6 Commits

Author	SHA1	Message	Date
John Smith	3471be4e56	add fast_4bit_matmul and auto switch 2 methods according to bottleneck	2023-03-21 08:43:07 +00:00
John Smith	dd0d5a31f7	add half support	2023-03-20 09:37:51 +00:00
John Smith	04f5575a23	reduced memory usage by a little	2023-03-20 00:51:52 +08:00
John Smith	2b84b32fbe	Update autograd_4bit.py	2023-03-18 22:13:11 +08:00
John Smith	6f4bbb40a9	Update autograd_4bit.py	2023-03-18 18:49:26 +08:00
John Smith	551f62a0e8	add patch for gptq and peft	2023-03-18 13:31:48 +08:00