:orphan: :py:mod:`fastchat.serve.compression` ==================================== .. py:module:: fastchat.serve.compression Module Contents --------------- Classes ~~~~~~~ .. autoapisummary:: fastchat.serve.compression.CompressionConfig fastchat.serve.compression.CLinear Functions ~~~~~~~~~ .. autoapisummary:: fastchat.serve.compression.compress fastchat.serve.compression.decompress .. py:class:: CompressionConfig Group-wise quantization. .. py:class:: CLinear(weight=None, bias=None, device=None) Compressed Linear Layer. .. py:function:: compress(tensor, config) Simulate group-wise quantization. .. py:function:: decompress(packed_data, config) Simulate group-wise dequantization.