3 Commits

Author	SHA1	Message	Date
Slightwind	3366d47694	[main][bugfix] Add 'layer_type' param to get_pergroup_param() for compatibility (#3682) ... Resolves a `TypeError: got an unexpected keyword argument 'layer_type'`. A recent change (PR #3311) started passing the `layer_type` argument when calling `get_pergroup_param()`. This specific implementation does not use this parameter, causing the error. This patch adds `layer_type=None` to the method signature to maintain API compatibility and ignore the unused argument. - vLLM version: v0.11.0rc3 - vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0 Signed-off-by: SlightwindSec <slightwindsec@gmail.com>	2025-10-23 21:26:33 +08:00
Slightwind	07ca1b9b78	[Refactor] Clean up w4a4_flatquant_dynamic implementation (#3440) ... Cleans up the initial implementation of `w4a4_flatquant_dynamic` for better readability and maintainability. - vLLM version: v0.11.0rc3 - vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0 --------- Signed-off-by: SlightwindSec <slightwindsec@gmail.com>	2025-10-17 23:53:19 +08:00
Slightwind	4f6d60eb06	[Feature] Add W4A4 Flat Quantization support (#3427) ... Introduce W4A4 Flat Quantization for better model compression and inference efficiency on Ascend devices. - vLLM version: v0.11.0rc3 - vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0 --------- Signed-off-by: SlightwindSec <slightwindsec@gmail.com>	2025-10-13 23:20:16 +08:00