Fine-tuning

Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.

10,336 stars+299 this week

Turn any repository into verifiable RL environments for coding agents - Harbor tasks you can train on, evaluate and share on the Hugging Face Hub

780 stars+19 this week

🎨 NeMo Data Designer: Generate high-quality synthetic data from scratch or from seed data.

2,402 stars+12 this week

Infrastructure for continually self‑improving agents

7,787 stars+27 this week

Scikit-learn for Recursive Self-improvement (RSI). A modular framework for recursive harness evolution.

423 stars+21 this week

Fizgig — LoRA & Fine-tune Studio for Klein 9B, Krea 2, MiniMax H3, Qwen Image 2.1, Anima & Z-Image Turbo and Ace-Step XL: train, fine-tune, profile, repair and extract LoRAs & LoKR for Image, Video & Music Models

695 stars+38 this week

Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs (Qwen3.8, DeepSeek-V4, GLM-5.1, InternLM3, Llama4, ...) and 300+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, GLM5.3, Gemma4, Llava, Phi4, ...) (AAAI 2025).

16,064 stars+45 this week

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

21,880 stars+18 this week

Official Python inference and LoRA trainer package for the LTX-2 audio–video generative model.

10,156 stars+110 this week

FineEnvs — RL Environments 101: building and scaling RL environments in the age of LLMs

554 stars+11 this week

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

75,952 stars+116 this week

A browser panel for the chores around ComfyUI: models with Civitai covers, trigger words and base models; Model Check finds missing models when a workflow opens; search outputs by prompt or LoRA and drag one back to restore its workflow; saved workflows, combos, prompt editing, an activity log with undo, and character voices with Anomalous TTS.

191 stars+34 this week

Distributed AI Model Training and LLM Fine-Tuning on Kubernetes

2,278 stars+6 this week

Easily fine-tune (SFT/RL), evaluate, and deploy Qwen, Gemma, or any open agentic LLM/VLM!

9,429 stars+8 this week

Code for Machine Learning for Trading, 3rd edition — from data sourcing to live execution.

21,760 stars+99 this week

Harness-Zero: Harness Distillation via Agent-as-Harness

139 stars+5 this week

The AI developer platform. Use Weights & Biases to train and fine-tune models, and manage models from experimentation to production.

11,314 stars+6 this week

Automated alignment adjustment for LLMs — direct steering, LoRA, and MoE expert-granular abliteration, optimized via multi-objective Optuna TPE.

286 stars+9 this week

A unified orchestration layer for heterogeneous AI compute. It standardizes how to manage compute and run training and inference on GPU clouds, Kubernetes, VMs, or bare-metal clusters.

2,308 stars+8 this week

PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.

13,996 stars+45 this week

《开源大模型食用指南》针对中国宝宝量身打造的基于Linux环境快速微调（全参数/Lora）、部署国内外开源大模型（LLM）/多模态大模型（MLLM）教程

32,732 stars+61 this week

Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.

31,447 stars+13 this week

20+ high-performance LLMs with recipes to pretrain, finetune and deploy at scale.

13,748 stars+9 this week

Multi-engine LoRA trainer for game characters and art styles. Train MOSS-Image, Qwen-Image-2.1, Krea 2, Anima and more in one workspace.

283 stars+4 this week

[ICML 2026] Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation" & Causal Forcing++

1,027 stars+9 this week

A high-throughput training framework for Speculative decoding models.

49 starsThis week: Not stated

Compile organizational policy into verifiable decision training data.

149 stars+99 this week

Governed neural weight editing and coding model adaptation. Native Rust patches, exact rollback, LoRA, Claude Code routing, and reproducible benchmarks.

13 stars+7 this week

Open-source code for training and scaling up energy-based foundation models.

301 stars+58 this week

MCore-Bridge: Providing Megatron-Core model definitions for state-of-the-art large models and making Megatron training as simple as Transformers — with support for 300+ large language models (Qwen3-Next, GLM-5.3, Deepseek-V4, MiniMax-2.7, ...) and 200+ multimodal large models (Qwen3.8, Qwen3-Omni, Gemma4, ...).

