Lasion's picture
Upload tokenizer
82b976e verified
|
raw
history blame
2.17 kB
metadata
license: cc-by-nc-4.0
tags:
  - generated_from_trainer
base_model: nguyenvulebinh/wav2vec2-base-vietnamese-250h
datasets:
  - common_voice_17_0
metrics:
  - wer
model-index:
  - name: wav2vec2-common-voice-17_0_vi
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: common_voice_17_0
          type: common_voice_17_0
          config: vi
          split: None
          args: vi
        metrics:
          - type: wer
            value: 0.43487928843710294
            name: Wer

wav2vec2-common-voice-17_0_vi

This model is a fine-tuned version of nguyenvulebinh/wav2vec2-base-vietnamese-250h on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7992
  • Wer: 0.4349

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: 8
  • 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: 1000
  • num_epochs: 30
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.261 4.3103 500 0.4182 0.3492
0.2061 8.6207 1000 0.5416 0.4044
0.1883 12.9310 1500 0.6796 0.4304
0.1336 17.2414 2000 0.8089 0.4378
0.1257 21.5517 2500 0.8244 0.4426
0.098 25.8621 3000 0.7992 0.4349

Framework versions

  • Transformers 4.40.0
  • Pytorch 2.2.1+cu121
  • Datasets 2.19.0
  • Tokenizers 0.19.1