import timm

# Loading https://huggingface.co/timm/eca_nfnet_l0
model = timm.create_model("hf-hub:timm/eca_nfnet_l0", pretrained=True)

import timm
import torch
from PIL import Image
from timm.data import resolve_data_config
from timm.data.transforms_factory import create_transform

# Load from Hub 🔥
model = timm.create_model(
    'hf-hub:nateraw/resnet50-oxford-iiit-pet',
    pretrained=True
)

# Set model to eval mode for inference
model.eval()

# Create Transform
transform = create_transform(**resolve_data_config(model.pretrained_cfg, model=model))

# Get the labels from the model config
labels = model.pretrained_cfg['label_names']
top_k = min(len(labels), 5)

# Use your own image file here...
image = Image.open('boxer.jpg').convert('RGB')

# Process PIL image with transforms and add a batch dimension
x = transform(image).unsqueeze(0)

# Pass inputs to model forward function to get outputs
out = model(x)

# Apply softmax to get predicted probabilities for each class
probabilities = torch.nn.functional.softmax(out[0], dim=0)

# Grab the values and indices of top 5 predicted classes
values, indices = torch.topk(probabilities, top_k)

# Prepare a nice dict of top k predictions
predictions = [
    {"label": labels[i], "score": v.item()}
    for i, v in zip(indices, values)
]
print(predictions)

[
    {'label': 'american_pit_bull_terrier', 'score': 0.9999998807907104},
    {'label': 'staffordshire_bull_terrier', 'score': 1.0000000149011612e-07},
    {'label': 'miniature_pinscher', 'score': 1.0000000149011612e-07},
    {'label': 'chihuahua', 'score': 1.0000000149011612e-07},
    {'label': 'beagle', 'score': 1.0000000149011612e-07}
]

python -m pip install huggingface_hub
hf auth login

from huggingface_hub import notebook_login
notebook_login()

import timm

# Build or load a model, e.g. timm's pretrained resnet18
model = timm.create_model('resnet18', pretrained=True, num_classes=4)

###########################
# [Fine tune your model...]
###########################

# Push it to the 🤗 Hub
timm.models.hub.push_to_hf_hub(
    model,
    'resnet18-random-classifier',
    model_config={'labels': ['a', 'b', 'c', 'd']}
)

# Load your model from the Hub
model_reloaded = timm.create_model(
    'hf-hub:<your-username>/resnet18-random-classifier',
    pretrained=True
)