>>105813980
chroma "tests" being posted on the lodestone hf
RL = "reinforced learning"
v41 when he started it, v42 current base model and 2x RL