Aarni Koskela
Aarni Koskela
Rebased for the new month. The finetuning example that was merged in #3296 was pretty messy...
If you want, I have a GTX 1070 (under WSL2, works surprisingly well) I can test on: ``` +---------------------------------------------------------------------------------------+ | NVIDIA-SMI 545.36 Driver Version: 546.33 CUDA Version: 12.3 | |-----------------------------------------+----------------------+----------------------+...
Did Tim say why it's "probably not safe"? Do we know of an actual situation where cublaslt isn't available? Is such a situation something we want to support?
Ah, I wasn't even aware of this conversation before opening #996 :)
Yeah, I guess an integrated solution will be better (and less maintenance for us, although more for @pennersr). I suppose we'd need to * take a look at feature parity...
Copying @pennersr's experimental migration management command here – we have his blessing to use this in this project: ```python import base64 from allauth.mfa.adapter import get_adapter from allauth.mfa.models import Authenticator from...
Yeah, that looks like your Git is somehow screwed up.
The peer dependency here is ``` "react": ">=15.0.0" ``` – is there something that's not working for you with React 18, @fergy-os?
We could support this by way of e.g. a [`!jinja2` YAML tag](https://stackoverflow.com/a/17192241/51685) in a YAML wildcard collection, so when such a wildcard is invoked, it's parsed using the Jinja engine......