
| 367 |
"learning_rate": 4.705745280752585e-07,
|
| 368 |
+
"logits/chosen": 0.5464267134666443,
|
| 369 |
+
"logits/rejected": 1.0262590646743774,
|
| 370 |
+
"logps/chosen": -444.06072998046875,
|
| 371 |
+
"logps/rejected": -483.55926513671875,
|
| 372 |
+
"loss": 0.7417,
|
| 373 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 374 |
+
"rewards/chosen": -1.5034980773925781,
|
| 375 |
+
"rewards/margins": 0.8456419706344604,
|
| 376 |
+
"rewards/rejected": -2.349139928817749,
|
| 377 |
"step": 230
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.25,
|
| 381 |
"learning_rate": 4.6612438066572555e-07,
|
| 382 |
+
"logits/chosen": 0.6961275935173035,
|
| 383 |
+
"logits/rejected": 1.6300386190414429,
|
| 384 |
+
"logps/chosen": -418.9384765625,
|
| 385 |
+
"logps/rejected": -454.033935546875,
|
| 386 |
+
"loss": 0.6919,
|
| 387 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 388 |
+
"rewards/chosen": -1.4168249368667603,
|
| 389 |
+
"rewards/margins": 0.863106369972229,
|
| 390 |
+
"rewards/rejected": -2.2799313068389893,
|
| 391 |
"step": 240
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 0.26,
|
| 395 |
"learning_rate": 4.6138518605333664e-07,
|
| 396 |
+
"logits/chosen": -0.6623570919036865,
|
| 397 |
+
"logits/rejected": -0.1930474489927292,
|
| 398 |
+
"logps/chosen": -347.79315185546875,
|
| 399 |
+
"logps/rejected": -429.45465087890625,
|
| 400 |
+
"loss": 0.7378,
|
| 401 |
+
"rewards/accuracies": 0.6875,
|
| 402 |
+
"rewards/chosen": -0.8020486831665039,
|
| 403 |
+
"rewards/margins": 0.6783057451248169,
|
| 404 |
+
"rewards/rejected": -1.4803544282913208,
|
| 405 |
"step": 250
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 0.27,
|
| 409 |
"learning_rate": 4.5636328249082514e-07,
|
| 410 |
+
"logits/chosen": -1.4540965557098389,
|
| 411 |
+
"logits/rejected": -0.7168424725532532,
|
| 412 |
+
"logps/chosen": -342.86102294921875,
|
| 413 |
+
"logps/rejected": -387.8767395019531,
|
| 414 |
+
"loss": 0.7447,
|
| 415 |
+
"rewards/accuracies": 0.78125,
|
| 416 |
+
"rewards/chosen": -0.5879959464073181,
|
| 417 |
+
"rewards/margins": 0.7121099233627319,
|
| 418 |
+
"rewards/rejected": -1.3001058101654053,
|
| 419 |
"step": 260
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 0.28,
|
| 423 |
"learning_rate": 4.510653863290871e-07,
|
| 424 |
+
"logits/chosen": -0.18550051748752594,
|
| 425 |
+
"logits/rejected": 0.8060259819030762,
|
| 426 |
+
"logps/chosen": -402.1506042480469,
|
| 427 |
+
"logps/rejected": -452.3603515625,
|
| 428 |
+
"loss": 0.7185,
|
| 429 |
+
"rewards/accuracies": 0.768750011920929,
|
| 430 |
+
"rewards/chosen": -1.002969741821289,
|
| 431 |
+
"rewards/margins": 1.018422245979309,
|
| 432 |
+
"rewards/rejected": -2.0213921070098877,
|
| 433 |
"step": 270
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 0.29,
|
| 437 |
"learning_rate": 4.4549858303465737e-07,
|
| 438 |
+
"logits/chosen": 0.21997830271720886,
|
| 439 |
+
"logits/rejected": 0.8937602043151855,
|
| 440 |
+
"logps/chosen": -420.73773193359375,
|
| 441 |
+
"logps/rejected": -503.6924743652344,
|
| 442 |
+
"loss": 0.6991,
|
| 443 |
+
"rewards/accuracies": 0.78125,
|
| 444 |
+
"rewards/chosen": -1.295493721961975,
|
| 445 |
+
"rewards/margins": 0.9076651334762573,
|
| 446 |
+
"rewards/rejected": -2.2031588554382324,
|
| 447 |
"step": 280
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 0.3,
|
| 451 |
"learning_rate": 4.396703177135261e-07,
|
| 452 |
+
"logits/chosen": 0.11293928325176239,
|
| 453 |
+
"logits/rejected": 0.5330738425254822,
|
| 454 |
+
"logps/chosen": -414.304931640625,
|
| 455 |
+
"logps/rejected": -458.75946044921875,
|
| 456 |
+
"loss": 0.7064,
|
| 457 |
+
"rewards/accuracies": 0.737500011920929,
|
| 458 |
+
"rewards/chosen": -1.2927477359771729,
|
| 459 |
+
"rewards/margins": 0.7714477777481079,
|
| 460 |
+
"rewards/rejected": -2.0641958713531494,
|
| 461 |
"step": 290
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 0.31,
|
| 465 |
"learning_rate": 4.335883851539693e-07,
|
| 466 |
+
"logits/chosen": 0.7407528162002563,
|
| 467 |
+
"logits/rejected": 1.589734435081482,
|
| 468 |
+
"logps/chosen": -405.79022216796875,
|
| 469 |
+
"logps/rejected": -488.19036865234375,
|
| 470 |
+
"loss": 0.7079,
|
| 471 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 472 |
+
"rewards/chosen": -1.223577857017517,
|
| 473 |
+
"rewards/margins": 1.1488986015319824,
|
| 474 |
+
"rewards/rejected": -2.372476816177368,
|
| 475 |
"step": 300
|
| 476 |
},
|
| 477 |
{
|
| 478 |
"epoch": 0.31,
|
| 479 |
+
"eval_logits/chosen": 0.87442946434021,
|
| 480 |
+
"eval_logits/rejected": 1.8370469808578491,
|
| 481 |
+
"eval_logps/chosen": -422.4754333496094,
|
| 482 |
+
"eval_logps/rejected": -510.4591064453125,
|
| 483 |
+
"eval_loss": 0.7289105653762817,
|
| 484 |
+
"eval_rewards/accuracies": 0.7559523582458496,
|
| 485 |
+
"eval_rewards/chosen": -1.383657455444336,
|
| 486 |
+
"eval_rewards/margins": 1.1031482219696045,
|
| 487 |
+
"eval_rewards/rejected": -2.4868052005767822,
|
| 488 |
+
"eval_runtime": 242.4112,
|
| 489 |
+
"eval_samples_per_second": 8.25,
|
| 490 |
"eval_steps_per_second": 0.26,
|
| 491 |
"step": 300
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 0.32,
|
| 495 |
"learning_rate": 4.272609194017105e-07,
|
| 496 |
+
"logits/chosen": 0.7091141939163208,
|
| 497 |
+
"logits/rejected": 1.7715873718261719,
|
| 498 |
+
"logps/chosen": -404.2896728515625,
|
| 499 |
+
"logps/rejected": -514.3563842773438,
|
| 500 |
+
"loss": 0.6597,
|
| 501 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 502 |
+
"rewards/chosen": -1.2207536697387695,
|
| 503 |
+
"rewards/margins": 1.231592059135437,
|
| 504 |
+
"rewards/rejected": -2.452345848083496,
|
| 505 |
"step": 310
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.33,
|
| 509 |
"learning_rate": 4.2069638288135547e-07,
|
| 510 |
+
"logits/chosen": 0.6078277826309204,
|
| 511 |
+
