gmalivenko/pytorch2keras issues and pull requests

#151 - padding bug

Issue - State: open - Opened by krgy12138 over 1 year ago - 1 comment
Labels: bug

#150 - Model's Accuracy Drops A Lot after converting

Issue - State: open - Opened by MT010104 almost 2 years ago
Labels: bug

#149 - AttributeError: Number of inputs is not equal 1 for unsqueeze layer

Issue - State: closed - Opened by MT010104 almost 2 years ago

#148 - Update README.md

Pull Request - State: open - Opened by a-maghan almost 2 years ago - 1 comment

#147 - ValueError: 'onnx::Ma/' is not a valid root scope name.

Issue - State: open - Opened by pranoyr about 2 years ago - 4 comments
Labels: bug

#146 - question

Issue - State: open - Opened by netilovefm1 about 2 years ago
Labels: bug

#145 - fix(requirements): lock onnx version

Pull Request - State: open - Opened by AmitMY over 2 years ago

#144 - How to import pytorch 2 keras?

Issue - State: open - Opened by vitoriarlima over 2 years ago - 1 comment
Labels: bug

#143 - Convert NCHW to NHWC

Issue - State: closed - Opened by JJLim99 over 2 years ago
Labels: enhancement

#142 - Fix README typo

Pull Request - State: open - Opened by gqgs over 2 years ago

#141 - TypeError: export() got an unexpected keyword argument 'do_constant_folding'

Issue - State: open - Opened by taylover-pei almost 3 years ago
Labels: bug

#140 - converts torch model with dtype double (float64) to keras with float32

Issue - State: open - Opened by HarHarLinks almost 3 years ago - 1 comment
Labels: bug

#139 - cannot import name 'optimizer' from 'onnx'

Issue - State: open - Opened by avani17101 almost 3 years ago - 5 comments
Labels: bug

#138 - adding new converters functions for onnx2keras

Issue - State: open - Opened by LamyaMohaned almost 3 years ago
Labels: enhancement

#137 - AttributeError: 'tuple' object has no attribute 'ndims'

Issue - State: open - Opened by LamyaMohaned almost 3 years ago
Labels: bug

#136 - Changed optional argument "names" to "name_policy" in README

Pull Request - State: open - Opened by balisujohn about 3 years ago

#135 - Batch Normalization Layers not present in the converted Keras model

Issue - State: open - Opened by rahman-mdatiqur about 3 years ago - 1 comment
Labels: bug

#134 - Put back line that was removed by accident

Pull Request - State: closed - Opened by wmvanvliet over 3 years ago

#133 - Remove optimizer as it is no longer supported

Pull Request - State: closed - Opened by wmvanvliet over 3 years ago

#132 - onnx.optimizer does not exist anymore

Issue - State: closed - Opened by londumas over 3 years ago - 2 comments

#131 - concatenate axis not getting changed when switching from NCHW mode to NHWC mode

Issue - State: open - Opened by yogeesh-alphaics over 3 years ago
Labels: bug

#130 - "image_data_format": "channels_first",

Issue - State: open - Opened by Gs928 over 3 years ago - 1 comment

#129 - pip install

Issue - State: open - Opened by dingguodong-826 over 3 years ago
Labels: bug

#128 - ValueError: Operands could not be broadcast together with shapes , when converting with change_ordering

Issue - State: open - Opened by shairoz-deci over 3 years ago - 1 comment
Labels: bug

#127 - Converted keras model has different parameters

Issue - State: open - Opened by mike-zyz over 3 years ago - 1 comment
Labels: bug

#126 - Type error while converting PyTorch model - Input 'y' of 'Mul' Op has type float32 that does not match type int32 of argument 'x'

Issue - State: open - Opened by jiteshm17 over 3 years ago

#125 - Convert convtranspose2d to channel_last

Issue - State: closed - Opened by HeaderZzj almost 4 years ago

#124 - converted model has different weights than the original model

Issue - State: open - Opened by mohamedabdallah1996 almost 4 years ago
Labels: bug

#123 - ValueError: Layer weight shape (112,) not compatible with provided weight shape (11,)

Issue - State: open - Opened by HeaderZzj almost 4 years ago

#122 - Keras model exception when predict "Default MaxpoolingOp only supports NHWC on device type CPU"

Issue - State: open - Opened by YangMingJerry about 4 years ago - 3 comments
Labels: bug

#121 - pytorch2keras convert prelu layer fail

Issue - State: open - Opened by luan1412167 about 4 years ago
Labels: bug

#120 - Batchnorm2d and Batchnormalization

Issue - State: closed - Opened by zhang-f about 4 years ago - 1 comment
Labels: bug

#119 - AttributeError: Can't gather from tf tensor.

