File tree 2 files changed +3
-3
lines changed
2 files changed +3
-3
lines changed Original file line number Diff line number Diff line change @@ -84,8 +84,8 @@ Optional arguments:
84
84
The group size to use for quantization. Recommended value is 128 and -1 uses per-column
85
85
quantization.
86
86
--backup-precision { none ,int8_sym ,int8_asym }
87
- Defines a backup precision for mixed-precision weight compression. Only valid for int4 weight
88
- format . If not provided, backup precision is int8_asym. ' none' stands for original floating-
87
+ Defines a backup precision for mixed-precision weight compression. Only valid for 4-bit weight
88
+ formats . If not provided, backup precision is int8_asym. ' none' stands for original floating-
89
89
point precision of the model weights, in this case weights are retained in their original
90
90
precision without any quantization. ' int8_sym' stands for 8-bit integer symmetric quantization
91
91
without zero point. ' int8_asym' stands for 8-bit integer asymmetric quantization with zero
Original file line number Diff line number Diff line change @@ -123,7 +123,7 @@ def parse_args_openvino(parser: "ArgumentParser"):
123
123
choices = ["none" , "int8_sym" , "int8_asym" ],
124
124
default = None ,
125
125
help = (
126
- "Defines a backup precision for mixed-precision weight compression. Only valid for int4 weight format . "
126
+ "Defines a backup precision for mixed-precision weight compression. Only valid for 4-bit weight formats . "
127
127
"If not provided, backup precision is int8_asym. 'none' stands for original floating-point precision of "
128
128
"the model weights, in this case weights are retained in their original precision without any "
129
129
"quantization. 'int8_sym' stands for 8-bit integer symmetric quantization without zero point. 'int8_asym' "
You can’t perform that action at this time.
0 commit comments