Instructions to use timm/davit_base.msft_in1k with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- timm
How to use timm/davit_base.msft_in1k with timm:
import timm model = timm.create_model("hf_hub:timm/davit_base.msft_in1k", pretrained=True) - Transformers
How to use timm/davit_base.msft_in1k with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-classification", model="timm/davit_base.msft_in1k") pipe("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("timm/davit_base.msft_in1k", dtype="auto") - Notebooks
- Google Colab
- Kaggle
timm version
#1
by ahbpp - opened
Hey!
Thank you for training the model with this new breakthrough architecture. However, can you please add a note in the README about timm lib version? It was not obvious to me that timm --pre (see: hf discussion) version is required.
@ahbpp it's near the top of the timm README (https://github.com/huggingface/pytorch-image-models#whats-new), but it's a temporary situation so didn't think it made sense to include (and then have to remove for 900 models here as I put them up), will be out of pre soon as getting near my 90% of models on hub for cleaning up some deprecations and moving out of pre-release