facebookresearch / mae_st

Official Open Source code for "Masked Autoencoders As Spatiotemporal Learners"

facebookresearch/mae_st Issues

Difference with VideoMAE
Updated 4 months ago
Finetuned checkpoints
Updated 5 months ago
Issue for generating output.
Updated 5 months ago
what is torch.fb.rendezvous.zeus responsible for?
Updated 5 months ago7
Can you please give the ViT-B pre-trained model checkpoint on K400, K600, K700?
Updated 7 months ago1
load model issue
Closed a year ago3
test.csv format
Updated 9 months ago1
'_OpNamespace' 'video_reader' object has no attribute 'probe_video_from_memory'
Updated 9 months ago
Question about the provided pre-trained weights.
Updated 9 months ago5
no_env
Closed a year ago1
Colab demo
Updated a year ago1
RuntimeError: can't start new thread
Updated a year ago
what is path_to_video_1 represents for? Does it mean the path to find the video in the folder? Eg. /dataset/video/video1.mp4
Closed a year ago
Dataset and actually running issues
Closed 2 years ago3
t_patch_size issue
Updated a year ago
Can you give the finetune model checkpoint on K400?
Closed a year ago1
Can you give the pre-trained model on IG-uncurated
Closed a year ago1
Epoch number and warmup epoch number
Updated a year ago
Encoder
Closed 2 years ago3
Finetune
Closed 2 years ago3
How to start multi-gpu training?
Closed 2 years ago1
Why repeated sampling does not affect the actual learning rate?
Closed 2 years ago1
403 error
Closed 2 years ago2