---
base_model: Qwen/Qwen2.5-VL-72B-Instruct
library_name: peft
pipeline_tag: image-text-to-text
tags:
- base_model:adapter:Qwen/Qwen2.5-VL-72B-Instruct
- lora
- transformers
- orthopaedic-surgery
- frame-captioning
---

# opscribe-captioner-ortho-tool-v2

SVD task arithmetic merge for Qwen2.5-VL-72B-Instruct surgical frame captioning.

| Adapter | Weight | Rank | Data |
|---|---|---|---|
| captioner_voiceover_v1 | 0.40 | 16 | Surgeon-narrated ortho voiceover dataset — general captioning |
| captioner_v2 | 0.60 | 32 | Targeted v2 dataset (793 train) — priority-tool pass for mallet, reamer, Bovie, saw |

Merged to rank 32 (alpha=32, scale=1.0) via efficient thin-SVD task arithmetic.

## Training Details
- captioner_v2 best val_loss: 0.0963 (3 epochs)
- Hardware: 4× NVIDIA H200, device_map=auto, bf16
- LoRA targets: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj