ColumbusAI

Any update on this?

> ``` > [Unit] > Description=cosmopolitan APE binfmt service > After=wsl-binfmt.service > > [Service] > Type=oneshot > ExecStart=/bin/sh -c "echo ':APE:M::MZqFpD::/usr/bin/ape:' >/proc/sys/fs/binfmt_misc/register" > > [Install] > WantedBy=multi-user.target > ``` >...

> Make sure to change line 29 of devika.py to your frontend URL. > > `CORS(app, resources={r"/*": {"origins": ["http://localhost:3001"]}}) # Change the origin to your frontend URL` Can we please...

thanks for sharing! I've added these variables now

I'm having this issue. I have a rtx 4090 + rtx 3090 and loading the gemma 27b model quantized to 8bpw. I have more than enough VRAM between these gpus,...

would like to see this as well! @rakshasa

> [v0.3.0](https://github.com/av/harbor/releases/tag/v0.3.0) enables removing all compose files from the root of the repo, however it'll be a breaking change as it won't be compatible with legacy mode for the CLI,...

@av here's what I see -- [https://youtu.be/9P6BLS3WLbU](https://youtu.be/9P6BLS3WLbU) I'm not sure why there's this big delay in 'harbor down' compared to 'harbor up'... perhaps it's as you mentioned and Harbor is...

``` admin@DESKTOP-PC:~$ harbor up 12:17:05 [DEBUG] Running 'up' for services: 12:17:06 [DEBUG] Matched compose files: 18 [+] Running 8/8 ✔ Network harbor_harbor-network Created 0.6s ✔ Container harbor.vllm Healthy 4.8s ✔...

Still getting this error when trying to run a llamafile in WSL2. @benhillis quick mention to see if this is on the team's radar?