Hyeongwon/P2-split5_only_answer_Qwen3-4B-Base_0501-bs64-epoch6
Hyeongwon/P2-split5_only_answer_Qwen3-4B-Base_0501-bs64-epoch6 is a 4 billion parameter language model fine-tuned from Hyeongwon/Qwen3-4B-Base. This model was trained using Supervised Fine-Tuning (SFT) with the TRL framework. It is specifically designed to generate direct answers to questions, as indicated by its fine-tuning objective. The model supports a context length of 32768 tokens, making it suitable for processing longer prompts and generating concise, relevant responses.
Loading preview...
Model Overview
Hyeongwon/P2-split5_only_answer_Qwen3-4B-Base_0501-bs64-epoch6 is a 4 billion parameter language model, fine-tuned from the base model Hyeongwon/Qwen3-4B-Base. This model has been specifically trained using Supervised Fine-Tuning (SFT) with the TRL framework, indicating an optimization for generating direct and concise answers to user questions.
Key Capabilities
- Question Answering: Optimized to provide direct answers to questions, making it suitable for conversational AI or information retrieval tasks where succinct responses are preferred.
- Base Model: Built upon the Qwen3-4B-Base architecture, leveraging its foundational language understanding capabilities.
- Context Length: Supports a substantial context window of 32768 tokens, allowing it to process and understand longer prompts before generating an answer.
Training Details
The model was trained using the SFT method, a common technique for instruction-tuning language models. The training utilized the TRL (Transformer Reinforcement Learning) library, with specific framework versions including TRL 0.25.1, Transformers 4.57.3, PyTorch 2.9.1, Datasets 3.6.0, and Tokenizers 0.22.2. Further details on the training run can be visualized via Weights & Biases.