CODEWITHSUNDEEP
pythonmacosstable-diffusionmps
lando2319
lando2319

Reputation: 1800

Upgrading to Stable Diffusion 3 from 2-1 on mac

I'm upgrading my stable diffusion from 2-1 to stable-diffusion-3-medium-diffusers

Here is my code which is working for version 2-1

# source venv/bin/activate

from diffusers import DiffusionPipeline

pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-2-1")
pipe = pipe.to("mps")

pipe.enable_attention_slicing()

print("Starting Process")

steps = 200
query = "Stormy Weather in Monte Carlo"

image = pipe(query, num_inference_steps=steps).images[0]

image.save("oneOffImage.jpg")

print("Successfully Created Image as oneOffImage.jpg")

I upgraded diffusers, signed up on hugging face for access to the gated repo, created and added the HF_TOKEN to my .env, and ran this code

# source venv/bin/activate

from diffusers import StableDiffusion3Pipeline
from dotenv import load_dotenv
import os

load_dotenv()

print("Starting Process")

pipe = StableDiffusion3Pipeline.from_pretrained("stabilityai/stable-diffusion-3-medium-diffusers")
pipe = pipe.to("mps")

# pipe.set_progress_bar_config(disable=True)
pipe.enable_attention_slicing()

print("Starting Process")

steps = 200
query = "Stormy Weather in Monte Carlo"

image = pipe(query, num_inference_steps=steps).images[0]

image.save("oneOffImage.jpg")

print("Successfully Created Image as oneOffImage.jpg")

I was able to download the model, also I logged the token and confirmed it's in the env vars, I tried adding torch and setting , torch_dtype=torch.float16) but that did nothing plus I think thats for cuda, I also tried adding an auth tag but that did nothing, I upgraded my transformers but I don't even thing that did anything. I'm running out of ideas.

Here is the current error

(venv) mikeland@mikes-mac-mini WeatherWindow % python3 oneOffGenStableDiffusion.py 
/Users/mikeland/WeatherWindow/venv/lib/python3.9/site-packages/diffusers/models/transformers/transformer_2d.py:34: FutureWarning: `Transformer2DModelOutput` is deprecated and will be removed in version 1.0.0. Importing `Transformer2DModelOutput` from `diffusers.models.transformer_2d` is deprecated and this will be removed in a future version. Please use `from diffusers.models.modeling_outputs import Transformer2DModelOutput`, instead.
  deprecate("Transformer2DModelOutput", "1.0.0", deprecation_message)
Starting Process
Loading pipeline components...:   0%|                                                                                  | 0/9 [00:00<?, ?it/s]
Traceback (most recent call last):
  File "/Users/mikeland/WeatherWindow/oneOffGenStableDiffusion.py", line 15, in <module>
    pipe = StableDiffusion3Pipeline.from_pretrained("stabilityai/stable-diffusion-3-medium-diffusers")
  File "/Users/mikeland/WeatherWindow/venv/lib/python3.9/site-packages/huggingface_hub/utils/_validators.py", line 114, in _inner_fn
    return fn(*args, **kwargs)
  File "/Users/mikeland/WeatherWindow/venv/lib/python3.9/site-packages/diffusers/pipelines/pipeline_utils.py", line 881, in from_pretrained
    loaded_sub_model = load_sub_model(
  File "/Users/mikeland/WeatherWindow/venv/lib/python3.9/site-packages/diffusers/pipelines/pipeline_loading_utils.py", line 703, in load_sub_model
    loaded_sub_model = load_method(os.path.join(cached_folder, name), **loading_kwargs)
  File "/Users/mikeland/WeatherWindow/venv/lib/python3.9/site-packages/transformers/modeling_utils.py", line 3122, in from_pretrained
    raise ImportError(
ImportError: Using `low_cpu_mem_usage=True` or a `device_map` requires Accelerate: `pip install accelerate`

I'm just looking for ANY example I can get to work on Mac at this point.

Upvotes: 0

Views: 358

Answers (1)

lando2319
lando2319

Reputation: 1800

Here is what ended up working for me

# source venv/bin/activate

from diffusers import StableDiffusion3Pipeline
import torch

from dotenv import load_dotenv
import os

load_dotenv()

os.environ["PYTORCH_MPS_HIGH_WATERMARK_RATIO"] = "0.0"

print("Starting Process")

pipe = StableDiffusion3Pipeline.from_pretrained(
        "stabilityai/stable-diffusion-3-medium-diffusers", 
        low_cpu_mem_usage=False,
        torch_dtype=torch.float16,
        variant="fp16",
        use_safttensors=True
        ).to("mps")


pipe.enable_attention_slicing()

print("Starting Process")

steps = 40
query = "Rain Weather in New York City, New York"

image = pipe(query, num_inference_steps=steps).images[0]

image.save("oneOffImage.jpg")

print("Successfully Created Image as oneOffImage.jpg")

Upvotes: 0

Related Questions