Upload app.py

app.py

@@ -0,0 +1,49 @@
+# Image captioning with ViT+GPT2
+from PIL import Image
+from transformers import VisionEncoderDecoderModel, ViTFeatureExtractor, PreTrainedTokenizerFast
+import requests
+model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+vit_feature_extactor = ViTFeatureExtractor.from_pretrained("google/vit-base-patch16-224-in21k")
+tokenizer = PreTrainedTokenizerFast.from_pretrained("distilgpt2")
+#url = 'https://d2gp644kobdlm6.cloudfront.net/wp-content/uploads/2016/06/bigstock-Shocked-and-surprised-boy-on-t-113798588-300x212.jpg'
+# with Image.open(requests.get(url, stream=True).raw) as img:
+#     pixel_values = vit_feature_extactor(images=img, return_tensors="pt").pixel_values
+# encoder_outputs = model.generate(pixel_values.to('cpu'),num_beams = 5)
+# generated_senetences = tokenizer.batch_decode(encoder_outputs, skip_special_tokens=True,)
+# generated_senetences
+# generated_senetences[0].split(".")[0]
+def vit2distilgpt2(img):
+    pixel_values = vit_feature_extactor(images=img, return_tensors="pt").pixel_values
+    encoder_outputs = generated_ids = model.generate(pixel_values.to('cpu'),num_beams=5)
+    generated_senetences = tokenizer.batch_decode(encoder_outputs, skip_special_tokens=True)
+
+    return(generated_senetences[0].split('.')[0])
+import gradio as gr
+inputs = [
+    gr.inputs.Image(type="pil",label="Original Images")
+]
+
+outputs = [
+    gr.outputs.Textbox(label = "Caption")
+]
+
+title = "Image Captioning using ViT + GPT2"
+description = "ViT and GPT2 are used to generate Image Caption for the uploaded image.COCO DataSet is used for Training"
+examples = [
+    ["Image1.png"],
+    ["Image2.png"],
+    ["Image3.png"]
+]
+
+
+
+
+gr.Interface(
+    vit2distilgpt2,
+    inputs,
+    outputs,
+    title=title,
+    description=description,
+    examples=examples,
+    theme="huggingface",
+).launch(debug=True, enable_queue=True)