zhaohq/PureRL-1.5B-v7-stage1-reasoning
TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kPublished:May 20, 2026Architecture:Transformer Warm
The zhaohq/PureRL-1.5B-v7-stage1-reasoning model is a 1.5 billion parameter language model, fine-tuned from Qwen/Qwen2.5-Math-1.5B. Developed by zhaohq, it utilizes the GRPO method for training, which is designed to enhance mathematical reasoning capabilities. With a context length of 32768 tokens, this model is optimized for complex reasoning tasks, particularly those involving mathematical problem-solving.
Loading preview...
Model Overview
The zhaohq/PureRL-1.5B-v7-stage1-reasoning model is a 1.5 billion parameter language model, fine-tuned from the base Qwen/Qwen2.5-Math-1.5B architecture. It was developed by zhaohq and trained using the TRL library.
Key Capabilities
- Enhanced Reasoning: This model is specifically fine-tuned for reasoning tasks, building upon the mathematical capabilities of its base model.
- GRPO Training Method: It leverages the GRPO (Gradient-based Reinforcement Learning with Policy Optimization) method, as introduced in the DeepSeekMath paper, to improve its performance in complex problem-solving.
- Extended Context Window: Supports a context length of 32768 tokens, allowing for processing longer inputs and maintaining coherence over extended dialogues or problem descriptions.
Good For
- Mathematical Reasoning: Ideal for applications requiring robust mathematical problem-solving and logical deduction.
- Complex Query Handling: Suitable for use cases where the model needs to process and reason over detailed or multi-step instructions.
- Research and Development: Provides a foundation for further research into reinforcement learning-based fine-tuning for reasoning tasks.