module `onnx_tools.model_checker`#

Short summary#

module mlprodict.onnx_tools.model_checker

Investigate issues happening with float32.

Classes#

class	truncated documentation
`MissingInputError`	Raised when an input is missing.

Functions#

function	truncated documentation
`astype_range`	Computes ranges for every number in an array once converted into float32. The function returns two matrices which …
`check_onnx`	Checks consistency of the model.
`enumerate_random_inputs`	Enumerates random matrices.
`onnx_shaker`	Shakes a model ONNX. Explores the ranges for every prediction. Uses `astype_range()`

Documentation#

Investigate issues happening with float32.

source on GitHub

exception mlprodict.onnx_tools.model_checker.MissingInputError#

Bases: RuntimeError

Raised when an input is missing.

mlprodict.onnx_tools.model_checker.astype_range(arr, dtype=<class 'numpy.float32'>, force=1)#

Computes ranges for every number in an array once converted into float32. The function returns two matrices which produces two numbers a et b, the number rounded to float32 is in interval $[a, b]$ .

Parameters:

arr – array
dtype – type to convert to
force – does something like [i] +/- force |i - [i]|

Returns:

minimum, maximum

source on GitHub

mlprodict.onnx_tools.model_checker.check_onnx(model, use_onnx=False, known_results=None, path=None)#

Checks consistency of the model.

Parameters:

model – onnx graph
use_onnx – calls onnx.checker.check_model
known_results – known results
path – path to a node (through subgraphs)

source on GitHub

mlprodict.onnx_tools.model_checker.enumerate_random_inputs(inputs, n=100, dtype=<class 'numpy.float32'>, force=1)#

Enumerates random matrices.

Parameters:

inputs – inputs (dictionary)
n – number of iterations
dtype – type to convert to
force – does something like [i] +/- force |i - [i]|

source on GitHub

mlprodict.onnx_tools.model_checker.onnx_shaker(oinf, inputs, output_fct, n=100, dtype=<class 'numpy.float32'>, force=1)#

Shakes a model ONNX. Explores the ranges for every prediction. Uses astype_range

Parameters:

oinf – object of type OnnxInference
inputs – inputs
output_fct – output function which extracts a single array from the output
dtype – type to convert to
force – does something like [i] +/- force |i - [i]|

Returns:

ranges for each predictions

See notebook Precision loss due to float32 conversion with ONNX for an example of use.

source on GitHub

mlprodict.onnx_tools.model_checker.randint(low, high=None, size=None, dtype=int)#

Return random integers from low (inclusive) to high (exclusive).

Return random integers from the “discrete uniform” distribution of the specified dtype in the “half-open” interval [low, high). If high is None (the default), then results are from [0, low).

Note

New code should use the integers method of a default_rng() instance instead; please see the Quick Start.

Parameters#

lowint or array-like of ints: Lowest (signed) integers to be drawn from the distribution (unless high=None, in which case this parameter is one above the highest such integer).
highint or array-like of ints, optional: If provided, one above the largest (signed) integer to be drawn from the distribution (see above for behavior if high=None). If array-like, must contain integer values
sizeint or tuple of ints, optional: Output shape. If the given shape is, e.g., (m, n, k), then m * n * k samples are drawn. Default is None, in which case a single value is returned.
dtypedtype, optional: Desired dtype of the result. Byteorder must be native. The default value is int.

New in version 1.11.0.

Returns#

outint or ndarray of ints: size-shaped array of random integers from the appropriate distribution, or a single such random int if size not provided.

Examples#

>>> np.random.randint(2, size=10)
array([1, 0, 0, 0, 1, 1, 0, 0, 1, 0]) # random
>>> np.random.randint(1, size=10)
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0])

Generate a 2 x 4 array of ints between 0 and 4, inclusive:

>>> np.random.randint(5, size=(2, 4))
array([[4, 0, 2, 1], # random
       [3, 2, 2, 0]])

Generate a 1 x 3 array with 3 different upper bounds

>>> np.random.randint(1, [3, 5, 10])
array([2, 2, 9]) # random

Generate a 1 by 3 array with 3 different lower bounds

>>> np.random.randint([1, 5, 7], 10)
array([9, 8, 7]) # random

Generate a 2 by 4 array using broadcasting with dtype of uint8

>>> np.random.randint([1, 3, 5, 7], [[10], [20]], dtype=np.uint8)
array([[ 8,  6,  9,  7], # random
       [ 1, 16,  9, 12]], dtype=uint8)

module onnx_tools.model_checker#

Short summary#

Classes#

Functions#

Documentation#

Parameters#

Returns#

See Also#

Examples#

module `onnx_tools.model_checker`#