Agent frameworks & infrastructure

Models, fine-tuning & training

2,504 agents · most starred first

Models, fine-tuning & training, page 4 of 26

CodeLlama-13b-Instruct-hf★ 161

transformers pytorch safetensors llama text-generation llama-2

Huihui-gemma-4-12B-coder-fable5-composer2.5-v1-abliterated★ 158

transformers safetensors gemma4_unified image-text-to-text abliterated uncensored

cartpole★ 157

OpenAI's cartpole env solver.

Qwen2.5-Coder-3B-Instruct★ 157

transformers safetensors qwen2 text-generation code codeqwen

open-jev★ 155

an open-source, from-first-principles reconstruction of the ideas behind TypeSafe AI's Jev, written in pytorch

TAIDE-LX-7B-Chat★ 154

transformers safetensors llama text-generation conversational license:other

Seed-Coder-8B-Reasoning★ 154

transformers safetensors llama text-generation conversational arxiv:2506.03524

rubra★ 153

Open Weight, tool-calling LLMs

Vistral-7B-Chat★ 152

transformers pytorch safetensors mistral text-generation LLMs

RedPajama-INCITE-Chat-3B-v1★ 151

transformers pytorch gpt_neox text-generation en dataset:togethercomputer/RedPajama-Data-1T

tf-agent★ 150

tensorflow reinforcement learning agents for OpenAI gym environments

gemma-4-12B-it-assistant★ 150

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

whitelightning★ 149

WhiteLightning distills massive, state-of-the-art language models into lightweight, hyper-efficient text classifiers. It's a command-line tool that lets you create specialized models that run anywhere—from the cloud to the edge—using the universal ONNX format for maximum compatibility.

PPO★ 149

PPO implementation for OpenAI gym environment based on Unity ML Agents

Yi-1.5-9B-Chat★ 149

transformers safetensors llama text-generation conversational arxiv:2403.04652

CodeLlama-7b-Python-hf★ 149

transformers pytorch safetensors llama text-generation llama-2

deepseek-coder-6.7b-base★ 149

transformers pytorch safetensors llama text-generation license:other

DeepSeek-Coder-V2-Lite-Base★ 149

transformers safetensors deepseek_v2 text-generation conversational custom_code

Kwaipilot_KAT-Coder-V2.5-Dev-GGUF★ 149

gguf code agent agentic-coding moe coding

agentic-rl★ 149

Agentic RL 中文零基础教程(33 章):从概念到 GRPO 实战,含 TRL 最小可跑示例;26–33 章附一套可运行的三方判别模型实证工程(encoder vs LLM-LoRA vs 规则基线)。第 25 章讲清 Jev / TypeSafe System One 与 RL 的能力边界 | Chinese Agentic RL tutorial (33 chapters) + a reproducible discriminative-model benchmark

CodeBooga-34B-v0.1★ 148

transformers safetensors llama text-generation license:llama2 text-generation-inference

Llama3-TAIDE-LX-8B-Chat-Alpha1★ 147

transformers safetensors llama text-generation conversational license:other

Multi-Turn-RL-Agent★ 146
ChatTS-14B★ 146

transformers pytorch chatts feature-extraction text-generation conversational

Contra-PPO-pytorch★ 144

Proximal Policy Optimization (PPO) algorithm for Contra

Multi-Agent-Distributed-PPO-Traffc-light-control★ 144

multi agent RL for traffic light control in Sumo using distributed PPO

AgentFlow★ 143

The First Unified Agent Data Synthesis Framework for Custom Agentic Task with all-in-one envrionment

conversation-tensorflow★ 142

TensorFlow implementation of Conversation Models

MiroTrain★ 142

MiroTrain is an efficient and algorithm-first framework research agent.

Qwen1.5-MoE-A2.7B-Chat★ 142

transformers safetensors qwen2_moe text-generation chat conversational

Gym.NET★ 140

openai/gym's popular toolkit for developing and comparing reinforcement learning algorithms port to C#.

