v0.10.0

• BugFix for FP8, pedantic-level 2
• Support python 3.12, 3.13

v0.9.0

• Support FP8, INT8, and AWQ for Qwen2-VL
• More accurate quantization for FP8, INT8, and AWQ.
• Deprecate Python3.8 support.

v0.8.0

• Support ExaoneForCausalLM
• Fix CohereForCausalLM with better quantization logic.

Release v0.7.0🚀

• Further optimization for running FP8, and INT8 quantization.
• Support searching automatic calibration dataset batch size for running FMO.
• Support [AWQ(Activation-aware Weight Quantization)].

Release v0.5.1🚀

Update Patch v0.5.1

Release v0.5.0🚀

Update Patch v0.5.0

Release v0.3.0🚀

Full Changelog: v0.2.1...v0.3.0

Release v0.2.1🚀

Full Changelog: v0.2.0...v0.2.1

Release v0.2.0🚀

Full Changelog: v0.1.0...v0.2.0

Release v0.1.0🚀

Update Patch v0.1.0