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

CUDA: refactor mmq, dmmv, mmvq (#7716)

Browse source 
* CUDA: refactor mmq, dmmv, mmvq

* fix out-of-bounds write

* struct for qk, qr, qi

* fix cmake build

* mmq_type_traits
This commit is contained in:
Johannes Gäßler 2024-06-05 16:53:00 +02:00 • committed by GitHub
parent 2b3389677a
commit 7d1a378b8f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
112 changed files with 1783 additions and 1767 deletions
Download patch file Download diff file Expand all files Collapse all files

16
ggml-cuda/template-instances/generate_cu_files.py
View file

@ -20,6 +20,18 @@ SOURCE_FATTN_WMMA_START = """// This file has been autogenerated by generate_cu_
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()
@ -57,3 +69,7 @@ for kq_acc_t in ["half", "float"]:
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))