Nguyen Tan Dat

moshi-rag. MoshiRAG is a compact full-duplex speech language model augmented with asynchronous knowledge retrieval to improve factuality without sacrificing real-time interactivity.

BFQ. Fast and Highly Expressive Policy Learning for Offline Reinforcement Learning via Bootstrapped Flow Q-Learning official source code

UTMOSv2. UTokyo-SaruLab MOS Prediction System

SeeandSniff. [ECCV 2026] Official Pytorch implementation for See & Sniff: Learning Visuo-Olfactory Representations

humanizer. Claude Code skill that removes signs of AI-generated writing from text

ponytail. Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.

mamtra_inference. [Interspeech 2026] The official inference code of MamTra

EcoSplat. [CVPR'26] EcoSplat: Efficiency-controllable Feed-forward 3D Gaussian Splatting from Multi-view Images (Highlight)

ITS-AVGen-LTX2. [TMLR] Inference-Time Scaling for Joint Audio–Video Generation

crossmodal-hub. [ICML 2026] Probing Cross-modal Information Hubs in Audio-Visual LLMs

ITS-AVGen. [TMLR] Inference-Time Scaling for Joint Audio–Video Generation

lst. Code for Latent Speech-Text Transformer (LST)

HumDial-FDBench. The Full-Duplex Interaction Track of the ICASSP 2026 Human-like Spoken Dialogue Systems Challenge aims to advance the evaluation of full-duplex dialogue systems by in- troducing a dual-channel dialogue dataset of real human- recorded conversations.

Qwen3-TTS. Qwen3-TTS is an open-source series of TTS models developed by the Qwen team at Alibaba Cloud, supporting stable, expressive, and streaming speech generation, free-form voice design, and vivid voice cloning.

In-Place-TTT. Python

VibeVoice. Open-Source Frontier Voice AI

Awesome-PhD-CV. Curated academic CV templates and guidelines for PhD students, researchers, and faculty job applicants.

zerobrew. A 5-20x faster experimental Homebrew alternative

3d-talking. Python

TTS-arxiv-daily. Automatically Update Text-to-speech (TTS) Papers Daily using Github Actions (Update Every 12th hours)

marvis-tts. Python

AI-Trader. "AI-Trader: 100% Fully-Automated Agent-Native Trading"

dllm. dLLM: Simple Diffusion Language Modeling

Fast-dLLM. Official implementation of "Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding"

lightly-studio. Curate, Annotate, and Manage Your Data in LightlyStudio.

plare. Policy Learning from Large Vision-Language Model Feedback Without Reward Modeling (IROS 2025)

neutts. On-device TTS model by Neuphonic

MAGE. JavaScript

seed-tts-eval. Python

vnstock. A beginner-friendly yet powerful Python toolkit for financial analysis and automation — built to make modern investing accessible to everyone

OpenCut. The open-source CapCut alternative

MDSGen. [ICLR'25] MDSGen: Fast and Efficient Masked Diffusion Temporal-Aware Transformers for Open-Domain Sound Generation

erlvlm. Enhancing Rating-Based Reinforcement Learning to Effectively Leverage Feedback from Large Vision-Language Models (ICML 2025)

vq_robust_rl. Mitigating Adversarial Perturbations for Deep Reinforcement Learning via Vector Quantization (IROS 2024)

pcdt. Predictive Coding for Decision Transformer (IROS 2024)

SplineGS. [CVPR'25] SplineGS: Robust Motion-Adaptive Spline for Real-Time Dynamic 3D Gaussians from Monocular Video

SSLalignment. Python

bd3lms. [ICLR 2025 Oral] Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Models

imm. Official implementation of Inductive Moment Matching

LLaSA_inference.

CosyVoice. Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.

librispeech-long. LibriSpeech-Long is a benchmark dataset for long-form speech generation and processing. Released as part of "Long-Form Speech Generation with Spoken Language Models" (arXiv 2024).

awesome-ssm-ml.

vall-e. PyTorch implementation of VALL-E(Zero-Shot Text-To-Speech), Reproduced Demo https://lifeiteng.github.io/valle/index.html

AudioLCM. PyTorch Implementation of AudioLCM (ACM-MM'24): a efficient and high-quality text-to-audio generation with latent consistency model.

libriheavy. Libriheavy: a 50,000 hours ASR corpus with punctuation casing and context

SpeechTokenizer. This is the code for the SpeechTokenizer presented in the SpeechTokenizer: Unified Speech Tokenizer for Speech Language Models. Samples are presented on

LLM-Codec. The open source code for LLM-Codec

USLM. Unified Speech Language Model for paper "SpeechTokenizer: Unified Speech Tokenizer for Speech Large Language Models"(ICLR 2024)

the-art-of-debugging. The Art of Debugging Open Book

ctm. Python

HiFi-GAN. HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis

fregrad. Code repository for FreGrad

fregrad. [ICASSP 2024] Official code for FreGrad

accelerate. 🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support

RAD-MMM. A TTS model that makes a speaker speak new languages

VoiceLDM. VoiceLDM: Text-to-Speech with Environmental Context

pronerf. [IEEE Access] ProNeRF: Learning Efficient Projection-Aware Ray Sampling for Fine-Grained Implicit Neural Radiance Fields

resume.github.com. Resumes generated using the GitHub informations

architecture-decision-record. Architecture decision record (ADR) examples for software planning, IT leadership, and template documentation

StyleSpeech. PyTorch Implementation of Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation

Daft-Exprt. PyTorch Implementation of Daft-Exprt: Robust Prosody Transfer Across Speakers for Expressive Speech Synthesis

survey. A Survey on Neural Speech Synthesis https://arxiv.org/pdf/2106.15561.pdf

espnet. End-to-End Speech Processing Toolkit

pyannote-audio. Neural building blocks for speaker diarization: speech activity detection, speaker change detection, overlapped speech detection, speaker embedding

lyra. A Very Low-Bitrate Codec for Speech Compression

etnlp. ETNLP: A toolkit to evaluate, extract, and visualize multiple embeddings

underthesea. Underthesea - AI Assistant

speech-synthesis-paper. List of speech synthesis papers.

d2l-vi. Một cuốn sách về Học Sâu đề cập đến nhiều framework phổ biến, được sử dụng trên 300 trường Đại học từ 55 đất nước bao gồm MIT, Stanford, Harvard, và Cambridge.

TensorFlowASR. :zap: TensorFlowASR: Almost State-of-the-art Automatic Speech Recognition in Tensorflow 2. Supported languages that can use characters or subwords

TensorFlowTTS. :stuck_out_tongue_closed_eyes: TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)

WaveRNN. WaveRNN Vocoder + TTS