mutisya/w2v-bert-adapters-3lang-e10-25_52-v5

W2V-BERT 2.0 ASR Adapters

This repository contains per-language bottleneck adapters for automatic speech recognition (ASR) trained on top of facebook/w2v-bert-2.0.

Model Description

• Base Model: facebook/w2v-bert-2.0 (600M parameters, frozen)
• Adapter Architecture: MMS-style bottleneck adapters (dim=64)
• Decoder: Lightweight transformer decoder (1 layer)
• Training: CTC loss with extended vocabulary for double vowels

Trained Adapters

Training in progress...

Adapter	Language	WER	Train Samples

Architecture

The model uses:

1. Frozen w2v-bert-2.0 encoder - Extracts audio representations
2. Bottleneck adapter - Language-specific adaptation (trainable)
3. Lightweight decoder - Transformer decoder block (trainable)
4. LM head - Per-language vocabulary projection (trainable)

Usage

Each adapter folder contains:

• adapter_weights.pt - Bottleneck adapter weights
• decoder_weights.pt - Decoder block weights
• lm_head_weights.pt - Language model head weights
• final_norm_weights.pt - Final layer norm weights
• vocab.json - Language-specific vocabulary
• adapter_config.json - Adapter configuration
• metrics.json - Training metrics

Loading an Adapter

import torch
from transformers import Wav2Vec2BertProcessor

# Load processor for specific language
processor = Wav2Vec2BertProcessor.from_pretrained(
    "mutisya/w2v-bert-adapters-3lang-e10-25_52-v5",
    subfolder="<adapter_id>"
)

# Load adapter weights
adapter_weights = torch.load(
    hf_hub_download("mutisya/w2v-bert-adapters-3lang-e10-25_52-v5", "<adapter_id>/adapter_weights.pt")
)

Training Configuration

• Epochs: 10
• Learning Rate: 0.0005
• Batch Size: 24 × 2 (effective: 48)
• Extended Vocabulary: True
• Adapter Dimension: 64

License

Apache 2.0

Citation

@misc{w2vbert-asr-adapters,
  author = {Mutisya},
  title = {W2V-BERT 2.0 ASR Adapters for African Languages},
  year = {2024},
  publisher = {HuggingFace},
  url = {https://huggingface.co/mutisya/w2v-bert-adapters-3lang-e10-25_52-v5}
}