Minimax H3 ref2va + fl2va Workflow
Simple toggle between ref2va and fl2va
Uses stock Minimax H3 sampler settings with Spectrum for a ~30-40% speed boost
I have yet to find a single other combination that doesn't have some kind of worse downside - The turbo LoRAs suck.
Motion Context video extension using save and load latents
How to use in workflow
Can lead to gamma shift after extending - may need to manually adjust brightness/contrast
FaceRefine for fixing distant garbled faces
required custom_nodes that aren't on the ComfyUI Manager included in archive
Simple postproduction settings
Interpolation using https://github.com/ethanfel/ComfyUI-Tween
GIMM VFI
SPEED VFI (seems better than RIFE at about the same speed)
Upscale
Tensorrt frame by frame upscale
RTX VSR
RTX Denoise/Deblur
Postproduction looping tools for fl2va using the same first and last frame
clip still end or starting frames + audio
Smoothen out the loop with rife generated frames using Seamless Loop
Generation Preview
Batch or single video upscale using GIMM VFI+Tensorrt as toggle below the main workflow
throw your videos into any folder inside ComfyUI/input/... and toggle the batch upscale to get everything running
No workflow will help, if your prompt doesn't follow the guidelines. I recommend looking through the following prompting guides if you have no idea about Minimax H3 prompt and then use other's or my prompts as help to get started.
I added the full workflows, including prompt and resources I used in the example videos, to the workflow archive.
https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/docs/VIDEO_PROMPT_WRITING_GUIDE_base_en.md
https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/docs/VIDEO_PROMPT_WRITING_GUIDE_ref_en.md
Resources
https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_audio_vae_fp32.safetensors
for low VRAM/RAM users
Thanks to
Description
I'll use this as my resource upload for my H3 videos - this is the easiest way to share it. I'll update the file every now and then with the resources of some of my videos on request or whenever I feel like it.
FAQ
Comments (11)
I have two issues or one, with either comfyui_fearnworksnodes
or
MiniMax-H3-NativeAudio-MusicVideo-Workflow
they appear as missing nodepacks, I hgave them both, try update and\or later versions.
You don't need to install Fearnworks nodes (I don't have them installed either), I don't know which but one of the nodes says it needs it but it doesn't.
For the "MiniMax-H3-NativeAudio-MusicVideo" you need to add the folder from the workflow archive inside the custom_nodes folder to your own custom_nodes folder. It's part of the FaceRefine but the author didn't include the nodes since they are part of another git but they are hidden inside another folder inside that git... so I added it to the wf archive to make it easier.
Thank for the reply! Somehow I managed to make it work. I just have another question, the final video was saved to xxx.safetensor, how can I make it save to video? I'm afraid to mess further things xD
The workflow saves it as latent under the "output/h3_context" (to use later as MotionContext video extensions) but also as mp4 in the "output/MiniMax_H3/current_date" folder.
@LastAssignment I know that was strange. I'm not getting the mp4 videos saved. Sorry I'm just too new to minimax. I've checked all the folders and even change the default path. But also noticed despite the model runs I'm getting these errors at the command line:
[ERROR] Failed to validate prompt for output 2199:
[ERROR] * MiniMaxH3ReferenceToVideo 2268:9:
[ERROR] - Exception when validating inner node: tuple index out of range
[ERROR] * BasicGuider 2268:14:
[ERROR] - Exception when validating inner node: tuple index out of range
[ERROR] * BasicScheduler 2268:16:
[ERROR] - Exception when validating inner node: tuple index out of range
[ERROR] * H3FaceTrackCrop 2268:2:
[ERROR] - Value not in list: select: 'largest_face' not in ['largest', 'most_central']
[ERROR] - Value not in list: detector: 'Looping Tools' not in ['face_yolov8m.pt', 'bbox\\face_yolov8m.pt', 'segm\\text\\person_yolov8m-seg.pt']
[ERROR] * SAMLoader 2268:20:
[ERROR] - Value not in list: model_name: 10 not in ['sam_vit_b_01ec64.pth']
[ERROR] Output will be ignored
@usernew87 It fails running the FaceRefine because it's missing the models, either disable FaceRefine or download the models (face_yolov8m.pt and sam_vit_b_01ec64.pth) using the ComfyUI Manager Model Manager.
Can you export the current workflow you're trying to run under File -> Export and upload it somewhere so I can see if you disconnected anything else by mistake?
@LastAssignment I keep getting these errors:
[INFO] got prompt
[ERROR] Failed to validate prompt for output 2199:
[ERROR] * BasicScheduler 2268:16:
[ERROR] - Exception when validating inner node: tuple index out of range
[ERROR] * MiniMaxH3ReferenceToVideo 2268:9:
[ERROR] - Exception when validating inner node: tuple index out of range
[ERROR] * BasicGuider 2268:14:
[ERROR] - Exception when validating inner node: tuple index out of range
[ERROR] Output will be ignored
I deleted my old workflow but even loading the default one I can't get the video to be saved, but I see the preview.
here's the failing workflow.
https://drive.google.com/file/d/1VIHtUpsNiky2omP1az-5VKVFmXOcE4J8/view?usp=sharing
@usernew87 It's still related to the FaceRefine not working, I honestly have no idea at this point. Does it work when you disable FaceRefine in the subgraph? What ComfyUI version are you running?
@LastAssignment Yes, I try disabling the FaceRefine and everything else. I always get errors at 'Conditioning + EMPTY AV latent', 'DENOISE - the key dial' and 'Guider'. The workflow worked perfectly fine, the preview show nice generations just that strange bug that didn't save the video. I messed a few nodes and managed to make it to generate and save the video. I'm not sure but it indeed seems a FaceRefine and maybe H3 Per-Frame Denoise bug. I removed also the preview , Idid a mess on the nodes xD
https://drive.google.com/file/d/1cNh7TAIlAYCGWm8SPc_DpS4eT-5zn1ao/view?usp=sharing
I made it work. I deleted all custom nodes and re-installed just the one needed for the workflow again, maybe there was a conflict or something. Thanks a lot for the help and your time, works like a charm now!
@usernew87 Good to hear that it worked out!