π₯ ComfyUI Workflow Release: Hunyuan Text-to-Video and Video-to-Video Generator
Added 2024-12-07 14:12:18 +0000 UTCHey Black Mixture Fam!
Transform your text descriptions into stunning videos with our latest Hunyuan-powered workflow! This implementation brings professional-quality video generation capabilities right to your ComfyUI setup.
π¦ Required Models:
Transformer & VAE Files
Download from: Hunyuan Video Comfy Repository
Place Transformer files in: ComfyUI/models/diffusion_models
Place VAE files in: ComfyUI/models/vae
Note: Manual download required (no autodownload)
LLM Text Encoder
Source: LLaVA-Llama Text Encoder
Install Path: ComfyUI/models/LLM/llava-llama-3-8b-text-encoder-tokenizer
Automatically downloads on first run
CLIP Text Encoder
Source: CLIP-ViT-Large
Install Path: ComfyUI/models/clip/clip-vit-large-patch14
Include: .safetensors weights and all config files
Supports autodownload
β¨ Features:
Text-driven video generation
High-quality motion synthesis
Customizable video parameters
Professional-grade output
π¨ Workflow Sections:
π₯ Load Models: Configure Hunyuan Video and text encoding components
π¨ Prompt: Input your video description text
π¦ Generation Settings: Adjust video parameters and quality
π© Save Video: Export your generated video
βοΈ Quick Start:
1. Update ComfyUI to latest version
2. Download and install all required models (Hunyuan)
3. Load the workflow
4. Install Missing Nodes
5.Enter your text description
6. Adjust generation parameters if needed
7. Generate and save your video
π‘ Important Notes:
Ensure your ComfyUI is up to date
Follow exact model placement instructions
Special thanks to Kijai for the wrappers and workflow examples
π HUGE Thanks to all of our amazing Patreon supporters and Happy Holidays!
Comments
This means you're missing a custom node required to run the workflow. Open the comfyui manager and click "install missing custom nodes" and it should appear with a list of the missing nodes. Click "install" and then click "restart" and it should work after it reboots.
Black Mixture
2024-12-29 06:33:34 +0000 UTCi tried this but i am getting this as an Missing Node Type - EmptyHunyuanLatentVideo can you tell me how i can get that working - its appearing in STEP 3 random Noise box as Red
Avinash Kumar
2024-12-29 06:30:50 +0000 UTCHi Steve, Thanks for your detailed feedback. I completely understand how frustrating it can be to troubleshoot issues like this, especially with the extra steps involved in setting up custom nodes. I wanted to let you know that Iβll be working on updating the workflow soon to utilize the native Hunyuan support in the latest version of ComfyUI. This should eliminate the need for additional custom nodes and make it much easier to get everything working without all the hassle. Thanks for your patience, and Iβll keep you posted when the updated version is ready!
Black Mixture
2024-12-21 18:06:34 +0000 UTCThis never worked for me. After 2 days of frustration, uninstalling/reinstalling dependencies, triple checking Alexander's post above, reading through Redit, asking ChatGPT for assistance, etc., I was never able to get this to work. I got Triton installed after much effort. And Sageattention. But Then I hit a CUDA DLL not loading issue and nothing - absolutely nothing - could get around that. Fortunately, the latest version of Comfy (out yesterday) has built in support for Hunyuan, so the custom nodes are no longer required. I did a fresh install and used their workflow and it works great. I really wanted to use this workflow, but unfortunately it errors out, even when used with the fresh Comfy install that runs Hunyuan natively.
Steve Warner
2024-12-21 17:55:40 +0000 UTC
