Files
Joseph IbrahimandClaude Opus 4.5 3bdb98b138 v1.1.0: AGPL-3.0 dual-license + artist-friendly naming
LICENSING:
- License: Apache 2.0 → AGPL-3.0 + Commercial dual-license
- Added COMMERCIAL_LICENSE.md with pricing tiers
- Added COPYRIGHT with IP documentation
- Added LICENSE_HEADER.txt for source files
- SPDX identifiers in all Python files

ARTIST UX:
- Category: deterministic → JI/Reproducible
- Display names: Technical jargon → Artist-friendly terms
  - Deterministic Sampler → Locked Sampler ⟳
  - Checksum Validator → Output Matcher
  - MoE Router → Expert Selector
  - Cascade Refiner → Multi-Pass Refiner
  - ECHO Context → Memory Recall

TECHNICAL:
- pyproject.toml: Updated license, classifiers
- README: Updated with new naming and workflow diagrams
- Consistent with ThinkingMachines batch-invariant research
- batch_size=1 enforcement documented

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-01-14 16:11:56 -05:00

38 lines
1.3 KiB
Plaintext

ComfyUI Deterministic Nodes
Copyright (C) 2025 Joseph Ibrahim <joseph@josephibrahim.com>
DUAL-LICENSE NOTICE
===================
This software is dual-licensed:
1. GNU Affero General Public License v3.0 (AGPL-3.0)
- See LICENSE file for full terms
- Free for open source use with copyleft requirements
- Network use triggers source disclosure requirements
2. Commercial License
- See COMMERCIAL_LICENSE.md for terms and pricing
- Available for proprietary/SaaS use without AGPL obligations
- Contact: joseph@josephibrahim.com
INTELLECTUAL PROPERTY
=====================
This software contains original implementations of:
- Batch-Invariant Inference: Per-item RNG reset methodology for guaranteed reproducibility
- Deterministic MoE Routing: Hash-based expert selection eliminating MCMC variance
- ECHO Context Integration: 4-tier hierarchical memory retrieval patterns
- Cascade Refinement: Sequential multi-stage deterministic processing
- Checksum Validation: Cryptographic reproducibility proof system
Based on ThinkingMachines research on defeating non-determinism in LLM inference.
ATTRIBUTION
===========
If you use this software, please cite:
Joseph Ibrahim, "ComfyUI Deterministic Nodes", 2025
https://github.com/joe002/comfyui-deterministic-nodes