ExternalVendorCode/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-06-17 14:28:07 +00:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

whisper : reorganize source code + improve CMake (#2256)

Browse Source 
* scripts : update sync [no ci]

* files : reorganize [no ci]

* sync : llama.cpp

* cmake : link math library

* cmake : build normal ggml library

* files : move headers to include

* objc : fix path to ggml-metal.h

* ci : fix WHISPER_CUDA -> GGML_CUDA

* scripts : sync LICENSE [no ci]
This commit is contained in:
Georgi Gerganov
2024-06-26 19:34:09 +03:00
committed by GitHub
parent bf4cb4abad
commit e30c679928
361 changed files with 176536 additions and 20466 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_F16, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(256, GGML_TYPE_F16, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(64, GGML_TYPE_F16, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f16.cuh"
DECL_FATTN_VEC_F16_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_F16, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_F16, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_0, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q4_1, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_0, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q5_1, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(128, GGML_TYPE_Q8_0, GGML_TYPE_Q8_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(256, GGML_TYPE_F16, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(64, GGML_TYPE_F16, GGML_TYPE_F16);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f32.cuh"
DECL_FATTN_VEC_F32_CASE(64, GGML_TYPE_F16, GGML_TYPE_Q8_0);

10
ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb16.cu Normal file
View File

@ -0,0 +1,10 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-wmma-f16.cuh"
DECL_FATTN_WMMA_F16_CASE(64, 16, float);
DECL_FATTN_WMMA_F16_CASE(80, 16, float);
DECL_FATTN_WMMA_F16_CASE(96, 16, float);
DECL_FATTN_WMMA_F16_CASE(112, 16, float);
DECL_FATTN_WMMA_F16_CASE(128, 16, float);
DECL_FATTN_WMMA_F16_CASE(256, 16, float);

9
ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb32.cu Normal file
View File

@ -0,0 +1,9 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-wmma-f16.cuh"
DECL_FATTN_WMMA_F16_CASE(64, 32, float);
DECL_FATTN_WMMA_F16_CASE(80, 32, float);
DECL_FATTN_WMMA_F16_CASE(96, 32, float);
DECL_FATTN_WMMA_F16_CASE(112, 32, float);
DECL_FATTN_WMMA_F16_CASE(128, 32, float);

10
ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb16.cu Normal file
View File

@ -0,0 +1,10 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-wmma-f16.cuh"
DECL_FATTN_WMMA_F16_CASE(64, 16, half);
DECL_FATTN_WMMA_F16_CASE(80, 16, half);
DECL_FATTN_WMMA_F16_CASE(96, 16, half);
DECL_FATTN_WMMA_F16_CASE(112, 16, half);
DECL_FATTN_WMMA_F16_CASE(128, 16, half);
DECL_FATTN_WMMA_F16_CASE(256, 16, half);

10
ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb32.cu Normal file
View File

@ -0,0 +1,10 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-wmma-f16.cuh"
DECL_FATTN_WMMA_F16_CASE(64, 32, half);
DECL_FATTN_WMMA_F16_CASE(80, 32, half);
DECL_FATTN_WMMA_F16_CASE(96, 32, half);
DECL_FATTN_WMMA_F16_CASE(112, 32, half);
DECL_FATTN_WMMA_F16_CASE(128, 32, half);
DECL_FATTN_WMMA_F16_CASE(256, 32, half);

8
ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb8.cu Normal file
View File

@ -0,0 +1,8 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-wmma-f16.cuh"
DECL_FATTN_WMMA_F16_CASE(64, 8, half);
DECL_FATTN_WMMA_F16_CASE(96, 8, half);
DECL_FATTN_WMMA_F16_CASE(128, 8, half);
DECL_FATTN_WMMA_F16_CASE(256, 8, half);

75
ggml/src/ggml-cuda/template-instances/generate_cu_files.py Executable file
View File

@ -0,0 +1,75 @@
#!/usr/bin/env python3
from glob import glob
import os
TYPES_KV = ["GGML_TYPE_Q4_0", "GGML_TYPE_Q4_1", "GGML_TYPE_Q5_0", "GGML_TYPE_Q5_1", "GGML_TYPE_Q8_0", "GGML_TYPE_F16"]
SOURCE_FATTN_VEC = """// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-vec-f{vkq_size}.cuh"
DECL_FATTN_VEC_F{vkq_size}_CASE({head_size}, {type_k}, {type_v});
"""
SOURCE_FATTN_WMMA_START = """// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../fattn-wmma-f16.cuh"
"""
SOURCE_FATTN_WMMA_CASE = "DECL_FATTN_WMMA_F16_CASE({head_size}, {cols_per_block}, {kq_acc_t});\n"
TYPES_MMQ = [
"GGML_TYPE_Q4_0", "GGML_TYPE_Q4_1", "GGML_TYPE_Q5_0", "GGML_TYPE_Q5_1", "GGML_TYPE_Q8_0",
"GGML_TYPE_Q2_K", "GGML_TYPE_Q3_K", "GGML_TYPE_Q4_K", "GGML_TYPE_Q5_K", "GGML_TYPE_Q6_K"
]
SOURCE_MMQ = """// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE({type});
"""
def get_short_name(long_quant_name):
return long_quant_name.replace("GGML_TYPE_", "").lower()
def get_head_sizes(type_k, type_v):
if type_k == "GGML_TYPE_F16" and type_v == "GGML_TYPE_F16":
return [64, 128, 256]
if type_k == "GGML_TYPE_F16":
return [64, 128]
return [128]
for filename in glob("*.cu"):
os.remove(filename)
for vkq_size in [16, 32]:
for type_k in TYPES_KV:
for type_v in TYPES_KV:
for head_size in get_head_sizes(type_k, type_v):
with open(f"fattn-vec-f{vkq_size}-instance-hs{head_size}-{get_short_name(type_k)}-{get_short_name(type_v)}.cu", "w") as f:
f.write(SOURCE_FATTN_VEC.format(vkq_size=vkq_size, head_size=head_size, type_k=type_k, type_v=type_v))
for kq_acc_t in ["half", "float"]:
for cols_per_block in [8, 16, 32]:
if kq_acc_t == "float" and cols_per_block == 8:
continue
with open(f"fattn-wmma-f16-instance-kq{kq_acc_t}-cpb{cols_per_block}.cu", "w") as f:
f.write(SOURCE_FATTN_WMMA_START)
for head_size in [64, 80, 96, 112, 128, 256]:
if cols_per_block == 8 and head_size % 32 != 0: # wmma fragment is 8x32
continue
if kq_acc_t == "float" and cols_per_block == 32 and head_size == 256: # register spilling, bad performance
continue
f.write(SOURCE_FATTN_WMMA_CASE.format(kq_acc_t=kq_acc_t, cols_per_block=cols_per_block, head_size=head_size))
for type in TYPES_MMQ:
with open(f"mmq-instance-{get_short_name(type)}.cu", "w") as f:
f.write(SOURCE_MMQ.format(type=type))

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q2_K);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q3_K);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q4_0);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q4_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q4_1);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q4_k.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q4_K);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q5_0.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q5_0);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q5_1.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q5_1);

5
ggml/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu Normal file
View File

@ -0,0 +1,5 @@
// This file has been autogenerated by generate_cu_files.py, do not edit manually.
#include "../mmq.cuh"
DECL_MMQ_CASE(GGML_TYPE_Q5_K);

Some files were not shown because too many files have changed in this diff Show More