"logits/rejected": 1.4416134357452393,
|
| 512 |
+
"logps/chosen": -417.4878845214844,
|
| 513 |
+
"logps/rejected": -491.2012634277344,
|
| 514 |
+
"loss": 0.7219,
|
| 515 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 516 |
+
"rewards/chosen": -1.358330249786377,
|
| 517 |
+
"rewards/margins": 0.8599346280097961,
|
| 518 |
+
"rewards/rejected": -2.2182650566101074,
|
| 519 |
"step": 320
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.35,
|
| 523 |
"learning_rate": 4.139035550786494e-07,
|
| 524 |
+
"logits/chosen": 0.2845739424228668,
|
| 525 |
+
"logits/rejected": 0.876905620098114,
|
| 526 |
+
"logps/chosen": -394.2958068847656,
|
| 527 |
+
"logps/rejected": -421.72393798828125,
|
| 528 |
+
"loss": 0.7567,
|
| 529 |
"rewards/accuracies": 0.7437499761581421,
|
| 530 |
+
"rewards/chosen": -1.147202730178833,
|
| 531 |
+
"rewards/margins": 0.7379333972930908,
|
| 532 |
+
"rewards/rejected": -1.8851358890533447,
|
| 533 |
"step": 330
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 0.36,
|
| 537 |
"learning_rate": 4.0689152079869306e-07,
|
| 538 |
+
"logits/chosen": -0.5365289449691772,
|
| 539 |
+
"logits/rejected": 0.34862279891967773,
|
| 540 |
+
"logps/chosen": -342.41485595703125,
|
| 541 |
+
"logps/rejected": -388.524658203125,
|
| 542 |
+
"loss": 0.776,
|
| 543 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 544 |
+
"rewards/chosen": -0.9666112065315247,
|
| 545 |
+
"rewards/margins": 0.6763932108879089,
|
| 546 |
+
"rewards/rejected": -1.6430044174194336,
|
| 547 |
"step": 340
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 0.37,
|
| 551 |
"learning_rate": 3.99669658015821e-07,
|
| 552 |
+
"logits/chosen": 0.45323339104652405,
|
| 553 |
+
"logits/rejected": 0.7332956194877625,
|
| 554 |
+
"logps/chosen": -399.76519775390625,
|
| 555 |
+
"logps/rejected": -506.49658203125,
|
| 556 |
+
"loss": 0.7213,
|
| 557 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 558 |
+
"rewards/chosen": -1.27635657787323,
|
| 559 |
+
"rewards/margins": 0.998461127281189,
|
| 560 |
+
"rewards/rejected": -2.2748172283172607,
|
| 561 |
"step": 350
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 0.38,
|
| 565 |
"learning_rate": 3.92247625331392e-07,
|
| 566 |
+
"logits/chosen": 0.6511309742927551,
|
| 567 |
+
"logits/rejected": 1.2110098600387573,
|
| 568 |
+
"logps/chosen": -422.53778076171875,
|
| 569 |
+
"logps/rejected": -472.4964294433594,
|
| 570 |
+
"loss": 0.7003,
|
| 571 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 572 |
+
"rewards/chosen": -1.3980082273483276,
|
| 573 |
+
"rewards/margins": 0.8997832536697388,
|
| 574 |
+
"rewards/rejected": -2.2977914810180664,
|
| 575 |
"step": 360
|
| 576 |
},
|
| 577 |
{
|
| 578 |
"epoch": 0.39,
|
| 579 |
"learning_rate": 3.846353490562664e-07,
|
| 580 |
+
"logits/chosen": 0.30174368619918823,
|
| 581 |
+
"logits/rejected": 0.9621099233627319,
|
| 582 |
+
"logps/chosen": -372.94635009765625,
|
| 583 |
+
"logps/rejected": -497.25714111328125,
|
| 584 |
+
"loss": 0.6659,
|
| 585 |
"rewards/accuracies": 0.7749999761581421,
|
| 586 |
+
"rewards/chosen": -1.3280470371246338,
|
| 587 |
+
"rewards/margins": 1.1047414541244507,
|
| 588 |
+
"rewards/rejected": -2.432788372039795,
|
| 589 |
"step": 370
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 0.4,
|
| 593 |
"learning_rate": 3.768430099352445e-07,
|
| 594 |
+
"logits/chosen": -0.1737074851989746,
|
| 595 |
+
"logits/rejected": 1.1676225662231445,
|
| 596 |
+
"logps/chosen": -460.0953063964844,
|
| 597 |
+
"logps/rejected": -537.3989868164062,
|
| 598 |
+
"loss": 0.6788,
|
| 599 |
+
"rewards/accuracies": 0.75,
|
| 600 |
+
"rewards/chosen": -1.6348745822906494,
|
| 601 |
+
"rewards/margins": 1.1145647764205933,
|
| 602 |
+
"rewards/rejected": -2.7494394779205322,
|
| 603 |
"step": 380
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 0.41,
|
| 607 |
"learning_rate": 3.6888102953122304e-07,
|
| 608 |
+
"logits/chosen": -0.2348991185426712,
|
| 609 |
+
"logits/rejected": 0.7535260915756226,
|
| 610 |
+
"logps/chosen": -421.128173828125,
|
| 611 |
+
"logps/rejected": -482.41656494140625,
|
| 612 |
+
"loss": 0.7027,
|
| 613 |
+
"rewards/accuracies": 0.762499988079071,
|
| 614 |
+
"rewards/chosen": -1.4536683559417725,
|
| 615 |
+
"rewards/margins": 1.019152045249939,
|
| 616 |
+
"rewards/rejected": -2.47282075881958,
|
| 617 |
"step": 390
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 0.42,
|
| 621 |
"learning_rate": 3.607600562872785e-07,
|
| 622 |
+
"logits/chosen": -0.24078145623207092,
|
| 623 |
+
"logits/rejected": 0.8144145011901855,
|
| 624 |
+
"logps/chosen": -454.4244689941406,
|
| 625 |
+
"logps/rejected": -500.9457092285156,
|
| 626 |
+
"loss": 0.6806,
|
| 627 |
+
"rewards/accuracies": 0.699999988079071,
|
| 628 |
+
"rewards/chosen": -1.5004165172576904,
|
| 629 |
+
"rewards/margins": 0.854039192199707,
|
| 630 |
+
"rewards/rejected": -2.3544554710388184,
|
| 631 |
"step": 400
|
| 632 |
},
|
| 633 |
{
|
| 634 |
"epoch": 0.42,
|
| 635 |
+
"eval_logits/chosen": 0.09922664612531662,
|
| 636 |
+
"eval_logits/rejected": 1.2713433504104614,
|
| 637 |
+
"eval_logps/chosen": -416.9630432128906,
|
| 638 |
+
"eval_logps/rejected": -503.67401123046875,
|
| 639 |
+
"eval_loss": 0.7040213346481323,
|
| 640 |
+
"eval_rewards/accuracies": 0.7698412537574768,
|
| 641 |
+
"eval_rewards/chosen": -1.3285325765609741,
|
| 642 |
+
"eval_rewards/margins": 1.0904221534729004,
|
| 643 |
+
"eval_rewards/rejected": -2.418954849243164,
|
| 644 |
+
"eval_runtime": 242.7344,
|
| 645 |
+
"eval_samples_per_second": 8.239,
|
| 646 |
+
"eval_steps_per_second": 0.26,
|
| 647 |
"step": 400
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.43,
|
| 651 |
"learning_rate": 3.5249095128531856e-07,
|
| 652 |
+
"logits/chosen": -0.06878291815519333,
|
| 653 |
+
"logits/rejected": 1.236665964126587,
|
| 654 |
+
"logps/chosen": -443.3794860839844,
|
| 655 |
+
