What's the pipeline for the following API call like?

[Sasopsy]

txt2img_payload = {
    "prompt": "Bald head <lora:bald_lora:1>",
    "sampler_name": "DPM++ 2M SDE",
    "steps": 20,
    "alwayson_scripts": {
        "controlnet": {
            "args": [
                {
                    "control_mode": "My prompt is more important",
                    "module": "inpaint_only",
                    "model": "control_v11p_sd15_inpaint [ebff9138]",
                    "enabled": True,
                    "image": {
                        "image": encoded_image, 
                        "mask": encoded_mask 
                    },
                    "low_vram": False,
                    "pixel_perfect": False,
                    "processor_res": -1,
                    "resize_mode": "Crop and Resize",
                    # "resize_mode": "Just Resize",
                    "save_detected_map": True,
                    "hr_option": "Both",
                    "guidance_start": 0,
                    "guidance_end": 1,
                    "weight": 1.0
                }]}},
                    "image": {
                        "image": encoded_image,
                    },
                    "height": 512,
                    "hr_upscaler": "Latent",
                    "n_iter": 1,
                    "seed": 10,
                    "width": 512,
                    "denoising_strength": "0.7",
                    "override_settings_restore_afterwards": True,
                    "mask_blur": 4
                }

Hi, the above is a txt2img payload for an inpaint task. I want to know what's happening to the mask. Are they added as an inpaint condition (removing the masked parts) or as a separate entity? If so, how does the txt2img pipeline process them?