Closed ArekKubinski closed 1 year ago
Also forgot to mention I was trying to run portable distribution and the errors happen when i click "enfugue" button on the web interface.
Hello @ArekKubinski, thank you for the report!
I can tell you that almost all of those "errors" are known and handled situations, with the actual showstopper being here:
Traceback (most recent call last):
File "transformers/__init__.py", line 3865, in <module>
raise OptionalDependencyNotAvailable()
transformers.utils.import_utils.OptionalDependencyNotAvailable
I can't be certain exactly which dependency it is saying is not available without some further digging, but I can say that Enfugue will almost certainly not work on your GPU, I am sorry. The 2GB of VRAM could have potentially been surmounted with a lot of optimization, but unfortunately that card was designed for CUDA 6, an architecture we have long since left behind - even if it does work, you will find unacceptably poor performance.
I'm very sorry to have to be the bearer of bad news! I can at least mention that as of today I've partnered with RunDiffusion.com to offer Enfugue in the cloud, and I know you can at least get a half-hour on an 8GB card there for free if you still want to try Enfugue.
Yeah, I was suspecting this to happen, thanks for answering, time to run cpu env with conda then 🫠
Hello, I am trying to run version 0.2.3 to no luck. I already tried to re download and reinstall it, clearing cache and db don't help also, here some errors:
This is a clear install and my OS is Linux Ubuntu 23.04 x86_64 GPU: Driver Version: 535.104.12 CUDA Version: 12.2 NVIDIA GeForce MX130 (yes it is 2GB VRAM card 😢)