"logps/rejected": -522.2283325195312,
|
| 656 |
+
"loss": 0.7085,
|
| 657 |
+
"rewards/accuracies": 0.75,
|
| 658 |
+
"rewards/chosen": -1.2952146530151367,
|
| 659 |
+
"rewards/margins": 1.0169426202774048,
|
| 660 |
+
"rewards/rejected": -2.312157154083252,
|
| 661 |
"step": 410
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.44,
|
| 665 |
"learning_rate": 3.4408477372034736e-07,
|
| 666 |
+
"logits/chosen": 0.041589152067899704,
|
| 667 |
+
"logits/rejected": 1.3509619235992432,
|
| 668 |
+
"logps/chosen": -390.40008544921875,
|
| 669 |
+
"logps/rejected": -440.0779724121094,
|
| 670 |
+
"loss": 0.7102,
|
| 671 |
+
"rewards/accuracies": 0.6875,
|
| 672 |
+
"rewards/chosen": -1.2629300355911255,
|
| 673 |
+
"rewards/margins": 0.8384572863578796,
|
| 674 |
+
"rewards/rejected": -2.1013875007629395,
|
| 675 |
"step": 420
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.45,
|
| 679 |
"learning_rate": 3.3555276610977276e-07,
|
| 680 |
+
"logits/chosen": 0.06685711443424225,
|
| 681 |
+
"logits/rejected": 1.2138116359710693,
|
| 682 |
+
"logps/chosen": -372.45196533203125,
|
| 683 |
+
"logps/rejected": -431.705078125,
|
| 684 |
+
"loss": 0.7376,
|
| 685 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 686 |
+
"rewards/chosen": -1.1025346517562866,
|
| 687 |
+
"rewards/margins": 0.8204299211502075,
|
| 688 |
+
"rewards/rejected": -1.9229644536972046,
|
| 689 |
"step": 430
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.46,
|
| 693 |
"learning_rate": 3.269063392575352e-07,
|
| 694 |
+
"logits/chosen": 0.9198445081710815,
|
| 695 |
+
"logits/rejected": 0.9934859275817871,
|
| 696 |
+
"logps/chosen": -396.35968017578125,
|
| 697 |
+
"logps/rejected": -476.847900390625,
|
| 698 |
+
"loss": 0.6966,
|
| 699 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 700 |
+
"rewards/chosen": -1.3736763000488281,
|
| 701 |
+
"rewards/margins": 0.9061026573181152,
|
| 702 |
+
"rewards/rejected": -2.2797789573669434,
|
| 703 |
"step": 440
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.47,
|
| 707 |
"learning_rate": 3.1815705699316964e-07,
|
| 708 |
+
"logits/chosen": 0.9360873103141785,
|
| 709 |
+
"logits/rejected": 1.5897537469863892,
|
| 710 |
+
"logps/chosen": -401.81207275390625,
|
| 711 |
+
"logps/rejected": -485.36041259765625,
|
| 712 |
+
"loss": 0.715,
|
| 713 |
"rewards/accuracies": 0.762499988079071,
|
| 714 |
+
"rewards/chosen": -1.3765710592269897,
|
| 715 |
+
"rewards/margins": 1.040223479270935,
|
| 716 |
+
"rewards/rejected": -2.416795015335083,
|
| 717 |
"step": 450
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.48,
|
| 721 |
"learning_rate": 3.0931662070620794e-07,
|
| 722 |
+
"logits/chosen": 0.9600452184677124,
|
| 723 |
+
"logits/rejected": 2.169283866882324,
|
| 724 |
+
"logps/chosen": -413.7256774902344,
|
| 725 |
+
"logps/rejected": -508.83892822265625,
|
| 726 |
+
"loss": 0.6982,
|
| 727 |
+
"rewards/accuracies": 0.75,
|
| 728 |
+
"rewards/chosen": -1.4869290590286255,
|
| 729 |
+
"rewards/margins": 1.0339877605438232,
|
| 730 |
+
"rewards/rejected": -2.5209171772003174,
|
| 731 |
"step": 460
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.49,
|
| 735 |
"learning_rate": 3.003968536966078e-07,
|
| 736 |
+
"logits/chosen": 1.2197582721710205,
|
| 737 |
+
"logits/rejected": 2.1600019931793213,
|
| 738 |
+
"logps/chosen": -441.291748046875,
|
| 739 |
+
"logps/rejected": -515.36083984375,
|
| 740 |
+
"loss": 0.6864,
|
| 741 |
+
"rewards/accuracies": 0.793749988079071,
|
| 742 |
+
"rewards/chosen": -1.4309029579162598,
|
| 743 |
+
"rewards/margins": 1.1178550720214844,
|
| 744 |
+
"rewards/rejected": -2.548758029937744,
|
| 745 |
"step": 470
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.5,
|
| 749 |
"learning_rate": 2.9140968536213693e-07,
|
| 750 |
+
"logits/chosen": 1.2262110710144043,
|
| 751 |
+
"logits/rejected": 2.440544843673706,
|
| 752 |
+
"logps/chosen": -372.9617919921875,
|
| 753 |
+
"logps/rejected": -461.28765869140625,
|
| 754 |
+
"loss": 0.7197,
|
| 755 |
+
"rewards/accuracies": 0.731249988079071,
|
| 756 |
+
"rewards/chosen": -1.4022343158721924,
|
| 757 |
+
"rewards/margins": 0.9221154451370239,
|
| 758 |
+
"rewards/rejected": -2.324349880218506,
|
| 759 |
"step": 480
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.51,
|
| 763 |
"learning_rate": 2.823671352438608e-07,
|
| 764 |
+
"logits/chosen": 1.2201405763626099,
|
| 765 |
+
"logits/rejected": 2.058537721633911,
|
| 766 |
+
"logps/chosen": -409.15814208984375,
|
| 767 |
+
"logps/rejected": -467.56573486328125,
|
| 768 |
+
"loss": 0.6978,
|
| 769 |
+
"rewards/accuracies": 0.706250011920929,
|
| 770 |
+
"rewards/chosen": -1.334566354751587,
|
| 771 |
+
"rewards/margins": 0.8473381996154785,
|
| 772 |
+
"rewards/rejected": -2.1819043159484863,
|
| 773 |
"step": 490
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.52,
|
| 777 |
"learning_rate": 2.73281296951072e-07,
|
| 778 |
+
"logits/chosen": 1.904309868812561,
|
| 779 |
+
"logits/rejected": 2.613346576690674,
|
| 780 |
+
"logps/chosen": -433.7508850097656,
|
| 781 |
+
"logps/rejected": -524.71044921875,
|
| 782 |
+
"loss": 0.7129,
|
| 783 |
+
"rewards/accuracies": 0.768750011920929,
|
| 784 |
+
"rewards/chosen": -1.621690034866333,
|
| 785 |
+
"rewards/margins": 1.130249261856079,
|
| 786 |
+
"rewards/rejected": -2.751939058303833,
|
| 787 |
"step": 500
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.52,
|
| 791 |
+
"eval_logits/chosen": 1.4090828895568848,
|
| 792 |
+
"eval_logits/rejected": 2.334313154220581,
|
| 793 |
+
"eval_logps/chosen": -430.316650390625,
|
| 794 |
+
"eval_logps/rejected": -514.4609375,
|
| 795 |
+
"eval_loss": 0.6979612112045288,
|
| 796 |
+
"eval_rewards/accuracies": 0.7440476417541504,
|
| 797 |
+
"eval_rewards/chosen": -1.4620689153671265,
|
| 798 |
+
"eval_rewards/margins": 1.0647554397583008,
|
| 799 |
+
"eval_rewards/rejected": -2.5268242359161377,
|
| 800 |
+
