[Docker] Add fastsafetensors to NVIDIA Dockerfile (#38950)
This commit is contained in:
@@ -15,6 +15,9 @@ flashinfer-cubin==0.6.7
|
||||
# breaking changes in 1.19.0
|
||||
nvidia-cudnn-frontend>=1.13.0,<1.19.0
|
||||
|
||||
# Required for faster safetensors model loading
|
||||
fastsafetensors >= 0.2.2
|
||||
|
||||
# QuACK and Cutlass DSL for FA4 (cute-DSL implementation)
|
||||
nvidia-cutlass-dsl>=4.4.2
|
||||
quack-kernels>=0.3.3
|
||||
|
||||
@@ -276,7 +276,9 @@ fastar==0.9.0
|
||||
fastparquet==2026.3.0
|
||||
# via genai-perf
|
||||
fastsafetensors==0.2.2
|
||||
# via -r requirements/rocm-test.in
|
||||
# via
|
||||
# -c requirements/rocm.txt
|
||||
# -r requirements/rocm-test.in
|
||||
filelock==3.25.2
|
||||
# via
|
||||
# -c requirements/common.txt
|
||||
|
||||
@@ -20,4 +20,6 @@ conch-triton-kernels==1.2.1
|
||||
timm>=1.0.17
|
||||
# amd-quark: required for Quark quantization on ROCm
|
||||
# To be consistent with test_quark.py
|
||||
amd-quark>=0.8.99
|
||||
amd-quark>=0.8.99
|
||||
# Required for faster safetensors model loading
|
||||
fastsafetensors >= 0.2.2
|
||||
Reference in New Issue
Block a user