7 Commits

Include renames

Author	SHA1	Message	Date
silveroxides	ee3d998a5d	Support w6a8_int8 quantization decoding and shape unpacking	2026-11-10 17:29:49 +01:00
silveroxides	9ebe34565c	Remove arbitrary low-bit tensor count cap and honor discard before preservation	2026-10-31 23:16:02 +01:00
silveroxides	1861d75aab	Match diffusion model layers across prefixes	2026-09-23 02:36:18 +02:00
silveroxides	57bd63e1eb	Support quantized diffusion model inputs	2026-09-23 01:53:25 +02:00
silveroxides	f93ca54c81	fix(io): stream UEL and normalize LoRA alpha	2026-08-14 08:43:53 +02:00
silveroxides	f390c12776	fix(merge): enforce LoRA tensor contracts ... - preserve direct diffs, precision, and guarded low-bit tensors - stream paired CWB layers with atomic CPU OOM fallback - save generated merge artifacts under models_dir	2026-08-02 18:11:19 +02:00
silveroxides	49b66ea64e	fix(lora): enforce adapter tensor contracts	2026-08-01 12:15:39 +02:00