"eval_runtime": 243.0213,
|
| 801 |
+
"eval_samples_per_second": 8.23,
|
| 802 |
"eval_steps_per_second": 0.259,
|
| 803 |
"step": 500
|
| 804 |
},
|
| 805 |
{
|
| 806 |
"epoch": 0.53,
|
| 807 |
"learning_rate": 2.641643219871597e-07,
|
| 808 |
+
"logits/chosen": 1.5503180027008057,
|
| 809 |
+
"logits/rejected": 2.431549072265625,
|
| 810 |
+
"logps/chosen": -441.70721435546875,
|
| 811 |
+
"logps/rejected": -510.1366271972656,
|
| 812 |
+
"loss": 0.6483,
|
| 813 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 814 |
+
"rewards/chosen": -1.403807282447815,
|
| 815 |
+
"rewards/margins": 1.1817026138305664,
|
| 816 |
+
"rewards/rejected": -2.585510015487671,
|
| 817 |
"step": 510
|
| 818 |
},
|
| 819 |
{
|
| 820 |
"epoch": 0.54,
|
| 821 |
"learning_rate": 2.550284034980507e-07,
|
| 822 |
+
"logits/chosen": 1.6724971532821655,
|
| 823 |
+
"logits/rejected": 2.7851836681365967,
|
| 824 |
+
"logps/chosen": -426.255859375,
|
| 825 |
+
"logps/rejected": -528.8610229492188,
|
| 826 |
+
"loss": 0.701,
|
| 827 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 828 |
+
"rewards/chosen": -1.6710426807403564,
|
| 829 |
+
"rewards/margins": 1.100687861442566,
|
| 830 |
+
"rewards/rejected": -2.771730661392212,
|
| 831 |
"step": 520
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 0.55,
|
| 835 |
"learning_rate": 2.4588575996495794e-07,
|
| 836 |
+
"logits/chosen": 1.8039824962615967,
|
| 837 |
+
"logits/rejected": 2.613548755645752,
|
| 838 |
+
"logps/chosen": -449.572265625,
|
| 839 |
+
"logps/rejected": -537.4893798828125,
|
| 840 |
+
"loss": 0.6894,
|
| 841 |
+
"rewards/accuracies": 0.737500011920929,
|
| 842 |
+
"rewards/chosen": -1.654245376586914,
|
| 843 |
+
"rewards/margins": 1.1508926153182983,
|
| 844 |
+
"rewards/rejected": -2.805138111114502,
|
| 845 |
"step": 530
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 0.57,
|
| 849 |
"learning_rate": 2.367486188632446e-07,
|
| 850 |
+
"logits/chosen": 1.2713382244110107,
|
| 851 |
+
"logits/rejected": 2.1325278282165527,
|
| 852 |
+
"logps/chosen": -456.769775390625,
|
| 853 |
+
"logps/rejected": -589.4976806640625,
|
| 854 |
+
"loss": 0.6831,
|
| 855 |
+
"rewards/accuracies": 0.75,
|
| 856 |
+
"rewards/chosen": -1.6260936260223389,
|
| 857 |
+
"rewards/margins": 1.2662837505340576,
|
| 858 |
+
"rewards/rejected": -2.8923773765563965,
|
| 859 |
"step": 540
|
| 860 |
},
|
| 861 |
{
|
| 862 |
"epoch": 0.58,
|
| 863 |
"learning_rate": 2.276292003092593e-07,
|
| 864 |
+
"logits/chosen": 1.5338995456695557,
|
| 865 |
+
"logits/rejected": 2.471559524536133,
|
| 866 |
+
"logps/chosen": -422.6114807128906,
|
| 867 |
+
"logps/rejected": -517.1149291992188,
|
| 868 |
+
"loss": 0.7116,
|
| 869 |
"rewards/accuracies": 0.75,
|
| 870 |
+
"rewards/chosen": -1.5288379192352295,
|
| 871 |
+
"rewards/margins": 1.1625785827636719,
|
| 872 |
+
"rewards/rejected": -2.6914165019989014,
|
| 873 |
"step": 550
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.59,
|
| 877 |
"learning_rate": 2.185397007170141e-07,
|
| 878 |
+
"logits/chosen": 1.3305785655975342,
|
| 879 |
+
"logits/rejected": 1.9697010517120361,
|
| 880 |
+
"logps/chosen": -409.57635498046875,
|
| 881 |
+
"logps/rejected": -467.7806091308594,
|
| 882 |
+
"loss": 0.7032,
|
| 883 |
"rewards/accuracies": 0.731249988079071,
|
| 884 |
+
"rewards/chosen": -1.4281054735183716,
|
| 885 |
+
"rewards/margins": 0.915216326713562,
|
| 886 |
+
"rewards/rejected": -2.3433218002319336,
|
| 887 |
"step": 560
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 0.6,
|
| 891 |
"learning_rate": 2.094922764865619e-07,
|
| 892 |
+
"logits/chosen": 1.1964863538742065,
|
| 893 |
+
"logits/rejected": 2.188833236694336,
|
| 894 |
+
"logps/chosen": -427.9024353027344,
|
| 895 |
+
"logps/rejected": -497.96917724609375,
|
| 896 |
+
"loss": 0.7036,
|
| 897 |
+
"rewards/accuracies": 0.737500011920929,
|
| 898 |
+
"rewards/chosen": -1.5176267623901367,
|
| 899 |
+
"rewards/margins": 0.8729730844497681,
|
| 900 |
+
"rewards/rejected": -2.3905997276306152,
|
| 901 |
"step": 570
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 0.61,
|
| 905 |
"learning_rate": 2.0049902774588797e-07,
|
| 906 |
+
"logits/chosen": 1.2527071237564087,
|
| 907 |
+
"logits/rejected": 2.2398314476013184,
|
| 908 |
+
"logps/chosen": -436.9583435058594,
|
| 909 |
+
"logps/rejected": -504.8207092285156,
|
| 910 |
+
"loss": 0.6924,
|
| 911 |
+
"rewards/accuracies": 0.768750011920929,
|
| 912 |
+
"rewards/chosen": -1.6303842067718506,
|
| 913 |
+
"rewards/margins": 1.0391124486923218,
|
| 914 |
+
"rewards/rejected": -2.669496774673462,
|
| 915 |
"step": 580
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 0.62,
|
| 919 |
"learning_rate": 1.9157198216806238e-07,
|
| 920 |
+
"logits/chosen": 0.6701461672782898,
|
| 921 |
+
"logits/rejected": 1.8555580377578735,
|
| 922 |
+
"logps/chosen": -405.43902587890625,
|
| 923 |
+
"logps/rejected": -496.9740295410156,
|
| 924 |
+
"loss": 0.6922,
|
| 925 |
+
"rewards/accuracies": 0.706250011920929,
|
| 926 |
+
"rewards/chosen": -1.323939561843872,
|
| 927 |
+
"rewards/margins": 0.8526731729507446,
|
| 928 |
+
"rewards/rejected": -2.176612615585327,
|
| 929 |
"step": 590
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 0.63,
|
| 933 |
"learning_rate": 1.8272307888529274e-07,
|
| 934 |
+
"logits/chosen": 0.462153822183609,
|
| 935 |
+
"logits/rejected": 1.8794240951538086,
|
| 936 |
+
"logps/chosen": -451.150634765625,
|
| 937 |
+
"logps/rejected": -545.4595947265625,
|
| 938 |
+
"loss": 0.6636,
|
| 939 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 940 |
+
"rewards/chosen": -1.2517478466033936,
|
| 941 |
+
"rewards/margins": 1.0819759368896484,
|
| 942 |
+
"rewards/rejected": -2.333723783493042,
|
| 943 |
"step": 600
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 0.63,
|
| 947 |
+
"eval_logits/chosen": 0.7470372319221497,
|
| 948 |
+
"eval_logits/rejected": 2.2081830501556396,
