happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

gguf-py : use copy-on-write mode for localtensor (#18162)

This commit is contained in:
Sigbjørn Skjæret 2025-12-18 13:45:38 +01:00 committed by GitHub
parent 54189c0d39
commit ec7b9329ae
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
gguf-py/gguf/utility.py
View File

@ -288,7 +288,7 @@ class LocalTensor:
data_range: LocalTensorRange data_range: LocalTensorRange
def mmap_bytes(self) -> np.ndarray: def mmap_bytes(self) -> np.ndarray:
return np.memmap(self.data_range.filename, mode='r', offset=self.data_range.offset, shape=self.data_range.size) return np.memmap(self.data_range.filename, mode='c', offset=self.data_range.offset, shape=self.data_range.size)
class SafetensorsLocal: class SafetensorsLocal: