img2img inpaint from api and webui result is different

[zlxie0106]

I used img2img inpaint upload two of same image call "cupofwater.png"
These are my settings:
prompt: street lora:LCM_LoRA_Weights_SDXL:1

then, I got this:

I tried to copy the env value to API and I expect I can get the same result:
the cupofwater.png is the transparent background.

`import webuiapi
from PIL import Image

api = webuiapi.WebUIApi(host='127.0.0.1', port=7860)

image = Image.open("cupofwater.png")
mask = Image.open("cupofwater.png")
inpainting_result = api.img2img(images=[image],
resize_mode=0,
mask_image=mask,
inpainting_fill=0,
mask_blur=0,
sampler_index = 'DPM++ SDE Karras',
inpainting_mask_invert=1,
prompt="street lora:LCM_LoRA_Weights_SDXL:1",
cfg_scale=2,
steps = 10,
denoising_strength=1
)

inpainting_result.image
`

but I got this:

plz tell me what am I missing.

[w-e-w]

I never used the https://github.com/mix1009/sdwebuiapi
it is not implemented by us so if you are issues using it then it's better that you go ask them
https://github.com/mix1009/sdwebuiapi

what I'm guessing is that because UI default vaules and the API default values are not the same
which means if you did not fill in all the arguments, it might default to some unexpected value you didn't expect

also assuming that default seed is -1 which means random seed
so even if you use the exact same parameter with random sees you're going to get different results

I suggest you implement your client yourself and don't use a pre-built Library by others
if you wish to do so you can look at the my example on the wiki and use API payload display to generate the payload https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/API

in my opinion this is a more reliable way of using the API

[zlxie0106]

thank you for your help.
as i know, webui sends JSON format api to the server. if I can catch my value when I send to API of payload from webui. maybe I can get something inspired. Would you tell me where may i catch these values?

[w-e-w]

put a breakpoint in

stable-diffusion-webui/modules/api/api.py

Lines 487 to 556 in ac8ffb3

	def img2imgapi(self, img2imgreq: models.StableDiffusionImg2ImgProcessingAPI):
	task_id = img2imgreq.force_task_id or create_task_id("img2img")
	init_images = img2imgreq.init_images
	if init_images is None:
	raise HTTPException(status_code=404, detail="Init image not found")
	mask = img2imgreq.mask
	if mask:
	mask = decode_base64_to_image(mask)
	script_runner = scripts.scripts_img2img
	infotext_script_args = {}
	self.apply_infotext(img2imgreq, "img2img", script_runner=script_runner, mentioned_script_args=infotext_script_args)
	selectable_scripts, selectable_script_idx = self.get_selectable_script(img2imgreq.script_name, script_runner)
	populate = img2imgreq.copy(update={ # Override __init__ params
	"sampler_name": validate_sampler_name(img2imgreq.sampler_name or img2imgreq.sampler_index),
	"do_not_save_samples": not img2imgreq.save_images,
	"do_not_save_grid": not img2imgreq.save_images,
	"mask": mask,
	})
	if populate.sampler_name:
	populate.sampler_index = None # prevent a warning later on
	args = vars(populate)
	args.pop('include_init_images', None) # this is meant to be done by "exclude": True in model, but it's for a reason that I cannot determine.
	args.pop('script_name', None)
	args.pop('script_args', None) # will refeed them to the pipeline directly after initializing them
	args.pop('alwayson_scripts', None)
	args.pop('infotext', None)
	script_args = self.init_script_args(img2imgreq, self.default_script_arg_img2img, selectable_scripts, selectable_script_idx, script_runner, input_script_args=infotext_script_args)
	send_images = args.pop('send_images', True)
	args.pop('save_images', None)
	add_task_to_queue(task_id)
	with self.queue_lock:
	with closing(StableDiffusionProcessingImg2Img(sd_model=shared.sd_model, **args)) as p:
	p.init_images = [decode_base64_to_image(x) for x in init_images]
	p.is_api = True
	p.scripts = script_runner
	p.outpath_grids = opts.outdir_img2img_grids
	p.outpath_samples = opts.outdir_img2img_samples
	try:
	shared.state.begin(job="scripts_img2img")
	start_task(task_id)
	if selectable_scripts is not None:
	p.script_args = script_args
	processed = scripts.scripts_img2img.run(p, *p.script_args) # Need to pass args as list here
	else:
	p.script_args = tuple(script_args) # Need to pass args as tuple here
	processed = process_images(p)
	finish_task(task_id)
	finally:
	shared.state.end()
	shared.total_tqdm.clear()
	b64images = list(map(encode_pil_to_base64, processed.images)) if send_images else []
	if not img2imgreq.include_init_images:
	img2imgreq.init_images = None
	img2imgreq.mask = None
	return models.ImageToImageResponse(images=b64images, parameters=vars(img2imgreq), info=processed.js())

[zlxie0106]

@w-e-w I added print(vars(img2imgreq)) in line 489. when I code and send api request, is truly to print these values, but when I used from webui to generate img2img, it did not execute print(vars(img2imgreq)). maybe there is another where to get webui requests?