From a0b82f98d657ccd148448e27d18922884b8a1632 Mon Sep 17 00:00:00 2001 From: JingyeChen <57484508+JingyeChen@users.noreply.github.com> Date: Thu, 7 Nov 2024 16:22:10 +0800 Subject: [PATCH] Add files via upload --- docs/index.html | 1170 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 1170 insertions(+) create mode 100644 docs/index.html diff --git a/docs/index.html b/docs/index.html new file mode 100644 index 0000000..3e614d2 --- /dev/null +++ b/docs/index.html @@ -0,0 +1,1170 @@ + + + + + + + + +
+ VideoTuna Logo +
+ + VideoTuna: Let's Finetune Video Generation Models! + + + + + + + + + + + + + + + + + + + + + + + + + +
+
+
+
+
+

VideoTuna: Let's Finetune Text-to-Video Models!

+
+ + VideoTuna is a pioneering codebase for video generation applications. It offers comprehensive pipelines covering pre-training, continuous training, post-training alignment, and fine-tuning. + + + + +
+ + + +
+
+ + + + + + + + + + + Code (v0.1.0) + + + + + + + + + Discord + + + + + + + + + WeChat + + + + +
+
+
+
+
+
+
+ + + + + + + + + + + + + +
+
+ + +
+
+
+

โœจHighlights: What is VideoTuna For?

+
+
+ +
+
+

๐ŸŒŸ

+

All in one framework

+

Inference and fine-tune state-of-the-art video generation models

+
+
+
+
+

๐ŸŒŸ

+

Pre-training

+

Build your own foundational text-to-video model

+
+
+
+
+

๐ŸŒŸ

+

Continuous training

+

Keep improving your model with new data

+
+
+
+
+

๐ŸŒŸ

+

Domain-specific

+

Adapt models to your specific scenario

+
+
+ + +
+
+

๐ŸŒŸ

+

Concept-specific

+

Teach your models with unique concepts

+
+
+
+
+

๐ŸŒŸ

+

Language understanding

+

Improve model's comprehension through training

+
+
+
+
+

๐ŸŒŸ

+

Post-processing

+

Enhance videos with enhancement model

+
+
+
+
+

๐ŸŒŸ

+

Human alignment

+

Post-training with RLHF for better results

+
+
+
+
+
+
+
+ + +
+
+ + +
+
+ +

๐ŸŒ†Gallery

+ +
+ +
+
+

All-in-one Model Inference
and Comparison

+
+
+ + +
+
+

3D Video VAE
for Video Compression

+
+
+ + +
+
+

Domain Finetuning
for Facial Video Generation

+
+
+ + +
+
+

Concept Finetuning
for Storytelling

+
+
+ +
+ + +
+ + +
+ +
+
+
+ Demo 1 +
+ Ground Truth +
+
+
+
+ +
+
+
+ Demo 2 +
+ VAE Reconstruction +
+
+
+
+ +
+
+
+ Demo 3 +
+ Ground Truth +
+
+
+
+ +
+
+
+ Demo 4 +
+ VAE Reconstruction +
+
+
+
+ + +
+
+
+ Demo 5 +
+ Ground Truth +
+
+
+
+ +
+
+
+ Demo 6 +
+ VAE Reconstruction +
+
+
+
+ +
+
+
+ Demo 7 +
+ Ground Truth +
+
+
+
+ +
+
+
+ Demo 8 +
+ VAE Reconstruction +
+
+
+
+ + +
+
+
+ Demo 9 +
+ Ground Truth +
+
+
+
+ +
+
+
+ Demo 10 +
+ VAE Reconstruction +
+
+
+
+ +
+
+
+ Demo 11 +
+ Ground Truth +
+
+
+
+ +
+
+
+ Demo 12 +
+ VAE Reconstruction +
+
+
+
+
+
+ + + + + + + + + + + + +
+
+ + + + + +
+ + +
+
+ + +
+
+ +

๐Ÿ“œLicense and Contact

+
+ Please follow CC-BY-NC-ND. If you want a license authorization, please contact + yhebm@connect.ust.hk and + yxingag@connect.ust.hk. +
+
+
+
+ + + +