End of training
Browse files
README.md
CHANGED
|
@@ -20,6 +20,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 20 |
# ADL_Gemma3
|
| 21 |
|
| 22 |
This model is a fine-tuned version of [zake7749/gemma-2-2b-it-chinese-kyara-dpo](https://huggingface.co/zake7749/gemma-2-2b-it-chinese-kyara-dpo) on an unknown dataset.
|
|
|
|
|
|
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
@@ -46,6 +48,16 @@ The following hyperparameters were used during training:
|
|
| 46 |
- lr_scheduler_type: cosine
|
| 47 |
- num_epochs: 4
|
| 48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
### Framework versions
|
| 50 |
|
| 51 |
- PEFT 0.13.2
|
|
|
|
| 20 |
# ADL_Gemma3
|
| 21 |
|
| 22 |
This model is a fine-tuned version of [zake7749/gemma-2-2b-it-chinese-kyara-dpo](https://huggingface.co/zake7749/gemma-2-2b-it-chinese-kyara-dpo) on an unknown dataset.
|
| 23 |
+
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 1.5088
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
|
|
| 48 |
- lr_scheduler_type: cosine
|
| 49 |
- num_epochs: 4
|
| 50 |
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 54 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 55 |
+
| 1.6494 | 1.0 | 1250 | 1.5550 |
|
| 56 |
+
| 1.6059 | 2.0 | 2500 | 1.5224 |
|
| 57 |
+
| 1.572 | 3.0 | 3750 | 1.5108 |
|
| 58 |
+
| 1.5713 | 4.0 | 5000 | 1.5088 |
|
| 59 |
+
|
| 60 |
+
|
| 61 |
### Framework versions
|
| 62 |
|
| 63 |
- PEFT 0.13.2
|