Agent frameworks & infrastructure

Models, fine-tuning & training

2,504 agents · most starred first

Models, fine-tuning & training, page 7 of 26

practical-rl★ 43

Reinforcement learning in Julia. Solving OpenAI gym.

LunarLander-v2_DeepRL★ 43

OpenAI LunarLander-v2 DeepRL-based solutions (DQN, DuelingDQN, D3QN)

QDPP★ 43

Multi-Agent Determinantal Q-Learning

ATLAS★ 43

A Continual Learning Framework for Production LLM Agents

afterimage★ 43

Generate conversational, tool-calling, structured-output, and preference datasets — easily and at scale

rs-bpe★ 43

A ridiculously fast Python BPE (Byte Pair Encoder) implementation written in Rust

RL.cu★ 43

RLVR training for LLM in CUDA/C++

TinyLlama-1.1B-Chat-v0.3★ 43

transformers pytorch safetensors llama text-generation en

Arch-Agent-7B★ 43

transformers safetensors qwen2 text-generation conversational en

persimmon-8b-chat★ 43

transformers pytorch persimmon text-generation license:apache-2.0 endpoints_compatible

codegen-2B-multi★ 43

transformers pytorch codegen text-generation arxiv:2203.13474 license:bsd-3-clause

Qwen-1_8B-Chat-Int4★ 43

transformers safetensors qwen text-generation custom_code zh

MARL★ 42

Implementation for mSAC methods in PyTorch

wink-naive-bayes-text-classifier★ 42

Naive Bayes Text Classifier

micronaut-optimizer★ 42
eval-protocol★ 42

Eval Protocol (EP) is an open solution for doing reinforcement learning fine-tuning on existing agents — across any language, container, or framework.

Yi-1.5-6B-Chat★ 42

transformers safetensors llama text-generation conversational arxiv:2403.04652

Everyone-Coder-4x7b-Base★ 42

transformers safetensors mixtral text-generation merge moe

Yi-Coder-1.5B-Chat★ 42

transformers safetensors llama text-generation conversational arxiv:2403.04652

Qwen1.5-72B-Chat-GPTQ-Int4★ 42

transformers safetensors qwen2 text-generation chat conversational

agentomics-ml★ 42

Autonomous agentic system for supervised machine learning model development.

deepseek-coder-33B-instruct-AWQ★ 42

transformers safetensors llama text-generation conversational base_model:deepseek-ai/deepseek-coder-33b-instruct

Tensorflow-Seq2Seq-Dialogs★ 41

Build conversation Seq2Seq models with TensorFlow

MAGDi★ 41

The code implementation of MAGDi: Structured Distillation of Multi-Agent Interaction Graphs Improves Reasoning in Smaller Language Models. Paper: https://arxiv.org/abs/2402.01620

ml-ralph★ 41

Autonomous ML agent for running experiments using Claude.

CodeGym★ 41

[ICLR2026] The official repository for the CodeGym project: "Generalizable End-to-End Tool-Use RL with Synthetic CodeGym"

Llama-2-7b-chat-hf-function-calling-v3★ 41

transformers safetensors llama text-generation facebook meta

limbic-tool-use-0.5B-32K★ 41

transformers safetensors qwen2 text-generation text-generation-inference unsloth

CFCQL★ 40

Code for NeurIPS2023 accepted paper: Counterfactual Conservative Q Learning for Offline Multi-agent Reinforcement Learning.

deepresearch-datagen-cli★ 40

Using deep research workflow to generate datasets for finetuning LLMs.

fast-reinforcement-learning★ 40

Important Note fastrl version 2 is being developed at fastrl. Note the link in the readme

EPO★ 40

The code for paper "EPO: Entropy-regularized Policy Optimization for LLM Agents Reinforcement Learning"

SPEAR★ 40

An RL Recipe for Building Agentic LLMs via Self-Imitation on Long-Horizon Agentic Tasks

tesserack★ 40

Compiling strategy guides into reward functions for reinforcement learning. Uses Claude Vision to extract unit tests from game guides, then trains agents with dense, interpretable rewards.

