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

Fix Vulkan F16 models

Browse source
This commit is contained in:
0cc4m 2024-01-30 18:22:18 +01:00
parent 8f8ddfcfad
commit eb76719eea
2 changed files with 692 additions and 1264 deletions
Download patch file Download diff file Expand all files Collapse all files

1952
ggml-vulkan-shaders.hpp
View file

File diff suppressed because it is too large Load diff

4
ggml_vk_generate_shaders.py
View file

@ -19,8 +19,8 @@ shader_int8_ext = """
# Type-specific defines # Type-specific defines
shader_f16_defines = """ shader_f16_defines = """
#define QUANT_K 32 #define QUANT_K 1
#define QUANT_R 2 #define QUANT_R 1
#define A_TYPE float16_t #define A_TYPE float16_t
""" """