Tool for generating high quality Synthetic datasets

An algorithm-focused interface for common llm training, continual learning, and reinforcement learning techniques

Synthetic Data Generation Toolkit for LLMs

Tools for merging pretrained large language models.

Python tool for converting files and office documents to Markdown.

Self-Adapting Language Models

(trying to) train llama2 7B fast

SkyRL: A Modular Full-stack RL Library for LLMs

Extract and convert data from any document, images, pdfs, word doc, ppt or URL into multiple formats (Markdown, JSON, CSV, HTML) with intelligent structured data extraction and advanced OCR.

The best ChatGPT that $100 can buy.

Why Low-Precision Transformer Training Fails: An Analysis on Flash Attention

QeRL enables RL for 32B LLMs on a single H100 GPU.

Structured Outputs

xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism

AgentFlow: In-the-Flow Agentic System Optimization

Introduction to PyTorch, covering tensor initialization, operations, indexing, and reshaping.

PIKE-RAG: sPecIalized KnowledgE and Rationale Augmented Generation

A simple package to automatically trace PyTorch training memory usage.

A tool for bandwidth measurements on NVIDIA GPUs.

NVIDIA Linux open GPU with P2P support

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory…

Ongoing research training transformer models at scale

Scalable toolkit for efficient model reinforcement

Deploy an AI Analyst in less than 2 mins — connect any LLM to any data source with centralized context management, observability, and control. Text-to-SQL, Text-to-Python, Text-to-Dashboard

Tina: Tiny Reasoning Models via LoRA

The official code implementation for "Cache-to-Cache: Direct Semantic Communication Between Large Language Models"