Quentin Fuxa

Simultaneous speech-to-text models

A self-hosted server that converts spoken audio to text in real time with very low delay, supports around 200 languages, identifies who is speaking, and works on NVIDIA GPUs, Apple Silicon, and standard CPUs.

Simultaneous translation model for 200 languages

Community maintained hardware plugin for vLLM on Apple Silicon

Standalone causal streaming runtime for Qwen3-ASR

Recursive analysis of python stack of packages using AWS Strands Agents

Research code for AlignAtt4LLM simultaneous speech translation.

Agentic RAG platform purpose-built for small language models (SLM). Robust PDF/SQL search

NVIDIA Canary ASR model optimized for Apple Silicon using MLX.

OCRmyPDF + FastAPI demo: track and display OCR progress in real-time via a custom plugin.

TopicGPT allows to integrate the benefits of LLMs into Topic Modelling

No description.

Interactive D3.js network graph component for Streamlit — force-directed layout, zone clustering, multiple shapes, search, actions, i18n

OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to be searched

No description.

No description.

Add a way to det

model_metric_analysis

Implementation of the Expectation Maximisation algorithm in python

No description.

No description.

No description.

LLM inference in C/C++

Fast inference engine for Transformer models

No description.

An open-source, code-first Python toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.

Nemo prunned for SortFormer only

Python utility to convert PyTorch model weights from '.bin' to '.safetensors' format.

No description.