-
The mask ratio for EEG pretraining in your paper is 75%, but in your code (config.py), the ratio is 10%. Could you please tell me the mask ratio you used for EEG pretrain? Many thanks.
HLSUD updated
11 months ago
-
Hi,
How did you end up with pretrain your 96k CT scan for 2000 epochs with only 8*40GB GPU?
How long it takes?
Our experience is that this is practically impossible. We have even more GPUs (32…
-
Has anybody else observed an increase in the amount of memory used by the GPU when using pretrained weights?
It's probably only a small increase, but for me it's taking me over the limits of my har…
-
Hi, I have been trying to run the run_evaluation.sh with the provided checkpoints downloaded and unzipped to the checkpoints directory. I am running into this error:
evaluate.py: error: argument -…
-
Hello, thank you for your sharing firstly!
Based on your tuition, I use create_pretraining_data.py to get a file of tf_example.tfrecord. Then I run run_pretraining.py. After these two steps, I can ge…
-
Hi Yuan,
Thanks again for this great work, I have been using both this and the original AST model for some downstream tasks. I am currently looking into some other time series data, and was wonder…
-
Hey there, my apologies if I got this wrong. I've been stepping through the Units model, and it looks like you never actually use the time encoding features within the model training. Some of the pa…
-
Hi! Thanks for providing such a wonderful work.
I wonder have you tried a ResNet backbone without ImageNet pretraining?
Is it possible that a pre-trained model might become one of the keys of the p…
-
Since the in-house images are all labeled in Chinese, I'm curious about how to enable the model to support detecting images like that. I simply translated the label from English to Chinese, but there …
-
## Feature request
After semi-supervised pretraining, can we do light-weighted fine-tunning or few-shot learning instead of classification?
**What is the expected behavior?**
Instead of fine-…