KMnP/vpt

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1244 stars · 105 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Official PyTorch implementation of Visual Prompt Tuning (VPT), an ECCV 2022 method for parameter-efficient fine-tuning of vision transformers by learning input prompts. It provides training and hyperparameter tuning scripts for ViT and Swin backbones on FGVC and VTAB datasets.

Use cases

• reproduce visual prompt tuning experiments from the ECCV 2022 paper
• fine-tune pretrained vision transformers without updating backbone weights
• evaluate prompt-based transfer learning on VTAB benchmarks
• compare shallow vs deep prompting on fine-grained classification tasks
• adapt ViT or Swin models to new image classification datasets efficiently

When to choose

• you need parameter-efficient adaptation of vision transformers
• you want the official reference implementation of VPT for research
• you are benchmarking prompt tuning against full fine-tuning on VTAB or FGVC datasets

When to avoid

• you need a production-ready training framework with broad model support
• you require non-ViT/Swin architectures or the latest PEFT methods
• you want a maintained library with frequent updates and long-term support

Facets

library · maturity maintenance

machine-learning deep-learning llm-training machine-learning computer-vision deep-learning python visual-prompt-tuning parameter-efficient-fine-tuning vision-transformer pytorch eccv-2022 transfer-learning research-code linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="KMnP/vpt")