Model save
Browse files- README.md +26 -33
- all_results.json +4 -18
- train_results.json +4 -4
- trainer_state.json +0 -0
README.md
CHANGED
|
@@ -3,15 +3,9 @@ library_name: transformers
|
|
| 3 |
license: other
|
| 4 |
base_model: trl-lib/qwen1.5-0.5b-sft
|
| 5 |
tags:
|
| 6 |
-
- alignment-handbook
|
| 7 |
- trl
|
| 8 |
- simpo
|
| 9 |
- generated_from_trainer
|
| 10 |
-
- trl
|
| 11 |
-
- simpo
|
| 12 |
-
- generated_from_trainer
|
| 13 |
-
datasets:
|
| 14 |
-
- yakazimir/ultrafeedback_binarized
|
| 15 |
model-index:
|
| 16 |
- name: qwen_qfUNL_entropy
|
| 17 |
results: []
|
|
@@ -22,18 +16,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 22 |
|
| 23 |
# qwen_qfUNL_entropy
|
| 24 |
|
| 25 |
-
This model is a fine-tuned version of [trl-lib/qwen1.5-0.5b-sft](https://huggingface.co/trl-lib/qwen1.5-0.5b-sft) on
|
| 26 |
It achieves the following results on the evaluation set:
|
| 27 |
-
- Loss:
|
| 28 |
-
- Rewards/chosen:
|
| 29 |
-
- Rewards/rejected:
|
| 30 |
-
- Rewards/accuracies: 0.
|
| 31 |
-
- Rewards/margins:
|
| 32 |
-
- Logps/rejected:
|
| 33 |
-
- Logps/chosen:
|
| 34 |
-
- Logits/rejected:
|
| 35 |
-
- Logits/chosen:
|
| 36 |
-
- Semantic Entropy: nan
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
@@ -66,22 +59,22 @@ The following hyperparameters were used during training:
|
|
| 66 |
|
| 67 |
### Training results
|
| 68 |
|
| 69 |
-
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 70 |
-
|
| 71 |
-
| 0.
|
| 72 |
-
| 0.
|
| 73 |
-
| 0.
|
| 74 |
-
| 0.
|
| 75 |
-
| 0.
|
| 76 |
-
| 0.
|
| 77 |
-
| 0.
|
| 78 |
-
| 0.
|
| 79 |
-
| 0.
|
| 80 |
-
| 0.
|
| 81 |
-
| 0.
|
| 82 |
-
| 0.
|
| 83 |
-
| 0.
|
| 84 |
-
| 0.
|
| 85 |
|
| 86 |
|
| 87 |
### Framework versions
|
|
|
|
| 3 |
license: other
|
| 4 |
base_model: trl-lib/qwen1.5-0.5b-sft
|
| 5 |
tags:
|
|
|
|
| 6 |
- trl
|
| 7 |
- simpo
|
| 8 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
model-index:
|
| 10 |
- name: qwen_qfUNL_entropy
|
| 11 |
results: []
|
|
|
|
| 16 |
|
| 17 |
# qwen_qfUNL_entropy
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [trl-lib/qwen1.5-0.5b-sft](https://huggingface.co/trl-lib/qwen1.5-0.5b-sft) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.6510
|
| 22 |
+
- Rewards/chosen: -1.7989
|
| 23 |
+
- Rewards/rejected: -2.5830
|
| 24 |
+
- Rewards/accuracies: 0.6736
|
| 25 |
+
- Rewards/margins: 0.7841
|
| 26 |
+
- Logps/rejected: -2.5830
|
| 27 |
+
- Logps/chosen: -1.7989
|
| 28 |
+
- Logits/rejected: 0.0192
|
| 29 |
+
- Logits/chosen: -0.0604
|
|
|
|
| 30 |
|
| 31 |
## Model description
|
| 32 |
|
|
|
|
| 59 |
|
| 60 |
### Training results
|
| 61 |
|
| 62 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 63 |
+
|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 64 |
+
| 0.6781 | 0.2141 | 400 | 0.6873 | -1.6444 | -1.8233 | 0.5475 | 0.1789 | -1.8233 | -1.6444 | 0.2857 | 0.1996 |
|
| 65 |
+
| 0.6757 | 0.4282 | 800 | 0.6641 | -1.6348 | -1.9815 | 0.6239 | 0.3467 | -1.9815 | -1.6348 | 0.3665 | 0.2730 |
|
| 66 |
+
| 0.6569 | 0.6422 | 1200 | 0.6602 | -1.7060 | -2.1644 | 0.6424 | 0.4584 | -2.1644 | -1.7060 | 0.2601 | 0.1749 |
