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_cfUNL_entropy
|
| 17 |
results: []
|
|
@@ -22,18 +16,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 22 |
|
| 23 |
# qwen_cfUNL_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 |
-
| 0.0 | 0.2141 | 400 |
|
| 72 |
-
| 0.0 | 0.4282 | 800 |
|
| 73 |
-
| 0.
|
| 74 |
-
| 0.0 | 0.8563 | 1600 |
|
| 75 |
-
| 0.0 | 1.0704 | 2000 |
|
| 76 |
-
| 0.0 | 1.2845 | 2400 |
|
| 77 |
-
| 0.0 | 1.4986 | 2800 |
|
| 78 |
-
| 0.0 | 1.7127 | 3200 |
|
| 79 |
-
| 0.0 | 1.9267 | 3600 |
|
| 80 |
-
| 0.0 | 2.1408 | 4000 |
|
| 81 |
-
| 0.0 | 2.3549 | 4400 |
|
| 82 |
-
| 0.0 | 2.5690 | 4800 |
|
| 83 |
-
| 0.0 | 2.7831 | 5200 |
|
| 84 |
-
| 0.0 | 2.9972 | 5600 |
|
| 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_cfUNL_entropy
|
| 11 |
results: []
|
|
|
|
| 16 |
|
| 17 |
# qwen_cfUNL_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: 0.0000
|
| 22 |
+
- Rewards/chosen: -45.0891
|
| 23 |
+
- Rewards/rejected: -46.1094
|
| 24 |
+
- Rewards/accuracies: 0.5682
|
| 25 |
+
- Rewards/margins: 1.0204
|
| 26 |
+
- Logps/rejected: -46.1094
|
| 27 |
+
- Logps/chosen: -45.0891
|
| 28 |
+
- Logits/rejected: 7.4245
|
| 29 |
+
- Logits/chosen: 7.7499
|
|
|
|
| 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 |
+
| 0.0 | 0.2141 | 400 | 0.0001 | -30.9341 | -32.7194 | 0.5697 | 1.7853 | -32.7194 | -30.9341 | 4.5291 | 4.5264 |
|
| 65 |
+
| 0.0 | 0.4282 | 800 | 0.0000 | -38.7534 | -40.2065 | 0.5593 | 1.4531 | -40.2065 | -38.7534 | 6.2341 | 6.3877 |
|
| 66 |
+
| 0.0009 | 0.6422 | 1200 | 0.0000 | -38.5460 | -39.9578 | 0.5512 | 1.4119 | -39.9578 | -38.5460 | 6.1244 | 6.2779 |
|
| 67 |
+
| 0.0 | 0.8563 | 1600 | 0.0000 | -40.0222 | -41.4115 | 0.5690 | 1.3893 | -41.4115 | -40.0222 | 6.5494 | 6.7346 |
|
| 68 |
+
| 0.0 | 1.0704 | 2000 | 0.0000 | -43.0566 | -44.2275 | 0.5653 | 1.1709 | -44.2275 | -43.0566 | 7.0818 | 7.3504 |
|
| 69 |
+
| 0.0 | 1.2845 | 2400 | 0.0000 | -43.5288 | -44.6477 | 0.5645 | 1.1189 | -44.6477 | -43.5288 | 7.1882 | 7.4775 |
|
| 70 |
+
| 0.0 | 1.4986 | 2800 | 0.0000 | -43.7383 | -44.8584 | 0.5660 | 1.1201 | -44.8584 | -43.7383 | 7.1745 | 7.4634 |
|
| 71 |
+
| 0.0 | 1.7127 | 3200 | 0.0000 | -44.4950 | -45.5556 | 0.5638 | 1.0605 | -45.5556 | -44.4950 | 7.2848 | 7.5950 |
|
| 72 |
+
| 0.0 | 1.9267 | 3600 | 0.0000 | -44.5958 | -45.6569 | 0.5645 | 1.0611 | -45.6569 | -44.5958 | 7.2814 | 7.5948 |
|
| 73 |
+
| 0.0 | 2.1408 | 4000 | 0.0000 | -44.8271 | -45.8411 | 0.5668 | 1.0140 | -45.8411 | -44.8271 | 7.4235 | 7.7436 |
|
| 74 |
+
| 0.0 | 2.3549 | 4400 | 0.0000 | -45.1344 | -46.1374 | 0.5653 | 1.0030 | -46.1374 | -45.1344 | 7.3526 | 7.6831 |
|
| 75 |
+
| 0.0 | 2.5690 | 4800 | 0.0000 | -45.0201 | -46.0501 | 0.5653 | 1.0300 | -46.0501 | -45.0201 | 7.3843 | 7.7103 |
|
| 76 |
+
| 0.0 | 2.7831 | 5200 | 0.0000 | -45.3432 | -46.3394 | 0.5653 | 0.9961 | -46.3394 | -45.3432 | 7.4499 | 7.7830 |
|
| 77 |
+
| 0.0 | 2.9972 | 5600 | 0.0000 | -45.0891 | -46.1094 | 0.5682 | 1.0204 | -46.1094 | -45.0891 | 7.4245 | 7.7499 |
|
| 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": 34.28,
|
| 13 |
-
"eval_samples": 1345,
|
| 14 |
-
"eval_samples_per_second": 39.236,
|
| 15 |
-
"eval_semantic_entropy": NaN,
|
| 16 |
-
"eval_steps_per_second": 9.831,
|
| 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": 0.00884905898078017,
|
| 5 |
+
"train_runtime": 29905.5175,
|
| 6 |
"train_samples": 59790,
|
| 7 |
+
"train_samples_per_second": 5.998,
|
| 8 |
+
"train_steps_per_second": 0.187
|
| 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": 0.00884905898078017,
|
| 5 |
+
"train_runtime": 29905.5175,
|
| 6 |
"train_samples": 59790,
|
| 7 |
+
"train_samples_per_second": 5.998,
|
| 8 |
+
"train_steps_per_second": 0.187
|
| 9 |
}
|
trainer_state.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|