#4040th of 79 in Image and videoFor image, video and design, image and design
hunyuandit
Bilingual Chinese-English text-to-image diffusion transformer with training and ControlNet
Live demo ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 4.3k
- License
- custom license
- Last commit
- Oct 2026
- Language
- Jupyter Notebook
Missed the latest GitHub check; these facts may be older.
Overview
PyTorch model definitions, pre-trained weights and inference code for Hunyuan-DiT, a latent diffusion transformer that uses a bilingual CLIP and multilingual T5 text encoder. The repo also covers full-parameter and LoRA training, ControlNet (canny, pose, depth), IP-Adapter, a captioner, and a DialogGen-based multi-turn Gradio demo. Diffusers, ComfyUI and Kohya integrations are provided.
Who it is for: Engineers self-hosting or fine-tuning text-to-image models
Strengths
- Handles Chinese and English prompts via bilingual CLIP plus mT5 encoders
- Ships training code: full-parameter and LoRA, plus ControlNet and IP-Adapter
- Integrations for Diffusers, ComfyUI and Kohya; Gradio and CLI demos
- Distillation and TensorRT versions reported at 50% and 47% faster; 6GB VRAM scripts
Weaknesses
- Requires an NVIDIA CUDA GPU; minimum 11GB peak memory for the standard pipeline
- Tested only on Linux with V100 and A100 GPUs
- Docker images are tar downloads, not a published registry image or compose file
- Hunyuan-DiT-S checkpoints and a WebUI are still unchecked in the plan
What it needs
- GPU required
- Needs PyTorch, CUDA, Conda, flash-attention, huggingface-cli
- Models: Hunyuan-DiT v1.1, Hunyuan-DiT v1.2, Distillation, ControlNet, LoRA
Also in Image and video
See all 79| Rank | Project | Score |
|---|---|---|
| 1 | frigateLocal NVR with realtime object detection for IP cameras | 93 out of 100 |
| 2 | MoneyPrinterTurboGenerates short videos from a topic, with script, footage, voiceover and subtitles | 88 out of 100 |
| 3 | toonflow-appInfinite-canvas tool for AI short drama and comic video production | 87 out of 100 |
| #4 | ComfyUINode-graph editor for running diffusion image, video, audio and 3D workflows | 86 out of 100 |
| #5 | InvokeAIWeb UI for diffusion image generation with inpainting canvas and node workflows | 86 out of 100 |