Virchow2
Successor to Virchow: ViT-H/14 pretrained on 3.1 million whole-slide images from about 225,000 patients across 45 countries, at mixed magnifications (5×–40×), with pathology-specific augmentations.
🤗 110 111 downloads / month ♥ 151 updated 2024-10-26 gatedAt a glance
What it does
Tile embeddings robust across magnifications and scanners, intended as a drop-in improvement over Virchow for every downstream slide-level task.
Tasks, data types and cancers
Architecture
Training data
3.1 million WSIs, ~225,000 patients, 45 countries; H&E plus a share of IHC; the largest multi-site pathology pretraining set reported at release.
Evaluation
| Benchmark / dataset | Metric | Value | External validation | Source |
|---|---|---|---|---|
| 12 zadan poziomu kafelka (tile-level) w patomorfologii obliczeniowej | porownanie z najlepszymi konkurencyjnymi modelami | wynik na poziomie stanu wiedzy (state of the art) na wszystkich 12 zadaniach w chwili publikacji; autorzy nie podaja w streszczeniu wartosci metryk ani nazw poszczegolnych zadan | no | Source |
| Skala modelu i pretreningu (dane deklarowane przez autorow) | liczba parametrow / rozmiar zbioru pretreningowego | Virchow2: 632 mln parametrow (vision transformer); pretrening na 3,1 mln preparatow histopatologicznych. W tej samej pracy opisano rowniez Virchow2G (1,9 mld parametrow) i Virchow2G Mini (22 mln, destylacja Virchow2G) — sa to ODREBNE modele, nie warianty tej karty | no | Source |
How to run
# generic timm loader — check the model card for the exact init args and image normalisation
import timm, torch
from huggingface_hub import login
login() # gated repos: accept the licence on huggingface.co first
model = timm.create_model('hf-hub:paige-ai/Virchow2', pretrained=True)
model.eval()
cfg = timm.data.resolve_data_config({}, model=model)
transform = timm.data.create_transform(**cfg)
# emb = model(transform(tile).unsqueeze(0)) # 1 x D tile embedding
Gated on Hugging Face; non-commercial. Same embedding recipe as Virchow (class token + mean patch tokens).
Regulatory status and intended use
Regulatory status is quoted from the source linked above and can change. Research-use-only models must not be used for clinical decisions.
Limitations and bias
- Preprint-level evaluation at release; independent benchmarks are still accumulating.
- Non-commercial licence.
Sources
This page is educational — it is not medical advice and does not replace consultation with an oncologist. Diagnostic and treatment decisions are made solely by specialist physicians.