README.md

metadata

license: apache-2.0
base_model: honzapucalek/p6_commonvoice_16_1
tags:
  - generated_from_trainer
datasets:
  - p6_impaired_v2
metrics:
  - wer
model-index:
  - name: p6_commonvoice_impaired
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: p6_impaired_v2
          type: p6_impaired_v2
          config: cs
          split: test
          args: cs
        metrics:
          - name: Wer
            type: wer
            value: 0.41563311220074994

p6_commonvoice_impaired

This model is a fine-tuned version of honzapucalek/p6_commonvoice_16_1 on the p6_impaired_v2 dataset. It achieves the following results on the evaluation set:

• Loss: 1.3476
• Wer: 0.4156

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• 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: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0078	14.93	1000	1.1002	0.4759
0.0007	29.85	2000	1.2277	0.4332
0.0012	44.78	3000	1.1811	0.4381
0.0001	59.7	4000	1.3258	0.4200
0.0001	74.63	5000	1.3476	0.4156

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1