onnxruntime

Fix large model check in Intel's Neural Compressor

#27746

Merged