Discussion-Phi-4-multimodal-instruct-audio-level-test
This model is a fine-tuned version of microsoft/Phi-4-multimodal-instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 14.8321
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: 4e-05
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-07 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 4.393 | 0.2235 | 10 | 15.2023 |
| 0.2619 | 0.4469 | 20 | 14.7785 |
| 0.2171 | 0.6704 | 30 | 14.8554 |
| 0.3195 | 0.8939 | 40 | 14.9321 |
| 0.2302 | 1.1117 | 50 | 14.8080 |
| 0.187 | 1.3352 | 60 | 14.8113 |
| 0.2844 | 1.5587 | 70 | 14.8289 |
| 0.177 | 1.7821 | 80 | 14.8365 |
| 0.1134 | 2.0 | 90 | 14.8888 |
| 0.1534 | 2.2235 | 100 | 14.8262 |
| 0.0892 | 2.4469 | 110 | 14.7385 |
| 0.0554 | 2.6704 | 120 | 14.8482 |
| 0.0261 | 2.8939 | 130 | 14.8321 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.4.1+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1