Version 1.22.0 Release

release-note-bugfix

• Correct the number of elements and byte size display in debug message
• fix sample missing problem for slice iterator
• Bugfix the higher-order gradient of gather_nd

release-note-build

• replace centos with ubuntu and use PEP600 tag
• fix problem by ignore invalid text file
• Ease protobuf versions

release-note-doc

• add a link to nnabla.org installation page.

release-note-op-layer

• Optimize normalization functions

release-note-utility

• Fix global seed setter for multi-device settings
• Fix pillow save image with wrong array shape in grayscale mode.
• add extension and type_config arguments to create_communicator.
• prompt user to install flatbuffers package

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.21.0 Release

release-note-break-compat

• Make inplace option in most of Function operations obsolete
• Inplace methods reverts to using inplace=False

release-note-bugfix

• fix status 9 error
• fix race condition for data iterator with cache dataset
• Fix duplicate calls in forward_all
• Properly close file handlers at exception
• Bugfix: fix nn.grad for floating variables
• fix parameter error for graph viewer
• fix indexing leaf variable problem
• Add unitest for graph/parameters issue

release-note-build

• Fix APT_OPTS for runtime dockerfiles
• Skip some tests on 32bit system.
• CI: Restructure dockerfile directory
• Fix build without http_proxy
• temporarily comment out a test which cause randomly crash.
• change image_utils imread behavior for pillow backend compatibility

release-note-core

• Support variadic inputs of nbla::functions::concatenate()

release-note-doc

• Fix doc error
• Add missing function modifiers to doc
• improve extension document
• modify contents of Math section and Array Manipulation
• Improve dropout explanation and add searchsorted to document
• update document for randint
• Fix generating tensorboard writer related document problem

release-note-format-converter

• fix: update onnx_opset_tensorrt, use opset11 version
• support ceil_mode for onnx exporter
• support weightnormalization exporter
• support nnabla to export weightstandardization & instancenormalization & layernormalization in onnx
• Add a quantized tflite converter to export nnp to int8 tflite
• support nnabla to export spectralnorm in onnx

release-note-op-layer

• Double backward for cumsum
• Modify cumprod backward to deal with zero input
• Support p=0 for dropout function
• Ignore label less than 0 in cross entropy
• add dot & matmul operator
• Boolean Indexing Functions
• Remove ReLUCudaCudnn
• Add roi_align function and tests
• Higher-order gradients for normalization functions

release-note-utility

• Reduce create_cache log for s3 dataset
• feat: reduce create cache log and print every 5%
• Support multiple callback
• Implement custom module nvml to replace pynvml
• refactor:remove redundant image backend switch

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.20.1 Release

release-note-bugfix

• install flatbuffer compiler for new tflite converter
• Fix type comparison from ndarray to list in ONNX Exporter

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.20.0 Release

release-note-break-compat

• L2-Norm initialization for weight normalization

release-note-bugfix

• Bug fix of fused binary operator
• Bugfix: forward(clear_no_need_grad=True) and nn.grad of Dropout
• Fix recompute to check if data is cleared before backward without setup_recompute
• fix: slove the bug that duplicated ID 331 and 332 in functions.yaml
• Sync between device and host after warm-up in GraphProfiler
• Fix RandomShift class initialization order issue
• Fix Conv1d of ONNX exporter and set the default opset to 11.
• Fix to produce incorrect graph/parameters issue
• Fix load BytesIO failure.

release-note-build

• support make nnabla-check-copyright

release-note-core

• Pass name attribute to parameter Variables
• Support forward recomputation in backward propagation
• No grad context and converter
• support sequential block to construct network

release-note-doc

• Update nnabla convert doc at 20210421

release-note-format-converter

• Add converter that directly convert nnp to tflite

release-note-op-layer

• Modify SoftPlus to accept beta argument
• Searchsorted and Meshgrid implementation

release-note-utility

• fix: pynvml.nvml has no attribute nvml_lib
• fix error when save network with F.mean()
• refactor: change audio_utils load resource behavior
• Allow parameter load and save in module
• Use urllib instead of requests
• Enable the support of module for viewers
• fix nnpload no variable name problem
• Add makefile target of install fileformat converter.

