Qwen2.5-14B-MegaFusion_V0
This model is a fine-tuned version of Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v8.7 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 4.9871
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: 2e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Use paged_adamw_8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 7.7644 | 0.2309 | 10 | 7.6364 |
| 7.5233 | 0.4618 | 20 | 7.4077 |
| 7.3034 | 0.6926 | 30 | 7.1544 |
| 7.0421 | 0.9235 | 40 | 6.8892 |
| 6.7811 | 1.1385 | 50 | 6.6204 |
| 6.4996 | 1.3694 | 60 | 6.3540 |
| 6.255 | 1.6003 | 70 | 6.0940 |
| 5.9838 | 1.8312 | 80 | 5.8364 |
| 5.7303 | 2.0462 | 90 | 5.5914 |
| 5.5161 | 2.2771 | 100 | 5.3681 |
| 5.3165 | 2.5079 | 110 | 5.1814 |
| 5.1162 | 2.7388 | 120 | 5.0485 |
| 5.0042 | 2.9697 | 130 | 4.9871 |
Framework versions
- PEFT 0.15.2
- Transformers 4.52.4
- Pytorch 2.7.1+cu126
- Datasets 3.6.0
- Tokenizers 0.21.2
- Downloads last month
- 9
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for Fayaz/Qwen2.5-14B-MegaFusion_V0
Base model
Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v8.7