Anh (Luka) Pham

icon company VerysellAI icon location Saint Petersburg, Russia

Results 15 comments of Anh (Luka) Pham

The results of LoRA are very bad, where did I go wrong?

@wen020 yes, I use 4090 to train, may the training requires at least more than 24Gb I think, otherwise it causes oom.

The results of LoRA are very bad, where did I go wrong?

The config of mine works on 4090, just switch to small model like 1B. The only problem I have is the lora result, it is bad.

The results of LoRA are very bad, where did I go wrong?

@wen020 yes, especially on faces :(((

The results of LoRA are very bad, where did I go wrong?

@dome272 Actually the 3B model trained on LoRAs is still bad, especially on faces. Have you ever tried to train faces with LoRA?

The results of LoRA are very bad, where did I go wrong?

@wen020 I can only train 1B model on 4090, otherwise it will cause oom.

Always get the Cuda Out Of Memory error when training LoRA despite fixing the batch_size

Thanks mate

[Feature Request]: Stable Video Diffusion access through API?

@DaveyBoy1970 You have any idea how we can call it?

Prediction failed for an unknown reason. It might have run out of memory? (exitcode 2)

Does anyone have solutions?

ERROR: Could not open requirements file: [Errno 2] No such file or directory: 'requirements.txt'

@evesloan have you run it? I always got the conflict errors for cuda and not able to install due to cuda, ... It would be nice if you can share...

Can we use multiple controlnet models

@CenekSanzak Hi, have you tried to implement your idea of multicontrolnet?