README.md

metadata

language:
  - sl
license: apache-2.0
tags:
  - generated_from_trainer
  - hf-asr-leaderboard
  - robust-speech-event
datasets:
  - common_voice
base_model: facebook/wav2vec2-xls-r-1b
model-index:
  - name: wav2vec2-large-xls-r-1B-common_voice-sl-ft
    results:
      - task:
          type: automatic-speech-recognition
          name: Speech Recognition
        dataset:
          name: Common Voice 7
          type: mozilla-foundation/common_voice_7_0
          args: lv
        metrics:
          - type: wer
            value: 23.26
            name: Test WER
          - type: cer
            value: 7.95
            name: Test CER
          - type: wer
            value: 13.59
            name: Test WER
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: Robust Speech Event - Dev Data
          type: speech-recognition-community-v2/dev_data
          args: sl
        metrics:
          - type: wer
            value: 62.71
            name: Test WER
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: Robust Speech Event - Test Data
          type: speech-recognition-community-v2/eval_data
          args: sl
        metrics:
          - type: wer
            value: 62.34
            name: Test WER

wav2vec2-large-xls-r-1B-common_voice-sl-ft

This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the common_voice dataset. It achieves the following results on the evaluation set:

• Loss: 0.2112
• Wer: 0.1404

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 400
• num_epochs: 100
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.8291	12.2	500	0.5674	0.7611
0.0416	24.39	1000	0.3093	0.2964
0.0256	36.59	1500	0.2224	0.2072
0.0179	48.78	2000	0.2274	0.1960
0.0113	60.98	2500	0.2078	0.1582
0.0086	73.17	3000	0.1898	0.1552
0.0059	85.37	3500	0.2054	0.1446
0.0044	97.56	4000	0.2112	0.1404

Framework versions

• Transformers 4.16.0.dev0
• Pytorch 1.10.1+cu102
• Datasets 1.17.1.dev0
• Tokenizers 0.10.3