I think the docker method fails due to the cpuinfo not find dot vector instructions on older pi cpus's
Then is you try the static or dynamic from watching the docker install it uses a specific commit that not sure what version or commit.
If you just install onnx into .venv
File "/home/stuart/gtcrn/ladspa/.venv/lib/python3.11/site-packages/onnxruntime/tools/ort_format_model/ort_flatbuffers_py/fbs/InferenceSession.py", line 25, in InferenceSessionBufferHasIdentifier
return flatbuffers.util.BufferHasIdentifier(buf, offset, b"\x4F\x52\x54\x4D", size_prefixed=size_prefixed)
^^^^^^^^^^^^^^^^
AttributeError: module 'flatbuffers' has no attribute 'util'
thread 'main' (9397) panicked at build.rs:139:17:
Model conversion failed for gtcrn_simple.onnx
note: run with RUST_BACKTRACE=1 environment variable to display a backtrace
I think the docker method fails due to the cpuinfo not find dot vector instructions on older pi cpus's
Then is you try the static or dynamic from watching the docker install it uses a specific commit that not sure what version or commit.
If you just install onnx into .venv
File "/home/stuart/gtcrn/ladspa/.venv/lib/python3.11/site-packages/onnxruntime/tools/ort_format_model/ort_flatbuffers_py/fbs/InferenceSession.py", line 25, in InferenceSessionBufferHasIdentifier
return flatbuffers.util.BufferHasIdentifier(buf, offset, b"\x4F\x52\x54\x4D", size_prefixed=size_prefixed)
^^^^^^^^^^^^^^^^
AttributeError: module 'flatbuffers' has no attribute 'util'
thread 'main' (9397) panicked at build.rs:139:17:
Model conversion failed for gtcrn_simple.onnx
note: run with
RUST_BACKTRACE=1environment variable to display a backtrace