README.md

metadata

library_name: transformers
language:
  - pt
license: apache-2.0
base_model: openai/whisper-tiny
tags:
  - generated_from_trainer
datasets:
  - RodrigoLimaRFL/nurc-sp_pseudo_labelled
metrics:
  - wer
model-index:
  - name: Whisper-Tiny-PTBR
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: nurc-sp_pseudo_labelled
          type: RodrigoLimaRFL/nurc-sp_pseudo_labelled
        metrics:
          - name: Wer
            type: wer
            value: 59.38036802234333

Whisper-Tiny-PTBR

This model is a fine-tuned version of openai/whisper-tiny on the nurc-sp_pseudo_labelled dataset. It achieves the following results on the evaluation set:

• Loss: 1.0137
• Wer: 59.3804

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: 16
• eval_batch_size: 8
• seed: 42
• 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
1.2522	0.5094	1000	1.1713	74.6895
1.0397	1.0188	2000	1.0796	68.5537
0.9879	1.5283	3000	1.0420	62.4686
0.9334	2.0377	4000	1.0195	59.7845
0.9834	2.5471	5000	1.0137	59.3804

Framework versions

• Transformers 4.44.2
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1