Fine-tuning Tools

Ultrafast serverless GPU inference, sandboxes, and background jobs

Scalable data pre processing and curation toolkit for LLMs

Nvidia GPU exporter for prometheus using nvidia-smi binary

Jackrong-llm-finetuning-guide — open-source AI/LLM project.

Hypernetworks that adapt LLMs for specific benchmark tasks using only textual task description as the input

Bringing the Unsloth experience to Mac users via Apple's MLX framework

DataDreamer: Prompt. Generate Synthetic Data. Train & Align Models.   🤖💤

GraphGen: Enhancing Supervised Fine-Tuning for LLMs with Knowledge-Driven Synthetic Data Generation

Byted PyTorch Distributed for Hyperscale Training of LLMs and RLs

A complete guide to start and improve your LLM skills in 2026 with little background in the field and stay up-to-date with the latest news and state-of-the-art techniques!

Browser automation system that uses AI-driven planning to navigate web pages and perform goals.

Hyper efficient storage for GPU workloads. Feed your GPUs at blazing fast speeds.

Adan: Adaptive Nesterov Momentum Algorithm for Faster Optimizing Deep Models

Framework for enhancing LLMs for RAG tasks using fine-tuning.

“AI-Compass”将为社区指引在 AI 技术海洋中航行的方向，无论你是初学者还是进阶开发者，都能在这里找到通往 AI 各大方向的路径。旨在帮助开发者系统性地了解 AI 的核心概念、主流技术、前沿趋势，并通过实践掌握从理论到落地的全过程。

LLM-PowerHouse: Unleash LLMs' potential through curated tutorials, best practices, and ready-to-use code for custom training and inferencing.

USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference

历年ICLR论文和开源项目合集，包含ICLR2021、ICLR2022、ICLR2023、ICLR2024、ICLR2025.

LLM (Large Language Model) FineTuning

irresponsible innovation. Try now at https://chat.dev/

Repo for fine-tuning Casual LLMs

The official repo of Aquila2 series proposed by BAAI, including pretrained & chat large language models.

Collection of best practices, reference architectures, model training examples and utilities to train large models on AWS.

InternEvo is an open-sourced lightweight training framework aims to support model pre-training without the need for extensive dependencies.