--- license: apache-2.0 language: - th metrics: - cer datasets: - rushai-dev/name_gen widget: - src: https://datasets-server.huggingface.co/assets/rushai-dev/name_gen/--/default/train/0/image/image.jpg example_title: นาง กชพร กระจ่างจิต - src: https://datasets-server.huggingface.co/assets/rushai-dev/name_gen/--/default/train/77/image/image.jpg example_title: นาง บวรรัช มั่นคง --- ```python from transformers import TrOCRProcessor, AutoTokenizer, ViTFeatureExtractor, VisionEncoderDecoderModel encode = 'rushai-dev/THAI-TrOCR' decode = "xlm-roberta-base" tokenizer = AutoTokenizer.from_pretrained(decode) feature_extractor = ViTFeatureExtractor.from_pretrained(encode) processor = TrOCRProcessor(feature_extractor=feature_extractor, tokenizer=tokenizer) model = VisionEncoderDecoderModel.from_pretrained(encode) ``` ```python from PIL import Image image = Image.open("xxxxxxx.png").convert("RGB") image pixel_values = processor(image, return_tensors="pt").pixel_values generated_ids = model.generate(pixel_values) generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0] generated_text ```