|
| 949 |
+
"eval_logps/chosen": -417.3849792480469,
|
| 950 |
+
"eval_logps/rejected": -513.6627197265625,
|
| 951 |
+
"eval_loss": 0.6876600980758667,
|
| 952 |
+
"eval_rewards/accuracies": 0.75,
|
| 953 |
+
"eval_rewards/chosen": -1.3327523469924927,
|
| 954 |
+
"eval_rewards/margins": 1.1860896348953247,
|
| 955 |
+
"eval_rewards/rejected": -2.5188419818878174,
|
| 956 |
+
"eval_runtime": 244.0904,
|
| 957 |
+
"eval_samples_per_second": 8.194,
|
| 958 |
"eval_steps_per_second": 0.258,
|
| 959 |
"step": 600
|
| 960 |
},
|
| 961 |
{
|
| 962 |
"epoch": 0.64,
|
| 963 |
"learning_rate": 1.7396415252139288e-07,
|
| 964 |
+
"logits/chosen": 1.345348834991455,
|
| 965 |
+
"logits/rejected": 2.903435707092285,
|
| 966 |
+
"logps/chosen": -418.60797119140625,
|
| 967 |
+
"logps/rejected": -482.93035888671875,
|
| 968 |
+
"loss": 0.6686,
|
| 969 |
+
"rewards/accuracies": 0.75,
|
| 970 |
+
"rewards/chosen": -1.4358808994293213,
|
| 971 |
+
"rewards/margins": 1.1490730047225952,
|
| 972 |
+
"rewards/rejected": -2.584954261779785,
|
| 973 |
"step": 610
|
| 974 |
},
|
| 975 |
{
|
| 976 |
"epoch": 0.65,
|
| 977 |
"learning_rate": 1.6530691736402316e-07,
|
| 978 |
+
"logits/chosen": 2.0353918075561523,
|
| 979 |
+
"logits/rejected": 3.255995512008667,
|
| 980 |
+
"logps/chosen": -455.14959716796875,
|
| 981 |
+
"logps/rejected": -525.127685546875,
|
| 982 |
+
"loss": 0.6894,
|
| 983 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 984 |
+
"rewards/chosen": -1.831215500831604,
|
| 985 |
+
"rewards/margins": 1.118740200996399,
|
| 986 |
+
"rewards/rejected": -2.949955701828003,
|
| 987 |
"step": 620
|
| 988 |
},
|
| 989 |
{
|
| 990 |
"epoch": 0.66,
|
| 991 |
"learning_rate": 1.5676295169786864e-07,
|
| 992 |
+
"logits/chosen": 2.8849244117736816,
|
| 993 |
+
"logits/rejected": 3.68190336227417,
|
| 994 |
+
"logps/chosen": -441.6712951660156,
|
| 995 |
+
"logps/rejected": -579.8939819335938,
|
| 996 |
+
"loss": 0.6633,
|
| 997 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 998 |
+
"rewards/chosen": -1.800830602645874,
|
| 999 |
+
"rewards/margins": 1.4441092014312744,
|
| 1000 |
+
"rewards/rejected": -3.2449398040771484,
|
| 1001 |
"step": 630
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
"epoch": 0.67,
|
| 1005 |
"learning_rate": 1.483436823197092e-07,
|
| 1006 |
+
"logits/chosen": 1.9812686443328857,
|
| 1007 |
+
"logits/rejected": 2.8578484058380127,
|
| 1008 |
+
"logps/chosen": -427.90460205078125,
|
| 1009 |
+
"logps/rejected": -523.0911865234375,
|
| 1010 |
+
"loss": 0.6953,
|
| 1011 |
"rewards/accuracies": 0.800000011920929,
|
| 1012 |
+
"rewards/chosen": -1.432652235031128,
|
| 1013 |
+
"rewards/margins": 1.2259643077850342,
|
| 1014 |
+
"rewards/rejected": -2.658616542816162,
|
| 1015 |
"step": 640
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
"epoch": 0.68,
|
| 1019 |
"learning_rate": 1.4006036925609243e-07,
|
| 1020 |
+
"logits/chosen": 1.2402979135513306,
|
| 1021 |
+
"logits/rejected": 2.491854190826416,
|
| 1022 |
+
"logps/chosen": -458.84344482421875,
|
| 1023 |
+
"logps/rejected": -508.12847900390625,
|
| 1024 |
+
"loss": 0.6884,
|
| 1025 |
+
"rewards/accuracies": 0.768750011920929,
|
| 1026 |
+
"rewards/chosen": -1.5252853631973267,
|
| 1027 |
+
"rewards/margins": 0.9278051257133484,
|
| 1028 |
+
"rewards/rejected": -2.4530904293060303,
|
| 1029 |
"step": 650
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
"epoch": 0.69,
|
| 1033 |
"learning_rate": 1.319240907040458e-07,
|
| 1034 |
+
"logits/chosen": 1.5803894996643066,
|
| 1035 |
+
"logits/rejected": 2.081526756286621,
|
| 1036 |
+
"logps/chosen": -454.36309814453125,
|
| 1037 |
+
"logps/rejected": -526.6981201171875,
|
| 1038 |
+
"loss": 0.6824,
|
| 1039 |
+
"rewards/accuracies": 0.75,
|
| 1040 |
+
"rewards/chosen": -1.5280418395996094,
|
| 1041 |
+
"rewards/margins": 1.0684349536895752,
|
| 1042 |
+
"rewards/rejected": -2.5964770317077637,
|
| 1043 |
"step": 660
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
"epoch": 0.7,
|
| 1047 |
"learning_rate": 1.239457282149695e-07,
|
| 1048 |
+
"logits/chosen": 1.9639816284179688,
|
| 1049 |
+
"logits/rejected": 3.2135062217712402,
|
| 1050 |
+
"logps/chosen": -435.85888671875,
|
| 1051 |
+
"logps/rejected": -534.4019775390625,
|
| 1052 |
+
"loss": 0.6736,
|
| 1053 |
+
"rewards/accuracies": 0.737500011920929,
|
| 1054 |
+
"rewards/chosen": -1.5270483493804932,
|
| 1055 |
+
"rewards/margins": 1.1297345161437988,
|
| 1056 |
+
"rewards/rejected": -2.656782627105713,
|
| 1057 |
"step": 670
|
| 1058 |
},
|
| 1059 |
{
|
| 1060 |
"epoch": 0.71,
|
| 1061 |
"learning_rate": 1.1613595214152711e-07,
|
| 1062 |
+
"logits/chosen": 2.091609477996826,
|
| 1063 |
+
"logits/rejected": 2.95405912399292,
|
| 1064 |
+
"logps/chosen": -404.2757873535156,
|
| 1065 |
+
"logps/rejected": -456.81732177734375,
|
| 1066 |
+
"loss": 0.7108,
|
| 1067 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1068 |
+
"rewards/chosen": -1.4280402660369873,
|
| 1069 |
+
"rewards/margins": 0.7583026885986328,
|
| 1070 |
+
"rewards/rejected": -2.186342716217041,
|
| 1071 |
"step": 680
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
"epoch": 0.72,
|
| 1075 |
"learning_rate": 1.0850520736699362e-07,
|
| 1076 |
+
"logits/chosen": 1.712774634361267,
|
| 1077 |
+
"logits/rejected": 3.230499267578125,
|
| 1078 |
+
"logps/chosen": -395.8936767578125,
|
| 1079 |
+
"logps/rejected": -499.32965087890625,
|
| 1080 |
+
"loss": 0.655,
|
| 1081 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 1082 |
+
"rewards/chosen": -1.3443926572799683,
|
| 1083 |
+
"rewards/margins": 1.2300078868865967,
|
| 1084 |
+
"rewards/rejected": -2.5744004249572754,
|
| 1085 |
"step": 690
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
"epoch": 0.73,
|
| 1089 |
"learning_rate": 1.0106369933615042e-07,
|
| 1090 |
+
"logits/chosen": 2.6943306922912598,
|
| 1091 |
+
"logits/rejected": 4.300943851470947,
