shulijia/MNLP_M3_mcqa_model_base_mathqa_cot_orig
TEXT GENERATIONConcurrency Cost:1Model Size:0.8BQuant:BF16Ctx Length:32kPublished:Jun 8, 2025Architecture:Transformer Warm
The shulijia/MNLP_M3_mcqa_model_base_mathqa_cot_orig is a 0.8 billion parameter language model, fine-tuned from Qwen/Qwen3-0.6B-Base, with a context length of 32768 tokens. This model is specifically trained using Supervised Fine-Tuning (SFT) with TRL, focusing on multiple-choice question answering tasks. Its primary strength lies in its ability to process and generate responses for complex reasoning problems, particularly those involving mathematical contexts.
Loading preview...
Model Overview
The shulijia/MNLP_M3_mcqa_model_base_mathqa_cot_orig is a 0.8 billion parameter language model, fine-tuned from the Qwen/Qwen3-0.6B-Base architecture. It leverages a substantial context window of 32768 tokens, enabling it to handle longer and more complex inputs.
Key Capabilities
- Multiple-Choice Question Answering (MCQA): The model is specifically fine-tuned for MCQA tasks, suggesting enhanced performance in selecting correct answers from a given set of options.
- Mathematical Reasoning: While not explicitly detailed, the model name "mathqa_cot_orig" implies a focus on mathematical question answering, potentially utilizing Chain-of-Thought (CoT) reasoning.
- Supervised Fine-Tuning (SFT): Trained using the TRL library with SFT, indicating a direct optimization for specific task performance based on labeled data.
Good For
- Mathematical Problem Solving: Ideal for applications requiring the model to understand and answer mathematical questions.
- Automated Assessment Systems: Can be integrated into systems that automatically evaluate multiple-choice responses, especially in technical or quantitative fields.
- Research in Reasoning: Useful for researchers exploring the effectiveness of fine-tuning smaller models for specialized reasoning tasks.