A GPU-accelerated library containing highly optimized building blocks and an execution engine for data processing to accelerate deep learning training and inference applications.
This DALI release includes the following key features and enhancements:
fn.readers.numpy
(#4796, #4848).iscrowd
entries from COCO (#4792).fn.experimental.remap
operator (#4790).fn.external_source
(#4793).iscrowd
entries from COCO (#4792)"depleted"
operator trace (#4794)There are no breaking changes in this DALI release.
No features were deprecated in this release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.NOTE: DALI builds for CUDA 12 dynamically link the CUDA toolkit. To use DALI, install the latest CUDA toolkit.
CUDA 11.0 and CUDA 12.0 builds use CUDA toolkit enhanced compatibility.
They are built with the latest CUDA 11.x/12.x toolkit respectively but they can run on the latest,
stable CUDA 11.0/CUDA 12.0 capable drivers (450.80 or later and 525.60 or later respectively).
However, using the most recent driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
Install via pip for CUDA 12.0:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda120==1.26.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda120==1.26.0
or for CUDA 11:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.26.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.26.0
Or use direct download links (CUDA 12.0):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
fn.experimental.readers.fits
) for the CPU backend (#4591).fn.experimental.equalize
) (#4742).fn.experimental.filter
) (#4764).Pipeline.run
(#4712).fn.readers.webdataset
performance (#4708).fn.readers.numpy
(#4745).fn.experimental.decoder.image
(#4727).fn.experimental.decoders.video
returning incorrect frames for high-resolution videos (#4717).fn.experimental.decoder.image
(#4723).math.abs
and math.floor
) incorrectly processing non-scalar samples (#4746).sample
to data
in automatic augmentation APIs (#4774)Pipeline.run()
(#4712)if
predicate and not
expression (#4715)There are no breaking changes in this DALI release.
No features were deprecated in this release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.NOTE: DALI builds for CUDA 12 dynamically link the CUDA toolkit. To use DALI, install the latest CUDA toolkit.
CUDA 11.0 and CUDA 12.0 builds use CUDA toolkit enhanced compatibility.
They are built with the latest CUDA 11.x/12.x toolkit respectively but they can run on the latest,
stable CUDA 11.0/CUDA 12.0 capable drivers (450.80 or later and 525.60 or later respectively).
However, using the most recent driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
Install via pip for CUDA 12.0:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda120==1.25.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda120==1.25.0
or for CUDA 11:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.25.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.25.0
Or use direct download links (CUDA 12.0):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
and
, or
, and not
boolean operators in pipelines (#4629, #4676).and
and or
, and not lazy not
support (#4629)There are no breaking changes in this DALI release.
No features were deprecated in this release.
experimental.decoder.image
may hang during a pipeline build or a teardown.
The issue has been fixed in nightly builds and will be fixed in release 1.25.privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.NOTE: DALI builds for CUDA 12 dynamically link the CUDA toolkit. To use DALI, install the latest CUDA toolkit.
CUDA 11.0 and CUDA 12.0 builds use CUDA toolkit enhanced compatibility.
They are built with the latest CUDA 11.x/12.x toolkit respectively but they can run on the latest,
stable CUDA 11.0/CUDA 12.0 capable drivers (450.80 or later and 525.60 or later respectively).
However, using the most recent driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
Install via pip for CUDA 12.0:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda120==1.24.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda120==1.24.0
or for CUDA 11:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.24.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.24.0
Or use direct download links (CUDA 12.0):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
experimental.inputs.video
operator that supports decoding large videos from memorybuffer across multiple iterations (#4613, #4584, #4603, #4564).fn.experimental.decoders.image
(#4625, #4600, #4587, #4572, #4592, #4548).fn.experimental.tensor_resize
operator (#4492).fn.experimental.equalize
operator (#4575, #4565).fn.constant
operator synchronization issue (#4643).fn.reshape
(#4631).VideoInput<MixedBackend>
(#4613)reshape
: restore the support for trailing wildcard in rel_shape
(#4623)DataId
mechanism for fn.inputs.video
operator (#4584)MixedBackend
support for InputOperator
(#4603)define_graph
argument from build
pipeline method (#4555)release_unused
function to memory pools. (#4556)constant
operator: Set proper stream in constant storage. (#4643)reshape
: Prevent out-of-bounds access with trailing wildcard in rel_shape
(#4631)rel_shape
length validation in reshape
(#4595)release_unused
. Don't rely on cudaGetMemInfo in preallocation tests. (#4596)There are no breaking changes in this DALI release.
No features were deprecated in this release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.NOTE: DALI builds for CUDA 12 dynamically link the CUDA toolkit. To use DALI, install the latest CUDA toolkit.
CUDA 11.0 and CUDA 12.0 builds use CUDA toolkit enhanced compatibility.
They are built with the latest CUDA 11.x/12.x toolkit respectively but they can run on the latest,
stable CUDA 11.0/CUDA 12.0 capable drivers (450.80 or later and 525.60 or later respectively).
