requirements.txt

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
nvidia-ml-py>=12
# Just a wrapper since nvidia-modelopt requires pynvml
pynvml>=12.0.0
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]~=0.27.0
nvidia-nccl-cu12
nvidia-cuda-nvrtc-cu12
transformers==4.51.0
pydantic>=2.9.1
pillow==10.3.0
wheel<=0.45.1
optimum
# evaluate needs datasets>=2.0.0 which triggers datasets>3.1.0 which is not stable: https://github.com/huggingface/datasets/issues/7467
datasets==3.1.0
evaluate
mpmath>=1.3.0
click
click_option_group
aenum
pyzmq
fastapi==0.115.4
uvicorn
httpx
setuptools
ordered-set
peft
einops
flashinfer-python~=0.2.3
opencv-python-headless
xgrammar==0.1.16
backoff
nvtx
matplotlib # FIXME: this is added to make nvtx happy