Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Appearance settings

Commit 6689d71

Browse filesBrowse files
abetlenabetlen
committed
Merge branch 'main' into configurable-chat-templates
2 parents 45a4188 + 6e167a2 commit 6689d71
Copy full SHA for 6689d71

File tree

Expand file treeCollapse file tree

6 files changed

+27
-7
lines changed
Filter options
Expand file treeCollapse file tree

6 files changed

+27
-7
lines changed

‎CHANGELOG.md

Copy file name to clipboardExpand all lines: CHANGELOG.md
+4Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -7,6 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
77

88
## [Unreleased]
99

10+
- Update llama.cpp to 8781013ef654270cbead3e0011e33a6d690fb168
11+
12+
## [0.2.6]
13+
1014
- Update llama.cpp to 80291a1d02a07f7f66666fb576c5b1e75aa48b46
1115

1216
## [0.2.5]

‎CMakeLists.txt

Copy file name to clipboardExpand all lines: CMakeLists.txt
+10-1Lines changed: 10 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
cmake_minimum_required(VERSION 3.4...3.22)
1+
cmake_minimum_required(VERSION 3.21)
22

33
project(llama_cpp)
44

@@ -33,4 +33,13 @@ if (LLAMA_BUILD)
3333
FRAMEWORK DESTINATION ${CMAKE_CURRENT_SOURCE_DIR}/llama_cpp
3434
RESOURCE DESTINATION ${CMAKE_CURRENT_SOURCE_DIR}/llama_cpp
3535
)
36+
# Workaround for Windows + CUDA https://github.com/abetlen/llama-cpp-python/issues/563
37+
install(
38+
FILES $<TARGET_RUNTIME_DLLS:llama>
39+
DESTINATION ${SKBUILD_PLATLIB_DIR}/llama_cpp
40+
)
41+
install(
42+
FILES $<TARGET_RUNTIME_DLLS:llama>
43+
DESTINATION ${CMAKE_CURRENT_SOURCE_DIR}/llama_cpp
44+
)
3645
endif()

‎llama_cpp/llama.py

Copy file name to clipboardExpand all lines: llama_cpp/llama.py
+2Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -437,6 +437,7 @@ def tokenize(self, text: bytes, add_bos: bool = True) -> List[int]:
437437
n_tokens = llama_cpp.llama_tokenize_with_model(
438438
self.model,
439439
text,
440+
len(text),
440441
tokens,
441442
n_ctx,
442443
add_bos,
@@ -447,6 +448,7 @@ def tokenize(self, text: bytes, add_bos: bool = True) -> List[int]:
447448
n_tokens = llama_cpp.llama_tokenize_with_model(
448449
self.model,
449450
text,
451+
len(text),
450452
tokens,
451453
n_tokens,
452454
add_bos,

‎llama_cpp/llama_cpp.py

Copy file name to clipboardExpand all lines: llama_cpp/llama_cpp.py
+9-4Lines changed: 9 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -58,7 +58,7 @@ def _load_shared_library(lib_base_name: str):
5858
if "CUDA_PATH" in os.environ:
5959
os.add_dll_directory(os.path.join(os.environ["CUDA_PATH"], "bin"))
6060
os.add_dll_directory(os.path.join(os.environ["CUDA_PATH"], "lib"))
61-
cdll_args["winmode"] = 0
61+
cdll_args["winmode"] = ctypes.RTLD_GLOBAL
6262

6363
# Try to load the shared library, handling potential errors
6464
for _lib_path in _lib_paths:
@@ -950,42 +950,47 @@ def llama_token_nl(ctx: llama_context_p) -> llama_token:
950950
# LLAMA_API int llama_tokenize(
951951
# struct llama_context * ctx,
952952
# const char * text,
953+
# int text_len,
953954
# llama_token * tokens,
954955
# int n_max_tokens,
955956
# bool add_bos);
956957
def llama_tokenize(
957958
ctx: llama_context_p,
958959
text: bytes,
960+
text_len: Union[c_int, int],
959961
tokens, # type: Array[llama_token]
960962
n_max_tokens: Union[c_int, int],
961963
add_bos: Union[c_bool, int],
962964
) -> int:
963-
return _lib.llama_tokenize(ctx, text, tokens, n_max_tokens, add_bos)
965+
return _lib.llama_tokenize(ctx, text, text_len, tokens, n_max_tokens, add_bos)
964966

965967

966-
_lib.llama_tokenize.argtypes = [llama_context_p, c_char_p, llama_token_p, c_int, c_bool]
968+
_lib.llama_tokenize.argtypes = [llama_context_p, c_char_p, c_int, llama_token_p, c_int, c_bool]
967969
_lib.llama_tokenize.restype = c_int
968970

969971

970972
# LLAMA_API int llama_tokenize_with_model(
971973
# const struct llama_model * model,
972974
# const char * text,
975+
# int text_len,
973976
# llama_token * tokens,
974977
# int n_max_tokens,
975978
# bool add_bos);
976979
def llama_tokenize_with_model(
977980
model: llama_model_p,
978981
text: bytes,
982+
text_len: Union[c_int, int],
979983
tokens, # type: Array[llama_token]
980984
n_max_tokens: Union[c_int, int],
981985
add_bos: Union[c_bool, bool],
982986
) -> int:
983-
return _lib.llama_tokenize_with_model(model, text, tokens, n_max_tokens, add_bos)
987+
return _lib.llama_tokenize_with_model(model, text, text_len, tokens, n_max_tokens, add_bos)
984988

985989

986990
_lib.llama_tokenize_with_model.argtypes = [
987991
llama_model_p,
988992
c_char_p,
993+
c_int,
989994
llama_token_p,
990995
c_int,
991996
c_bool,

‎pyproject.toml

Copy file name to clipboardExpand all lines: pyproject.toml
+1-1Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -54,7 +54,7 @@ all = [
5454
[tool.scikit-build]
5555
wheel.packages = ["llama_cpp"]
5656
cmake.verbose = true
57-
cmake.minimum-version = "3.12"
57+
cmake.minimum-version = "3.21"
5858
minimum-version = "0.5"
5959
sdist.exclude = [".git", "vendor/llama.cpp/.git"]
6060

‎vendor/llama.cpp

Copy file name to clipboard

0 commit comments

Comments
0 (0)
Morty Proxy This is a proxified and sanitized view of the page, visit original site.