Issue - State: open - Opened by nestyme over 4 years ago - 3 comments

#118 - Tensorflow, Keras, PyTorch versions which work for this project

Issue - State: open - Opened by IgorSusmelj over 4 years ago - 1 comment
Labels: enhancement

#117 - Package crashes on import with protobuf==3.12.3 and tensorflow==2.2.0 on macOS

Issue - State: open - Opened by farzadab over 4 years ago
Labels: bug

#116 - TypeError: tuple indices must be integers or slices, not list

Issue - State: closed - Opened by zhang-f over 4 years ago - 4 comments
Labels: bug

#115 - all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)

Issue - State: open - Opened by Tian14267 over 4 years ago - 4 comments
Labels: bug

#114 - cudaGetDevice() failed. Status: CUDA driver version is insufficient for CUDA runtime version

Issue - State: open - Opened by lawo123 over 4 years ago
Labels: bug

#113 - how to set onnx opset_version=11

Issue - State: open - Opened by lawo123 over 4 years ago
Labels: bug

#112 - AttributeError: Cannot convert ConvTranspose2d with output_padding != 0

Issue - State: open - Opened by jaggernaut007 over 4 years ago - 1 comment
Labels: bug

#111 - Inaccurate keras model output after converting a finetuned Resnet50 Pyotrch model.

Issue - State: closed - Opened by SKRohit over 4 years ago - 3 comments
Labels: bug

#110 - Does it enable me to convert PyTorch model (NCHW) to TFlite (NHWC)?

Issue - State: open - Opened by MaratZakirov over 4 years ago - 8 comments
Labels: enhancement

#109 - tensorflow.python.framework.errors_impl.InvalidArgumentError: cannot compute ConcatV2 as input #1(zero-based) was expected to be a int64 tensor but is a int32 tensor [Op:ConcatV2] name: concat

Issue - State: open - Opened by suven2019 over 4 years ago - 2 comments
Labels: enhancement

#108 - HDF5 library version mismatched error

Issue - State: closed - Opened by aviaisr over 4 years ago - 1 comment
Labels: bug

#107 - Add support for Conv1D

Issue - State: open - Opened by CyanideBoy over 4 years ago - 2 comments
Labels: enhancement

#106 - pytorch version model transfer :AttributeError: Unsupported number of inputs

Issue - State: open - Opened by lipeng1109 over 4 years ago
Labels: bug

#105 - I made the change_ordering=True in the pytorch2keras function

Issue - State: open - Opened by shashidhar100 over 4 years ago
Labels: bug

#104 - keras model's summary , first output shape is [(None, 1, 28,28)]

Issue - State: open - Opened by sunpengfei1122 over 4 years ago
Labels: enhancement

#103 - TypeError: pytorch_to_keras() got an unexpected keyword argument 'names

Issue - State: open - Opened by daquinoaldo almost 5 years ago - 1 comment
Labels: bug

#102 - Weight data has no target variable for Resnet18

Issue - State: open - Opened by kdmcdrm almost 5 years ago - 1 comment
Labels: bug

#101 - Trouble importing Roberta

Issue - State: open - Opened by mortonjt almost 5 years ago - 1 comment
Labels: enhancement

#100 - IndexError: list index (1) out of range

Issue - State: open - Opened by Brave731 about 5 years ago - 2 comments
Labels: bug

#99 - Is it possible to support prelu?

Issue - State: open - Opened by tzm1003306213 about 5 years ago - 2 comments
Labels: enhancement

#98 - RuntimeError: Failed to export an ONNX attribute, since it's not constant, please try to make things (e.g., kernel size) static if possible

Issue - State: closed - Opened by Brave731 about 5 years ago - 1 comment
Labels: bug

#97 - 'Tensor' object has no attribute '_keras_shape'

Issue - State: open - Opened by Guocode about 5 years ago - 4 comments
Labels: bug

#96 - Input names mismatch while converting ResNet-18(number of input names provided (1) did not match number of inputs (103))

Issue - State: closed - Opened by vamsidharmuthireddy about 5 years ago - 4 comments
Labels: bug

#95 - The error of the converted model is large

Issue - State: closed - Opened by liuxin81 about 5 years ago - 1 comment
Labels: bug

#94 - ValueError: Layer weight shape (1, 512) not compatible with provided weight shape (2048, 512)

Issue - State: open - Opened by poormag about 5 years ago - 2 comments
Labels: bug

#93 - KeyError: 'Floor'

Issue - State: open - Opened by livecle about 5 years ago - 4 comments
Labels: bug

#92 - upsampling_nearest test fails

Issue - State: open - Opened by MichaelRamamonjisoa about 5 years ago - 3 comments
Labels: bug

#91 - Typos in readme.md

Issue - State: open - Opened by qureshizawar over 5 years ago - 1 comment
Labels: bug

#90 - Update README.md to highlight code blocks and fix missing typo

Pull Request - State: closed - Opened by kittinan over 5 years ago - 1 comment

#89 - Key error" 'Slice'

Issue - State: open - Opened by Jagadish-Kumaran over 5 years ago
Labels: bug

#88 - Expose onnx optimizer

Pull Request - State: closed - Opened by SleepProgger over 5 years ago - 1 comment

#87 - Implement Split

Issue - State: closed - Opened by kilroythethird over 5 years ago - 6 comments
Labels: enhancement

#86 - Fix layer naming

Pull Request - State: closed - Opened by PS1TD over 5 years ago

#85 - Update readme and documentation.

