Create app.py

app.py

@@ -0,0 +1,42 @@
+import torch
+import gradio as gr
+from transformers import AutoTokenizer, VisionEncoderDecoderModel, ViTImageProcessor, pipeline
+
+# Initialize device and models for captioning
+device = 'cpu'
+encoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+decoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+model_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+feature_extractor = ViTImageProcessor.from_pretrained(encoder_checkpoint)
+tokenizer = AutoTokenizer.from_pretrained(decoder_checkpoint)
+caption_model = VisionEncoderDecoderModel.from_pretrained(model_checkpoint).to(device)
+
+# Initialize the image generation model (e.g., Stable Diffusion)
+image_gen_model = pipeline("text-to-image", model="CompVis/stable-diffusion-v1-4")
+
+def predict(image):
+    # Generate a caption from the image
+    image = image.convert('RGB')
+    image_tensor = feature_extractor(images=image, return_tensors="pt").pixel_values.to(device)
+    caption_ids = caption_model.generate(image_tensor, max_length=128, num_beams=3)[0]
+    caption_text = tokenizer.decode(caption_ids, skip_special_tokens=True)
+
+    # Generate an image from the caption
+    generated_images = image_gen_model(caption_text, num_images=1)
+
+    return caption_text, generated_images[0]
+
+# Set up Gradio interface
+input = gr.Image(label="Upload any Image", type='pil')
+outputs = [gr.Textbox(label="Caption"), gr.Image(label="Generated Image")]
+examples = [f"example{i}.jpeg" for i in range(1, 3)]
+
+title = "Image Captioning and Generation"
+interface = gr.Interface(
+    fn=predict,
+    inputs=input,
+    outputs=outputs,
+    examples=examples,
+    title=title,
+)
+interface.launch(debug=True)