Blake Wyatt

Does anyone know how to do this? I would love to use this library for a simple Minecraft chat I'm making.

I've been able to reproduce this with the latest versions of `torch` and `transformers`. torch==1.10.1 transformers==4.15.0

To confirm this is a problem with `rq` and not the Bert model itself, I ran 4 Python instances in parallel executing the code below for 10 minutes and was...

The problem seems to be reduced with `torch.set_num_threads(1)` indicating workers could be communicating or sharing memory somehow.

@Perfare any chance we will get this feature soon? Would love to be able to load a folder and extract all assets via the command line. Doesn't have to be...

@lowercase00 I did, but I believe it made no difference. My memory is hazy now that it's been over a year. I ended up not using Redis at all and...

@PedroMTQ I just went the multiprocessing route and spun up multiple processes to do things in parallel instead. There's a way to share the model across all those processes without...

Can you verify that a folder starting with "quant_cuda" exists in your `installer_files/env/libs/site-packages` folder? If it doesn't, some fixes I've made to the installer might help https://github.com/xNul/one-click-installers

@HB-Stratos you're right, I typo'd there. `quant_cuda` should be under `installer_files/env/Lib/site-packages`. This is very strange. Under the new install using my one-click installer, could you double check that there isn't...

@HB-Stratos this is exactly what we're looking for. Cool, now let's see if it runs a model. Get to the download model menu via the install or you can open...