acomagu/matsuollm2025-advancedcompe-1

＜【課題】ここは自分で記入して下さい＞

This repository provides a LoRA adapter fine-tuned from Qwen/Qwen2.5-3B-Instruct.

Training Data

• HF dataset id: (not set)
• Local dataset path: out_dagger_alfworld_replay/iter_001/aggregate_messages_all.jsonl

Training Configuration

• Max sequence length: 1024
• Epochs: 1
• Learning rate: 2e-04
• LoRA: r=16, alpha=32

Notes

• Upload source adapter is expected to be the model trained after ALFWorld DAgger replay.