【HELP】I meet a collapse
like this:
PS F:\2024\my_homework\gwsj\finalWork\stable-diffusion-3.5-large> python
Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
import torch
from diffusers import StableDiffusion3Pipeline
pipe = StableDiffusion3Pipeline.from_pretrained(
... "stabilityai/stable-diffusion-3.5-large",
... torch_dtype=torch.float16
... )
Loading checkpoint shards: 100%|███████████████████████████████████| 2/2 [00:03<00:00, 1.51s/it]
Loading pipeline components...: 67%|████████████████████ | 6/9 [00:03<00:02, 1.39it/s]You setadd_prefix_space
. The tokenizer needs to be converted from the slow tokenizers
Loading pipeline components...: 78%|███████████████████████▎ | 7/9 [00:04<00:01, 1.68it/s]
PS F:\2024\my_homework\gwsj\finalWork\stable-diffusion-3.5-large>
even if I change the "bfloat16" to "float16"
sry , Maybe my 16G Memory can run this ** large** model