Issues with diffusers and deepspeed's init_inference() in latest version

dyedd · December 4, 2023, 2:30pm

I need everyone’s help, I had started an issue but so far I haven’t found out how to make the fix:

[BUG]init_inference() fails for diffusers

opened 07:48AM - 28 Nov 23 UTC

bug inference

**Describe the bug** Code source: https://github.com/microsoft/DeepSpeedExample…s/blob/master/inference/huggingface/stable-diffusion/test-stable-diffusion.py **To Reproduce** I've tried both the latest version, and the requirements.txt version as supplied with the code. Running it through a single gpu this is no problem, however when running it with multiple gpus, it reports an error. **Expected behavior** Supports multi-gpu **Screenshots** ![20231127084144](https://github.com/microsoft/DeepSpeed/assets/38277217/256375c6-834a-4c65-865f-d44a36496817)

Topic		Replies	Views
Infrence time increase when using multi-GPU DeepSpeed	1	766	November 28, 2023
I have a question about multi-GPU inference DeepSpeed	0	1253	March 9, 2023
Can a diffuser pipeline run on multiple GPUs? Amazon SageMaker	2	768	May 31, 2023
Stable diffusion `train_text_to_image.py` only on one gpu 🧨 Diffusers	5	974	May 2, 2023
Multiple gpu not properly parallelized during model.generate() 🤗Transformers	4	1285	October 9, 2022

Issues with diffusers and deepspeed's init_inference() in latest version

Related Topics