Hey all trying to play with the api and assumed that it wanted a base 64 image for the img2img post request but that seems not to be the case? can anyone point me in the right direction?

I'm doing an API call however when I leave the seed value empty, I'm getting an error. How can I randomize it?

There doesn't seem to be a way to load local diffuser models, is there?
I'd like to load my local fine tuned models, without having to push them to huggingface

import subprocess
import os; os.makedirs("images", exist_ok=True)
import time
import streamlit as st

Run the diffuzers app

diffuzers_process = subprocess.Popen(["diffuzers", "app", "--host", "localhost", "--port", "8001", "--device", "cuda","--output","/content/images"], stdout=subprocess.PIPE)

time.sleep(30)

Run localtunnel to expose the app to the public

lt_process = subprocess.Popen(["lt", "--port", "8001","--subdomain","myapp"], stdout=subprocess.PIPE)
#lt_process = subprocess.Popen(["ngrok", "http", "8501"], stdout=subprocess.PIPE)

Wait for the localtunnel URL to be generated

lt_url = None
while lt_url is None:
line = lt_process.stdout.readline()
if "https" in line.decode("utf-8"):
lt_url = line.decode("utf-8").strip()

Print the localtunnel URL

print("LocalTunnel URL:", lt_url)

Wait for user input to stop the processes

input("Press enter to stop the processes...")

Terminate the processes

lt_process.terminate()
diffuzers_process.terminate()

i want to use without ngork key or anything else so wrote this, this does not work sometime says connection error and when it does works and i reload the model or page, session state error comes, https://docs.streamlit.io/library/api-reference/session-state

can you help or add defroum exnetion api ? i can hire for it

Error message: "subprocess-exited-with-error"

The error message "subprocess-exited-with-error" indicates that a subprocess being run by pip has failed. Specifically, in this case, it seems to be related to an issue with the metadata generation of a Python package.

The error message further explains that when attempting to read the long_description element in the setup.py file, a UnicodeDecodeError occurred. This usually happens when the character encoding is not specified correctly.

To resolve this issue, make sure that the character encoding is specified correctly when reading the long_description element in the setup.py file. For instance, if your file is assumed to be encoded with UTF-8, you can use the following code:

with open('README.md', encoding='utf-8') as f: long_description = f.read()

Using python 3.10 on Windows and Ubuntu 22.04 I get the error:

ModuleNotFoundError: No module named 'altair.vegalite.v4'

I was able to resolve this by manually running

pip install altair==4.2.2

This should probably be added to the requirements.txt?

Thank you for making it public, it's pretty awesome.
I have however a question that do you have any suggestion about the host that I can have the access to GPU?
Thanks,
Ai

Hi, I really appreciate this project.

In colab, in pip install step, it occurs an error for resolving dependencies.

Is there a pipeline for outpainting?

Hi, Abhi!
Thanks for the great work. Could you please add a text generation feature as well? This will help lots of researchers and AI aspirants.

Thanks,

from diffuzers.inpainting import Inpainting

File "/home/ubuntu/diffuzers/./diffuzers/inpainting.py", line 9, in
import streamlit as st
File "/home/ubuntu/.local/lib/python3.8/site-packages/streamlit/init.py", line 55, in
from streamlit.delta_generator import DeltaGenerator as _DeltaGenerator
File "/home/ubuntu/.local/lib/python3.8/site-packages/streamlit/delta_generator.py", line 45, in
from streamlit.elements.arrow_altair import ArrowAltairMixin
File "/home/ubuntu/.local/lib/python3.8/site-packages/streamlit/elements/arrow_altair.py", line 36, in
from altair.vegalite.v4.api import Chart
ModuleNotFoundError: No module named 'altair.vegalite.v4'

While running on macOS M1 I get

ERROR: Command errored out with exit status 1: /Users/loreto/opt/miniconda3/bin/python -u -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'/private/var/folders/dr/lzd3czf922qcgj0xpdyg3dc40000gn/T/pip-install-3lhdxoh9/pycryptodome_3d5818659ba54c90872d4993a0d9e451/setup.py'"'"'; __file__='"'"'/private/var/folders/dr/lzd3czf922qcgj0xpdyg3dc40000gn/T/pip-install-3lhdxoh9/pycryptodome_3d5818659ba54c90872d4993a0d9e451/setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__file__) if os.path.exists(__file__) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' install --record /private/var/folders/dr/lzd3czf922qcgj0xpdyg3dc40000gn/T/pip-record-4rdpmt_6/install-record.txt --single-version-externally-managed --compile --install-headers /Users/loreto/opt/miniconda3/include/python3.9/pycryptodome Check the logs for full command output.

that seems to be connected to this issue
Nitrokey/pynitrokey#226

I have tried to update/install Xcode additional components - as suggested here

/Applications/Xcode.app/Contents/MacOS/Xcode -installComponents

with no success so far.