open-trajectory-gym★ 40

Agentic RL post-training framework

llm-merge-adapter★ 40

gradio region:us

granite-3b-code-instruct-2k★ 40

transformers safetensors llama text-generation code granite

Qwen1.5-32B-Chat-GPTQ-Int4★ 40

transformers safetensors qwen2 text-generation chat conversational

LatentSkill★ 40

LatentSkill: From In-Context Textual Skills to In-Weight Latent Skills for LLM Agents

Mosaic★ 40

Exploring Multi-Session Agentic Reinforcement Learning

maddpg★ 39

Implementation of Multi-Agent Deep Deterministic Policy Gradients

MAD-OPD★ 39

Official code for "Breaking the Ceiling in On-Policy Distillation via Multi-Agent Debate" (arXiv:2605.01347).

genetic-algorithm★ 39

A dependency-free Python genetic-algorithm engine with pluggable fitness criteria — the reusable search core for an LLM/AI autoresearcher.

Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-NVFP4-MTP★ 39

transformers safetensors qwen3_5 image-text-to-text abliterated uncensored

Trendyol-LLM-7b-chat-v1.0★ 39

transformers safetensors mistral text-generation conversational tr

Deep-Reinforcement-Learning★ 38

A collection of several Deep Reinforcement Learning techniques (Deep Q Learning, Policy Gradients, ...), gets updated over time.

granite-3b-code-base-2k★ 38

transformers safetensors llama text-generation code granite

Agents-A1-Q8_0-GGUF★ 38

gguf quantized moe vlm vision agentic

CodeLlama-34B-Python-GGUF★ 38

transformers gguf llama llama-2 text-generation code

GLM-5.3-Flash-Coder-RCO-152E-GGUF★ 38

gguf moe expert-pruning rco code agentic

keras-seq2seq-chatbot-with-attention★ 37

It is a seq2seq encoder decoder chatbot using keras and with attention

ferret★ 37

An extensible RL framework for training LLM agents with advanced search capabilities, built on VERL and supporting state-of-the-art search strategies.

Trendyol-LLM-7B-chat-v4.1.0★ 37

safetensors qwen2 text-generation conversational tr license:apache-2.0

Yi-1.5-9B-Chat-16K★ 37

transformers safetensors llama text-generation conversational arxiv:2403.04652

Qwen2.5-Coder-32B-Instruct-abliterated★ 37

transformers safetensors qwen2 text-generation code codeqwen

Qwen3.8-KAT-AgentWorld-35B-A3B-Coder-GGUF★ 37

transformers gguf merge moe a3b ties

Agent-Omit★ 36

Agent-Omit: Training Efficient LLM Agents for Adaptive Thought and Observation Omission via Reinforcement Learning

granite-20b-functioncalling★ 36

transformers safetensors gpt_bigcode text-generation conversational arxiv:2407.00121

internlm2-chat-1_8b★ 36

transformers safetensors internlm2 text-generation conversational custom_code

ModelRepo★ 35

reproduce some RL or Multi-Agent models

ChatGLM2-With-Rua-Tutorial★ 35

无需预算,使用你的个人数据克隆自己——赛博飞升!Clone yourself by tuning a LLM using your own data.

T3★ 35

(ICLR 2026 Oral) Code for the paper: Reducing Belief Deviation in Reinforcement Learning for Active Reasoning of LLM Agents

function-calling-data-synthesizer★ 35

Automated pipeline for generating, verifying, and preparing high-quality function calling datasets for model fine-tuning. Reduces manual annotation through multi-stage validation and conversion to training formats. Built for Azure ML with permissive license models (Phi3.5 MoE, Phi4 etc).

