File size: 1,405 Bytes
beebda2 e095ac0 beebda2 e095ac0 beebda2 e095ac0 227a02e e095ac0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 |
import gradio as gr
from fastai.vision.all import PILImage, load_learner
categories = ('Cat', 'Dog')
interface_title = "FastAI Cat vs. Dog Classifier 🐶"
interface_description = """
The model was built on resnet18 and trained on Kaggle's Dogs vs Cats competition.
The full code is available here: https://www.kaggle.com/code/sagsan/dogs-vs-cats-fastai
To use it, upload an image or select one of the examples below.
The output shows model's confidence scores for each category.
"""
def get_labels(fn):
fn_str = str(fn).lower() # Get the filename as a lowercase string
if 'dog' in fn_str:
return 'dog'
elif 'cat' in fn_str:
return 'cat'
else:
# Crucial for safety, even if you assume the data is clean
# It catches any unexpected file that has neither 'dog' nor 'cat'
raise ValueError(f"File must be labeled 'dog' or 'cat', but is not: {fn}")
def classify_image(img):
pred, idx, probs = learn.predict(img)
return dict(zip(categories, map(float, probs)))
learn = load_learner("./model.pkl")
image = gr.Image(width=192, height=192)
label = gr.Label()
examples = ["3.jpg", "5.jpg", "6.jpg", "7.jpg", "8.jpg", "10.jpg", "14.jpg", "17.jpg", "23.jpg", "44.jpg"]
intf = gr.Interface(fn=classify_image, examples=examples, inputs=image, outputs=label, title=interface_title, description=interface_description)
intf.launch(inline=False)
|