Fetching metadata from the HF Docker repository...
Oct 25 Yesterday 06:34 Update account and repository links to haiyangliu0 Apr 9 08:46 Update README.md Haiyang Liu
Mar 16 12:28 feat: add Bo Zheng homepage link 12:27 feat: add author homepage links for Yiyi Cai, Yuhan Wu, Kunhang Li 12:22 fix: correct author order per paper 12:20 feat: add FloodDiffusion title, CVPR 2026, authors, paper/code badges 12:16 feat: spectators see 'Take Over' button to seamlessly take control 12:11 fix: syntax error in get_frame else branch indentation 12:04 feat: spectator mode - all visitors see the same streaming generation 11:50 fix: move k_lens to GPU in SDPA fallback (tested locally) 11:44 fix: build proper attention mask in SDPA fallback for text cross-attention 11:37 perf: increase server frame buffer target to 16 for batch fetching 11:34 revert: remove float16 override, bfloat16 is safer (float16 overflow risk) 11:25 perf: batch frame fetching to overcome HTTP round-trip latency 11:20 perf: use float16 instead of bfloat16 in SDPA for T4 tensor core acceleration 11:14 fix: use model_manager.get_buffer_status() for config endpoint instead of model attrs 10:47 fix: adapt model_manager to HF model API (no schedule_config/cfg_config dicts) 10:39 fix: cast SDPA output back to original dtype + idempotent patching 10:34 fix: patch flash_attention with SDPA fallback for T4 (no flash-attn) 10:22 fix: correct CUDA base image tag to 12.4.1 10:19 fix: upgrade to CUDA 12.4 + torch>=2.6 for CVE-2025-32434 compat 10:14 fix: pin transformers<5 for google/umt5-base compat (no model_type in config) 10:08 fix: remove pre-download step, model downloads at runtime startup 10:04 fix: add diffusers, ftfy, regex deps from model dependency tree 10:00 fix: add omegaconf and safetensors deps required by model 09:57 fix: add lightning dependency required by model 09:53 fix: split pip install to avoid index-url conflict for non-torch packages 09:45 feat: initial FloodDiffusion streaming demo for HF Space haiyangliu0 +1: Claude Opus 4.6 (1M context) Docker-based Space with T4 GPU for real-time streaming motion generation.
Model loaded from ShandaAI/FloodDiffusionTiny via HF Hub.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> 09:45 initial commit Haiyang Liu
