qwen2_5_omni_text_only
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7071
- Token Acc: 0.7822
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: 0.0001
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- total_eval_batch_size: 8
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.05
- num_epochs: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Token Acc |
|---|
| 0.9310 | 0.2494 | 50 | 0.9598 | 0.7252 |
| 0.9040 | 0.4988 | 100 | 0.9260 | 0.7308 |
| 0.8718 | 0.7481 | 150 | 0.9006 | 0.7390 |
| 0.8514 | 0.9975 | 200 | 0.8722 | 0.7447 |
| 0.7619 | 1.2444 | 250 | 0.8436 | 0.7509 |
| 0.7307 | 1.4938 | 300 | 0.8114 | 0.7579 |
| 0.7084 | 1.7431 | 350 | 0.7834 | 0.7644 |
| 0.6884 | 1.9925 | 400 | 0.7524 | 0.7711 |
| 0.5837 | 2.2394 | 450 | 0.7358 | 0.7757 |
| 0.5543 | 2.4888 | 500 | 0.7175 | 0.7793 |
| 0.5535 | 2.7382 | 550 | 0.7093 | 0.7814 |
| 0.5640 | 2.9875 | 600 | 0.7072 | 0.7821 |
| 0.5640 | 3.0 | 603 | 0.7071 | 0.7822 |
Framework versions
- Transformers 4.57.0
- Pytorch 2.8.0+cu128
- Datasets 3.6.0
- Tokenizers 0.22.1
