Description
This model is a fine-tuned version of ishmamzarif/finetuned-modelv4 . ishmamzarif/finetuned-modelv4 is a finetuned version of bangla-speech-processing/BanglaASR on Bangla speech data.
Environment:
- Python version: 3.12.12
- PyTorch version: 2.8.0+cu126
- Librosa version: 0.10.1
- NumPy version: 1.26.4
Training Parameters:
- BATCH_SIZE = 4
- GRADIENT_ACCUMULATION_STEPS = 4 # Effective batch size = 16
- LEARNING_RATE = 2e-5
- WARMUP_STEPS = 200
- NUM_TRAIN_EPOCHS = 5
- LOGGING_STEPS = 50
Validation Set Evaluation:
| Epoch | Training Loss | Validation Loss | WER (%) | Normalized Levenshtein Similarity (%) |
|---|---|---|---|---|
| 0 | 1.422500 | 1.484552 | 21.104972 | 85.282828 |
| 2 | 1.419100 | 1.467892 | 15.524862 | 88.919192 |
| 4 | 1.412700 | 1.465618 | 14.585635 | 89.000000 |
- Downloads last month
- 2