2 Commits

Include renames

Author	SHA1	Message	Date
ThereforeGames	dc1c2cc498	Add MiniMax H3 support and ConvRot validation ... Introduce MODEL_TYPE_MINIMAX_H3, keep-float/int4-sensitive presets, and automatic adapter detection; bump package version to 2.1.0 and update changelog. Add robust ConvRot group-size validation (_validate_convrot_group_size), stricter Hadamard shape checks, and improved error handling when loading/saving native INT8 ConvRot metadata. Add unit tests covering ConvRot metadata round-trips, invalid metadata rejection, save-time consistency checks, and MiniMax H3 detection.	2026-08-07 06:51:09 -04:00
ThereforeGames	722dadcbc5	Rebrand, add INT4 support & quantization modes ... Rebrand the project as "ComfyUI Quantization Toolkit" and introduce native INT4 (ConvRot W4A4) support, mixed INT4/INT8 modes, and expanded quantization policies. Add an INT4 torch.compile compatibility shim, INT4-aware lazy-compile changes, richer LoRA stack nodes with autogrow, and a Quantized LoRA entry/apply pair. Replace ad-hoc prints with structured logging, bump package metadata to 2.0.0, and add documentation (advanced usage, checkpoints, policies) plus many unit tests for compile, lazy-compile, LoRA, adapter caching, saving, and modes. Misc: kernel config, model save/load, and numerous runtime/adapter refactors and robustness fixes.	2026-08-04 07:11:56 -04:00