|
| 67 |
+
| 0.6562 | 0.8563 | 1600 | 0.6584 | -1.8368 | -2.3836 | 0.6513 | 0.5468 | -2.3836 | -1.8368 | 0.1796 | 0.0944 |
|
| 68 |
+
| 0.6883 | 1.0704 | 2000 | 0.6545 | -1.7098 | -2.2986 | 0.6639 | 0.5888 | -2.2986 | -1.7098 | 0.2146 | 0.1248 |
|
| 69 |
+
| 0.6581 | 1.2845 | 2400 | 0.6533 | -1.7444 | -2.3861 | 0.6691 | 0.6417 | -2.3861 | -1.7444 | 0.1530 | 0.0644 |
|
| 70 |
+
| 0.6444 | 1.4986 | 2800 | 0.6537 | -1.7815 | -2.4833 | 0.6684 | 0.7018 | -2.4833 | -1.7815 | 0.0665 | -0.0145 |
|
| 71 |
+
| 0.6575 | 1.7127 | 3200 | 0.6520 | -1.7922 | -2.5114 | 0.6654 | 0.7192 | -2.5114 | -1.7922 | 0.1107 | 0.0260 |
|
| 72 |
+
| 0.6481 | 1.9267 | 3600 | 0.6507 | -1.7358 | -2.4632 | 0.6736 | 0.7275 | -2.4632 | -1.7358 | 0.0939 | 0.0113 |
|
| 73 |
+
| 0.607 | 2.1408 | 4000 | 0.6506 | -1.7686 | -2.5161 | 0.6751 | 0.7475 | -2.5161 | -1.7686 | 0.0842 | 0.0005 |
|
| 74 |
+
| 0.6294 | 2.3549 | 4400 | 0.6514 | -1.8215 | -2.5986 | 0.6714 | 0.7771 | -2.5986 | -1.8215 | 0.0008 | -0.0778 |
|
| 75 |
+
| 0.6098 | 2.5690 | 4800 | 0.6507 | -1.7918 | -2.5693 | 0.6766 | 0.7775 | -2.5693 | -1.7918 | 0.0735 | -0.0103 |
|
| 76 |
+
| 0.6302 | 2.7831 | 5200 | 0.6507 | -1.7943 | -2.5780 | 0.6751 | 0.7837 | -2.5780 | -1.7943 | 0.0395 | -0.0418 |
|
| 77 |
+
| 0.6181 | 2.9972 | 5600 | 0.6510 | -1.7989 | -2.5830 | 0.6736 | 0.7841 | -2.5830 | -1.7989 | 0.0192 | -0.0604 |
|
| 78 |
|
| 79 |
|
| 80 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,23 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 2.999297541394882,
|
| 3 |
-
"eval_logits/chosen": NaN,
|
| 4 |
-
"eval_logits/rejected": NaN,
|
| 5 |
-
"eval_logps/chosen": NaN,
|
| 6 |
-
"eval_logps/rejected": NaN,
|
| 7 |
-
"eval_loss": NaN,
|
| 8 |
-
"eval_rewards/accuracies": 0.0,
|
| 9 |
-
"eval_rewards/chosen": NaN,
|
| 10 |
-
"eval_rewards/margins": NaN,
|
| 11 |
-
"eval_rewards/rejected": NaN,
|
| 12 |
-
"eval_runtime": 34.2826,
|
| 13 |
-
"eval_samples": 1345,
|
| 14 |
-
"eval_samples_per_second": 39.233,
|
| 15 |
-
"eval_semantic_entropy": NaN,
|
| 16 |
-
"eval_steps_per_second": 9.83,
|
| 17 |
"total_flos": 0.0,
|
| 18 |
-
"train_loss": 0.
|
| 19 |
-
"train_runtime":
|
| 20 |
"train_samples": 59790,
|
| 21 |
-
"train_samples_per_second":
|
| 22 |
-
"train_steps_per_second": 0.
|
| 23 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 2.999297541394882,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.6513626721822561,
|
| 5 |
+
"train_runtime": 30253.6444,
|
| 6 |
"train_samples": 59790,
|
| 7 |
+
"train_samples_per_second": 5.929,
|
| 8 |
+
"train_steps_per_second": 0.185
|
| 9 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 2.999297541394882,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 59790,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 2.999297541394882,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.6513626721822561,
|
| 5 |
+
"train_runtime": 30253.6444,
|
| 6 |
"train_samples": 59790,
|
| 7 |
+
"train_samples_per_second": 5.929,
|
| 8 |
+
"train_steps_per_second": 0.185
|
| 9 |
}
|
trainer_state.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|