RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding.

Implementing a ChatGPT-like LLM in PyTorch from scratch, step by step

Fast and customizable framework for automatic ML model creation (AutoML)

18 Lessons, Get Started Building with Generative AI 🔗 https://microsoft.github.io/generative-ai-for-beginners/

Source code for all Elastic connectors, developed by the Search team at Elastic, and home of our Python connector development framework

Efficient implementations of state-of-the-art linear attention models in Pytorch and Triton

Multi-LoRA inference server that scales to 1000s of fine-tuned LLMs

Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads

Benchmarking long-form factuality in large language models. Original code for our paper "Long-form factuality in large language models".

Libraries for applying sparsification recipes to neural networks with a few lines of code, enabling faster and smaller models

Sparsity-aware deep learning inference runtime for CPUs

LLM Workshop by Sourab Mangrulkar

A Native-PyTorch Library for LLM Fine-tuning

Open-Sora: Democratizing Efficient Video Production for All

llama.cpp gguf file parser for javascript

PyTorch code and models for V-JEPA self-supervised learning from video.

Go ahead and axolotl questions

Description Describes the IndicNLP corpus and associated datasets

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

Custom data types and layouts for training and inference

Faster Pytorch bitsandbytes 4bit fp4 nn.Linear ops

🐙 Guides, papers, lecture, notebooks and resources for prompt engineering

TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficie…

Temporary anonymous version

Detect file content types with deep learning

Official repository for 'GaussianHead: High-fidelity Head Avatars with Learnable Gaussian Derivation'

Set-of-Mark Prompting for LMMs

Minimal, clean code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.