MM-EUREKA: Exploring the Frontiers of Multimodal Reasoning with Rule-based Reinforcement Learning
Les fonctionnalités principales de modalminds/mm-eureka sont : Critic-Free Algorithms, Reasoning Models, Reinforcement Learning Frameworks.
Les alternatives open-source à modalminds/mm-eureka incluent : sail-sg/understand-r1-zero. jiayi-pan/tinyzero — TinyZero is a reinforcement learning framework and implementation designed to train language models to develop… agentica-project/rllm — 🚀 Reinforcement Learning for Language Agents🌟. deep-agent/r1-v. hiyouga/easyr1 — EasyR1 is a distributed model training system and reinforcement learning framework for large language and… inclusionai/areal — AReaL is a system for agent orchestration, distributed model training, and parameter-efficient tuning. It provides a…
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
🚀 Reinforcement Learning for Language Agents🌟
