Havuz

STEER — AI4R (AI for Research) experiment framework for auto-research and autonomous research. Connect existing research projects and let an

Tiny, fast Brazilian Portuguese text-to-speech (21M params)

GleamLM大语言模型全栈训练平台——核心代码用 PyTorch 从零实现，最低仅需 12GB 消费级显卡即可完整训练，另有一套工业代码与手写实现消费同一份数据逐层对照。模型训练支持CLI和可视化训练。

A practical lab for exploring Apple's Core AI framework, model assets, specialization, and on-device inference.

Composable LLM decoding on the probability simplex: library and paper code.

🌹 Rose: Range-Of-Slice Equilibration PyTorch optimizer. Stateless optimization through range-normalized gradient updates.

Build world models from scratch — from perception and latent dynamics to planning, JEPA, VLA, and 3D/4D intelligence.

First open-source implementation of Google TurboQuant (ICLR 2026) -- near-optimal KV cache compression for LLM inference. 5x compression wit

Pyramis-L3D: hierarchical sparse attention research prototype (code only; weights on Hugging Face)

Official implementation of "Unifying Distributional Training for One-Step Visual Generation" (MGFlow).

Local text-to-speech: 28 voices, 10 languages, voice cloning, SDKs for Python, Swift, Go, Rust and TypeScript

Production-ready, bleeding-edge, upstream-first vLLM Docker image for the NVIDIA DGX Spark (GB10 / sm_121a), verified against real models on

AI watermark removal built for Apple Silicon. SynthID-class pixel regeneration, visible vendor labels and provenance metadata, with every pi

AnyFlow - flow-matching video generation: training, inference and VBench evaluation for t2v and i2v models.

Open source release of Data Science Agents that can analyze and transform datasets as well as train and infer models to augment your data sc

Self-tuning multi-agent AI trading system. 8-source signal fusion, Bull/Bear/Judge debate on Claude Opus 4.7, Kelly + ATR position sizing. P

Evaan — a fully local, offline-capable AI companion chatbot built with Python & Hugging Face Transformers, running on Qwen2.5-0.5B-Instruct

结合预训练编码器、BiLSTM 与 GlobalPointer 的实体关系联合抽取模型，支持标注校验、训练、评估及三元组预测。

Inference engine for large language models on consumer GPUs, with deep optimization for older hardware

Open-source Environment toolkit of claw-like agents, support task/harness generation and evaluation

🚀 Engram-PEFT: An unofficial implementation of DeepSeek Engram. Inject high-capacity conditional memory into LLMs via sparse retrieval PEFT

Implementation of LatentMoE,Toward Optimal Accuracy per FLOP and Parameter in Mixture of Experts (Elango et al., NVIDIA 2026) — in Pytorch.

Unofficial fairseq-free PyTorch implementation of UTMOS (v1, 2022), matching the original system.

Train linear embedding adapters with triplet loss to align retrieval embeddings with your queries (RAG).

One GPU. Full LLM workflow. Real benchmarks. No cloud required.

a LLM inference engine to run on consumer hardware

First open-source TurboQuant KV cache compression for LLM inference. Drop-in for HuggingFace. pip install turboquant.

🔥 AI agent that ports CUDA code to AMD ROCm using local LLMs — FastAPI + React + Ollama | Live: rocmporter-agent.vercel.app

🧠 One LLM, split across a Mac (Apple MPS) + a Windows PC (NVIDIA CUDA) — heterogeneous pipeline inference over a framework-neutral wire, bit

A tiny language model that talks like a house cat named Miso.

Recomputation-free, context-independent KV caching for large language models

A comprehensive hands-on, AI-first Python curriculum with 80+ interactive Jupyter notebooks for AI Engineers from core basics to LLMs, RAG,

โคลนเสียงพูดภาษาไทยด้วย JaiTTS (F5-TTS) — ติดตั้งคำสั่งเดียว ใช้งานผ่านหน้าเว็บได้ทันที · Thai voice cloning made easy: one-command setup, l

End-to-end pipeline for seeing how LLMs actually process your prompts. Capture attention across every layer, render heatmaps and cooking cur

My complete journey to becoming an Agentic AI Engineer through structured learning, projects, experiments, and production-ready implementati

MoE expert compression toolkit — prune, merge, and forge specialist models for local inference.

Watch a real AI model think - live 3D visualization of a transformer's actual forward pass: tokens, embeddings, attention, neurons, sampling

I switched phones and found thousands of WhatsApp photos waiting for me, mostly memes and screenshots. This finds the ones worth keeping: se

Building an Image-to-Video (I2V) Model from Scratch

[ICML 2026] NanoQuant: Efficient Sub-1-Bit Quantization of Large Language Models

Reproducible OpenEnv-compatible reinforcement-learning environments for social-media integrity moderation.

Production-ready batch inference for Cohere’s Arabic/English ASR model, with optimized Silero VAD, multi-file GPU batching, subtitles, and o

Interactive, animated architecture maps for any Hugging Face model — no weights downloaded

A modern, hands-on, self-paced course that takes you from your first line of Python to shipping a real AI-powered automation.

Lynn 原生 LLM 推理引擎 · W4A8/NVFP4 量化 · 自写 CUDA/Triton kernel · MoE · 投机解码 | Lynn-native LLM inference engine for NVIDIA Blackwell

Compose neural architectures as typed, executable graph cards with two-way PyTorch sync, step-by-step local execution, and a constrained AI

Taiwanese Hokkien (Taigi) speech-to-text transcriber - MediaTek Breeze-ASR-26 with faster-whisper, tuned for RTX 3050 4GB low-VRAM GPUs. Gra

Unified KV-cache compression for LLM inference: 12 Python-native methods, guarded add-on composition and routing, analytical capacity simula

Build and train a mini Kimi K3 from scratch — a pure-PyTorch playground from a custom 88M model to 2T-scale LLM profiles.