Chaoqi Wang
Chaoqi Wang
You can divide `outputs` (which is the logit) in the `sample` function of models.py by the desired temperature, the feed it `torch.multinomial` for sampling. But if you do greedy decoding,...
sorry, I did not back up it. You can simply run `utils/build_vocab.py` for generating the vocab.pkl.
Yes, you are right. But I did not implement the validation part, which is test() in main.py. In general, it will be tested on validation set for computing scores such...
How did you compute the validation loss? Feed the caption in a teacher-forcing manner, and then compute the loss, just like what we did for training?
I think the training loss is a bit higher than expected. It should be at 1.5 or so. Have you checked the caption generated on training set? You might need...