|
| 1092 |
+
"logps/chosen": -449.5069274902344,
|
| 1093 |
+
"logps/rejected": -574.8884887695312,
|
| 1094 |
+
"loss": 0.6217,
|
| 1095 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 1096 |
+
"rewards/chosen": -1.9530900716781616,
|
| 1097 |
+
"rewards/margins": 1.3479530811309814,
|
| 1098 |
+
"rewards/rejected": -3.3010432720184326,
|
| 1099 |
"step": 700
|
| 1100 |
},
|
| 1101 |
{
|
| 1102 |
"epoch": 0.73,
|
| 1103 |
+
"eval_logits/chosen": 2.593170166015625,
|
| 1104 |
+
"eval_logits/rejected": 3.816297769546509,
|
| 1105 |
+
"eval_logps/chosen": -473.18865966796875,
|
| 1106 |
+
"eval_logps/rejected": -579.6353759765625,
|
| 1107 |
+
"eval_loss": 0.6762357354164124,
|
| 1108 |
+
"eval_rewards/accuracies": 0.7698412537574768,
|
| 1109 |
+
"eval_rewards/chosen": -1.8907891511917114,
|
| 1110 |
+
"eval_rewards/margins": 1.2877792119979858,
|
| 1111 |
+
"eval_rewards/rejected": -3.1785686016082764,
|
| 1112 |
+
"eval_runtime": 244.2095,
|
| 1113 |
+
"eval_samples_per_second": 8.19,
|
| 1114 |
+
"eval_steps_per_second": 0.258,
|
| 1115 |
"step": 700
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.74,
|
| 1119 |
"learning_rate": 9.382138040640714e-08,
|
| 1120 |
+
"logits/chosen": 2.803864002227783,
|
| 1121 |
+
"logits/rejected": 3.797267198562622,
|
| 1122 |
+
"logps/chosen": -505.75677490234375,
|
| 1123 |
+
"logps/rejected": -552.4486083984375,
|
| 1124 |
+
"loss": 0.6429,
|
| 1125 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 1126 |
+
"rewards/chosen": -2.0083601474761963,
|
| 1127 |
+
"rewards/margins": 1.1107932329177856,
|
| 1128 |
+
"rewards/rejected": -3.1191532611846924,
|
| 1129 |
"step": 710
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.75,
|
| 1133 |
"learning_rate": 8.678793653740632e-08,
|
| 1134 |
+
"logits/chosen": 3.4752883911132812,
|
| 1135 |
+
"logits/rejected": 4.102308750152588,
|
| 1136 |
+
"logps/chosen": -513.325439453125,
|
| 1137 |
+
"logps/rejected": -616.204345703125,
|
| 1138 |
+
"loss": 0.6616,
|
| 1139 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 1140 |
+
"rewards/chosen": -2.2190704345703125,
|
| 1141 |
+
"rewards/margins": 1.2776607275009155,
|
| 1142 |
+
"rewards/rejected": -3.4967312812805176,
|
| 1143 |
"step": 720
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.76,
|
| 1147 |
"learning_rate": 7.997277433690983e-08,
|
| 1148 |
+
"logits/chosen": 3.118082046508789,
|
| 1149 |
+
"logits/rejected": 4.068647861480713,
|
| 1150 |
+
"logps/chosen": -515.4818725585938,
|
| 1151 |
+
"logps/rejected": -578.458984375,
|
| 1152 |
+
"loss": 0.6692,
|
| 1153 |
+
"rewards/accuracies": 0.78125,
|
| 1154 |
+
"rewards/chosen": -2.115870475769043,
|
| 1155 |
+
"rewards/margins": 1.1491236686706543,
|
| 1156 |
+
"rewards/rejected": -3.2649941444396973,
|
| 1157 |
"step": 730
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.77,
|
| 1161 |
"learning_rate": 7.338500848029602e-08,
|
| 1162 |
+
"logits/chosen": 3.267651319503784,
|
| 1163 |
+
"logits/rejected": 4.356374263763428,
|
| 1164 |
+
"logps/chosen": -517.941650390625,
|
| 1165 |
+
"logps/rejected": -612.4710693359375,
|
| 1166 |
+
"loss": 0.6626,
|
| 1167 |
+
"rewards/accuracies": 0.831250011920929,
|
| 1168 |
+
"rewards/chosen": -2.0364277362823486,
|
| 1169 |
+
"rewards/margins": 1.4974032640457153,
|
| 1170 |
+
"rewards/rejected": -3.5338311195373535,
|
| 1171 |
"step": 740
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.78,
|
| 1175 |
"learning_rate": 6.70334495204884e-08,
|
| 1176 |
+
"logits/chosen": 2.9188966751098633,
|
| 1177 |
+
"logits/rejected": 3.9505248069763184,
|
| 1178 |
+
"logps/chosen": -495.83734130859375,
|
| 1179 |
+
"logps/rejected": -612.5488891601562,
|
| 1180 |
+
"loss": 0.6476,
|
| 1181 |
+
"rewards/accuracies": 0.762499988079071,
|
| 1182 |
+
"rewards/chosen": -2.124872922897339,
|
| 1183 |
+
"rewards/margins": 1.2153751850128174,
|
| 1184 |
+
"rewards/rejected": -3.3402485847473145,
|
| 1185 |
"step": 750
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.8,
|
| 1189 |
"learning_rate": 6.092659210462231e-08,
|
| 1190 |
+
"logits/chosen": 3.145782709121704,
|
| 1191 |
+
"logits/rejected": 3.7394192218780518,
|
| 1192 |
+
"logps/chosen": -498.886474609375,
|
| 1193 |
+
"logps/rejected": -587.1813354492188,
|
| 1194 |
+
"loss": 0.6186,
|
| 1195 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 1196 |
+
"rewards/chosen": -2.2077949047088623,
|
| 1197 |
+
"rewards/margins": 1.1572462320327759,
|
| 1198 |
+
"rewards/rejected": -3.3650412559509277,
|
| 1199 |
"step": 760
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.81,
|
| 1203 |
"learning_rate": 5.507260361320737e-08,
|
| 1204 |
+
"logits/chosen": 3.364577531814575,
|
| 1205 |
+
"logits/rejected": 4.243520259857178,
|
| 1206 |
+
"logps/chosen": -541.8561401367188,
|
| 1207 |
+
"logps/rejected": -659.7272338867188,
|
| 1208 |
+
"loss": 0.6917,
|
| 1209 |
+
"rewards/accuracies": 0.731249988079071,
|
| 1210 |
+
"rewards/chosen": -2.4785799980163574,
|
| 1211 |
+
"rewards/margins": 1.123822569847107,
|
| 1212 |
+
"rewards/rejected": -3.602402448654175,
|
| 1213 |
"step": 770
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.82,
|
| 1217 |
"learning_rate": 4.947931323697982e-08,
|
| 1218 |
+
"logits/chosen": 3.392416477203369,
|
| 1219 |
+
"logits/rejected": 4.268471717834473,
|
| 1220 |
+
"logps/chosen": -480.16455078125,
|
| 1221 |
+
"logps/rejected": -572.0571899414062,
|
| 1222 |
+
"loss": 0.6843,
|
| 1223 |
"rewards/accuracies": 0.706250011920929,
|
| 1224 |
+
"rewards/chosen": -2.2396440505981445,
|
| 1225 |
+
"rewards/margins": 1.116714596748352,
|
| 1226 |
+
"rewards/rejected": -3.356358766555786,
|
| 1227 |
"step": 780
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.83,
|
| 1231 |
"learning_rate": 4.415420150605398e-08,
|
| 1232 |
+
"logits/chosen": 2.9198615550994873,
|
| 1233 |
+
"logits/rejected": 3.857909679412842,
|
