Galician Xlsr

🚀 Akashpb13/Galician_xlsr

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_7_0 - hu dataset. It offers high - performance in automatic speech recognition tasks, especially for Galician language processing.

✨ Features

• Automatic Speech Recognition: Capable of accurately transcribing speech, as demonstrated by its performance on multiple datasets.
• Trained on Specific Datasets: Utilizes the mozilla - foundation/common_voice_8_0 dataset for training, ensuring robustness in real - world scenarios.

📦 Installation

No installation steps were provided in the original document, so this section is skipped.

📚 Documentation

Model description

The base model "facebook/wav2vec2 - xls - r - 300m" was finetuned to adapt to the Galician language and specific speech recognition tasks.

Intended uses & limitations

More information needed.

Training and evaluation data

The training data includes Common voice Galician train.tsv, dev.tsv, invalidated.tsv, reported.tsv, and other.tsv. Only those points were considered where upvotes were greater than downvotes, and duplicates were removed after concatenating all the datasets from common voice 7.0.

Training procedure

For creating the training dataset, all possible datasets were appended, and a 90 - 10 split was used.

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.000096
• train_batch_size: 16
• eval_batch_size: 16
• seed: 13
• gradient_accumulation_steps: 2
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_steps: 500
• num_epochs: 100
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.16.0.dev0
• Pytorch 1.10.0+cu102
• Datasets 1.18.3
• Tokenizers 0.10.3

Evaluation Commands

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

python eval.py --model_id Akashpb13/Galician_xlsr --dataset mozilla - foundation/common_voice_8_0 --config gl --split test

🔧 Technical Details

The model was fine - tuned on a specific dataset, using a set of well - defined hyperparameters. The training process involved appending multiple datasets and performing a 90 - 10 split. The learning rate, batch size, and other hyperparameters were carefully selected to optimize the model's performance.

📄 License

The model is licensed under the Apache - 2.0 license.