vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

gguf-py : add support for I8, I16 and I32 (#6045)

Browse source 
* Refactor dtype handling to be extensible

This code is equivalent as before, but now it is prepared to easily add
more NumPy dtypes.

* Add support for I8, I16 and I32

These types are allowed in the GGUF specification.

* Add support for I8, I16 and I32 to gguf_writer

* Add support for I8, I16, I32 to gguf_reader
This commit is contained in:
Ondřej Čertík 2024-03-14 04:40:14 -06:00 • committed by GitHub
parent 3fe8d7a17f
commit 3ca23481dd
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 27 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

6
gguf-py/gguf/constants.py
View file

@ -661,6 +661,9 @@ class GGMLQuantizationType(IntEnum):
IQ3_S = 21
IQ2_S = 22
IQ4_XS = 23
I8 = 24
I16 = 25
I32 = 26
class GGUFEndian(IntEnum):
@ -727,6 +730,9 @@ GGML_QUANT_SIZES = {
GGMLQuantizationType.IQ3_S: (256, 2 + QK_K // 4 + QK_K // 8 + QK_K // 32 + 4),
GGMLQuantizationType.IQ2_S: (256, 2 + QK_K // 4 + QK_K // 16),
GGMLQuantizationType.IQ4_XS: (256, 2 + 2 + QK_K // 2 + QK_K // 64),
GGMLQuantizationType.I8: (1, 1),
GGMLQuantizationType.I16: (1, 2),
GGMLQuantizationType.I32: (1, 4),
}