lakshya

img_norm_cfg and img_scale

1

In the configs/soft_teacher.py/base.py. Following statistcs have been used for img_norm `img_norm_cfg = dict(mean=[103.530, 116.280, 123.675], std=[1.0, 1.0, 1.0], to_rgb=False)` However, the coco_instance.py in the mmdet has following img_norm_cfg `mean=[123.675, 116.28,...

--resume-from functionality

1

I would recommend following changes, which 1) Currently, even if auto_resume is not mentioned in the config, find_latest_checkpoint command will still run as cfg.get("auto_resume", True) would return True if not...

I believe the gradient checkpointing can be very useful if you have to maintain some minimum batch size, and you can't do that with your hardware. I was training a...

Increase output video length and fps for eval.

1

Hello, excellent work. Really liked the inversion idea. Can you suggest how can I increase the output video length for evaluation and increase fps?

[BUG]

5

**Describe the bug** The sequence length during training is different than specified, in the configs, I've specified seq-len 50016, which is divisible by the tensor-model-parallel-size=4, however, during multinode training I'm...

Fps for 8 frame input to the encoder

1

Hello, is there any recommended fps for the 8 frame input to the intern video 2 stage 2 model, paper mentioned sparsely sampled but didn't mention much details