| 1234 |
+
"logps/chosen": -536.2203369140625,
|
| 1235 |
+
"logps/rejected": -663.2730102539062,
|
| 1236 |
+
"loss": 0.6706,
|
| 1237 |
+
"rewards/accuracies": 0.8062499761581421,
|
| 1238 |
+
"rewards/chosen": -2.358766794204712,
|
| 1239 |
+
"rewards/margins": 1.428501844406128,
|
| 1240 |
+
"rewards/rejected": -3.787268877029419,
|
| 1241 |
"step": 790
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.84,
|
| 1245 |
"learning_rate": 3.9104390285376374e-08,
|
| 1246 |
+
"logits/chosen": 2.381801128387451,
|
| 1247 |
+
"logits/rejected": 4.157925605773926,
|
| 1248 |
+
"logps/chosen": -550.6214599609375,
|
| 1249 |
+
"logps/rejected": -629.8056030273438,
|
| 1250 |
+
"loss": 0.6418,
|
| 1251 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 1252 |
+
"rewards/chosen": -2.0941646099090576,
|
| 1253 |
+
"rewards/margins": 1.3854446411132812,
|
| 1254 |
+
"rewards/rejected": -3.4796090126037598,
|
| 1255 |
"step": 800
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.84,
|
| 1259 |
+
"eval_logits/chosen": 2.6092491149902344,
|
| 1260 |
+
"eval_logits/rejected": 3.865464448928833,
|
| 1261 |
+
"eval_logps/chosen": -494.04217529296875,
|
| 1262 |
+
"eval_logps/rejected": -602.0606689453125,
|
| 1263 |
+
"eval_loss": 0.6711603403091431,
|
| 1264 |
"eval_rewards/accuracies": 0.7678571343421936,
|
| 1265 |
+
"eval_rewards/chosen": -2.0993239879608154,
|
| 1266 |
+
"eval_rewards/margins": 1.3034968376159668,
|
| 1267 |
+
"eval_rewards/rejected": -3.4028208255767822,
|
| 1268 |
+
"eval_runtime": 243.1214,
|
| 1269 |
+
"eval_samples_per_second": 8.226,
|
| 1270 |
"eval_steps_per_second": 0.259,
|
| 1271 |
"step": 800
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
"epoch": 0.85,
|
| 1275 |
"learning_rate": 3.433663324986208e-08,
|
| 1276 |
+
"logits/chosen": 3.123400926589966,
|
| 1277 |
+
"logits/rejected": 4.341358184814453,
|
| 1278 |
+
"logps/chosen": -507.13848876953125,
|
| 1279 |
+
"logps/rejected": -566.4569091796875,
|
| 1280 |
+
"loss": 0.6669,
|
| 1281 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 1282 |
+
"rewards/chosen": -2.2766828536987305,
|
| 1283 |
+
"rewards/margins": 0.9975360035896301,
|
| 1284 |
+
"rewards/rejected": -3.274219036102295,
|
| 1285 |
"step": 810
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 0.86,
|
| 1289 |
"learning_rate": 2.9857306851953897e-08,
|
| 1290 |
+
"logits/chosen": 3.36864972114563,
|
| 1291 |
+
"logits/rejected": 3.826308488845825,
|
| 1292 |
+
"logps/chosen": -454.4519958496094,
|
| 1293 |
+
"logps/rejected": -553.3275756835938,
|
| 1294 |
+
"loss": 0.7048,
|
| 1295 |
"rewards/accuracies": 0.762499988079071,
|
| 1296 |
+
"rewards/chosen": -2.0178139209747314,
|
| 1297 |
+
"rewards/margins": 1.1636488437652588,
|
| 1298 |
+
"rewards/rejected": -3.1814627647399902,
|
| 1299 |
"step": 820
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
"epoch": 0.87,
|
| 1303 |
"learning_rate": 2.567240179368185e-08,
|
| 1304 |
+
"logits/chosen": 2.7587532997131348,
|
| 1305 |
+
"logits/rejected": 3.9622387886047363,
|
| 1306 |
+
"logps/chosen": -461.407470703125,
|
| 1307 |
+
"logps/rejected": -577.1260986328125,
|
| 1308 |
+
"loss": 0.6538,
|
| 1309 |
+
"rewards/accuracies": 0.78125,
|
| 1310 |
+
"rewards/chosen": -2.125457286834717,
|
| 1311 |
+
"rewards/margins": 1.223615050315857,
|
| 1312 |
+
"rewards/rejected": -3.3490726947784424,
|
| 1313 |
"step": 830
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
"epoch": 0.88,
|
| 1317 |
"learning_rate": 2.1787515014630357e-08,
|
| 1318 |
+
"logits/chosen": 3.194129228591919,
|
| 1319 |
+
"logits/rejected": 3.398770809173584,
|
| 1320 |
+
"logps/chosen": -526.1636962890625,
|
| 1321 |
+
"logps/rejected": -602.7716064453125,
|
| 1322 |
+
"loss": 0.7025,
|
| 1323 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 1324 |
+
"rewards/chosen": -2.1302103996276855,
|
| 1325 |
+
"rewards/margins": 1.0573285818099976,
|
| 1326 |
+
"rewards/rejected": -3.1875391006469727,
|
| 1327 |
"step": 840
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
"epoch": 0.89,
|
| 1331 |
"learning_rate": 1.820784220652766e-08,
|
| 1332 |
+
"logits/chosen": 2.8333277702331543,
|
| 1333 |
+
"logits/rejected": 3.949988842010498,
|
| 1334 |
+
"logps/chosen": -497.5380859375,
|
| 1335 |
+
"logps/rejected": -547.4320068359375,
|
| 1336 |
+
"loss": 0.6689,
|
| 1337 |
"rewards/accuracies": 0.7437499761581421,
|
| 1338 |
+
"rewards/chosen": -1.9726041555404663,
|
| 1339 |
+
"rewards/margins": 1.0516941547393799,
|
| 1340 |
+
"rewards/rejected": -3.0242981910705566,
|
| 1341 |
"step": 850
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
"epoch": 0.9,
|
| 1345 |
"learning_rate": 1.4938170864468636e-08,
|
| 1346 |
+
"logits/chosen": 2.8082051277160645,
|
| 1347 |
+
"logits/rejected": 4.054238796234131,
|
| 1348 |
+
"logps/chosen": -488.4153747558594,
|
| 1349 |
+
"logps/rejected": -592.6768798828125,
|
| 1350 |
+
"loss": 0.6703,
|
| 1351 |
+
"rewards/accuracies": 0.731249988079071,
|
| 1352 |
+
"rewards/chosen": -2.0998573303222656,
|
| 1353 |
+
"rewards/margins": 1.3178421258926392,
|
| 1354 |
+
"rewards/rejected": -3.4176993370056152,
|
| 1355 |
"step": 860
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 0.91,
|
| 1359 |
"learning_rate": 1.1982873884064465e-08,
|
| 1360 |
+
"logits/chosen": 2.500764846801758,
|
| 1361 |
+
"logits/rejected": 3.581740140914917,
|
| 1362 |
+
"logps/chosen": -425.3038024902344,
|
| 1363 |
+
"logps/rejected": -571.4251708984375,
|
| 1364 |
+
"loss": 0.6579,
|
| 1365 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 1366 |
+
"rewards/chosen": -1.7831649780273438,
|
| 1367 |
+
"rewards/margins": 1.468294382095337,
|
| 1368 |
+
"rewards/rejected": -3.2514591217041016,
|
| 1369 |
"step": 870
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.92,
|
| 1373 |
"learning_rate": 9.345903713082304e-09,
|
| 1374 |
+
"logits/chosen": 2.309072971343994,
|
| 1375 |