106 stars+1 this week

H2O LLM Studio - a framework and no-code GUI for fine-tuning LLMs. Documentation: https://docs.h2o.ai/h2o-llmstudio/

5,194 stars+3 this week

DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation

250 stars+10 this week

Low-code framework for building custom LLMs, neural networks, and other AI models

11,793 stars+2 this week

Democratizing Reinforcement Learning for LLMs

5,913 stars+6 this week

Generate High-Quality Synthetics, Train, Measure, and Evaluate in a Single Pipeline

895 stars+4 this week

A general fine-tuning kit geared toward image/video/audio diffusion models.

2,955 stars+4 this week

Synthetic data curation for post-training and structured data extraction

1,764 stars+3 this week

🧑‍🏫 60+ Implementations/tutorials of deep learning papers with side-by-side notes 📝; including transformers (original, xl, switch, feedback, vit, ...), optimizers (adam, adabelief, sophia, ...), gans(cyclegan, stylegan2, ...), 🎮 reinforcement learning (ppo, dqn), capsnet, distillation, ... 🧠

67,729 stars+83 this week

Example 📓 Jupyter notebooks that demonstrate how to build, train, and deploy machine learning models using 🧠 Amazon SageMaker.

11,003 stars+4 this week

The complete post-training stack for LLMs, plus RL for any problem. SFT, DPO and multi-turn agentic RL, with evolutionary hyperparameter optimization built in.

966 stars+2 this week

Modular Reinforcement Learning (RL) library (implemented in PyTorch, JAX, and NVIDIA Warp) with support for Gymnasium/Gym, NVIDIA Isaac Lab, MuJoCo Playground and other environments

1,105 stars+3 this week

Techniques for deep learning with satellite & aerial imagery

10,301 stars+7 this week

One-click Portable Windows installation of 'AI-Toolkit by Ostris'

542 stars+3 this week

TFDS is a collection of datasets ready to use with TensorFlow, Jax, ...

4,610 stars+3 this week

High-throughput synthetic and pretraining dataset sifter for TypeSafe Jev. Rust streaming core, Parquet and JSONL I/O, typed Choice/Score/Noul judgments, speculative fan-out, 24.0 rows/sec measured.

110 stars+5 this week

verl for a single consumer GPU. PPO, GRPO and on-policy distillation on NVIDIA GPUs.

310 stars+3 this week

MedicalGPT: Training Your Own Medical GPT Model with ChatGPT Training Pipeline. 训练医疗大模型，实现了包括增量预训练(PT)、有监督微调(SFT)、RLHF、DPO、ORPO、GRPO。

5,888 stars+10 this week

An open-source robotics workshop for agents: simulate complete robots on high-fidelity physics, build digital twins, make the parts on CNC, laser and 3D-printer twins, connect simulation to real robots, and let AI agents program, test and debug it all, just by talking to it.

192 stars+3 this week

Lora beYond Conventional methods, Other Rank adaptation Implementations for Stable diffusion.

2,539 stars+2 this week

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

5,211 stars+3 this week

A high-performance RL training-inference weight synchronization framework, designed to enable second-level parameter updates from training to inference in RL workflows

179 stars0 this week

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

105 stars+1 this week

🏗️ Fine-tune, build, and deploy open-source LLMs easily!

543 stars+1 this week

Reinforcement Learning via Self-Distillation (SDPO)

1,143 stars+4 this week

SD-Trainer. LoRA & Dreambooth training scripts & GUI use kohya-ss's trainer, for diffusion model.

6,220 stars+17 this week

AI-powered synthetic data generation — structured tables, unstructured documents, multi-provider LLM support, referential integrity, and code-gen mode for millions of rows

178 stars0 this week

WildlifeDatasets: An open-source toolkit for animal re-identification

216 stars+2 this week

Reinforcement learning on general 2D physics environments in JAX. ICLR 2025 Oral.

284 stars+2 this week

Synthetic Data Generation Toolkit for LLMs

170 stars+1 this week

Agent-R1: Training Powerful LLM Agents with End-to-End Reinforcement Learning

1,717 stars+6 this week

A native Mac App for LLM fine-tuning on Apple Silicon — fully on-device, fully open source.

287 stars+3 this week

Deep learning training framework for image super resolution and restoration.

