metadata
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: sambit-workacc
results: []
sambit-workacc
This model is a fine-tuned version of facebook/wav2vec2-base on the PolyAI/minds14 dataset. It achieves the following results on the evaluation set:
- Loss: 24.8037
- Wer: 1.0
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: 1e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 40
- training_steps: 200
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 0.8 | 20 | 34.5760 | 1.0063 |
| 47.3781 | 1.6 | 40 | 24.8037 | 1.0 |
| 27.6882 | 2.4 | 60 | 8.2192 | 1.0 |
| 27.6882 | 3.2 | 80 | 4.0556 | 1.0 |
| 6.5092 | 4.0 | 100 | 3.6545 | 1.0 |
| 4.1312 | 4.8 | 120 | 3.5571 | 1.0 |
| 4.1312 | 5.6 | 140 | 3.5060 | 1.0 |
| 3.929 | 6.4 | 160 | 3.4571 | 1.0 |
| 3.8668 | 7.2 | 180 | 3.4330 | 1.0 |
| 3.8668 | 8.0 | 200 | 3.4136 | 1.0 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.2
- Datasets 2.15.0
- Tokenizers 0.15.2