4_bar_lmd_clean_custom_epochs10
This model is a fine-tuned version of MikeMpapa/4_bar_lmd_clean_custom_epochs10 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.7890
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: 0.005
- train_batch_size: 48
- eval_batch_size: 32
- seed: 1
- gradient_accumulation_steps: 2
- total_train_batch_size: 96
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.01
- num_epochs: 81
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.9313 | 9.42 | 12000 | 0.9207 |
| 0.9129 | 18.84 | 24000 | 0.9018 |
| 0.8847 | 28.26 | 36000 | 0.8786 |
| 0.8537 | 37.68 | 48000 | 0.8541 |
| 0.8177 | 47.1 | 60000 | 0.8412 |
| 0.7778 | 56.51 | 72000 | 0.8111 |
| 0.7385 | 65.93 | 84000 | 0.7931 |
| 0.7079 | 75.35 | 96000 | 0.7890 |
Framework versions
- Transformers 4.36.0
- Pytorch 2.1.0
- Datasets 2.15.0
- Tokenizers 0.15.1