3 Commits

Include renames

Author	SHA1	Message	Date
limbicnation	ebbb25a2a3	Add dual-stream prompt refiner node ... Add PromptDualStreamRefinerNode that produces a positive and negative prompt pair in a single pass via Ollama, intended for the shipped Q8 GGUF of qwen2-5-7b-dual-stream-prompt-lora. Reuses OllamaClient for streaming, timeouts, progress, and llama-runner crash handling, and parses Positive/Negative output defensively across label variants. Includes config/Modelfile.dualstream, unit tests, node registration, and a corrected implementation plan replacing the invalid local transformers/PEFT approach.	2026-06-10 22:10:16 +02:00
limbicnation	52dcfbcb17	fix: address PR #8 review issues ... - Fix _weighted_average deduplication bug (now uses weight-ratio emphasis markers) - Replace bare except Exception with specific exception handling in OllamaClient - Propagate API contract mismatches (TypeError/AttributeError) as RuntimeError - Replace print() with logging.getLogger(__name__) in all new nodes - Expose top_p in PromptRefinerNode INPUT_TYPES for consistency - Update PR-8-REVIEW.md with fix log and merge recommendation	2026-04-28 05:01:03 +02:00
limbicnation	89c8341612	feat: add prompt chain nodes + test infrastructure ... Phase 1 — Foundation Hardening: - Extract ollama_client.py adapter from PromptGeneratorNode - Add pytest suite (29 tests, 70% coverage gate) - Unify style sources into config/styles.yaml - Extend CI with pytest + coverage Phase 2 — Prompt Chain Nodes: - PromptRefinerNode: iterative LLM refinement (1-3 passes) - NegativePromptNode: style-aware negative prompt generation - PromptCombinerNode: blend/concat/weighted_average modes New graph target: PromptGenerator -> PromptRefiner -> PromptCombiner -> CLIP ^ NegativePrompt ---------+ Closes roadmap phases 1.1-2.3	2026-04-27 07:00:58 +02:00