metadata
base_model: Qwen/Qwen2-0.5B-Instruct
library_name: peft
license: apache-2.0
metrics:
- accuracy
tags:
- trl
- reward-trainer
- generated_from_trainer
model-index:
- name: Qwen2-0.5B-Reward
results: []
Qwen2-0.5B-Reward
This model is a fine-tuned version of Qwen/Qwen2-0.5B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6423
- Accuracy: 0.628
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- total_train_batch_size: 64
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.9852 | 0.0516 | 50 | 0.9181 | 0.478 |
| 0.811 | 0.1032 | 100 | 0.8217 | 0.52 |
| 0.7701 | 0.1548 | 150 | 0.7529 | 0.56 |
| 0.7239 | 0.2064 | 200 | 0.7145 | 0.59 |
| 0.723 | 0.2580 | 250 | 0.6917 | 0.597 |
| 0.6912 | 0.3096 | 300 | 0.6812 | 0.615 |
| 0.6577 | 0.3612 | 350 | 0.6702 | 0.626 |
| 0.645 | 0.4128 | 400 | 0.6632 | 0.616 |
| 0.6749 | 0.4644 | 450 | 0.6584 | 0.623 |
| 0.6497 | 0.5160 | 500 | 0.6541 | 0.625 |
| 0.659 | 0.5676 | 550 | 0.6526 | 0.626 |
| 0.634 | 0.6192 | 600 | 0.6495 | 0.626 |
| 0.6393 | 0.6708 | 650 | 0.6463 | 0.624 |
| 0.6263 | 0.7224 | 700 | 0.6456 | 0.629 |
| 0.6428 | 0.7740 | 750 | 0.6440 | 0.625 |
| 0.6335 | 0.8256 | 800 | 0.6431 | 0.634 |
| 0.6313 | 0.8772 | 850 | 0.6425 | 0.638 |
| 0.6323 | 0.9288 | 900 | 0.6419 | 0.636 |
| 0.6313 | 0.9804 | 950 | 0.6424 | 0.645 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.19.1