nodes/inboxabhishek-ComfyUI-VdoFlash

T

Abhishek Tripathi b5809c2943 Polish: Explicitly provide all VHS inputs to silence warnings.

2026-04-14 17:11:57 +05:30

Polish: Explicitly provide all VHS inputs to silence warnings.

2026-04-14 17:11:57 +05:30

Final stabilization: Added OUTPUT_NODE=True and IS_CHANGED, and finalized API test script.

2026-04-14 17:04:35 +05:30

__init__.py

Initial commit ✨ VdoFlash scaffold with validation layer

2026-04-14 14:55:12 +05:30

.gitignore

Initial commit ✨ VdoFlash scaffold with validation layer

2026-04-14 14:55:12 +05:30

director_nodes.py

Final stabilization: Added OUTPUT_NODE=True and IS_CHANGED, and finalized API test script.

2026-04-14 17:04:35 +05:30

nodes.py

Final: Professional-Grade Native Refactor based on expert audit. Launcher pattern, portable VHS-independence, and zero-deadlock DAG orchestration.

2026-04-14 16:43:01 +05:30

README.md

Initial commit ✨ VdoFlash scaffold with validation layer

2026-04-14 14:55:12 +05:30

README.md

ComfyUI-VdoFlash

A robust ComfyUI Custom Node for orchestrating fully automated image-to-video pipelines with dynamic fallbacks, memory management, and intelligent model validation handling.

Features

Validation Engine: Built-in VRAM constraint checks and dimension auto-correction.
Fail-Safe Video Orchestration: Will intelligently fallback to image frame-repetition if video graphs fail to compile or OOM.
VRAM Aware Continuity Mode (WIP)

Installation

cd ComfyUI/custom_nodes/
git clone <repository_url>