add: README.md

README.md

@@ -1,4 +1,5 @@
-```
+## EXAMPLE
+```python
 import requests
 from PIL import Image
 from transformers import (
@@ -7,18 +8,25 @@ from transformers import (
     PreTrainedTokenizerFast,
 )
 
+# device setting
 device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+# load feature extractor and tokenizer
 encoder_model_name_or_path = "ddobokki/vision-encoder-decoder-vit-gpt2-coco-ko"
 feature_extractor = ViTFeatureExtractor.from_pretrained(encoder_model_name_or_path)
 tokenizer = PreTrainedTokenizerFast.from_pretrained(encoder_model_name_or_path)
+
+# load model
 model = VisionEncoderDecoderModel.from_pretrained(encoder_model_name_or_path)
 model.to(device)
 
+# inference
 url = 'http://images.cocodataset.org/val2017/000000039769.jpg'
 with Image.open(requests.get(url, stream=True).raw) as img:
     pixel_values = feature_extractor(images=img, return_tensors="pt").pixel_values
 
 generated_ids = model.generate(pixel_values.to(device),num_beams=5)
 generated_text = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
-print(generated_text)
+
+>> ['고양이 두마리가 소파에 누워 있다.']
 ```