metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- new_dataset
metrics:
- accuracy
model-index:
- name: wav2vec2-base-finetuned-manthan-gujarati-digits
results: []
wav2vec2-base-finetuned-manthan-gujarati-digits
This model is a fine-tuned version of facebook/wav2vec2-base on the new_dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.5613
- Accuracy: 0.9923
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.3392 | 0.98 | 12 | 1.1315 | 0.9665 |
| 1.2319 | 1.98 | 24 | 0.9487 | 0.9716 |
| 1.0824 | 2.98 | 36 | 0.8338 | 0.9820 |
| 0.9995 | 3.98 | 48 | 0.7533 | 0.9845 |
| 0.8175 | 4.98 | 60 | 0.6759 | 0.9923 |
| 0.8015 | 5.98 | 72 | 0.6425 | 0.9845 |
| 0.7417 | 6.98 | 84 | 0.6048 | 0.9871 |
| 0.7181 | 7.98 | 96 | 0.5850 | 0.9923 |
| 0.6907 | 8.98 | 108 | 0.5687 | 0.9897 |
| 0.6511 | 9.98 | 120 | 0.5613 | 0.9923 |
Framework versions
- Transformers 4.19.0
- Pytorch 1.11.0+cu113
- Datasets 1.14.0
- Tokenizers 0.12.1