-
Just a quick question. I was wondering if the masks which get passed to the model at inference are for the purposes of masked tokens for self-supervision or if they are a different mask? Thanks
-
Plugin NameMinecraft Development
Plugin Version2021.2-1.5.12
OS NameLinux
Java Version11.0.11
Java VM VendorJetBrains s.r.o.
App NameIDEA
App Full NameIntelliJ IDEA
App Version NameIntelliJ IDEA
Is …
-
I want to reproduce the result of BYOL by myself. However, I'm not sure how to use multi machine to train the ImageNet Dataset in batch size 4096, can you give a simple example?
-
**Is your feature request related to a problem? Please describe.**
DeiT ([https://github.com/facebookresearch/deit](url)) heavily depends on `timm`, and the loader implemented by the original DeiT…
-
The loss function for the BYOL model seems to be wrong. Below is a code snippet to demonstrate this issue.
```python
import torch
# Settings for BYOL.
use_negative_examples_from_batch = False
…
-
I am using pytorch lightning with Adam optmiser to train a BYOL model, the model and pipeline works fine while training, when I stop and resume from checkpoint, it raise this error. I didn't do anyt…
-
Hi,
Thanks for proving such a cool implementation! When I ran your code with multiple GPUs, I ran into the below error.
GPU available: True, used: True
TPU available: False, using: 0 TPU cores
…
-
The SimCLR and MoCo architectures are currently implemented as standalone architectures.
Before adding new architectures like BYOL or SimSiam we should probably make use of the existing code, e.g.
-…
-
Thanks for your excellent work! But i have trouble in linear evaluation.
I get the model from Model Zoo (BYOL (@XiaohangZhan) | selfsup/byol/r50_bs4096_ep200.py (@xieenze) | byol_r50-e3b0c442.pth), t…
-
## Description
Describe the problem you're having or the enhancement you'd like to request.
There is no CFT that provides for HA across AZs, 3 NICs, and BYOL that doesn't also require a public mgm…