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.19.0 Release

release-note-bugfix

• Bug fix: Correct default arguments of F.norm_normalization
• Make data_iterator_simple simpler
• fix the problem when save repeated float arguments
• Fix error of saving model and enable variable_batch_size.
• fix overflow in reshape
• fix nnp_graph load hang up
• Fix numerical instability for log_sigmoid and softplus
• remove inplace_grad() related code
• Bugfix: grad computation of linear functions
• fix nnp refactor memory problem
• Fix issues of the converter.

release-note-build

• Fix download failure on Windows
• remove conda

release-note-core

• Clear grad after propagation
• clears input buffer not needed for after propagation

release-note-doc

• Update nnabla converter document for release nnabla
• Update nnabla converter document for release nnabla

release-note-format-converter

• Support export to tensorflow frozen graph again.

release-note-op-layer

• fused binary operator
• F.cumsum and F.cumprod

release-note-utility

• Add stop_exhausted option for data_iterator

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.18.0 Release

release-note-bugfix

• Remove redundant input for some funcs.
• fix -1 axis error and grad not found
• fix coding error for mutex setting
• improve grad() with filter out dead parameters
• Increase gpu speed threshold
• Fix nnabla Nan error in training.
• fix: solve the error that pynvml cannot solve the path to nvml.dll on windows

release-note-core

• Higher-order Gradients

release-note-op-layer

• Memory optimization for SpectralNorm

release-note-format-converter

• Support conversion tflite to nnabla.

release-note-utility

• Fix to support utf-8 enconding with BOM
• Support BatchNorm1d for ONNX importer.
• Add feature for nnabla_cli
• Add OOC option to nnabla_cli train
• Support comment columns in dataset CSV files

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.17.0 Release

release-note-bugfix

• Fix async copy in NNabla::ones and zeros
• Fix a bug in function reference count
• pynvml cannot solve the path to nvml.dll on windows
• Fix generate nnb config file error
• Fix error in using VGG
• Fix segmentation fault caused by use of uninitilized variable

release-note-build

• Limit numpy version to 1.19.5
• fix link error for extra NBLA_API
• Remove ONNX from NNabla requirement and Update document.
• Update for ARM build

release-note-core

• Add seed function
• add error message when parameter is illegal

release-note-doc

• Add Citation for Arxiv Paper
• add incompatible GPU information for installation doc
• Update nnabla converter document for release nnabla
• Fix documentation error in SwapInOutScheduler

release-note-op-layer

• Change dtype of initializer as float32
• Deformable Convolution
• Negative axis support for nnabla functions
• Memory optimization for STFT/iSTFT
• Optimize normalization functions implementation
• Pad memory optimization and add repeat mode
• Constant mode optoin for F.random_shift

release-note-utility

• NNP Network Refactor
• add alternative cli callback
• Attach initializer info to Variable.

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.16.0 Release

release-note-bugfix

• fix backward graph expansion by nn.grad
• update gnumakefile in examples/cpp
• Fix code typo
• support multi-thread training
• Give correct names to parameters conv_cos and conv_sin

release-note-build

• fix window build bug
• Fix test shift error

release-note-core

• Support out-of-core training

release-note-doc

• Update nnabla convert doc at 20201228
• Add document and trivial modification for using tensorboard

release-note-format-converter

• Support tf2.x for nnabla converter.

release-note-utility

• Set colab default accelerator to gpu
• Enable cache option for slice method of DataItertor.
• support the file-like object in nnploader

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.15.0 Release

release-note-bugfix

• Fix Slice output
• Fix bugs in clear buffer during forward propagation
• Correct the copy index generation when unfolding.

release-note-doc

• Update nnabla converter document for release nnabla

release-note-format-converter

• Fix GRU repeated parameters.
• Separate ONNX converter from nnabla.
• improve optimize pb by add yacc rules.

release-note-op-layer

• BatchLogdet Function

release-note-utility

• Adjust atol values for rnn/gru/lstm
• Update of cli/forward for bayesian DNN

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.