Qwen1.5-110B-Chat★ 139

transformers safetensors qwen2 text-generation chat conversational

gemma-4-E4B-it-assistant★ 138

transformers safetensors gemma4_assistant text-generation any-to-any arxiv:2607.02770

codegen25-7b-multi_P★ 138

transformers pytorch llama text-generation code dataset:bigcode/starcoderdata

Llama-2-7b-chat-hf-function-calling-v2★ 137

transformers safetensors gguf llama text-generation facebook

Gemma4-31b-uncensored-assistant★ 137

gradio region:us

RSI-Jev★ 137

Typed-decision models (noul / choice / score) trained by a self-improving loop of AI agents — checkpoints, the code that produced them, and every version that failed.

deepseek-coder-1.3b-base★ 136

transformers pytorch llama text-generation license:other text-generation-inference

gdmix★ 135

A deep ranking personalization framework

dc-rl★ 135

SustainDC is a set of Python environments for Data Center simulation and control using Heterogeneous Multi Agent Reinforcement Learning. Includes customizable environments for workload scheduling, cooling optimization, and battery management, with integration into Gymnasium.

TxT360★ 134

docker region:us

internlm-chat-20b★ 134

transformers pytorch internlm feature-extraction text-generation custom_code

openhermes-function-calling★ 133
Qwen3.8-27B-OBLITERATED-Mythos-Class-Agentic★ 133

transformers safetensors qwen3_5 image-text-to-text qwen qwen3.8

Athene-V2-Agent★ 132

transformers safetensors qwen2 text-generation RLHF Nexusflow

Transformer-in-generating-dialogue★ 131

An Implementation of 'Attention is all you need' with Chinese Corpus

tf2-transformer-chatbot★ 130

Transformer Chatbot in TensorFlow 2 with TPU support.

wine_variety_classification★ 129

Examples on how to use various LLM providers with a Wine Classification problem

TiktokenSharp★ 128

Token calculation for OpenAI models, using `o200k_base` `cl100k_base` `p50k_base` encoding.

CodeLlama-7b-hf★ 128

transformers pytorch safetensors llama text-generation facebook

agentinstruct★ 126

Code repo for "Agent Instructs Large Language Models to be General Zero-Shot Reasoners"

Hammer★ 126

Hammer: Robust Function-Calling for On-Device Language Models via Function Masking

AgentFly★ 123

Scalable and extensible reinforcement learning for LM agents.

agentcore-rl-toolkit★ 123

Scalable, Composable Reinforcement Learning for Real-World Agents with Bedrock AgentCore.

llama-2-7b-chat-hf★ 123

transformers pytorch llama text-generation text-generation-inference endpoints_compatible

DQN-Atari-Agents★ 122

DQN-Atari-Agents: Modularized & Parallel PyTorch implementation of several DQN Agents, i.a. DDQN, Dueling DQN, Noisy DQN, C51, Rainbow, and DRQN

DREAM★ 122

Scalable implementation of DREAM - Deep RL for multi-agent imperfect information games

Seed-Coder-8B-Instruct★ 120

transformers safetensors llama text-generation conversational arxiv:2506.03524

PPP-Agent★ 119

Training Proactive and Personalized LLM Agents

llm-agent-rl-lab★ 119

Reproducing and studying RL algorithms for LLM agents, including PPO, GRPO, GSPO, DAPO, OPD and beyond.

Chatbot-Training-Corpus★ 118

总结了一些可以用作聊天机器人训练实作的文字语聊,包含中英文不同语言

dialogue-generation★ 118

Generating responses with pretrained XLNet and GPT-2 in PyTorch.

