houzhijian / CONQUER

[2021 MultiMedia] CONQUER: Contextual Query-aware Ranking for Video Corpus Moment Retrieval
33 stars 6 forks source link

Model config for training on DiDeMo #5

Closed Aarontncl closed 8 months ago

Aarontncl commented 1 year ago

Hi, zhijian

Thanks for your great work and code! I notice the provided config files don't include the model config for training on DiDeMo, and the command lines for this dataset also seem to be missing. Could you please share all these original configurations to make experiments on DiDeMo dataset more accessible? Thanks again!

zhanglong-ustc commented 10 months ago

Hi, zhijian

Thanks for your great work and code! I notice the provided config files don't include the model config for training on DiDeMo, and the command lines for this dataset also seem to be missing. Could you please share all these original configurations to make experiments on DiDeMo dataset more accessible? Thanks again!