Wav2vec2 Large Xls R 300m Hi CV7

🚀 wav2vec2-large-xls-r-300m-hi-CV7

This model is designed for automatic speech recognition, fine - tuned from facebook/wav2vec2-xls-r-300m on the MOZILLA - FOUNDATION/COMMON_VOICE_7_0 - HI dataset. It offers high - performance speech recognition capabilities, achieving excellent results on evaluation sets.

✨ Features

• Fine - tuned Model: Based on the pre - trained facebook/wav2vec2-xls-r-300m, fine - tuned on the HI dataset of MOZILLA - FOUNDATION/COMMON_VOICE_7_0.
• Multiple Evaluation Metrics: Evaluated on multiple datasets with metrics such as WER (Word Error Rate) and CER (Character Error Rate).

📦 Installation

The README does not provide specific installation steps, so this section is skipped.

💻 Usage Examples

The README does not provide code examples, so this section is skipped.

📚 Documentation

Evaluation Results

This model achieves the following results on the evaluation set:

• Loss: 0.6588
• Wer: 0.2987

Evaluation Commands

1. To evaluate on mozilla - foundation/common_voice_8_0 with test split

python eval.py --model_id DrishtiSharma/wav2vec2-large-xls-r-300m-hi-CV7 --dataset mozilla-foundation/common_voice_7_0 --config hi --split test --log_outputs

2. To evaluate on speech - recognition - community - v2/dev_data

NA

Training Hyperparameters

The following hyperparameters were used during training:

• 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: 2000
• num_epochs: 60
• mixed_precision_training: Native AMP

Training Results

Framework Versions

• Transformers 4.16.2
• Pytorch 1.10.0+cu111
• Datasets 1.18.3
• Tokenizers 0.11.0

🔧 Technical Details

The README does not provide in - depth technical details, so this section is skipped.

📄 License

• License: apache - 2.0