Generate an image based on a json via terminal without web gui?

[Barkoczy]

Is it possible to generate an image based on a json subfile that can be copied to the clipboard from the history of generating images via web ui directly from the command lines?

[mashb1t]

Hey @Barkoczy, this is currently not possible without tweaking the code yourself. Gradio is an essential part of Fooocus and console only use cases are currently not supported.
What you can already do is to use the Gradio websockets API which you can find the docs for in the footer.

[Barkoczy]

Gradio API does not work. I've looked through all the GitHub threads about this problem and followed the API documentation and other various solutions I found. The result is that it is not possible to generate an image using the Gradio API. Is there anything else you would like to try? Thank you

from gradio_client import Client

# Client
client = Client("http://127.0.0.1:7865/", serialize=False)

# Temporary Image
tmp_image = "https://raw.githubusercontent.com/gradio-app/gradio/main/test/test_files/bus.png"

# Model
base_model = "sd_xl_base_1.0_0.9vae.safetensors"
refiner_model = "sd_xl_refiner_1.0.safetensors"

# Prompt
positive = """
  a cat
"""

negative = """
  watermark:1.1, deformed hand
"""

# Styles
styles = ["Fooocus V2", "Fooocus Sharp", "Fooocus Enhance", "Fooocus Masterpiece"]

# Performance [Quality, Speed, Extreme Speed, Lightning]
performance =  "Speed"

# Aspect Ration
aspect_ratio = "1024x1024"

# Image Mime Type [png, jpg, webp]
image_mime_type = "webp"

# Image Number
n_images = 1

# Seed
seed = "0"

# Generate
client.predict(
  False, positive, negative, styles, performance, aspect_ratio,
  n_images,
  image_mime_type,
  seed,	# str in 'Seed' Textbox component
  True, 2,
  5,  #(numeric value between 1.0 and 30.0)in 'Guidance Scale' Slider component
  base_model, # str 'Base Model (SDXL only)' Dropdown component
  refiner_model, 0.8, 
  True, "sd_xl_offset_example-lora_1.0.safetensors", 0.1 ,#Lora1
  True, "Hyper-SDXL-8steps-lora.safetensors", 1 ,#Lora2
  False, "None", 0 ,#Lora3
  False, "None", 0 ,#Lora4
  False, "None", 0 ,#Lora5
  False, "", "Disabled", tmp_image, ["Left"], tmp_image, "", tmp_image, 
  False, # bool in 'Disable Preview' Checkbox component
  False, # bool in 'Disable Intermediate Results' Checkbox component
  True, # bool in 'Disable seed increment' Checkbox component
  0.1, 0.1, 0, 1,
  "dpmpp_2m_sde_gpu",
  "karras", # str (Option from: ['normal', 'karras', 'exponential', 'sgm_uniform', 'simple', 'ddim_uniform', 'lcm', 'turbo'])
  -1,-1,-1,-1,-1,-1,
  False, False, False, False, 1, 1,
  "joint", # str (Option from: ['joint', 'separate', 'vae'])
  0, False, 0, 0, 0, 0, False, False, "None", 0, 0, False, False, -64,
  False, # bool in 'Save Metadata to Images' Checkbox component
  "fooocus", # str in 'Metadata Scheme' Radio component
  tmp_image, 0, 0, "ImagePrompt",
  tmp_image, 0, 0, "ImagePrompt",
  tmp_image, 0, 0, "ImagePrompt",
  tmp_image, 0, 0, "ImagePrompt",
  fn_index=40
)
result = client.predict(fn_index=41)

print(result)

