My (old) go-to workflows for use in comfy UI. Get the new one HERE
You will need some additional nodes, in the comfy manager go to 'manager'>install missing custom nodes and that should do the trick.
Try my Flux prompt craft GPT
Get the GGUF CLIPs here
My SDXL/PONY workflow uses Clip Vector Sculpting to improve image quality
The FLUX AIO and FLUX + Upscale AIO V1 is as simple as possible for users using one of the AIO (All In One) safetensors checkpoints here on Civit and not the Official DiT only version from Hugging Face.
The FLUX + UPSCALE + Lora v3 is the one I use most, and is compatible with NepotismFUX V2, it's as simple as possible while still allowing you to change whatever you need on the fly. perfect for intermediate and expert users alike.
Description
more better 👍
FAQ
Comments (10)
You didn't made the model_name field visible for the UpscaleModelLoader in the group node. So if you don't have 4x_NMKD-Siax_200k.pth in your upscale_models folder you can't change it and it will create an error.
good catch, thanks. i updated the file
I got the image in preview window,but it upscale failed.
error message:
Error occurred when executing UpscaleModelLoader: 'NoneType' object has no attribute 'lower' File "E:\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\ComfyUI_windows_portable\ComfyUI\comfy_extras\nodes_upscale_model.py", line 29, in load_model sd = comfy.utils.load_torch_file(model_path, safe_load=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\ComfyUI_windows_portable\ComfyUI\comfy\utils.py", line 14, in load_torch_file if ckpt.lower().endswith(".safetensors") or ckpt.lower().endswith(".sft"): ^^^^^^^^^^
then I download 4x_NMKD-Siax_200k.pth to models/upscale_models ，it work fine.
please redownload, i just updated it, you prob just dont have the same upscale model im using
Love this workflow! Very compact and easy to use, the upscaled images come out fantastic!! Thanks for sharing!
this dont work after execute got this error:
Error occurred when executing CheckpointLoaderSimple: cannot access local variable 'dtype_t5' where it is not associated with a value File "F:\Ai\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\Ai\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\Ai\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\Ai\ComfyUI_windows_portable\ComfyUI\nodes.py", line 518, in load_checkpoint out = comfy.sd.load_checkpoint_guess_config(ckpt_path, output_vae=True, output_clip=True, embedding_directory=folder_paths.get_folder_paths("embeddings")) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\Ai\ComfyUI_windows_portable\ComfyUI\comfy\sd.py", line 544, in load_checkpoint_guess_config clip_target = model_config.clip_target(state_dict=sd) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\Ai\ComfyUI_windows_portable\ComfyUI\comfy\supported_models.py", line 659, in clip_target return supported_models_base.ClipTarget(comfy.text_encoders.flux.FluxTokenizer, comfy.text_encoders.flux.flux_clip(dtype_t5=dtype_t5))
you need to update comfyui for the flux support update, also make sure you have the t5xxl_fp8_e4m3fn clip downloaded and in your clip folder
@BobsBlazed I have all setup because i use the normal way of flux and work but this workflow and model always give the error!
@smereces_ai what workflow version of mine are you using and what version of Flux are you using
i use the upscaler + lora and it works ok, when i use just flux i cant change the checkpoint