File size: 884 Bytes
1867c2e
 
80725e9
 
1867c2e
80725e9
 
 
 
 
 
1867c2e
80725e9
 
 
1867c2e
 
 
 
 
80725e9
 
 
 
 
1867c2e
4b6d380
 
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
import gradio as gr
from PIL import Image
from transformers import ViTImageProcessor, ViTForImageClassification
import torch

# Load the image processor and model
processor = ViTImageProcessor.from_pretrained('wambugu1738/crop_leaf_diseases_vit')
model = ViTForImageClassification.from_pretrained(
    'wambugu1738/crop_leaf_diseases_vit',
    ignore_mismatched_sizes=True
)

# Define a function to make predictions
def classify_image(image):
    inputs = processor(images=image, return_tensors="pt")
    outputs = model(**inputs)
    logits = outputs.logits
    predicted_class_idx = logits.argmax(-1).item()
    return model.config.id2label[predicted_class_idx]

# Create the Gradio interface
app = gr.Interface(
    fn=classify_image,
    inputs=gr.Image(type="numpy"),  # Corrected input type
    outputs="text"
)
# Launch the Gradio app with a public link
app.launch(share=True)