qwen2_5_omni_wo_text_1016
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5971
- Token Acc: 0.8167
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 |
|---|
| 0.9245 | 0.2 | 50 | 0.9275 | 0.7326 |
| 0.8908 | 0.4 | 100 | 0.9021 | 0.7424 |
| 0.8486 | 0.6 | 150 | 0.8778 | 0.7488 |
| 0.8094 | 0.8 | 200 | 0.8473 | 0.7570 |
| 0.8086 | 1.0 | 250 | 0.8194 | 0.7621 |
| 0.6808 | 1.2 | 300 | 0.7861 | 0.7708 |
| 0.6688 | 1.4 | 350 | 0.7513 | 0.7789 |
| 0.6371 | 1.6 | 400 | 0.7212 | 0.7850 |
| 0.6053 | 1.8 | 450 | 0.6915 | 0.7929 |
| 0.5887 | 2.0 | 500 | 0.6602 | 0.7994 |
| 0.4825 | 2.2 | 550 | 0.6381 | 0.8059 |
| 0.4620 | 2.4 | 600 | 0.6171 | 0.8110 |
| 0.4564 | 2.6 | 650 | 0.6043 | 0.8144 |
| 0.4562 | 2.8 | 700 | 0.5991 | 0.8164 |
| 0.4196 | 3.0 | 750 | 0.5971 | 0.8167 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.8.0+cu128
- Datasets 3.6.0
- Tokenizers 0.22.1
