An implementation of GRPO for Unsloth's VLMs training
-
Updated
Aug 7, 2025 - Python
#
grpotrainer
Here are 10 public repositories matching this topic...
PARL (Parallel-Agent Reinforcement Learning) is a training paradigm that teaches models to decompose complex tasks into parallel subtasks and coordinate multiple agents simultaneously.
reinforcement-learning ai ml multi-agent rl swarms agents kimi rlhf agentic deepseek grpo grpotrainer subagents moonshotai
-
Updated
Mar 24, 2026 - Python
Your efficient and accurate answer verification system for RL training.
-
Updated
Jun 23, 2025 - Python
simpleR1: A Simple Framework for Training R1-like Models
reinforcement-learning ppo trl deepseek-r1 grpo r1-zero grpotrainer reinforcement-learning-with-verifiable-rewards
-
Updated
Aug 12, 2025 - Python
Recreating the minimal training methods of DeepSeek-R1 for small langauge models.
-
Updated
Feb 10, 2025 - Python
-
Updated
May 8, 2025 - Python
Public showcase of NovaLiveSystem: a biomimetic cognitive architecture with interoception and distributed intelligence.
computer-science ai computer-vision ml developer-tools adaptive-learning autonomous-agents cognitive-systems human-centered-ai sota-technique ai-research-and-development grpotrainer biomimetic-ai
-
Updated
Mar 5, 2026
This project implements the reasoning training pipeline introduced in the DeepSeek-R1 paper, applying Group Relative Policy Optimization (GRPO) to teach a 7B language model to reason step-by-step through mathematical problems.
-
Updated
May 13, 2026 - Python
Low-cost GRPO fine-tuning pipeline for valence-arousal state estimation and tone-aligned responses in Llama 3.2 1B.
emotional-intelligence supervised-machine-learning finetuning self-awareness emotion-classifier large-language-models llm supervised-finetuning finetuning-llms finetuning-large-language-models llama3-2 grpo grpotrainer self-aware-ai grpo-training
-
Updated
Jul 14, 2026 - Jupyter Notebook
Add this topic to your repo
To associate your repository with the grpotrainer topic, visit your repo's landing page and select "manage topics."