llm2025_main_rev2
This is a LoRA adapter fine-tuned on structured output tasks (REV2 SFT). It focuses on extracting pure Output only, removing CoT/Reasoning/Approach.
Base Model
Qwen/Qwen3-4B-Instruct-2507
Training Data
- u-10bei/structured_data_with_cot_dataset_512_v4
- daichira/structured-hard-sft-4k
Method
- Output-only extraction (Robust Regex Filtering)
- Fence stripping
- Unsloth optimized training
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for fieldvalley-llm2025/llm2025_main_rev2
Base model
Qwen/Qwen3-4B-Instruct-2507