However, using the most recent driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
Install via pip for CUDA 12.0:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda120==1.23.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda120==1.23.0
or for CUDA 11:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.23.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.23.0
Or use direct download links (CUDA 12.0):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
experimental.inputs.video
operator that supports decoding video from memorybuffer across multiple iterations to reduce memory usage (#4519).fn.experimental.filter
(convolution) operator (#4298, #4525).No major issues were fixed in this release.
VideoInput
operator for the CPU (#4519)VideoInput
operator (#4513)InputOperator
from ExternalSource
(#4505)Operator
inheritance from VideoDecoderBase
(#4508)#include <optional>
. (#4520)No features were deprecated in this release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.NOTE: DALI builds for CUDA 12 dynamically link the CUDA toolkit. To use DALI, install the latest CUDA toolkit.
CUDA 11.0 and CUDA 12.0 builds use CUDA toolkit enhanced compatibility.
They are built with the latest CUDA 11.x/12.x toolkit respectively but they can run on the latest,
stable CUDA 11.0/CUDA 12.0 capable drivers (450.80 or later and 525.60 or later respectively).
However, using the most recent driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
Install via pip for CUDA 12.0:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda120==1.22.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda120==1.22.0
or for CUDA 11:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.22.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.22.0
Or use direct download links (CUDA 12.0):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
experimental.decoders.image
experimental.decoders.image_crop
experimental.decoders.image_random_crop
experimental.decoders.image_slice
The following issues were fixed in this release:
There are no breaking changes in this DALI release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.Install via pip for CUDA 10.2:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda102==1.21.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda102==1.21.0
or for CUDA 11:
CUDA 11.0 build uses CUDA toolkit enhanced compatibility. It is built with the latest CUDA 11.x toolkit
while it can run on the latest, stable CUDA 11.0 capable drivers (450.80 or later).
Using the latest driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.21.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.21.0
Or use direct download links (CUDA 10.2):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
fn.experimental.remap
operator for generic geometric transformation of images and video (#4379, #4419, #4365, #4374, #4425).fn.experimental.inflate
operator that enables decompression of LZ4 compressed input (#4366).The following issues were fixed in this release:
fn.experimental.remap
optimizations (#4419)fn.experimental.remap
operator (#4379)cuh
files to linter (#4384)privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.Install via pip for CUDA 10.2:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda102==1.20.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda102==1.20.0
or for CUDA 11:
CUDA 11.0 build uses CUDA toolkit enhanced compatibility. It is built with the latest CUDA 11.x toolkit
while it can run on the latest, stable CUDA 11.0 capable drivers (450.80 or later).
Using the latest driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.20.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.20.0
Or use direct download links (CUDA 10.2):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
experimental.decoders.video
stand-alone video decoder to decode video on GPU and CPU provided as an in-memory buffer (for example, through an external source) (#4354, #4296).The following issues were fixed in this release:
There are no breaking changes in this DALI release.
DALI will drop support for CUDA 10.2 in an upcoming release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.Install via pip for CUDA 10.2:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda102==1.19.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda102==1.19.0
or for CUDA 11:
CUDA 11.0 build uses CUDA toolkit enhanced compatibility. It is built with the latest CUDA 11.x toolkit
while it can run on the latest, stable CUDA 11.0 capable drivers (450.80 or later).
Using the latest driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.19.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.19.0
Or use direct download links (CUDA 10.2):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
fill_value
argument for each sample in the fn.erase
operator (#4182).FramesDecoder
(#4184).audio_resample
operator out of experimental module (#4194).The following issues were fixed in this release:
WebDataset integration using External Source
example (#4240)fill_value
argument in Erase operator (#4182)audio_resample
out of experimental module (#4194)There are no breaking changes in this DALI release.
There are no deprecated features in this DALI release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.Install via pip for CUDA 10.2:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda102==1.18.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda102==1.18.0
or for CUDA 11:
CUDA 11.0 build uses CUDA toolkit enhanced compatibility. It is built with the latest CUDA 11.x toolkit
while it can run on the latest, stable CUDA 11.0 capable drivers (450.80 or later).
Using the latest driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.18.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.18.0
Or use direct download links (CUDA 10.2):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code:
This DALI release includes the following key features and enhancements:
The following issues were fixed in this release:
nose2
(#4037)\
(#4123)There are no breaking changes in this DALI release.
There are no deprecated features in this DALI release.
privileged=yes
in Extra Settings for AWS data points--privileged
or --security-opt seccomp=unconfined
for bare Docker.Install via pip for CUDA 10.2:
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda102==1.17.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda102==1.17.0
or for CUDA 11:
CUDA 11.0 build uses CUDA toolkit enhanced compatibility. It is built with the latest CUDA 11.x toolkit
while it can run on the latest, stable CUDA 11.0 capable drivers (450.80 or later).
Using the latest driver may enable additional functionality.
More details can be found in enhanced CUDA compatibility guide.
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-cuda110==1.17.0
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/ nvidia-dali-tf-plugin-cuda110==1.17.0
Or use direct download links (CUDA 10.2):
Or use direct download links (CUDA 11.0):
FFmpeg source code:
Libsndfile source code: