Skip to content
Permalink
master
Switch branches/tags

Commits on Apr 25, 2022

  1. fix in ONNX conversion of slice operation

    Viktor Gyenes committed Apr 25, 2022

Commits on Apr 22, 2022

  1. add option to dump all tensors when executing a model

    Viktor Gyenes committed Apr 22, 2022

Commits on Apr 13, 2022

Commits on Apr 7, 2022

  1. fixing compiler warnings for exception handling

    Viktor Gyenes committed Apr 7, 2022

Commits on Mar 28, 2022

  1. Merge branch 'master' into channelwise_quantization

    Viktor Gyenes committed Mar 28, 2022

Commits on Mar 24, 2022

  1. Merge branch 'master' into channelwise_quantization

    Viktor Gyenes committed Mar 24, 2022
  2. Merge branch 'master' into array_length_eval

    Viktor Gyenes committed Mar 24, 2022
  3. fix padding size in pytorch execution of multilinear upsample using d…

    …econv
    Viktor Gyenes committed Mar 24, 2022
  4. add conversion of Erf op to ONNX converter

    fix default value of axis in conversion of Gather op
    Viktor Gyenes committed Mar 24, 2022

Commits on Feb 24, 2022

Commits on Feb 16, 2022

  1. fix dtype of float numpy arrays in when converted from python float a…

    …rrays (python wrapper)
    Viktor Gyenes committed Feb 16, 2022

Commits on Feb 14, 2022

  1. propagate tensor constants of quantization algorithms to python side …

    …as numpy arrays, handle them in shape inference and quantized execution of interpreter
    Viktor Gyenes committed Feb 14, 2022
  2. allow nested array syntax for assigning multi-dimensional tensor cons…

    …tants to parameters of quantization algorithms
    
    update zero_point_linear_quantize operator to accept tensor parameters
    Viktor Gyenes committed Feb 14, 2022

Commits on Jan 28, 2022

  1. add option to dequantize quantized tflite networks during conversion

    check agains per channel quantization attributes before writing out NNEF model
    Viktor Gyenes committed Jan 28, 2022

Commits on Jan 27, 2022

Commits on Jan 24, 2022

  1. fix conversion of leaky relu op from Caffe

    bug fix in conversion of argmax op in Caffe
    Viktor Gyenes committed Jan 24, 2022

Commits on Jan 21, 2022

  1. bug fix for when using convert.py in optimization mode

    Viktor Gyenes committed Jan 21, 2022

Commits on Jan 17, 2022

  1. better way to set C++11 standard in cmake (not inherited by project t…

    …hat uses NNEF parser)
    Viktor Gyenes committed Jan 17, 2022

Commits on Jan 6, 2022

  1. fix unintentionally committed sample.cpp

    Viktor Gyenes committed Jan 6, 2022
  2. add conversion of static forward LSTM operation from ONNX to NNEF usi…

    …ng custom compound fragments
    Viktor Gyenes committed Jan 6, 2022
  3. fix some broken identity tflite tests that result in empty graph with…

    … the latest tf version
    Viktor Gyenes committed Jan 6, 2022
  4. add option to denote fragment dependencies by custom fragments define…

    …d by the converters
    Viktor Gyenes committed Jan 6, 2022

Commits on Jan 5, 2022

  1. bug fix in optimizing squeeze op into variables

    Viktor Gyenes committed Jan 5, 2022
  2. fixes in generating output tensors for fragments with tuple/array res…

    …ults
    Viktor Gyenes committed Jan 5, 2022
  3. fix in exception handling in shape propagation to print the right err…

    …or message
    Viktor Gyenes committed Jan 5, 2022

Commits on Jan 3, 2022

  1. add missing limits header to shape.h

    Viktor Gyenes committed Jan 3, 2022

Commits on Nov 17, 2021

  1. add memory requirement computation to gmac calculator

    Viktor Gyenes committed Nov 17, 2021

Commits on Nov 9, 2021

Commits on Nov 8, 2021

  1. add io-transpose flag to ONNX to NNEF conversion, to facilitate fixin…

    …g models that were converter to ONNX from TF and contain NXC input/output dimension order
    Viktor Gyenes committed Nov 8, 2021

Commits on Oct 11, 2021

Commits on Sep 14, 2021

  1. enhance printing/dumping of internal graph representation to a more r…

    …eadable form
    Viktor Gyenes committed Sep 14, 2021
Older