Initial upload of the application and requirements files

app.py

@@ -0,0 +1,38 @@
+import gradio as gr
+from transformers import CLIPProcessor, CLIPModel
+
+model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
+processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
+
+
+def inference(input_img, captions):
+    captions_list = captions.split(",")
+    inputs = processor(text=captions_list, images=input_img, return_tensors="pt", padding=True)
+    outputs = model(**inputs)
+    # this is the image-text similarity score
+    logits_per_image = outputs.logits_per_image  
+    probs = logits_per_image.softmax(dim=1)
+    probabilities_percentages = ', '.join(['{:.2f}%'.format(prob.item() * 100) for prob in probs[0]])
+    return probabilities_percentages
+
+title = "CLIP Inference: Application using a pretrained CLIP model"
+description = "An application using Gradio interface that accepts an image and some captions, and displays a probability score with which each caption describes the image "
+
+examples = [["12863.jpg","photo of water, a photo of pizza, photo of a smiling lady, photo of luggage, photo of bank"],
+            ["12659.jpg","person riding bicycle, person driving car, photo of traffic lights, photo of vehicle, photo of light"],
+            ["12291.jpg","photo of a cat, a photo of a cat sleeping on keyboard, photo of desktop monitor, photo of typing keyboard, photo of computer mouse, photo of water glass"],
+            ["12272.jpg","person playing base ball, person holding bat, photo of a net, photo of audience behind net, photo of currency"],
+            ["9309.jpg","a photo of cows, a photo of grass, group of cows grazing grass, photo of electric pole, photo of trees"],
+            ["3805.jpg","a photo of water, Zebras drinking water, photo of a bird swimming, photo of grass"],
+            ["2788.jpg","a photo of a man dropping pigeon feed, a photo of pigeons, photo of a man feeding pigeons, photo of water, people walking, a photo of old building in the background"]
+           ]
+
+demo = gr.Interface(
+    inference,
+    inputs = [gr.Image(shape=(416, 416), label="Input Image"), gr.Textbox(placeholder="Enter different captions for image, separated by comma")],
+    outputs = [gr.Textbox(label="Probability score of captions")],
+    title = title,
+    description = description,
+    examples = examples,
+)
+demo.launch()

requirements.txt

@@ -0,0 +1,7 @@
+torch
+torchvision
+pillow
+numpy
+gradio
+pytorch-lightning
+transformers