BestHub
Sign in
Tag

Diffusion Transformer

0 views collected around this technical thread.

Bilibili Tech
Bilibili Tech
Mar 4, 2025 · Artificial Intelligence

Engineering Practices and Optimizations for Text‑to‑Video Generation Models (OpenSora, CogVideoX) on Bilibili TTV Team

The Bilibili TTV team optimized OpenSora and CogVideoX text‑to‑video models by redesigning data storage with Alluxio, parallelizing VAE encoding, applying dynamic sequence‑parallel and DeepSpeed‑Ulysses attention, adapting GPU code for NPU execution, leveraging profiling‑driven kernel fusion, FlashAttention, and expandable memory to dramatically increase training efficiency and frame throughput, while outlining future pipeline‑parallel and ZeRO‑3 scaling plans.

Diffusion TransformerFlashAttentionNPU
0 likes · 26 min read
Engineering Practices and Optimizations for Text‑to‑Video Generation Models (OpenSora, CogVideoX) on Bilibili TTV Team
Tencent Cloud Developer
Tencent Cloud Developer
May 15, 2024 · Artificial Intelligence

Tencent Open-Sources HunYuan DiT: First Chinese-Native Text-to-Image Model with 1.5B Parameters

Tencent has open‑sourced its upgraded 1.5‑billion‑parameter HunYuan DiT model—the first Chinese‑native, bilingual (Chinese‑English) text‑to‑image diffusion‑with‑transformer system—delivering about 20% visual quality improvement, multi‑round generation, video‑generation potential, and free commercial use, with full weights, inference code, and algorithms available on Hugging Face and GitHub for developers and enterprises.

Chinese-native AIDiT architectureDiffusion Transformer
0 likes · 6 min read
Tencent Open-Sources HunYuan DiT: First Chinese-Native Text-to-Image Model with 1.5B Parameters
Tencent Cloud Developer
Tencent Cloud Developer
Feb 21, 2024 · Artificial Intelligence

OpenAI Sora: Technical Principles and Industry Impact Analysis

OpenAI’s Sora, a text‑to‑video model released during Chinese New Year, combines a VAE encoder, latent diffusion with a DiT transformer, and a VAE decoder to generate videos from prompts, supporting flexible durations and resolutions, language understanding, and uses in creation, editing, and entertainment, though it struggles with physical consistency and long‑term coherence, and its debut is reshaping short‑form video, digital‑human, gaming, and graphics industries.

AI video generationDiffusion TransformerOpenAI
0 likes · 14 min read
OpenAI Sora: Technical Principles and Industry Impact Analysis