CORAA-MUPE-ASR distil-whisper fine-tuned

This model is a fine-tuned version of RodrigoLimaRFL/distil-large-nurc-sp on the CORAA-MUPE-ASR dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3067
  • Wer: 16.9068

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: 18000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.3322 0.1734 3000 0.3687 19.8514
0.3245 0.3467 6000 0.3466 18.9951
0.3021 0.5201 9000 0.3320 18.0409
0.2852 0.6934 12000 0.3220 18.0697
0.2819 0.8668 15000 0.3095 17.4081
0.2062 1.0402 18000 0.3067 16.9068

Framework versions

  • Transformers 4.42.4
  • Pytorch 2.3.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
55
Safetensors
Model size
756M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for sidleal/distil-whisper-coraa-mupe-asr-2

Finetuned
(2)
this model

Dataset used to train sidleal/distil-whisper-coraa-mupe-asr-2

Evaluation results