jayleicn / TVQAplus

[ACL 2020] PyTorch code for TVQA+: Spatio-Temporal Grounding for Video Question Answering

http://tvqa.cs.unc.edu

jayleicn/TVQAplus Issues

Would you like to add preprocess script(s)?
Updated 9 months ago
I can't download the dataset
Updated a year ago
Requests for tvqa_plus_stage_features
Closed 3 years ago
How to invoke load_jsonl_as_dict?
Closed 3 years ago1
Regional visual feature的提取是否需要原图呀？
Closed 3 years ago3
Feature Extraction Steps
Closed 3 years ago2
data preprocess
Closed 4 years ago1
Performance without BERT fine-tuning
Closed 4 years ago3
Fine-tuning format and data loading
Closed 4 years ago7
Train the full STAGE model
Closed 4 years ago
Extra vocab used for tokenization and feature extraction with bert?
Closed 4 years ago8
use multi GPUs probelm == attributeError: 'tuple' object has no attribute 'item
Closed 4 years ago
Use multi gpu ,then ts_label zip error
Closed 4 years ago1
AttributeError: 'tqdm' object has no attribute 'n'
Closed 4 years ago
Big difference between training accuracy and validation accuracy
Closed 4 years ago5
ETA publication verification code and test data
Closed 5 years ago1
Confusion finetuning format for BERT
Closed 5 years ago1
Script to finetune bert and extract features on TVQA+
Closed 5 years ago8
Unable to unzip your feature file
Closed 5 years ago2
I can not download the datasets
Closed 5 years ago1
When the code will be released?
Closed 5 years ago3
some confusions about this paper
Closed 5 years ago4
Impossible to download the dataset
Closed 5 years ago1