Merge pull request #18 from DarthAffe/Quantizations

Added new quanization-types
This commit is contained in:
DarthAffe 2024-07-28 14:14:48 +02:00 committed by GitHub
commit f21a3d876c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -8,7 +8,7 @@ public enum Quantization
Q4_0 = 2,
Q4_1 = 3,
// Q4_2 = 4, support has been removed
// Q4_3 (5) support has been removed
// Q4_3 = 5, support has been removed
Q5_0 = 6,
Q5_1 = 7,
Q8_0 = 8,
@ -18,7 +18,7 @@ public enum Quantization
Q4_K = 12,
Q5_K = 13,
Q6_K = 14,
Q8_K = 15,
Q8_K = 15,
IQ2_XXS = 16,
IQ2_XS = 17,
IQ3_XXS = 18,
@ -27,7 +27,14 @@ public enum Quantization
IQ3_S = 21,
IQ2_S = 22,
IQ4_XS = 23,
I8,
I16,
I32,
I8 = 24,
I16 = 25,
I32 = 26,
I64 = 27,
F64 = 28,
IQ1_M = 29,
BF16 = 30,
Q4_0_4_4 = 31,
Q4_0_4_8 = 32,
Q4_0_8_8 = 33,
}