Describe the bug

Hey dear diffusers team,

thanks a lot for all your hard work!

I would like to make use of the torch_dtype="auto" keyword argument when loading a model/pipeline as specified here, but the usage does not work as expected (see example below). Can you help me out with some guidance on how to use it correctly or let me know whether there is something wrong with the handling of this argument?

Thank you!

Reproduction

from diffusers import StableDiffusionPipeline

model = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", torch_dtype="auto")

Logs

Passed `torch_dtype` torch.float32 is not a `torch.dtype`. Defaulting to `torch.float32`.

System Info

• 🤗 Diffusers version: 0.33.1
• Platform: Linux-5.15.0-136-generic-x86_64-with-glibc2.35
• Running on Google Colab?: No
• Python version: 3.10.17
• PyTorch version (GPU?): 2.7.0+cu126 (True)
• Flax version (CPU?/GPU?/TPU?): not installed (NA)
• Jax version: not installed
• JaxLib version: not installed
• Huggingface_hub version: 0.30.2
• Transformers version: 4.51.3
• Accelerate version: 1.6.0
• PEFT version: 0.15.2
• Bitsandbytes version: 0.45.5
• Safetensors version: 0.5.3
• xFormers version: not installed
• Accelerator: NVIDIA H100 PCIe, 81559 MiB
• Using GPU in script?: Yes
• Using distributed or parallel set-up in script?: No

Who can help?

No response