asr_mind_model_colab_v4
This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 3.1781
- 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: 0.0001
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 12.5749 | 34.4956 | 1000 | 3.3044 | 1.0 |
| 12.3430 | 68.9912 | 2000 | 3.0759 | 1.0 |
| 12.4295 | 103.4602 | 3000 | 3.2682 | 1.0 |
| 12.6956 | 137.9558 | 4000 | 3.1781 | 1.0 |
Framework versions
- Transformers 5.8.1
- Pytorch 2.10.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2
- Downloads last month
- 1
Model tree for cdb-0/asr_mind_model_colab_v4
Base model
facebook/wav2vec2-base