PipiZong

Hi, It seems we cannot get the same results even when we fix the seed, set num_worker to be 0 in dataloader and disable cuDNN. Do you have any idea...

Does Oscar work for other languages?

2

Hi, Thanks for sharing your work! If I want to fine-tune the pre-trained model for a small Chinese dataset, is it possible and flexible?

Hi, I noticed your paper showed the results by DILP, but I couldn't find the corresponding code about the DILP implementation. Could you please share the code with me? Many...

Question about pretraining

1

Thanks for sharing your great work. Just wondering which model do you select for the downstream task since you save pertaining models every 40 epochs. Thank you.

Hi, Thanks for sharing your work! Does this pretrained model work for Chinese?

Hi, I observed the regularizer will become negative, how to explain this? Is it a potential issue as a regularizer which should be positive?

Hi, Thanks for sharing your work! Could you please share the related code for the graph construction, like using the Stanford parser to identify all nouns and a visual grounding...

Hello, Thanks for your recent updates! And sorry to bother you again. Could you please also share with us the specific files for data/test/vocab_path required in 'python run_summarization.py --mode=train --data_path=*...

Hi Eric, I just found the computation of vid_h and vid_c is the same as attached. Is this a potential bug? 

Hello, Thanks for sharing this great work. Could you please also share the code for Multilingual Video Captioning? It will be very helpful for us.