minsu0567/Uni-IAD-R2-Qwen3.5-answer-last
The minsu0567/Uni-IAD-R2-Qwen3.5-answer-last is a 4.5 billion parameter language model, fine-tuned from unsloth/Qwen3.5-4B. This model is specifically optimized for tasks requiring the generation of the 'answer_last' format, leveraging a 32768 token context length. It is designed for applications where structured, final answer extraction or generation is a primary requirement.
Loading preview...
Model Overview
The minsu0567/Uni-IAD-R2-Qwen3.5-answer-last is a 4.5 billion parameter language model, fine-tuned from the unsloth/Qwen3.5-4B base model. It was trained on the PA_SFT_2_answer_last dataset, indicating a specialization in generating responses that prioritize the final answer.
Key Training Details
This model was fine-tuned with the following hyperparameters:
- Learning Rate: 1e-05
- Batch Size: 1 (train), 8 (eval)
- Gradient Accumulation Steps: 2
- Optimizer: ADAMW_BNB
- LR Scheduler: Cosine with 100 warmup steps
- Epochs: 1.0
Intended Use Cases
While specific use cases are not detailed in the original model card, the fine-tuning on an "answer_last" dataset suggests its utility in scenarios where the model needs to extract or formulate a concise, final answer from given context or instructions. This could include question-answering systems, summarization tasks focused on key outcomes, or structured data extraction where the ultimate answer is paramount.