metadata
license: apache-2.0
datasets:
- atasoglu/flickr8k-dataset
language:
- en
metrics:
- rouge
pipeline_tag: image-to-text
tags:
- image
- vision
Vision Encoder Decoder (ViT + GPT2) model that fine-tuned on flickr8k-dataset for image-to-text task.
Example:
from transformers import VisionEncoderDecoderModel, ViTImageProcessor, AutoTokenizer
import torch
from PIL import Image
# load models
feature_extractor = ViTImageProcessor.from_pretrained("atasoglu/vit-gpt2-flickr8k")
tokenizer = AutoTokenizer.from_pretrained("atasoglu/vit-gpt2-flickr8k")
model = VisionEncoderDecoderModel.from_pretrained("atasoglu/vit-gpt2-flickr8k")
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)
# load image
img = Image.open("example.jpg")
# encode (extracting features)
pixel_values = feature_extractor(images=[img], return_tensors="pt").pixel_values
pixel_values = pixel_values.to(device)
# generate caption
output_ids = model.generate(pixel_values)
# decode
preds = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
print(preds)
For more, see this awesome blog.