onnx-diagnostic 0.1.0

Investigate ONNX models

Getting started

git clone https://github.com/sdpython/onnx-diagnostic.git
cd onnx-diagnostic
pip install -e .

or

pip install onnx-diagnostic

Snapshot of usefuls tools

string_type

import torch
from onnx_diagnostic.helpers import string_type

inputs = (
    torch.rand((3, 4), dtype=torch.float16),
    [
        torch.rand((5, 6), dtype=torch.float16),
        torch.rand((5, 6, 7), dtype=torch.float16),
    ]
)

# with shapes
print(string_type(inputs, with_shape=True))

>>> (T10s3x4,#2[T10s5x6,T10s5x6x7])

onnx_dtype_name

import onnx
from onnx_diagnostic.helpers import onnx_dtype_name

itype = onnx.TensorProto.BFLOAT16
print(onnx_dtype_name(itype))
print(onnx_dtype_name(7))

>>> BFLOAT16
>>> INT64

max_diff

Returns the maximum discrancies accross nested containers containing tensors.

Documentation

See onnx-diagnostic.