model_qint8_arm64.onnx is not working for aarch64 architecture. #23533
Unanswered
narendra9079
asked this question in
Other Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I am trying to run the quantized version of https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2/blob/main/onnx/model_qint8_arm64.onnx. on aarch64 architecture. But is not able to generate embedding. It is failing when I run sessin.run() function. But the avx2.onnx version is running fine on linux. I am using ort = { version = "=2.0.0-rc.0", default-features = false, features=["ndarray"], optional = true }
ort-sys = { version = "=2.0.0-rc.0", optional = true } these version for onnx run time.
Can you please provide some opinion ??
Beta Was this translation helpful? Give feedback.
All reactions