Projects sharing features with DLLM RL

This software project accompanies the research paper, DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation.

Reinforcement Learning in Go

[Paper](https://arxiv.org/pdf/2508.13148) [Project](https://cli212.github.io/MDPO/)

Blockwise SFT for Diffusion Language Models: Reconciling Bidirectional Attention and Autoregressive Decoding

This is the official implementation of the paper d3LLM: Ultra-Fast Diffusion LLM using Pseudo-Trajectory Distillation (ICML'26), where we introduce a novel recipe for building an ultra-fast diffusion language model named d3LLM (pseuDo-Distilled Diffusion LLM) 🚀.

ChainerRL is a deep reinforcement learning library built on top of Chainer.

The repository contains the code for the DRAKES method presented in the paper: Fine-Tuning Discrete Diffusion Models via Reward Optimization with Applications to DNA and Protein design (ICLR 2025). DRAKES is a fine-tuning method for reward optimization or alignment in discrete diffusion models,…

d1: Scaling Reasoning in Diffusion Large Language Models via Reinforcement Learning A two-stage approach combining masked SFT with diffu -GRPO—a novel policy gradient method based on GRPO that features efficient log probability estimation designed for masked dLLMs—to scale reasoning capabilities…

Training Optimal Large Diffusion Language Models Jinjie Ni†, Qian Liu, Chao Du, Longxu Dou, Hang Yan, Zili Wang, Tianyu Pang, Michael Qizhe Shieh

SPG: Sandwiched Policy Gradient for Masked Diffusion Language Models

```bash conda create -n tad python=3.10 -y conda activate tad

Are Autoregressive Large Language Models Implicit Teachers for Diffusion Large Language Models? A comprehensive framework for transferring alignment knowledge from AR-LLMs to Diffusion Models

EasyR1 is a distributed model training system and reinforcement learning framework for large language and vision-language models. It functions as a multimodal trainer and an implementation of a Proximal Policy Optimization pipeline designed to refine the reasoning and perception capabilities of models that process both text and images. The system specializes in distributing reinforcement learning workloads across multiple compute nodes to manage high memory requirements. It optimizes hardware utilization through padding-free training and fine-tuning to fit large models onto available graphics

Current Diffusion Language Models (DLMs) have been studied at a smaller scale compared to their autoregressive (AR) counterparts and lack fair comparison on language modeling benchmarks. Additionally, training diffusion models from scratch at scale remains challenging. We propose adapting…

NeurIPS 2024 Code for the paper "Diffusion of Thoughts: Chain-of-Thought Reasoning in Diffusion Language Models"

](https://huggingface.co/Dream-org/Dream-v0-Base-7B)

simpleRL-reason is a training framework designed to improve mathematical and logical deduction in large language models. It utilizes reinforcement learning and policy optimization to enhance the accuracy and transparency of step-by-step deduction chains. The project implements a pipeline that establishes baseline capabilities through supervised fine-tuning before applying reinforcement learning to maximize deductive accuracy. It features a reward modeling toolkit that calculates scalar feedback by comparing generated reasoning steps against verified mathematical ground truths. The framework

This library provides a comprehensive framework for fine-tuning, aligning, and distilling transformer-based language models. It serves as a toolkit for adapting models to specialized domains through supervised learning, while offering advanced methodologies to improve output quality and reasoning capabilities. The project distinguishes itself through specialized alignment and optimization techniques, including direct preference optimization and reinforcement learning, which allow models to be tuned against human preferences without complex reward modeling. It further supports training efficie

AReaL is a system for agent orchestration, distributed model training, and parameter-efficient tuning. It provides a framework for developing multi-turn reasoning agents and training large models using reinforcement learning from human feedback. The project implements a toolkit for improving the visual reasoning and geometry problem solving capabilities of vision-language models. It utilizes a memory-efficient tuning system to optimize mathematical and reasoning models across different inference backends. The infrastructure supports large-scale training through tensor, pipeline, and expert p

🕹️ A diverse suite of scalable reinforcement learning environments in JAX

TinyZero is a reinforcement learning framework and implementation designed to train language models to develop reasoning and self-verification abilities. It provides a training pipeline to optimize model performance on mathematical and logical tasks. The project serves as a minimal reproduction of the DeepSeek R1 architectural and training approach. It focuses on creating reasoning models that can solve structured problems through autonomous chain-of-thought discovery. The framework incorporates group relative policy optimization and reward-based self-correction to improve accuracy on logica

MegaDLMs

The implementation of "Addressing the Training-Inference Discrepancy in Discrete Diffusion for Text Generation"

Julia code for the book Reinforcement Learning An Introduction

Persistent advantage learning dueling double DQN for the Arcade Learning Environment

🚀 Reinforcement Learning for Language Agents🌟