A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 8, 12:30 UTC
[Misc] Bump `fastsafetensors` version for latest fixes (#34273)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Nick Hill committed
Feb 11, 2026 at 00:30 UTC
79504027ef93a742846856e81fc25de369dc5e22
4 files changed
+4
-5
requirements/nightly_torch_test.txt
+1
-1
@@ -43,5 +43,5 @@ tritonclient>=2.51.0
43
numba == 0.61.2 # Required for N-gram speculative decoding
44
numpy
45
runai-model-streamer[s3,gcs]==0.15.3
46
-fastsafetensors>=0.1.10
46
+fastsafetensors>=0.2.2
47
pydantic>=2.12 # 2.11 leads to error on python 3.13
requirements/test.in
+1
-1
@@ -53,7 +53,7 @@ arctic-inference == 0.1.1 # Required for suffix decoding test
53
numba == 0.61.2 # Required for N-gram speculative decoding
54
numpy
55
runai-model-streamer[s3,gcs]==0.15.3
56
-fastsafetensors>=0.1.10
56
+fastsafetensors>=0.2.2 # 0.2.2 contains important fixes for multi-GPU mem usage
57
pydantic>=2.12 # 2.11 leads to error on python 3.13
58
decord==0.6.0
59
terratorch @ git+https://github.com/IBM/terratorch.git@1.1.rc3 # required for PrithviMAE test
requirements/test.txt
+1
-2
@@ -224,7 +224,7 @@ fastparquet==2024.11.0
224
# via genai-perf
225
fastrlock==0.8.2
226
# via cupy-cuda12x
227
-fastsafetensors==0.1.10
227
+fastsafetensors==0.2.2
228
# via -r requirements/test.in
229
filelock==3.16.1
230
# via
@@ -1174,7 +1174,6 @@ torch==2.10.0+cu129
1174
# bitsandbytes
1175
# efficientnet-pytorch
1176
# encodec
1177
- # fastsafetensors
1177
# kornia
1178
# lightly
1179
# lightning
setup.py
+1
-1
@@ -1035,7 +1035,7 @@ setup(
1035
extras_require={
1036
"bench": ["pandas", "matplotlib", "seaborn", "datasets", "scipy"],
1037
"tensorizer": ["tensorizer==2.10.1"],
1038
- "fastsafetensors": ["fastsafetensors >= 0.1.10"],
1038
+ "fastsafetensors": ["fastsafetensors >= 0.2.2"],
1039
"runai": ["runai-model-streamer[s3,gcs] >= 0.15.3"],
1040
"audio": [
1041
"librosa",