Error response

May 20 12:48:25 homelab server.sh[47008]: Traceback (most recent call last):
May 20 12:48:25 homelab server.sh[47008]:   File "/home/ubuntu/miniconda3/envs/fooocus/lib/python3.10/site-packages/gradio/routes.py", line 488, in run_predict
May 20 12:48:25 homelab server.sh[47008]:     output = await app.get_blocks().process_api(
May 20 12:48:25 homelab server.sh[47008]:   File "/home/ubuntu/miniconda3/envs/fooocus/lib/python3.10/site-packages/gradio/blocks.py", line 1429, in process_api
May 20 12:48:25 homelab server.sh[47008]:     inputs = self.preprocess_data(fn_index, inputs, state)
May 20 12:48:25 homelab server.sh[47008]:   File "/home/ubuntu/miniconda3/envs/fooocus/lib/python3.10/site-packages/gradio/blocks.py", line 1239, in preprocess_data
May 20 12:48:25 homelab server.sh[47008]:     processed_input.append(block.preprocess(inputs[i]))
May 20 12:48:25 homelab server.sh[47008]:   File "/home/ubuntu/servers/Fooocus/modules/gradio_hijack.py", line 279, in preprocess
May 20 12:48:25 homelab server.sh[47008]:     im = processing_utils.decode_base64_to_image(x)
May 20 12:48:25 homelab server.sh[47008]:   File "/home/ubuntu/miniconda3/envs/fooocus/lib/python3.10/site-packages/gradio/processing_utils.py", line 59, in decode_base64_to_image
May 20 12:48:25 homelab server.sh[47008]:     img = Image.open(BytesIO(base64.b64decode(image_encoded)))
May 20 12:48:25 homelab server.sh[47008]:   File "/home/ubuntu/miniconda3/envs/fooocus/lib/python3.10/base64.py", line 87, in b64decode
May 20 12:48:25 homelab server.sh[47008]:     return binascii.a2b_base64(s)
May 20 12:48:25 homelab server.sh[47008]: binascii.Error: Incorrect padding

[mashb1t]

@Barkoczy Please find a working example in https://github.com/toutjavascript/FoooXus-Fooocus-Extender/blob/main/src/api.py#L250-L344. Make sure to encode empty images as base64 or send the encoded empty value https://github.com/toutjavascript/FoooXus-Fooocus-Extender/blob/main/src/api.py#L39

[Barkoczy]

I have tried all the options and these sent constantly it reported an error. Finally I used the Fooocus API project, which solved all my needs. Thank you for your help and good luck.

[TriptSharma]

@mashb1t I am passing the encoded empty value (from https://github.com/toutjavascript/FoooXus-Fooocus-Extender/blob/main/src/api.py#L39) as an api call from my local system to the colab hosted gradio app. However, it's giving me the following error:

Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\client.py", line 468, in predict
    ).result()
      ^^^^^^^^
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\client.py", line 1499, in result
    return super().result(timeout=timeout)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Python311\Lib\concurrent\futures\_base.py", line 456, in result
    return self.__get_result()
           ^^^^^^^^^^^^^^^^^^^
  File "C:\Python311\Lib\concurrent\futures\_base.py", line 401, in __get_result
    raise self._exception
  File "C:\Python311\Lib\concurrent\futures\thread.py", line 58, in run
    result = self.fn(*self.args, **self.kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\compatibility.py", line 63, in _inner
    data = self.serialize(*data)
           ^^^^^^^^^^^^^^^^^^^^^
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\compatibility.py", line 244, in serialize
    o = tuple([s.serialize(d) for s, d in zip(self.serializers, data)])
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\compatibility.py", line 244, in <listcomp>
    o = tuple([s.serialize(d) for s, d in zip(self.serializers, data)])
               ^^^^^^^^^^^^^^
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\serializing.py", line 188, in serialize
    return utils.encode_file_to_base64(Path(load_dir) / x)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\work\gen-pipe\gen-env\Lib\site-packages\gradio_client\utils.py", line 689, in encode_file_to_base64
    with open(f, "rb") as file:
         ^^^^^^^^^^^^^
OSError: [Errno 22] Invalid argument: 'data:image\\png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVQImWNgYGBgAAAABQABh6FO1AAAAABJRU5ErkJggg=='

Is it supposed to read it as a file?

[IPv6]

@mashb1t I am passing the encoded empty value (from https://github.com/toutjavascript/FoooXus-Fooocus-Extender/blob/main/src/api.py#L39)

Better ask this in FoooXus github, it is a different project