cjiao/goldengoose-divsweep_goose_n128_random_seed100-25grp
The cjiao/goldengoose-divsweep_goose_n128_random_seed100-25grp model is a 1.5 billion parameter instruction-tuned language model, fine-tuned from Qwen/Qwen2.5-1.5B-Instruct by cjiao. It was trained using the TRL framework and incorporates the GRPO method, which is designed to enhance mathematical reasoning capabilities. With a context length of 32768 tokens, this model is optimized for tasks requiring robust mathematical problem-solving and general instruction following.
Loading preview...
Model Overview
The cjiao/goldengoose-divsweep_goose_n128_random_seed100-25grp is a 1.5 billion parameter instruction-tuned language model. It is a fine-tuned variant of the Qwen/Qwen2.5-1.5B-Instruct base model, developed by cjiao.
Key Capabilities & Training
This model's primary differentiator lies in its training methodology. It was fine-tuned using the TRL (Transformer Reinforcement Learning) framework and specifically leveraged the GRPO (Gradient-based Reward Policy Optimization) method. GRPO is a technique introduced in the research paper "DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models" (arXiv:2402.03300), indicating a focus on improving mathematical reasoning abilities.
- Base Model: Qwen/Qwen2.5-1.5B-Instruct
- Parameter Count: 1.5 billion
- Context Length: 32768 tokens
- Training Method: Fine-tuned with TRL, incorporating GRPO for enhanced mathematical reasoning.
Use Cases
Given its specialized training with GRPO, this model is particularly well-suited for:
- Mathematical Reasoning Tasks: Applications requiring problem-solving, logical deduction, and numerical understanding.
- General Instruction Following: As an instruction-tuned model, it can handle a wide range of natural language prompts.
- Research and Development: Exploring the impact of GRPO on smaller language models for specific domains.