Create app.py

app.py

@@ -0,0 +1,38 @@
+import gradio as gr
+from transformers import AutoImageProcessor, AutoModelForImageClassification
+from PIL import Image
+import torch
+import numpy as np
+import space
+# Load model and processor from the Hugging Face Hub
+MODEL_REPO = "Rausda6/autotrain-yh172-uui7d"  # Replace with your actual model repo name
+model = AutoModelForImageClassification.from_pretrained(MODEL_REPO)
+processor = AutoImageProcessor.from_pretrained(MODEL_REPO)
+
+labels = model.config.id2label
+@spaces.GPU
+def classify_image(img: Image.Image):
+    inputs = processor(images=img, return_tensors="pt")
+    with torch.no_grad():
+        outputs = model(**inputs)
+        logits = outputs.logits
+        probs = torch.nn.functional.softmax(logits, dim=-1)[0]
+
+    # Build result dictionary with confidence values
+    probs_dict = {labels[i]: float(probs[i]) for i in range(len(probs))}
+    # Sort and format nicely
+    sorted_probs = sorted(probs_dict.items(), key=lambda x: x[1], reverse=True)
+    top_label, top_score = sorted_probs[0]
+
+    return {"Prediction": top_label, "Confidence": f"{top_score:.2%}"}, dict(sorted_probs)
+
+# Gradio interface
+demo = gr.Interface(
+    fn=classify_image,
+    inputs=gr.Image(type="pil"),
+    outputs=[gr.Label(label="Top Prediction"), gr.Label(num_top_classes=6, label="Class Probabilities")],
+    title="Image Classification with AutoTrain Model",
+    description="Upload a JPG image to classify it using the fine-tuned model."
+)
+
+demo.launch()