CodeLlama-13b-hf★ 118

transformers pytorch safetensors llama text-generation llama-2

TinyLlama-1.1B-Chat-v0.6★ 118

transformers safetensors gguf llama text-generation conversational

Qwen1.5-14B-Chat★ 118

transformers safetensors qwen2 text-generation chat conversational

TCOD★ 116

[COLM 2026]TCOD: Exploring Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agents

Qwen1.5-32B-Chat★ 116

transformers safetensors qwen2 text-generation chat conversational

granite-8b-code-instruct-4k★ 115

transformers safetensors llama text-generation code granite

aibuildai-llm-posttrain-agent★ 115

AIBuildAI LLM-Post-Train Agent: a recursive self-improving (RSI) agent for autonomous LLM post-training

astraflow★ 114

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

gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2★ 112

transformers safetensors gemma4_unified image-text-to-text gemma4 coding

aarambh-studio★ 112

🦀 Decoder-only LLM built from scratch in pure Rust using Candle — no Python, no PyTorch. Gated DeltaNet + sparse attention, fine-grained MoE, native video/document understanding, long-horizon tool agents, quantization-aware training. Scales: Tiny (25M) to Large (1.3B).

Agents-A1-4B★ 111

transformers safetensors qwen3_5 image-text-to-text vlm vision

Distributed-MADDPG★ 108

Distributed Multi-Agent Cooperation Algorithm based on MADDPG with prioritized batch data.

reasoning-bank-slm★ 107

An experiment that applies Google Research's `ReasoningBank` technique to Small Language Models. This experiment hopes to show that the same gains from the ReasoningBank paper also applies to much smaller, less capable models.

Qwen-14B-Chat-Int4★ 107

transformers safetensors qwen text-generation custom_code zh

OpenGPTAndBeyond★ 107

Open efforts to implement ChatGPT-like models and beyond.

weight-space-programming-autoresearch★ 107

🧠 Let the agent write the weights, why not?

Multi-Agent-Mec_Offloading-Use_DRL★ 106

使用Drl来解决多智能体卸载问题

CodeQwen1.5-7B★ 105

transformers safetensors qwen2 text-generation pretrained en

CodeNinja-1.0-OpenChat-7B★ 104

transformers safetensors mistral text-generation code text-generation-inference

hanabi_SAD★ 103

Simplified Action Decoder for Deep Multi-Agent Reinforcement Learning

Pytorch-RL-CPP★ 103

A Repository with C++ implementations of Reinforcement Learning Algorithms (Pytorch)

CNN_Intent_Classification★ 102

CNN for intent classification task in a Chatbot

internlm-chat-7b★ 101

transformers pytorch internlm feature-extraction text-generation custom_code

codegen-350M-mono★ 101

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

Aegis-agent★ 101
EnvFactory★ 100

The official paper for EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL.

Lego-RL★ 100

Lego-RL: Harness-Native Reinforcement Learning for Coding Agents

axion1_0★ 100

Frontier-level agent capabilities with nothing more than high-quality data and a simple training recipe: 2 SFT + RL, no mid-training/DPO/OPD. Full recipe released.

Harness-R1★ 99

Official Repository of Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectories

training_hub★ 98

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

training-agents★ 98

A repo on resources for training agents.

MATD3implementation★ 97

Code for the RL method MATD3 described in the paper "Reducing Overestimation Bias in Multi-Agent Domains Using Double Centralized Critics"

MAT-Agent★ 97

MAT: Multi-modal Agent Tuning 🔥 ICLR 2025 (Spotlight)

Qwen2.5-Coder-0.5B-Instruct★ 97

transformers safetensors qwen2 text-generation code codeqwen

Step-5-Preview★ 97

StepFun's flagship model for agentic work

Okapi★ 96

Okapi: Instruction-tuned Large Language Models in Multiple Languages with Reinforcement Learning from Human Feedback

level-replay★ 96

This code implements Prioritized Level Replay, a method for sampling training levels for reinforcement learning agents that exploits the fact that not all levels are equally useful for agents to learn from during training.

torque★ 96

Torque is a Declarative, typesafe DSL for building synthetic LLM datasets — compose conversations like React components

More in Agent frameworks & infrastructure