how to use foocus with gradio client api

[ishangrotra]

hey so i have been trying to use foocus with a gradio client api but but i always seem to get the error
'Could not find 'data' key in response. Response received: {'detail': [{'type': 'model_attributes_type', 'loc': ['body'], 'msg': 'Input should be a valid dictionary or object to extract fields from', 'input': '{"data": [null], "fn_index": 0, "session_hash": "7642beb0-f999-42f7-83aa-03483b560420"}', 'url': '[https://errors.pydantic.dev/2.6/v/model_attributes_type](https://errors.pydantic.dev/2.6/v/model_attributes_type/)'}]}'
can any one help me and tell me what is going wrong
code used
from gradio_client import Client

client = Client("link of the hosted gradio app")
result = client.predict(
"ImagePrompt", # str in 'Type' Radio component
fn_index=2
)
print(result)

[t1hq]

so i'm not the only one ：）, I'm suffering this issue for a week, and then founded, I'm not python expert so I thought this might not be that hard.
the key api in focus 2.3.0 is fn 40, and fn 41:

• 40 sets all kinds of params
• 41 makes fooocus to draw

first call 40, then 41, done.

client = Client("http://127.0.0.1:7865/")
result = client.predict(
	...
	fn_index=40
)
print(result)
result= client.predict(fn_index=41)

set params according to values in UI, not in sample.
a demo code is posted here, hopes chinese is not a problem.

[dentro-innovation]

This seemed to work in March, however I tried it yesterday and got the error Expected 1 arguments, got 91.

Anybody knows how to do it nowdays?

[dinasenpai]

Hii, I am facing the same ValueError: Expected 1 arguments, got 92, Please revert back if found any solution.

[dentro-innovation]

Hi I switched over to use https://github.com/mrhan1993/Fooocus-API when I need to use the api.

The instructions in the readme aren't that clear, but here's how to do it:

1. Clone the fooocus-api repo (git clone https://github.com/mrhan1993/Fooocus-API.git)
2. inside the cloned folder create a new virtual environment and activate it (python3 -m venv .venv; source .venv/bin/activate)
3. install the requirements.txt (pip install -r requirements.txt)
4. if you already have Fooocus installed on your machine, go to the Fooocus folder and look for the config.txt file. Copy this file and paste it into the root folder of Fooocus-API
5. run Fooocus-API (inside the Fooocus-API root folder run python3 main.py)
6. go to http://localhost:8888/docs to see the swagger specification of the API
7. Generate an image, e.g.:

url = http://localhost:8888/v2/generation/text-to-image-with-ip # or e.g. "http://localhost:8888/v2/generation/image-prompt"

# Request headers
headers = {"accept": "image/png", "Content-Type": "application/json"}

# Request payload
payload = {
    # "prompt": "a happy customer of DentroAI who can't believe how much more productive his middle sized business is thanks to Dentro",
    "negative_prompt": "",
    "style_selections": [
        "Fooocus V2",
        # "Fooocus Enhance",
        # "Fooocus Sharp"
    ],
    "performance_selection": "Speed",
    "aspect_ratios_selection": "896*1152",
    "image_number": 1,
    "image_seed": -1,
    "sharpness": 2,
    "guidance_scale": 4,
    "base_model_name": "juggernautXL_v8Rundiffusion.safetensors",
    "base_model_name": "juggernautXL_v8Rundiffusion.safetensors",
    "refiner_switch": 0.5,
    "loras": [
        {"enabled": True, "model_name": "None", "weight": 1},
        {"enabled": True, "model_name": "None", "weight": 1},
        {"enabled": True, "model_name": "None", "weight": 1},
        {"enabled": True, "model_name": "None", "weight": 1},
        {"enabled": True, "model_name": "None", "weight": 1},
    ],
    "advanced_params": {
        "adaptive_cfg": 7,
        "adm_scaler_end": 0.3,
        "adm_scaler_negative": 0.8,
        "adm_scaler_positive": 1.5,
        "black_out_nsfw": False,
        "canny_high_threshold": 128,
        "canny_low_threshold": 64,
        "clip_skip": 2,
        "controlnet_softness": 0.25,
        "debugging_cn_preprocessor": False,
        "debugging_dino": False,
        "debugging_enhance_masks_checkbox": False,
        "debugging_inpaint_preprocessor": False,
        "dino_erode_or_dilate": 0,
        "disable_intermediate_results": False,
        "disable_preview": False,
        "disable_seed_increment": False,
        "freeu_b1": 1.01,
        "freeu_b2": 1.02,
        "freeu_enabled": False,
        "freeu_s1": 0.99,
        "freeu_s2": 0.95,
        "inpaint_advanced_masking_checkbox": True,
        "inpaint_disable_initial_latent": False,
        "inpaint_engine": "v2.6",
        "inpaint_erode_or_dilate": 0,
        "inpaint_respective_field": 1,
        "inpaint_strength": 1,
        "invert_mask_checkbox": False,
        "mixing_image_prompt_and_inpaint": False,
        "mixing_image_prompt_and_vary_upscale": False,
        "overwrite_height": -1,
        "overwrite_step": -1,
        "overwrite_switch": -1,
        "overwrite_upscale_strength": -1,
        "overwrite_vary_strength": -1,
        "overwrite_width": -1,
        "refiner_swap_method": "joint",
        "sampler_name": "dpmpp_2m_sde_gpu",
        "scheduler_name": "karras",
        "skipping_cn_preprocessor": False,
        "vae_name": "Default (model)",
    },
    "save_meta": True,
    "meta_scheme": "fooocus",
    "save_extension": "png",
    "save_name": "",
    "read_wildcards_in_order": False,
    "require_base64": False,
    "async_process": False,
    "webhook_url": "",
    "image_prompts": [],
}

response = requests.post(url, json=payload, headers=headers)
# Save the generated image
with open("happy_customer_of_dentroAI.png", "wb") as f:
    f.write(response.content)

Things to know:

• you don't need to run your other Fooocus instance at the same time while running Fooocus-API
• for manual image generation I still use the normal Fooocus, just running it as normal

[t1hq]

the api changed from 2.3.0 to 2.5.5, it looks like the key func is now fn 67 and fn 68, but more params are added to fn 67, the params need reviewing if you want to use fooocus 2.5.5 api

[dinasenpai]

Thanks for the detailed approach, worked like a charm for me.

[mashb1t]

also see #2608

[ishangrotra]

hey thanks for taking time i wanted to ask will the provided code work in the link for gradio live link ?
or just for the local link
example gradio live link : https://4dc12345678.gradio.live/

[saibabu10]

I have a question how to connect with the javascript api ? please send me the code if you know

[mashb1t]

docs see https://www.gradio.app/guides/getting-started-with-the-js-client, same parameters and endpoint as above