Create app.py

app.py

@@ -0,0 +1,31 @@
+from huggingface_hub import list_models
+import gradio as gr
+from transformers import pipeline
+
+# Fetch timm models from Hugging Face Hub
+timm_models = list_models(filter="timm", sort="downloads", limit=20)  # Fetch top 20 based on downloads
+model_ids = [model.modelId for model in timm_models]
+
+# Initialize a pipeline with a default model
+default_model = model_ids[0]
+pipe = pipeline("image-classification", model=default_model)
+
+# Function for classification
+def classify(image, model_name):
+    pipe.model = model_name  # Update model dynamically
+    results = pipe(image)
+    return {result["label"]: round(result["score"], 2) for result in results}
+
+# Gradio Interface
+demo = gr.Interface(
+    fn=classify,
+    inputs=[
+        gr.Image(type="pil", label="Upload an Image"),
+        gr.Dropdown(choices=model_ids, label="Select timm Model", value=default_model)
+    ],
+    outputs=gr.Label(num_top_classes=3, label="Top Predictions"),
+    title="timm Model Image Classifier",
+    description="Select a timm model and upload an image for classification."
+)
+
+demo.launch()