gemma-3-27b-it-lora-pitvis-instruments

Gemma 3 27B-it LoRA — PitVis-2023 instrument recognition

LoRA adapter plus an 18-way linear classification head on google/gemma-3-27b-it, fine-tuned for multi-label instrument presence on PitVis-2023 endoscopic pituitary frames.

This is the “Gemma 3 27B fine-tuned” row on the Instruments tab of the SDSC × Chicago Booth surgical video understanding leaderboard.

Prompt example

This closed-set example mirrors the leaderboard format, not a text-input API for this checkpoint.

[surgical frame]

Which instruments are visible in this endoscopic pituitary-surgery frame?
Select every matching label.

- bipolar forceps
- cottle
- cup forceps
- dural scissors
- freer elevator
- haemostatic foam
- irrigation syringe
- kerrisons
- micro doppler
- nasal cutting forceps
- pituitary rongeurs
- retractable knife
- ring curette
- spatula dissector
- stealth pointer
- suction
- surgical drill
- tissue glue

Files

• adapter_model.safetensors / adapter_config.json — LoRA (r=128, alpha=256)
• classifier.pt — linear head (18 tools)
• model_config.json — tool vocabulary and training hyperparameters

You need the adapter and the classification head.

Training

• 10 epochs, lr 5e-6, effective batch size 2, seed 42
• 84,666 training frames / 30,896 validation frames

Evaluation

Full PitVis-2023 validation split (95% bootstrap CI):

References

• Skobelev, K., Fithian, E., Baranovski, Y., et al. A Comparative Study in Surgical AI: Potential and Limitations of Data, Compute, and Scaling. arXiv:2603.27341, 2026.
• Dataset: Das, A., Khan, D. Z., Psychogyios, D., et al. PitVis-2023 Challenge: Workflow Recognition in Videos of Endoscopic Pituitary Surgery. Medical Image Analysis, 106, 103716 (2025).

Limitations

Research baseline only. Not a medical device.