Update app.py

app.py

@@ -9,21 +9,38 @@ import tensorflow as tf
 os.environ['TF_ENABLE_ONEDNN_OPTS'] = '0'
 
 device='cpu'
-encoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
-decoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
-model_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+# encoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+# decoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+# model_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
 
-feature_extractor = ViTFeatureExtractor.from_pretrained(encoder_checkpoint)
-tokenizer = AutoTokenizer.from_pretrained(decoder_checkpoint)
-model = VisionEncoderDecoderModel.from_pretrained(model_checkpoint).to(device)
+# feature_extractor = ViTFeatureExtractor.from_pretrained(encoder_checkpoint)
+# tokenizer = AutoTokenizer.from_pretrained(decoder_checkpoint)
+# model = VisionEncoderDecoderModel.from_pretrained(model_checkpoint).to(device)
 
-def predict(image, max_length=64, num_beams=4):
-  image = image.convert('RGB')
-  image = feature_extractor(image, return_tensors="pt").pixel_values.to(device)
-  clean_text = lambda x: x.replace('<|endoftext|>','').split('\n')[0]
-  caption_ids = model.generate(image, max_length = max_length)[0]
-  caption_text = clean_text(tokenizer.decode(caption_ids))
-  return caption_text 
+# def predict(image, max_length=64, num_beams=4):
+#   image = image.convert('RGB')
+#   image = feature_extractor(image, return_tensors="pt").pixel_values.to(device)
+#   clean_text = lambda x: x.replace('<|endoftext|>','').split('\n')[0]
+#   caption_ids = model.generate(image, max_length = max_length)[0]
+#   caption_text = clean_text(tokenizer.decode(caption_ids))
+#   return caption_text 
+
+model_id = "nttdataspain/vit-gpt2-coco-lora"
+model = VisionEncoderDecoderModel.from_pretrained(model_id)
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+feature_extractor = ViTFeatureExtractor.from_pretrained(model_id)
+
+# Predict function
+def predict_prompts(image):
+    img = Image.open(image).convert('RGB')
+    model.eval()
+    pixel_values = feature_extractor(images=[img], return_tensors="pt").pixel_values
+    with torch.no_grad():
+        output_ids = model.generate(pixel_values, max_length=16, num_beams=4, return_dict_in_generate=True).sequences
+
+    preds = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
+    preds = [pred.strip() for pred in preds]
+    return preds[0]
 
 input = gr.inputs.Image(label="Upload any Image", type = 'pil', optional=True)
 output = gr.outputs.Textbox(type="text",label="Captions")