HiPER-agent★ 35

[ICML 2026] This is the official implementation for paper HiPER: Hierarchical Reinforcement Learning with Explicit Credit Assignment for Large Language Model Agents

Step-RL★ 35

基于强化学习的 LLM Agent 长链路决策优化系统

simple-llm-finetuner★ 35

gradio region:us

codeparrot-small★ 35

transformers pytorch tensorboard gpt2 text-generation code

CodeLlama-13B-Python-GGUF★ 35

transformers gguf llama llama-2 text-generation code

sharpening-tax★ 35

Official codebase for the paper "Sharpening Tax in Post-Training"

Qwen3-Coder-Next★ 35

transformers safetensors qwen3_next text-generation unsloth qwen

SDEN-Pytorch★ 34

Pytorch Implementation of "Sequential Dialogue Context Modeling for Spoken Language Understanding( https://arxiv.org/abs/1705.03455 )"

Multi-Agent-Diverse-Generative-Adversarial-Networks★ 34

Easy-to-follow Pytorch tutorial Notebook for Multi-Agent-Diverse-Generative-Adversarial-Networks

Matlab_PPO★ 34

This is a MATLAB-based reinforcement learning framework that includes the Proximal Policy Optimization (PPO) algorithm and its multi-agent extension (MAPPO). It supports GPU acceleration and parallel computing, making it suitable for research and engineering applications in control systems.

AutoAdapt★ 34

AutoAdapt: An Automated Domain Adaptation Framework for Large Language Models

stablelm-2-1_6b-chat★ 34

transformers safetensors stablelm text-generation causal-lm conversational

vinallama-7b-chat★ 34

transformers pytorch safetensors llama text-generation conversational

CodeNinja-1.0-OpenChat-7B-GGUF★ 34

transformers gguf mistral code text-generation-inference text-generation

mmi_anti_pytorch★ 33

A pytorch implementation for the MMI-anti model

magnet★ 33

the small distributed language model toolkit; fine-tune state-of-the-art LLMs anywhere, rapidly

Mistral-7B-Instruct-v0.1-function-calling-v2★ 33

transformers safetensors gguf mistral text-generation pytorch

Llama-2-13b-chat-hf★ 33

transformers pytorch safetensors llama text-generation facebook

Llama-2-7B-Chat-fp16★ 33

transformers pytorch safetensors llama text-generation facebook

gemma-4-31B-it-qat-q4_0-unquantized-assistant★ 33

transformers safetensors gemma4_assistant text-generation image-text-to-text conversational

gemma-4-12B-it-qat-q4_0-unquantized-assistant★ 33

transformers safetensors gemma4_unified_assistant text-generation any-to-any arxiv:2607.02770

Jais-2-8B-Chat★ 33

transformers safetensors jais2 text-generation conversational ar

Llama-3-8B-Instruct-Coder-v2-GGUF★ 33

transformers gguf text-generation-inference unsloth llama trl

Mixtral-8x7B-Instruct-v0.1-function-calling-v3★ 32

transformers safetensors mixtral text-generation function calling function-calling

Llama-3-Trendyol-LLM-8b-chat-v2.0★ 32

transformers safetensors llama text-generation conversational tr

gpt2-multiplication★ 32

gradio region:us

Phi-4-multimodal-instruct-abliterated★ 32

transformers safetensors phi4mm text-generation nlp code

tora-70b-v1.0★ 32

transformers pytorch llama text-generation code math

TinyMistral-248M-Chat-v4★ 32

transformers safetensors mistral text-generation conversational en

Building-Math-Agents-with-Multi-Turn-Iterative-Preference-Learning★ 31

This is an official implementation of the paper ``Building Math Agents with Multi-Turn Iterative Preference Learning'' with multi-turn DPO and KTO.

MultiAgent-PPO★ 31

Proximal Policy Optimization with Beta distribution - uses multi agent Unity ML Tennis

Lifelong-Hanabi★ 31

A Continual Multi-agent RL testbed based on Hanabi

dfac★ 31

[ICML 2021] DFAC Framework: Factorizing the Value Function via Quantile Mixture for Multi-Agent Distributional Q-Learning

ARIA★ 31

Source code for our paper: "ARIA: Training Language Agents with Intention-Driven Reward Aggregation".

ARC-Neuron-LLMBuilder★ 31

A governed local AI build-and-memory system that trains small brains, compares them, protects the better one, archives the worse one, and preserves the evidence of why. v1.0.0/governed-v2.2.0+

Llama-2-13b-chat-norwegian★ 31

transformers pytorch safetensors llama text-generation llama-2

Xe-Forge★ 31

Multi-stage LLM agent pipeline for optimizing AI kernels on Intel GPU — from analysis to autotuning.

More in Agent frameworks & infrastructure