|
|
--- |
|
|
base_model: |
|
|
- tencent/HunyuanVideo-1.5 |
|
|
license: other |
|
|
license_name: tencent-hunyuan-community |
|
|
license_link: https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5/blob/master/LICENSE |
|
|
pipeline_tag: text-to-video |
|
|
tags: |
|
|
- t2v |
|
|
- i2v |
|
|
- hunyuan |
|
|
- rapid |
|
|
- aio |
|
|
--- |
|
|
This is a combination of HunyuanVideo 1.5, VAE and accelerators to make using the video model easy and fast. |
|
|
|
|
|
Supports both T2V and I2V in 1 model. 1 CFG, 4 steps (but see releases below, as generally 6-8 gives far better results). FP8 precision (except the VAE, which is FP16). |
|
|
|
|
|
 |
|
|
|
|
|
**I2V usage:** you will need the "clip vision" model: https://huggingface.co/Comfy-Org/HunyuanVideo_1.5_repackaged/tree/main/split_files/clip_vision |
|
|
|
|
|
**T2V usage:** bypass the "input image" optional group. |
|
|
|
|
|
**V1:** 50/50 mix of the 720p I2V and T2V models. Lightx2v 480p T2V 4-step LORA added at 0.5 strength, as more seems to kill motion. Use 6+ steps for far better results. **er_sde/beta recommended**. |
|
|
|
|
|
**LICENSE COMPLIANCE:** |
|
|
|
|
|
License: TENCENT HUNYUAN COMMUNITY LICENSE (https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5/blob/main/LICENSE) |
|
|
|
|
|
Tencent is not affiliated with, associated with, sponsoring, or endorsing anything related to this repository. |
|
|
|
|
|
I have changed the files. |
|
|
|
|
|
Powered by Tencent Hunyuan |
|
|
|
|
|
Tencent Hunyuan is licensed under the Tencent Hunyuan Community License Agreement, Copyright © 2025 Tencent. All Rights Reserved. The trademark rights of “Tencent Hunyuan” are owned by Tencent or its affiliate. |
