mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-13 22:18:36 +08:00
45 lines
750 B
Plaintext
45 lines
750 B
Plaintext
accelerate>=0.25.0
|
|
build
|
|
colored
|
|
cuda-python # Do not override the custom version of cuda-python installed in the NGC PyTorch image.
|
|
diffusers>=0.27.0
|
|
lark
|
|
mpi4py
|
|
numpy<2
|
|
onnx>=1.12.0
|
|
onnx_graphsurgeon>=0.5.2
|
|
openai
|
|
polygraphy
|
|
psutil
|
|
pynvml>=11.5.0
|
|
pulp
|
|
pandas
|
|
h5py==3.12.1
|
|
StrEnum
|
|
sentencepiece>=0.1.99
|
|
tensorrt~=10.9.0
|
|
# https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-25-03.html#rel-25-03 uses 2.7.0a0.
|
|
torch>=2.6.0,<=2.7.0a0
|
|
torchvision
|
|
nvidia-modelopt[torch]~=0.25.0
|
|
nvidia-nccl-cu12
|
|
nvidia-cuda-nvrtc-cu12
|
|
transformers==4.48.3
|
|
pydantic>=2.9.1
|
|
pillow==10.3.0
|
|
wheel
|
|
optimum
|
|
evaluate
|
|
mpmath>=1.3.0
|
|
click
|
|
click_option_group
|
|
aenum
|
|
pyzmq
|
|
fastapi==0.115.4
|
|
uvicorn
|
|
httpx
|
|
setuptools
|
|
ordered-set
|
|
einops
|
|
flashinfer-python~=0.2.0
|