+
"logits/rejected": 3.692427158355713,
|
| 1376 |
+
"logps/chosen": -488.32049560546875,
|
| 1377 |
+
"logps/rejected": -586.3157958984375,
|
| 1378 |
+
"loss": 0.6445,
|
| 1379 |
+
"rewards/accuracies": 0.737500011920929,
|
| 1380 |
+
"rewards/chosen": -2.1420960426330566,
|
| 1381 |
+
"rewards/margins": 1.187731146812439,
|
| 1382 |
+
"rewards/rejected": -3.329827070236206,
|
| 1383 |
"step": 880
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 0.93,
|
| 1387 |
"learning_rate": 7.030787065396865e-09,
|
| 1388 |
+
"logits/chosen": 2.455681800842285,
|
| 1389 |
+
"logits/rejected": 3.950096607208252,
|
| 1390 |
+
"logps/chosen": -487.016357421875,
|
| 1391 |
+
"logps/rejected": -593.2111206054688,
|
| 1392 |
+
"loss": 0.701,
|
| 1393 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 1394 |
+
"rewards/chosen": -2.2053864002227783,
|
| 1395 |
+
"rewards/margins": 1.07200026512146,
|
| 1396 |
+
"rewards/rejected": -3.277386426925659,
|
| 1397 |
"step": 890
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 0.94,
|
| 1401 |
"learning_rate": 5.04062020432286e-09,
|
| 1402 |
+
"logits/chosen": 2.607274293899536,
|
| 1403 |
+
"logits/rejected": 3.9918441772460938,
|
| 1404 |
+
"logps/chosen": -505.3465881347656,
|
| 1405 |
+
"logps/rejected": -603.4442138671875,
|
| 1406 |
+
"loss": 0.6678,
|
| 1407 |
"rewards/accuracies": 0.71875,
|
| 1408 |
+
"rewards/chosen": -2.0696864128112793,
|
| 1409 |
+
"rewards/margins": 1.1221836805343628,
|
| 1410 |
+
"rewards/rejected": -3.1918704509735107,
|
| 1411 |
"step": 900
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 0.94,
|
| 1415 |
+
"eval_logits/chosen": 2.4517972469329834,
|
| 1416 |
+
"eval_logits/rejected": 3.733224630355835,
|
| 1417 |
+
"eval_logps/chosen": -487.1844482421875,
|
| 1418 |
+
"eval_logps/rejected": -594.1102905273438,
|
| 1419 |
+
"eval_loss": 0.6716480851173401,
|
| 1420 |
+
"eval_rewards/accuracies": 0.7638888955116272,
|
| 1421 |
+
"eval_rewards/chosen": -2.030747175216675,
|
| 1422 |
+
"eval_rewards/margins": 1.2925708293914795,
|
| 1423 |
+
"eval_rewards/rejected": -3.323317766189575,
|
| 1424 |
+
"eval_runtime": 244.2155,
|
| 1425 |
+
"eval_samples_per_second": 8.189,
|
| 1426 |
+
"eval_steps_per_second": 0.258,
|
| 1427 |
"step": 900
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 0.95,
|
| 1431 |
"learning_rate": 3.3780648016376866e-09,
|
| 1432 |
+
"logits/chosen": 2.913886308670044,
|
| 1433 |
+
"logits/rejected": 4.131613731384277,
|
| 1434 |
+
"logps/chosen": -459.2386779785156,
|
| 1435 |
+
"logps/rejected": -578.1640014648438,
|
| 1436 |
+
"loss": 0.6622,
|
| 1437 |
+
"rewards/accuracies": 0.793749988079071,
|
| 1438 |
+
"rewards/chosen": -2.1169772148132324,
|
| 1439 |
+
"rewards/margins": 1.2496932744979858,
|
| 1440 |
+
"rewards/rejected": -3.3666698932647705,
|
| 1441 |
"step": 910
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 0.96,
|
| 1445 |
"learning_rate": 2.0453443778310766e-09,
|
| 1446 |
+
"logits/chosen": 2.6672825813293457,
|
| 1447 |
+
"logits/rejected": 3.8567919731140137,
|
| 1448 |
+
"logps/chosen": -503.61395263671875,
|
| 1449 |
+
"logps/rejected": -597.9525756835938,
|
| 1450 |
+
"loss": 0.6413,
|
| 1451 |
"rewards/accuracies": 0.8187500238418579,
|
| 1452 |
+
"rewards/chosen": -2.079343557357788,
|
| 1453 |
+
"rewards/margins": 1.3192239999771118,
|
| 1454 |
+
"rewards/rejected": -3.3985676765441895,
|
| 1455 |
"step": 920
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
"epoch": 0.97,
|
| 1459 |
"learning_rate": 1.0442413283435758e-09,
|
| 1460 |
+
"logits/chosen": 2.472501754760742,
|
| 1461 |
+
"logits/rejected": 3.5483956336975098,
|
| 1462 |
+
"logps/chosen": -481.59515380859375,
|
| 1463 |
+
"logps/rejected": -590.560791015625,
|
| 1464 |
+
"loss": 0.6391,
|
| 1465 |
+
"rewards/accuracies": 0.8125,
|
| 1466 |
+
"rewards/chosen": -2.0882575511932373,
|
| 1467 |
+
"rewards/margins": 1.377394676208496,
|
| 1468 |
+
"rewards/rejected": -3.4656529426574707,
|
| 1469 |
"step": 930
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
"epoch": 0.98,
|
| 1473 |
"learning_rate": 3.760945397705828e-10,
|
| 1474 |
+
"logits/chosen": 2.6073191165924072,
|
| 1475 |
+
"logits/rejected": 3.743886947631836,
|
| 1476 |
+
"logps/chosen": -536.0969848632812,
|
| 1477 |
+
"logps/rejected": -624.2562255859375,
|
| 1478 |
+
"loss": 0.6442,
|
| 1479 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 1480 |
+
"rewards/chosen": -2.063697338104248,
|
| 1481 |
+
"rewards/margins": 1.2653493881225586,
|
| 1482 |
+
"rewards/rejected": -3.3290467262268066,
|
| 1483 |
"step": 940
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
"epoch": 0.99,
|
| 1487 |
"learning_rate": 4.17975992204056e-11,
|
| 1488 |
+
"logits/chosen": 2.321948766708374,
|
| 1489 |
+
"logits/rejected": 3.977466583251953,
|
| 1490 |
+
"logps/chosen": -496.20025634765625,
|
| 1491 |
+
"logps/rejected": -609.7675170898438,
|
| 1492 |
+
"loss": 0.6933,
|
| 1493 |
+
"rewards/accuracies": 0.768750011920929,
|
| 1494 |
+
"rewards/chosen": -2.116821527481079,
|
| 1495 |
+
"rewards/margins": 1.348793387413025,
|
| 1496 |
+
"rewards/rejected": -3.4656143188476562,
|
| 1497 |
"step": 950
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 1.0,
|
| 1501 |
"step": 955,
|
| 1502 |
"total_flos": 0.0,
|
| 1503 |
+
"train_loss": 0.7263204834224042,
|
| 1504 |
+
"train_runtime": 20734.7169,
|
| 1505 |
+
"train_samples_per_second": 2.948,
|
| 1506 |
+
"train_steps_per_second": 0.046
|
| 1507 |
}
|
| 1508 |
],
|
| 1509 |
"logging_steps": 10,
|
| 1510 |
"max_steps": 955,
|
| 1511 |
"num_input_tokens_seen": 0,
|
| 1512 |
"num_train_epochs": 1,
|
| 1513 |
+
"save_steps": 100000000,
|
| 1514 |
"total_flos": 0.0,
|
| 1515 |
"train_batch_size": 8,
|
| 1516 |
"trial_name": null,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96176aeb701ce092111002b23ff886e7082dc5b7b988ccce9913225c1f4e42a5
|
| 3 |
+
size 6008
|