Predict with all models

app.py

@@ -14,34 +14,35 @@ model_names = [
 ]
 
 
-def process(image_file, top_k, model_name):
-    p = pipeline("image-classification", model=model_name)
-    pred = p(image_file)
-    return {x["label"]: x["score"] for x in pred[:top_k]}
+def process(image_file, top_k):
+    labels = []
+    for m in model_names:
+        p = pipeline("image-classification", model=m)
+        pred = p(image_file)
+        labels.append({x["label"]: x["score"] for x in pred[:top_k]})
+    return labels
 
 
 # Inputs
 image = gr.Image(type="filepath", label="Upload an image")
 top_k = gr.Slider(minimum=1, maximum=5, step=1, value=5, label="Top k classes")
-model_selection = gr.Dropdown(
-    model_names, value="google/vit-base-patch16-224", label="Pick a model"
-)
 
 # Output
-labels = gr.Label()
+labels = [gr.Label(label=m) for m in model_names]
 
 description = "This Space lets you quickly compare the most popular image classifiers available on the hub, including the recent NAT and DINAT models. All of them have been fine-tuned on the ImageNet-1k dataset. Anecdotally, the three sample images have been generated with a Stable Diffusion model :)"
 
 iface = gr.Interface(
     theme="huggingface",
     description=description,
+    layout="horizontal",
     fn=process,
-    inputs=[image, top_k, model_selection],
-    outputs=[labels],
+    inputs=[image, top_k],
+    outputs=labels,
     examples=[
-        ["bike.jpg", 5, "google/vit-base-patch16-224"],
-        ["car.jpg", 5, "microsoft/swin-base-patch4-window7-224"],
-        ["food.jpg", 5, "facebook/convnext-base-224"],
+        ["bike.jpg", 5],
+        ["car.jpg", 5],
+        ["food.jpg", 5],
     ],
     allow_flagging="never",
 )