A query regarding Image-Cap about torch-light HOT 4 OPEN

https://github.com/ne7ermore/torch-light/tree/master/Image-Cap
Hi @ne7ermore,
I found this code very educational. Thank you. But, I do not have a GPU. When I run train.py on a CPU, the estimated time of training is about 3-4 weeks. So, could you please share the trained model. It would be of great help. My email id - [email protected].
Thank You

Hello, did you get the trained model? I also have the same problem. If you do get it, could you please share the trained model with me ? My email is [email protected]. Thanks a lot.

from torch-light.

@ne7ermore
Thanks for your sharing code for Image Cap. I find there might be some bug in the train.py. There isn't definition of "speak" in model.py. Or I misunderstand, could you please give me some advice?

def eval():
actor.eval()
eval_score = .0
for imgs, labels in tqdm(validation_data,
mininterval=1,
desc="Actor-Critic Eval",
leave=False):
enc = actor.encode(imgs)

    hidden = actor.feed_enc(enc)
    ### Bug to fix??
    words, _ = actor.speak(hidden)

    scores = rouge_l(words, labels)
    scores = scores.sum()

    eval_score += scores.data

eval_score = eval_score[0] / validation_data.sents_size

return eval_score

from torch-light.

I met the same issue as you, have you solved it ? @Michael-Hsu

from torch-light.

from torch-light.