Version 1.14.0 Release

release-note-bugfix

• Fix norm api level and function id
• Fix issue that BatchMatmul adds new features to cause the converter t…
• Use the new h5py v3 data access method.
• Add missing format character in function parameter check.
• Fix the docs related to the output maps

release-note-build

• Change libarchive location
• Improve building dockerfiles
• Limit onnx version to 1.6.0
• Build with chocolatey environment.

release-note-doc

• update cuda cudnn table
• correct nnabla-ext-cuda package name in installation
• update converter pkg name
• fix document error for to_dlpack

release-note-format-converter

• Fix missing parameters from nnp to nnb.
• update dockerfile to install nnabla_tf_converter
• Enabla GRU for c-runtime.
• Feature/20200921 separate tf converter from nnabla
• fix error when no .protobuf file and parameter [] operator.

release-note-op-layer

• Norm and NormNormalization
• Add Adabelief solver proposed by Juntang Zhuang
• Add ScatterAdd Function
• Inplace for unary/binary arithmetic functions

Install the latest nnabla by:

pip install nnabla
pip install nnabla-ext-cuda102 # For CUDA version 10.2 users

You can also install the cuda extension with specific versions from one of the following. See also FAQ

• nnabla-ext-cuda100 (CUDA 10.0 x cuDNN 7.6)
• nnabla-ext-cuda102 (CUDA 10.2 x cuDNN 8.0)
• nnabla-ext-cuda110 (CUDA 11.0 x cuDNN 8.0)

pip install nnabla
pip install nnabla_ext_cuda102  # For CUDA 10.2 x cuDNN 8.0 users

For distributed training, You need to install the correct package that
matches the version of MPI installed on your system.

We prepared following packages.

• nnabla-ext-cuda100-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda100-nccl2-mpi3-1-6
• nnabla-ext-cuda102-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda102-nccl2-mpi3-1-6
• nnabla-ext-cuda110-nccl2-mpi2-1-1 (Ubuntu18.04 default)
• nnabla-ext-cuda110-nccl2-mpi3-1-6

If you want to use a version of MPI not listed above, you need to build it from the source.

Additional setup may be required depending on your OS or environment. Please check Python Package Installation Guide for details.

Install Guide
To use C++ inference feature, follow the demonstration on MNIST inference in C++.

Demonstration

Build Guide
The "nnabla-ext-cuda" package is temporarily unavailable. Use of this package is not recommended. Please install nnabla-ext-cuda110, nnabla-ext-cuda102, nnabla-ext-cuda100 instead.
The following nnabla CUDA extension packages have been deprecated and the PyPi repository has been (or going to be) closed.

• nnabla-ubuntu16
• nnabla-ubuntu18
• nnabla-ext-cuda
• nnabla-ext-cuda80
• nnabla-ext-cuda91
• nnabla-ext-cuda92
• nnabla-ext-cuda101
• nnabla-ext-cuda90-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu16
• nnabla-ext-cuda100-nccl2-ubuntu18
• nnabla-ext-cuda101-nccl2-ubuntu16
• nnabla-ext-cuda101-nccl2-ubuntu18

The following "nnabla-ext-cuda" docker images have been deprecated.

• py27-cuda92
• py36-cuda92
• py37-cuda92
• py27-cuda92-v1.0.xx
• py36-cuda92-v1.0.xx
• py37-cuda92-v1.0.xx

We've decided to change nnabla's versioning policy to semantic versioning.
This change has been applied from version 1.1.0.

Python 2 is no longer be supported from v1.5.0.
CUDA8.0 is no longer be supported from v1.6.0.
Python3.5 is no longer be supported from v1.14.0.
CUDA9.0 is no longer be supported from v1.14.0.