maxious

/

embeddinggemma-300m-onnx

like 1

Hi! I tried to reproduce your method, but optimum-cli export onnx throws ImportError: cannot import name 'ExportCommand' from 'optimum.commands.export' (unknown location). Please note that I can only use the CPU or XPU pytorch distro because I do not have an NVIDIA GPU on my machine. Here are the pip install commands I used from a fresh venv:

pip install -U pip
pip install torch --index-url https://download.pytorch.org/whl/cpu
pip install git+https://github.com/simondanielsson/optimum-onnx.git@feature/add-gemma3-export
pip install git+https://github.com/huggingface/transformers@v4.56.0-Embedding-Gemma-preview
pip install onnx onnxscript sentence-transformers onnxruntime

Could you please provide an updated procedure I can reproduce? Thanks!