knowing-when-to-look-adaptive-attention
knowing-when-to-look-adaptive-attention copied to clipboard
PyTorch Implementation of Knowing When to Look: Adaptive Attention via a Visual Sentinal for Image Captioning
why blue4 is always 0? setting up scorers... computing Bleu score... {'testlen': 7600, 'reflen': 5064, 'guess': [7600, 7200, 6800, 6400], 'correct': [891, 39, 0, 0]} ratio: 1.5007898894151854 Bleu_1: 0.117 Bleu_2:...
hello,the error is: PTBTokenizer tokenized 307821 tokens at 312344.45 tokens per second. PTBTokenizer tokenized 51126 tokens at 83364.85 tokens per second. setting up scorers... computing Bleu score... {'testlen': 46127, 'reflen':...
Can you help me how to solve this issue ? 
Hello, is your code suitable for Flickr 30K images?
Hi, I tried to train the model with Flickr30K dataset, but when I tested, I got Bleu-4 score of 0.16137 which is away lower than the results that reported in...
Can you please tell me how to upload image dataset to link in this project ? Is their any python file ?
Please help me to find these dataset files below mentioned filename .. TRAIN_CAPLENS_coco.json TRAIN_CAPTIONS_coco.json TRAIN_IMAGES_coco.hdf5 VAL_IMAGES_coco'.hdf5 and another help where can I add train and val image dataset in the...
Hi!I am an undergraduate student who is just starting to learn about image description. I've been reading this paper recently. Your code has helped me a lot. My computer is...