Issue - State: closed - Opened by gmalivenko over 5 years ago

#84 - Update tracing module

Issue - State: closed - Opened by gmalivenko over 5 years ago
Labels: enhancement

#83 - Can't convert the official mobilenet model

Issue - State: open - Opened by Apollo-XI over 5 years ago - 6 comments
Labels: bug, confirmed

#82 - onnx:ELU support

Pull Request - State: closed - Opened by PS1TD over 5 years ago

#81 - my output keras model is channel first?

Issue - State: closed - Opened by hackeritchy over 5 years ago - 9 comments
Labels: bug

#80 - when Converting convolution ...， get error: if len(weights[weights_name].numpy().shape) == 5: # 3D conv , KeyError: 'context_path.banch1.0.weight'

Issue - State: closed - Opened by titi1109 over 5 years ago - 1 comment
Labels: bug

#79 - Tensorflow requirement

Issue - State: open - Opened by gokceneraslan over 5 years ago
Labels: bug

#78 - Layer weight shape don't match

Issue - State: open - Opened by Gasp34 over 5 years ago - 2 comments
Labels: bug

#77 - fix flatten issue in converted tensorflowjs model

Pull Request - State: open - Opened by ShuangLiu1992 over 5 years ago

#76 - reshape after view is wrong

Issue - State: open - Opened by ShuangLiu1992 over 5 years ago - 2 comments
Labels: bug

#75 - error message typo

Pull Request - State: open - Opened by ShuangLiu1992 over 5 years ago

#74 - keyError:'31_np'

Issue - State: open - Opened by Dreamgang over 5 years ago - 3 comments
Labels: bug

#73 - "Getting Started" documentation unclear

Issue - State: closed - Opened by pavelgrib over 5 years ago - 1 comment
Labels: bug

#72 - names of layers

Issue - State: closed - Opened by gurjeet8197 over 5 years ago - 1 comment
Labels: enhancement

#71 - KeyError in convert_gather

Issue - State: open - Opened by lhfowl over 5 years ago - 1 comment

#70 - BatchNorm2d config not transferring properly

Issue - State: closed - Opened by jezza770 over 5 years ago - 2 comments

#69 - Converting clip with a min value of 0 as a ReLU

Pull Request - State: closed - Opened by ElteHupkes over 5 years ago - 1 comment

#68 - Use DepthwiseConv2D rather than SeparableConv2D with no-op pointwise conv

Pull Request - State: closed - Opened by ElteHupkes over 5 years ago

#67 - Does it also support GPU version models?

Issue - State: closed - Opened by pride0723 over 5 years ago - 1 comment
Labels: bug

#66 - Explicit Tensorflow import in lambda layer

Pull Request - State: closed - Opened by ElteHupkes over 5 years ago - 1 comment

#65 - `KeyError` with nested transposed convolution in `nn.Sequential`

Issue - State: closed - Opened by ElteHupkes over 5 years ago - 5 comments
Labels: bug

#64 - Pretrained Resnet 34 "KeyError: '0.1.weight'"

Issue - State: closed - Opened by I-C-Karakozis over 5 years ago - 1 comment
Labels: bug

#61 - Slicing with dynamic input shapes

Issue - State: closed - Opened by Kreiswolke almost 6 years ago - 2 comments
Labels: bug

#59 - Problem with converting BatchNorm layer

Issue - State: open - Opened by Golbstein almost 6 years ago - 3 comments
Labels: bug

#58 - It's not working with Pytorch = 1.0

Issue - State: closed - Opened by Golbstein almost 6 years ago - 4 comments
Labels: bug

#57 - model.save error

Issue - State: closed - Opened by lith0613 almost 6 years ago - 3 comments
Labels: bug

#54 - Update conditions on pytorch versions

Pull Request - State: closed - Opened by EloiZalczer almost 6 years ago

#53 - Error when importing torch._C.ListType

Issue - State: closed - Opened by EloiZalczer almost 6 years ago - 1 comment
Labels: bug

#52 - Dense layer conversion bug with multiple inputs

Issue - State: closed - Opened by calclavia almost 6 years ago - 2 comments
Labels: bug

#49 - can't pickle SwigPyObject objects

Issue - State: closed - Opened by NoushNabi almost 6 years ago - 5 comments
Labels: bug

#46 - Error in AlexNet test

Issue - State: closed - Opened by bayesrule almost 6 years ago - 1 comment
Labels: bug

#39 - Test MENet / Inception v3 with a new version of the converter.

Issue - State: closed - Opened by gmalivenko about 6 years ago

#38 - MaxPooling / AveragePooling issue

Issue - State: closed - Opened by gmalivenko about 6 years ago

#20 - Multiple Inputs

Issue - State: closed - Opened by rchavezj over 6 years ago - 10 comments
Labels: enhancement

GitHub / gmalivenko/pytorch2keras issues and pull requests