Spaces:

---

haodongli

/

Lotus-2_Normal

like 9

Running on Zero

Hi @haodongli , thanks for the request. Since A100 capacity is limited, it's not always available on demand. I've asked our infra team if there's an available A100. As it's the weekend, their response might take a bit longer, but I'll get back to you once I get a reply.

Hi @haodongli , thanks for the request. Since A100 capacity is limited, it's not always available on demand. I've asked our infra team if there's an available A100. As it's the weekend, their response might take a bit longer, but I'll get back to you once I get a reply.

I too have been looking forword to trying it out properly

@haodongli We do have available A100 now. Have you requested access to A100?

BTW, out of curiosity, what error did you get with ZeroGPU? When I duplicate this Space, it seems to run fine on ZeroGPU.

We've enabled access to the A100 for you. This isn't a free grant, so you'll be billed as usual, but you can switch hardware to A100.

Thanks @hysts , I've switched to A100.

For the ZeroGPU, it said: ZeroGPU worker error AttributeError.

Also, since you've tested this demo on ZeroGPU (I've merged your modifications), can you help me switch to ZeroGPU to save some money for me? I deeply appreciate that!

Thanks for the details. The error sometimes happens on ZeroGPU Spaces and the cause isn't always the same. So it's hard to tell why it happened in this Space. Anyway, I've re-assigned ZeroGPU to this Space. Let me know if it happens again.

Thanks! @hysts I'm so glad to see your prompt response!

This normal demo works well in my side!

By the way, can you please switch the depth demo (https://huggingface.co/spaces/haodongli/Lotus-2_Depth) onto ZeroGPU? I've removed the duplicated @spaces.GPU in app.py following your modifications here.

Thanks again for your assistance and the amazing huggingface community!

Best,

I've reassigned ZeroGPU to the depth demo as well. Thanks!