hellokitty/image-captioning

1import torch2import gradio as gr 3import re 4from transformers import AutoTokenizer, ViTFeatureExtractor, VisionEncoderDecoderModel5 6device='cpu'7encoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"8decoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"9model_checkpoint = "nlpconnect/vit-gpt2-image-captioning"10feature_extractor = ViTFeatureExtractor.from_pretrained(encoder_checkpoint)11tokenizer = AutoTokenizer.from_pretrained(decoder_checkpoint)12model = VisionEncoderDecoderModel.from_pretrained(model_checkpoint).to(device)13 14def predict(image,max_length=64, num_beams=4):15  image = image.convert('RGB')16  image = feature_extractor(image, return_tensors="pt").pixel_values.to(device)17  clean_text = lambda x: x.replace('<|endoftext|>','').split('\n')[0]18  caption_ids = model.generate(image, max_length = max_length)[0]19  caption_text = clean_text(tokenizer.decode(caption_ids))20  return caption_text 21 22css = '''23h1#title {24  text-align: center;25}26h3#header {27  text-align: center;28}29img#overview {30  max-width: 800px;31  max-height: 600px;32}33img#style-image {34  max-width: 1000px;35  max-height: 600px;36}37'''38 39input = gr.inputs.Image(label="Upload your Image", type = 'pil', optional=True)40output = gr.outputs.Textbox(type="auto",label="Captions")41examples = [f"example{i}.jpg" for i in range(1,7)]42 43description= "Image captioning application made using transformers"44title = "Image Captioning 🖼️"45 46interface = gr.Interface(47        fn=predict,48        inputs = input,49        theme="grass",50        outputs=output,51        examples = examples,52        title=title,53        description=description,54        article = article,55    )56interface.launch(debug=True)