139 stars+2 this week

The Best TypeScript framework for fine-tuning

146 stars0 this week

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

1,231 stars+2 this week

Not stated

1,731 stars+9 this week

Modular reinforcement learning framework for algorithmic trading

430 stars+1 this week

Minimal and Clean Reinforcement Learning Examples

3,678 stars+1 this week

Single-GPU LLM fine-tuning preflight: memory estimates, runnable LoRA/QLoRA recipes, and local measurements.

795 stars0 this week

Build, personalize and control your own LLMs. From data pre-processing to fine-tuning, xTuring provides an easy way to personalize open-source LLMs. Join our discord community: https://discord.gg/TgHXuSJEk6

2,679 stars+2 this week

From-scratch PyTorch replications of classic and SOTA AI/ML papers: transformers, attention, GANs, diffusion, CLIP, LoRA, MoE, plus SFT/GRPO post-training experiments.

429 stars0 this week

Decentralized LLMs fine-tuning and inference with offloading

135 stars0 this week

Deep Reinforcement Learning: Zero to Hero!

2,293 stars0 this week

🦞 Just talk to your agent — it learns and EVOLVES 🧬.

3,467 stars+2 this week

Source code of paper "RLCSD: Reinforcement Learning with Contrastive On-Policy Self-Distillation"

116 stars+1 this week

LoRA (Low-Rank Adaptation) inspector for Stable Diffusion

101 stars0 this week

Model Training for ADI's MAX78000 and MAX78002 Edge AI Devices

117 stars0 this week

从 MiniMind 源码读起,再延伸到现代大模型技术体系的中文学习笔记。主线逐行精读预训练 / SFT / DPO / PPO / GRPO 与训练机制;附录 17 篇进阶卷覆盖量化、投机解码、RLHF 全景、模型代际史等 MiniMind 没涉及、但进阶绕不开的主题。

217 stars0 this week

Pairwise LLM judges (A/B/tie): budget-aware multi-turn packing, position-bias correction, pseudo-label distillation. Generalized from the 4th-place (gold) solution to Kaggle LMSYS Chatbot Arena.

170 stars0 this week

[CVPR 2026] Official release of "Spatial-SSRL: Enhancing Spatial Understanding via Self-Supervised Reinforcement Learning"

141 stars0 this week

[ICLR 2026] An official implementation of "CapRL: Stimulating Dense Image Caption Capabilities via Reinforcement Learning"

236 stars0 this week

LightRFT: Light, Efficient, Omni-modal & Reward-model Driven Reinforcement Fine-Tuning Framework

379 stars0 this week

A companion to CMU professor Zico Kolter's Intro to Modern AI. Learn the basics of machine learning, then train your own LLM from scratch.

117 stars0 this week

collection of text2cypher datasets, evaluations, and finetuning instructions

245 stars+1 this week

Scalable and extensible reinforcement learning for LM agents.

123 stars+1 this week

Local AI runtime for training & running small LLMs directly on Apple Neural Engine (ANE). No CoreML. No Metal. Offline, on-device fine-tuning & inference on M-series silicon.

133 stars0 this week

World's most accurate password guessing AI tool. A PyTorch implementation of PassLLM (USENIX 2025) that leverages PII and LoRA fine-tuning to outperform existing tools by over 45% on consumer hardware.

130 stars+1 this week

Code For All-Quadrant Bounded Clipping GRPO. arxiv.org/pdf/2601.03895

438 stars0 this week

AI-assisted media curator for large image/video datasets. Streamlined captioning, cropping, masking for LoRA/diffusion training workflows.

162 stars0 this week

[ICML 2026] a unified reinforcement learning toolbox for joint RL on language models and diffusion models

102 stars+1 this week

Human-feedback and data-governance infrastructure for RSI research: annotation, review, provenance and frozen dataset candidates.

100 stars0 this week

Hypernetworks that update LLMs to remember factual information

841 stars+3 this week

Dataflow-Oriented Reinforcement Learning for (Multi-)Agentic LLMs

114 stars0 this week

Finetune Falcon, LLaMA, MPT, and RedPajama on consumer hardware using PEFT LoRA

106 stars0 this week

Speedrunning LoRA fine-tuning: frozen task, frozen hardware, public wall-clock leaderboard. modded-nanogpt for fine-tuning.

151 stars+1 this week

Phi-3, -3.5, and -4 for Mac: Locally-run Vision and Language Models for Apple Silicon

280 stars0 this week