thinkonward

/

geophysical-foundation-model

like 6

Follow

ThinkOnward 22

Hi there!

Niels from the open-source team at Hugging Face. Glad to see you adopted the PyTorchModelHubMixin class. Just wondering why snapshot_download is used?

The point of the Mixin class is to allow to do just this:

from GFM import ElasticViTMAE

model = ElasticViTMAE.ElasticViTMAE.from_pretrained("thinkonward/geophysical-foundation-model")

instead of:

import torch
from huggingface_hub import snapshot_download

MODEL_REPO_ID = "thinkonward/geophysical-foundation-model"
snapshot_download(repo_id=MODEL_REPO_ID, repo_type="model", local_dir="./gfm-weights")

# import the architecture from the GitHub repository
from GFM import ElasticViTMAE
model = ElasticViTMAE.ElasticViTMAE()

The from_pretrained method will automatically equip the model with the weights.