Xls R Hausa 40

🚀 Automatic Speech Recognition Model

This model is a fine - tuned speech recognition model that addresses the challenges of transcribing speech in the Hausa language. It offers high - quality automatic speech recognition capabilities, leveraging a pre - trained model and fine - tuning on a specific dataset to achieve good performance.

🚀 Quick Start

This model is a fine - tuned version of [facebook/wav2vec2 - xls - r - 300m](https://huggingface.co/facebook/wav2vec2 - xls - r - 300m) on the MOZILLA - FOUNDATION/COMMON_VOICE_8_0 - HA dataset. It achieves the following results on the evaluation set:

• Loss: 0.4998
• Wer: 0.5153

✨ Features

• Fine - Tuned: Based on the pre - trained [facebook/wav2vec2 - xls - r - 300m](https://huggingface.co/facebook/wav2vec2 - xls - r - 300m), fine - tuned on the HA dataset of MOZILLA - FOUNDATION/COMMON_VOICE_8_0.
• Multilingual Adaptability: Leveraged the power of the pre - trained model, potentially adaptable to other languages with further fine - tuning.
• Performance Metrics: Achieved specific loss and WER (Word Error Rate) on the evaluation set, demonstrating its effectiveness.

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

🔧 Technical Details

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 9.6e - 05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• 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: 80.0
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.17.0.dev0
• Pytorch 1.10.2+cu113
• Datasets 1.18.4.dev0
• Tokenizers 0.11.0

📄 License

This model is licensed under the Apache - 2.0 license.