README.md

metadata

license: apache-2.0
language: sah
tags:
  - generated_from_trainer
  - robust-speech-event
  - hf-asr-leaderboard
datasets:
  - common_voice
model-index:
  - name: wav2vec2-large-xlsr-53-sah-CV8
    results:
      - task:
          name: Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice sah
          type: common_voice
          args: sah
        metrics:
          - name: Test WER
            type: wer
            value: 56.06
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 8.0
          type: mozilla-foundation/common_voice_8_0
          args: sah
        metrics:
          - name: Test WER
            type: wer
            value: 43.75

wav2vec2-large-xlsr-53-sah-CV8

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice dataset. It achieves the following results on the evaluation set:

• Loss: 0.5089
• Wer: 0.5606

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: 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_steps: 300
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
4.6849	16.67	500	1.1135	0.9344
0.8223	33.33	1000	0.5148	0.5686
0.5477	50.0	1500	0.5089	0.5606

Framework versions

• Transformers 4.11.3
• Pytorch 1.10.0+cu111
• Datasets 1.18.1
• Tokenizers 0.10.3