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_fUNL_entropy
|
| 17 |
results: []
|
|
@@ -22,18 +16,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 22 |
|
| 23 |
# qwen_fUNL_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_fUNL_entropy
|
| 11 |
results: []
|
|
|
|
| 16 |
|
| 17 |
# qwen_fUNL_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: -42.7794
|
| 23 |
+
- Rewards/rejected: -43.9149
|
| 24 |
+
- Rewards/accuracies: 0.5668
|
| 25 |
+
- Rewards/margins: 1.1356
|
| 26 |
+
- Logps/rejected: -43.9149
|
| 27 |
+
- Logps/chosen: -42.7794
|
| 28 |
+
- Logits/rejected: 7.2567
|
| 29 |
+
- Logits/chosen: 7.5393
|
|
|
|
| 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 | -29.4740 | -31.2238 | 0.5690 | 1.7498 | -31.2238 | -29.4740 | 4.5258 | 4.5083 |
|
| 65 |
+
| 0.0 | 0.4282 | 800 | 0.0000 | -37.0465 | -38.5277 | 0.5579 | 1.4811 | -38.5277 | -37.0465 | 6.2307 | 6.3643 |
|
| 66 |
+
| 0.001 | 0.6422 | 1200 | 0.0000 | -38.7942 | -40.1267 | 0.5668 | 1.3324 | -40.1267 | -38.7942 | 6.5149 | 6.7000 |
|
| 67 |
+
| 0.0 | 0.8563 | 1600 | 0.0000 | -38.5913 | -40.0107 | 0.5668 | 1.4194 | -40.0107 | -38.5913 | 6.5708 | 6.7471 |
|
| 68 |
+
| 0.0 | 1.0704 | 2000 | 0.0000 | -40.7799 | -42.0174 | 0.5675 | 1.2374 | -42.0174 | -40.7799 | 7.0075 | 7.2451 |
|
| 69 |
+
| 0.0 | 1.2845 | 2400 | 0.0000 | -40.9809 | -42.2090 | 0.5645 | 1.2280 | -42.2090 | -40.9809 | 6.9425 | 7.1883 |
|
| 70 |
+
| 0.0 | 1.4986 | 2800 | 0.0000 | -41.7185 | -42.9016 | 0.5631 | 1.1831 | -42.9016 | -41.7185 | 7.2071 | 7.4629 |
|
| 71 |
+
| 0.0 | 1.7127 | 3200 | 0.0000 | -41.7373 | -42.9487 | 0.5675 | 1.2115 | -42.9487 | -41.7373 | 7.0907 | 7.3464 |
|
| 72 |
+
| 0.0 | 1.9267 | 3600 | 0.0000 | -42.3165 | -43.4863 | 0.5668 | 1.1698 | -43.4863 | -42.3165 | 7.2080 | 7.4815 |
|
| 73 |
+
| 0.0 | 2.1408 | 4000 | 0.0000 | -43.0385 | -44.1473 | 0.5697 | 1.1088 | -44.1473 | -43.0385 | 7.2552 | 7.5548 |
|
| 74 |
+
| 0.0 | 2.3549 | 4400 | 0.0000 | -42.9448 | -44.0525 | 0.5705 | 1.1077 | -44.0525 | -42.9448 | 7.2918 | 7.5836 |
|
| 75 |
+
| 0.0 | 2.5690 | 4800 | 0.0000 | -43.0768 | -44.1767 | 0.5675 | 1.0999 | -44.1767 | -43.0768 | 7.3794 | 7.6690 |
|
| 76 |
+
| 0.0 | 2.7831 | 5200 | 0.0000 | -43.1227 | -44.2291 | 0.5690 | 1.1064 | -44.2291 | -43.1227 | 7.2960 | 7.5933 |
|
| 77 |
+
| 0.0 | 2.9972 | 5600 | 0.0000 | -42.7794 | -43.9149 | 0.5668 | 1.1356 | -43.9149 | -42.7794 | 7.2567 | 7.5393 |
|
| 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": 34.2578,
|
| 13 |
-
"eval_samples": 1345,
|
| 14 |
-
"eval_samples_per_second": 39.261,
|
| 15 |
-
"eval_semantic_entropy": NaN,
|
| 16 |
-
"eval_steps_per_second": 9.837,
|
| 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": 0.010826070491851839,
|
| 5 |
+
"train_runtime": 30412.7115,
|
| 6 |
"train_samples": 59790,
|
| 7 |
+
"train_samples_per_second": 5.898,
|
| 8 |
+
"train_steps_per_second": 0.184
|
| 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": 0.010826070491851839,
|
| 5 |
+
"train_runtime": 30412.7115,
|
| 6 |
"train_samples": 59790,
|
| 7 |
+
"train_samples_per_second": 5.898,
|
| 8 |
+
"train_steps_per_second": 0.184
|
| 9 |
}
|
trainer_state.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|