From a7340d378909cadbfd72dbd1f5b756f907c2a3e0 Mon Sep 17 00:00:00 2001 From: skal Date: Fri, 13 Feb 2026 17:46:09 +0100 Subject: CNN v2 training: Add --grayscale-loss option for luminance-based loss computation Add option to compute loss on grayscale (Y = 0.299*R + 0.587*G + 0.114*B) instead of full RGBA channels. Useful for training models that prioritize luminance accuracy over color accuracy. Changes: - training/train_cnn_v2.py: Add --grayscale-loss flag and grayscale conversion in loss computation - scripts/train_cnn_v2_full.sh: Add --grayscale-loss parameter support - doc/CNN_V2.md: Document grayscale loss in training configuration and checkpoint format - doc/HOWTO.md: Add usage examples for --grayscale-loss flag Co-Authored-By: Claude Sonnet 4.5 --- scripts/gemini_start.bat | 8 -------- 1 file changed, 8 deletions(-) delete mode 100644 scripts/gemini_start.bat (limited to 'scripts/gemini_start.bat') diff --git a/scripts/gemini_start.bat b/scripts/gemini_start.bat deleted file mode 100644 index 22bca70..0000000 --- a/scripts/gemini_start.bat +++ /dev/null @@ -1,8 +0,0 @@ -@echo off -REM Start a Gemini session with core context - -gemini --files ^ - PROJECT_CONTEXT.md ^ - BUILD.md ^ - PHASE2_COMPRESSION.md ^ - "Read the project context carefully. Summarize the project goals and current phase. Wait." -- cgit v1.2.3