I don't see any method to set it while running the app on Windows

Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/streamlit/runtime/scriptrunner/script_runner.py", line 565, in _run_script
exec(code, module.dict)
File "/usr/local/lib/python3.10/dist-packages/diffuzers/Home.py", line 7, in
from diffuzers.x2image import X2Image
File "/usr/local/lib/python3.10/dist-packages/diffuzers/x2image.py", line 14, in
from diffusers import (
File "", line 1075, in _handle_fromlist
File "/usr/local/lib/python3.10/dist-packages/diffusers/utils/import_utils.py", line 675, in getattr
value = getattr(module, name)
File "/usr/local/lib/python3.10/dist-packages/diffusers/utils/import_utils.py", line 675, in getattr
value = getattr(module, name)
File "/usr/local/lib/python3.10/dist-packages/diffusers/utils/import_utils.py", line 674, in getattr
module = self._get_module(self._class_to_module[name])
File "/usr/local/lib/python3.10/dist-packages/diffusers/utils/import_utils.py", line 686, in _get_module
raise RuntimeError(
RuntimeError: Failed to import diffusers.pipelines.alt_diffusion.pipeline_alt_diffusion_img2img because of the following error (look up to see its traceback):
cannot import name 'CpuOffload' from 'accelerate.hooks' (/usr/local/lib/python3.10/dist-packages/accelerate/hooks.py)

• ``

hey all just checking if pix2pix is supposed to be working as I'm getting some odd errors
ValueError: Incorrect configuration settings! The config of pipeline.unet: FrozenDict([('sample_size', 64), ('in_channels', 4), ('out_channels', 4), ('center_input_sample', False), ('flip_sin_to_cos', True), ('freq_shift', 0), ('down_block_types', ['CrossAttnDownBlock2D', 'CrossAttnDownBlock2D', 'CrossAttnDownBlock2D', 'DownBlock2D']), ('mid_block_type', 'UNetMidBlock2DCrossAttn'), ('up_block_types', ['UpBlock2D', 'CrossAttnUpBlock2D', 'CrossAttnUpBlock2D', 'CrossAttnUpBlock2D']), ('only_cross_attention', False), ('block_out_channels', [320, 640, 1280, 1280]), ('layers_per_block', 2), ('downsample_padding', 1), ('mid_block_scale_factor', 1), ('act_fn', 'silu'), ('norm_num_groups', 32), ('norm_eps', 1e-05), ('cross_attention_dim', 1024), ('attention_head_dim', [5, 10, 20, 20]), ('dual_cross_attention', False), ('use_linear_projection', True), ('class_embed_type', None), ('num_class_embeds', None), ('upcast_attention', False), ('resnet_time_scale_shift', 'default'), ('_class_name', 'UNet2DConditionModel'), ('_diffusers_version', '0.8.0'), ('_name_or_path', '/home/jerrick/.cache/huggingface/diffusers/models--stabilityai--stable-diffusion-2-base/snapshots/d28fc8045793886e512c5389771d3b3d560f9575/unet')]) expects 4 but received num_channels_latents: 4 + num_channels_image: 4 = 8. Please verify the config of pipeline.unetor yourimageinput.

demo.launch(share = True) enables running this app on Colab. Is it possible to add that as a CLI argument?

Hi,
When i build a basic streamlit app like you, click a button GENERATE to start generate images. But when click this button multiple times, a numbers of processes will be spawned, run parallel, fill GPU utility and all stuck. With your app, this problem is not happen, every processes is spawn and wait in a queue. I have read your code, but don't see how to fix mine. Can you help me?
Thanks.

Ran in venv, followed instructions to run diffuzers app

Error

python3.10/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 565, in _run_script

Hardware
MacBook Pro (13-inch, 2020
16 GB 3733 MHz LPDDR4X
Intel Iris Plus Graphics 1536 MB
2 GHz Quad-Core Intel Core i5

Each time I run it with a new conda env, it always say all the dependencies/modules that were installed are not found

It is fixed when I move out of the env and do a pip install -r requirements.txt (not ideal as I have other apps conflicting)

Error happens when I run diffuzers api

Hi, Abshishek!
Thanks for amazing work. I was just trying and found the API has got some issues (or I am using it incorrectly, don't know). The screen shots are attached for your perusal, have a look at them, thanks!
The commands to run the API (Am I using these environment variables correctly?):

The link result:

And also could you please update the diffuzers.ipynb file as well?

Thanks again!

Hi @abhishekkrthakur 👋 thanks for the repo ❤ and maybe you like this huggingface/diffusers#1904 (comment)

ngrok sometimes so slow, so I used other solutions.
so I need edit "cli/run_app.py"
L105
... "--server.enableCORS", "false", "--server.enableXsrfProtection", "false", "--server.enableWebsocketCompression", "false",
...
But hard coding isn't a good solution either, so maybe we can do it in the configuration,maybe we can add args：
diffuzers app ..... --cors http://xxx.xx.xx

thanks.

I run the commond in the linux server:
diffuzers api --port 8000 --device cuda:6

it shows:

What I can see in the browser is:

Does anyone know why this is?

Can you simplify diffuzers so it meets the desired intended audience of everyone? Or at least, spin a new version made for everyone from the ground up?