# Imports import json import os import random import sys from typing import Sequence, Mapping, Any, Union def get_value_at_index(obj: Union[Sequence, Mapping], index: int) -> Any: """Return a sequence or mapping result item by index.""" try: return obj[index] except KeyError: return obj["result"][index] def get_comfyui_path() -> str: """Return the configured ComfyUI path, preferring COMFYUI_PATH when set.""" comfyui_path = os.environ.get("COMFYUI_PATH") if comfyui_path: return comfyui_path return find_path("ComfyUI") def find_path(name: str, path: str = None) -> str: """Recursively search parent folders until the named entry is found.""" if path is None: path = os.getcwd() if name in os.listdir(path): path_name = os.path.join(path, name) print(f"{name} found: {path_name}") return path_name parent_directory = os.path.dirname(path) if parent_directory == path: return None return find_path(name, parent_directory) def add_comfyui_directory_to_sys_path() -> None: """Add the ComfyUI checkout to sys.path.""" comfyui_path = get_comfyui_path() if comfyui_path is not None and os.path.isdir(comfyui_path): if comfyui_path in sys.path: sys.path.remove(comfyui_path) sys.path.insert(0, comfyui_path) print(f"'{comfyui_path}' added to sys.path") def add_extra_model_paths() -> None: """Load ComfyUI extra model paths configuration when available.""" try: from main import load_extra_path_config except ImportError: print( "Could not import load_extra_path_config from main.py. Looking in utils.extra_config instead." ) from utils.extra_config import load_extra_path_config extra_model_paths = find_path("extra_model_paths.yaml") if extra_model_paths is not None: load_extra_path_config(extra_model_paths) else: print("Could not find the extra_model_paths config file.") def bootstrap_comfyui_runtime() -> None: """Mirror the allocator-related ComfyUI startup steps before torch import.""" add_comfyui_directory_to_sys_path() import comfy.options comfy.options.enable_args_parsing() from comfy.cli_args import args if os.name == "nt": os.environ["MIMALLOC_PURGE_DELAY"] = "0" if args.default_device is not None: default_dev = args.default_device devices = list(range(32)) devices.remove(default_dev) devices.insert(0, default_dev) devices = ",".join(map(str, devices)) os.environ["CUDA_VISIBLE_DEVICES"] = str(devices) os.environ["HIP_VISIBLE_DEVICES"] = str(devices) if args.cuda_device is not None: os.environ["CUDA_VISIBLE_DEVICES"] = str(args.cuda_device) os.environ["HIP_VISIBLE_DEVICES"] = str(args.cuda_device) os.environ["ASCEND_RT_VISIBLE_DEVICES"] = str(args.cuda_device) if args.oneapi_device_selector is not None: os.environ["ONEAPI_DEVICE_SELECTOR"] = args.oneapi_device_selector if args.deterministic and "CUBLAS_WORKSPACE_CONFIG" not in os.environ: os.environ["CUBLAS_WORKSPACE_CONFIG"] = ":4096:8" import cuda_malloc if "rocm" in cuda_malloc.get_torch_version_noimport(): os.environ["OCL_SET_SVM_SIZE"] = "262144" def cleanup_comfyui_runtime(unload_models: bool | None = None) -> None: """Best-effort cleanup for embedded or repeated generated-script execution.""" import gc def run_cleanup_hook(name: str, should_run: bool = True) -> None: if not should_run or not hasattr(model_management, name): return cleanup_fn = getattr(model_management, name) try: cleanup_fn() except Exception as exc: warnings.warn( f"ComfyUI cleanup hook {name} failed during teardown: {exc}", RuntimeWarning, stacklevel=2, ) should_unload = unload_models if should_unload is None: should_unload = os.environ.get( "COMFYUI_TOPYTHON_UNLOAD_MODELS", "" ).lower() in { "1", "true", "yes", "on", } try: import comfy.model_management as model_management except ModuleNotFoundError: gc.collect() return run_cleanup_hook("cleanup_models_gc") run_cleanup_hook("unload_all_models", should_run=should_unload) run_cleanup_hook("soft_empty_cache") gc.collect() # Workflow data def build_workflow() -> dict[str, Any]: return { "1": { "class_type": "CheckpointLoaderSimple", "inputs": {"ckpt_name": "v1-5-pruned-emaonly-fp16.safetensors"}, }, "2": { "class_type": "CLIPTextEncode", "inputs": { "text": "a small cottage in a meadow, soft daylight", "clip": ["1", 1], }, }, "3": { "class_type": "CLIPTextEncode", "inputs": {"text": "blurry, low quality", "clip": ["1", 1]}, }, "4": { "class_type": "EmptyLatentImage", "inputs": {"width": 512, "height": 512, "batch_size": 1}, }, "5": { "class_type": "KSampler", "inputs": { "seed": 1, "steps": 4, "cfg": 7, "sampler_name": "euler", "scheduler": "normal", "denoise": 1, "model": ["1", 0], "positive": ["2", 0], "negative": ["3", 0], "latent_image": ["4", 0], }, }, "6": { "class_type": "VAEDecode", "inputs": {"samples": ["5", 0], "vae": ["1", 2]}, }, "7": { "class_type": "SaveImage", "inputs": {"filename_prefix": "E2E_text_to_image", "images": ["6", 0]}, }, } def build_extra_pnginfo() -> dict[str, Any] | None: return None workflow = build_workflow() prompt = json.loads(json.dumps(workflow)) extra_pnginfo = build_extra_pnginfo() # Workflow execution def main(unload_models: bool | None = None): bootstrap_comfyui_runtime() add_extra_model_paths() # Node imports from nodes import ( CLIPTextEncode, CheckpointLoaderSimple, EmptyLatentImage, KSampler, NODE_CLASS_MAPPINGS, SaveImage, VAEDecode, ) import torch try: with torch.inference_mode(): checkpointloadersimple = CheckpointLoaderSimple() checkpointloadersimple_1 = checkpointloadersimple.load_checkpoint( ckpt_name="v1-5-pruned-emaonly-fp16.safetensors" ) cliptextencode = CLIPTextEncode() cliptextencode_2 = cliptextencode.encode( text="a small cottage in a meadow, soft daylight", clip=get_value_at_index(checkpointloadersimple_1, 1), ) cliptextencode_3 = cliptextencode.encode( text="blurry, low quality", clip=get_value_at_index(checkpointloadersimple_1, 1), ) emptylatentimage = EmptyLatentImage() emptylatentimage_4 = emptylatentimage.generate( width=512, height=512, batch_size=1 ) ksampler = KSampler() vaedecode = VAEDecode() saveimage = SaveImage() for q in range(1): node_5_seed = prompt["5"]["inputs"]["seed"] = random.randint(1, 2**64) ksampler_5 = ksampler.sample( seed=node_5_seed, steps=4, cfg=7, sampler_name="euler", scheduler="normal", denoise=1, model=get_value_at_index(checkpointloadersimple_1, 0), positive=get_value_at_index(cliptextencode_2, 0), negative=get_value_at_index(cliptextencode_3, 0), latent_image=get_value_at_index(emptylatentimage_4, 0), ) vaedecode_6 = vaedecode.decode( samples=get_value_at_index(ksampler_5, 0), vae=get_value_at_index(checkpointloadersimple_1, 2), ) saveimage_7 = saveimage.save_images( filename_prefix="E2E_text_to_image", images=get_value_at_index(vaedecode_6, 0), prompt=prompt, extra_pnginfo=extra_pnginfo, ) finally: cleanup_comfyui_runtime(unload_models=unload_models) # Entrypoint if __name__ == "__main__": main()