---
license: apache-2.0
base_model: openbmb/MiniCPM5-1B
tags:
- build-small-hackathon
- minicpm
- lora
- structured-output
- eyewitness
- trl
- sft
language:
- en
- es
---

# MiniCPM5-1B-eyewitness — testimony→attribute parser

LoRA fine-tune of **openbmb/MiniCPM5-1B** that turns messy bilingual (EN/ES) eyewitness
descriptions ("caterpillar eyebrows, pinta de no haber dormido, un gorro de lana") into
strict attribute JSON over a closed vocabulary. Powers the
[EYEWITNESS](https://huggingface.co/spaces/build-small-hackathon/eyewitness) game.

- **Training data**: 4,000 synthetic (testimony → labels) pairs generated by the game
  engine itself — ground truth by construction, no human labeling
  (generator: `train/gen_dataset.py` in the Space repo).
- **Recipe**: LoRA r=16 α=32 on all-linear, 2 epochs, bf16, TRL SFTTrainer, trained on Modal (A10G).
- **Eval loss**: 0.2846 (2% held-out split).

Built for the Build Small Hackathon (Thousand Token Wood track).