lucidrains/TimeSformer-pytorch issues and pull requests

#21 - fix runtime error in SpaceTime Attention

Pull Request - State: closed - Opened by adam-mehdi over 3 years ago - 1 comment

#20 - How can plot the heatmap of the self-attentions?

Issue - State: open - Opened by YNawal over 3 years ago

#19 - Please provide example code on how to experiment on new dataset

Issue - State: open - Opened by Altaf-hucn over 3 years ago

#18 - Fine-tune with new datasets

Issue - State: open - Opened by Jeba-create over 3 years ago

#17 - Training a joint Space time model

Issue - State: open - Opened by Vidyaranya over 3 years ago - 1 comment

#16 - Update timesformer_pytorch.py

Pull Request - State: closed - Opened by aarti9 over 3 years ago

#15 - problem in timesformer_pytorch.py

Issue - State: open - Opened by Weizhongjin over 3 years ago - 2 comments

#14 - Set frame_mask and cls_attn_mask if mask is None

Pull Request - State: closed - Opened by quaertym almost 4 years ago - 1 comment

#13 - TimSformer in TF, thanks

Issue - State: open - Opened by junyongyou almost 4 years ago - 7 comments

#12 - 最后得到的是整个视频的语义向量？

Issue - State: open - Opened by lonngxiang almost 4 years ago

#11 - How to deal with varying length video? Thanks

Issue - State: closed - Opened by junyongyou almost 4 years ago - 2 comments

#10 - Is there anyone try to train the TimeSformer on Something-Something-V2?

Issue - State: closed - Opened by jianghaojun almost 4 years ago

#9 - Imagenet Pretrained Weights

Issue - State: open - Opened by RaivoKoot almost 4 years ago - 5 comments

#8 - Discussion on training issues I have encountered

Issue - State: open - Opened by zmy1116 almost 4 years ago - 33 comments

#7 - using nn.Emdedding ecounter following problem

Issue - State: open - Opened by henanjun almost 4 years ago

#6 - Model does not converge

Issue - State: open - Opened by duduheihei almost 4 years ago - 1 comment

#5 - how to train the model? can you provide train code?

Issue - State: open - Opened by dotsonliu almost 4 years ago - 2 comments

#4 - Request for Training & Evaluation Code

Issue - State: open - Opened by wanghao9610 almost 4 years ago - 5 comments

#3 - How to deal with Long time video data which has1000 frames