aquiffoo's picture

In a Training Loop 🔄

aquiffoo

aquiffoo

·

https://aquiffoo.is-a.dev/

AI & ML interests

thanks for everything.

Recent Activity

liked a model about 23 hours ago

zai-org/GLM-OCR

liked a model 1 day ago

stepfun-ai/Step-3.5-Flash

reacted to yuriyvnv's post with 👍 1 day ago

🎯 WAVe: 1B Multimodal Embedding Model for Word-Level Speech Quality Multimodal embeddings for speech + transcript that verify quality at the word level, not just sentence level. Catches mispronunciations, timing errors, and prosody issues that sentence-level filters miss. 📊 Impact on Portuguese ASR: • 34% reduction in training steps • 50% better cross-domain generalization • 30% less synthetic data needed • Word-aligned attention finds errors other methods miss 🏗️ Architecture: • Text: XLM-RoBERTa (278M params) • Audio: Wav2Vec2-BERT 2.0 (581M params) • Word Alignment: Multi-head attention + GLU (14M params) • Total: 1B parameters ``` from transformers import AutoModel, AutoProcessor processor = AutoProcessor.from_pretrained( "yuriyvnv/WAVe-1B-Multimodal-PT", trust_remote_code=True ) model = AutoModel.from_pretrained( "yuriyvnv/WAVe-1B-Multimodal-PT", trust_remote_code=True ) ``` # Assess speech-transcript alignment ``` inputs = processor(text="Olá, como está?", audio=audio_array, sampling_rate=16000, return_tensors="pt") quality = model(**inputs).quality_score.item() ``` Perfect for filtering synthetic speech datasets before ASR training. Model: https://huggingface.co/yuriyvnv/WAVe-1B-Multimodal-PT Code to create WAVe : https://github.com/yuriyvnv/WAVe #multimodal #speech #embeddings #asr #syntheticdata #qualityassessment

View all activity

Organizations

aquiffoo 's datasets 2

aquiffoo/foo-euler-1m

Viewer • Updated May 10, 2025 • 1M • 3

aquiffoo/foo-euler-100k

Viewer • Updated May 10, 2025 • 100k • 1