opscribe-phase-detector-v1

LoRA adapter for surgical phase detection, fine-tuned on the EgoSurgery dataset.

  • Base model: Qwen/Qwen2.5-VL-7B-Instruct
  • Task: Surgical phase classification from video frames
  • Dataset: EgoSurgery (24,925 train / 2,769 val frames)
  • Framework: LLaMA-Factory, LoRA rank 32, 3 epochs, lr=1e-4
  • Train loss: 0.0286
  • Usage: OpScribe pipeline phase detection stage
Downloads last month
130
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for rbryant19/opscribe-phase-detector-v1

Adapter
(239)
this model