Wav2vec2 Large Xlsr 53 Spanish

🚀 XLSR Wav2Vec2 Spanish by Jonatas Grosman

This project adds a custom language model to the Spanish version of Wav2Vec2-Large-XLSR-53. It fine - tunes the facebook/wav2vec2-large-xlsr-53 model on Spanish data from Common Voice. The model is designed for automatic speech recognition, and it's important to ensure that the speech input is sampled at 16kHz.

✨ Features

• Fine - tuned for Spanish: Trained specifically on Spanish data from Common Voice to improve performance in Spanish speech recognition.
• GPU - supported training: Thanks to the GPU credits from OVHcloud, the model has been fine - tuned effectively.
• Multiple evaluation metrics: Evaluated using Word Error Rate (WER) and Character Error Rate (CER), both with and without a language model.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

Basic Usage

Using the ASRecognition library:

from asrecognition import ASREngine

asr = ASREngine("es", model_path="jonatasgrosman/wav2vec2-large-xlsr-53-spanish")

audio_paths = ["/path/to/file.mp3", "/path/to/another_file.wav"]
transcriptions = asr.transcribe(audio_paths)

Advanced Usage

Writing your own inference script:

import torch
import librosa
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

LANG_ID = "es"
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-spanish"
SAMPLES = 10

test_dataset = load_dataset("common_voice", LANG_ID, split=f"test[:{SAMPLES}]")

processor = Wav2Vec2Processor.from_pretrained(MODEL_ID)
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)

# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
    speech_array, sampling_rate = librosa.load(batch["path"], sr=16_000)
    batch["speech"] = speech_array
    batch["sentence"] = batch["sentence"].upper()
    return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)
inputs = processor(test_dataset["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)

with torch.no_grad():
    logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits

predicted_ids = torch.argmax(logits, dim=-1)
predicted_sentences = processor.batch_decode(predicted_ids)

for i, predicted_sentence in enumerate(predicted_sentences):
    print("-" * 100)
    print("Reference:", test_dataset[i]["sentence"])
    print("Prediction:", predicted_sentence)

Prediction Results

📚 Documentation

Evaluation

1. To evaluate on mozilla-foundation/common_voice_6_0 with split test

python eval.py --model_id jonatasgrosman/wav2vec2-large-xlsr-53-spanish --dataset mozilla-foundation/common_voice_6_0 --config es --split test

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

python eval.py --model_id jonatasgrosman/wav2vec2-large-xlsr-53-spanish --dataset speech-recognition-community-v2/dev_data --config es --split validation --chunk_length_s 5.0 --stride_length_s 1.0

📄 License

This model is licensed under the Apache 2.0 license.

📚 Model Index

• Name: XLSR Wav2Vec2 Spanish by Jonatas Grosman
• Results:
  • Task: Automatic Speech Recognition
    • Dataset: Common Voice es
      • Metrics:
        • Test WER: 8.82
        • Test CER: 2.58
        • Test WER (+LM): 6.27
        • Test CER (+LM): 2.06
  • Task: Automatic Speech Recognition
    • Dataset: Robust Speech Event - Dev Data
      • Metrics:
        • Dev WER: 30.19
        • Dev CER: 13.56
        • Dev WER (+LM): 24.71
        • Dev CER (+LM): 12.61

📖 Citation

If you want to cite this model you can use this:

@misc{grosman2021wav2vec2-large-xlsr-53-spanish,
  title={XLSR Wav2Vec2 Spanish by Jonatas Grosman},
  author={Grosman, Jonatas},
  publisher={Hugging Face},
  journal={Hugging Face Hub},
  howpublished={\url{https://huggingface.co/jonatasgrosman/wav2vec2-large-xlsr-53-spanish}},
  year={2021}
}

📋 Information Table