ndeclarke's picture
Upload tokenizer
b4894aa verified
|
raw
history blame
3 kB
metadata
base_model: facebook/mms-1b-all
datasets:
  - common_voice_17_0
library_name: transformers
license: cc-by-nc-4.0
metrics:
  - wer
  - bleu
tags:
  - generated_from_trainer
model-index:
  - name: wav2vec2-mms-1b-CV17.0-training_set_variations
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: common_voice_17_0
          type: common_voice_17_0
          config: ta
          split: validation
          args: ta
        metrics:
          - type: wer
            value: 0.4958060228262364
            name: Wer
          - type: bleu
            value: 0.2629057639184852
            name: Bleu

wav2vec2-mms-1b-CV17.0-training_set_variations

This model is a fine-tuned version of facebook/mms-1b-all on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5099
  • Wer: 0.4958
  • Cer: 0.0885
  • Bleu: 0.2629

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: 0.001
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.15
  • training_steps: 2000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer Cer Bleu
11.4061 50.0 50 4.6415 1.0007 0.9640 0.0
1.789 100.0 100 0.3026 0.4457 0.0734 0.3064
0.08 150.0 150 0.3223 0.4304 0.0711 0.3275
0.0473 200.0 200 0.3547 0.4426 0.0742 0.3156
0.0364 250.0 250 0.3786 0.4556 0.0761 0.2972
0.0298 300.0 300 0.4070 0.4629 0.0800 0.2875
0.0279 350.0 350 0.4190 0.4688 0.0799 0.2864
0.0253 400.0 400 0.4353 0.4755 0.0818 0.2757
0.0198 450.0 450 0.4808 0.5066 0.0887 0.2432
0.0216 500.0 500 0.4699 0.4780 0.0815 0.2777
0.0194 550.0 550 0.4745 0.4895 0.0877 0.2643
0.0201 600.0 600 0.5035 0.4971 0.0881 0.2647
0.0153 650.0 650 0.5099 0.4958 0.0885 0.2629

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.0
  • Tokenizers 0.19.1