IntenseWebs/openvino
3
0 Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update prerelease_information.md (#16898)

Browse Source
This commit is contained in:
Karol Blaszczak authored and GitHub committed 2023-04-12 20:04:25 +00:00
1 parent 418f70abb0
commit 65a49e903c
1 file changed
+18 -13
Download Patch File Download Diff File Expand all files Collapse all files
docs/resources/prerelease_information.md
+18 -13
View File
@@ -7,6 +7,7 @@ OpenVINO developers continue to roll out pre-release software. On this page you
a general changelog for each version published under the current cycle.
.. note::
These versions are pre-release software and have not undergone full validation or qualification. OpenVINO™ toolkit pre-release is:
* NOT to be incorporated into production software/solutions.
@@ -16,26 +17,25 @@ a general changelog for each version published under the current cycle.
.. dropdown:: OpenVINO Toolkit 2023.0.0.dev20230229
.. dropdown:: OpenVINO Toolkit 2023.0.0.dev20230407
:open:
:animate: fade-in-slide-down
:color: primary
* Added remote tensor support (accepting tensor located in graph memory) in C API 2.0
* Introduced model caching on GPU. Model Caching which reduces First Inference Latency (FIL) is
* Enabled remote tensor in C API 2.0 (accepting tensor located in graph memory)
* Introduced model caching on GPU. Model Caching, which reduces First Inference Latency (FIL), is
extended to work as a single method on both CPU and GPU plug-ins.
* Added the post-training Accuracy-Aware Quantization mechanism for OpenVINO IR. By using this mechanism
the user can define the accuracy drop criteria and NNCF will consider it during the quantization.
* CPU plugin migrated to OneDNN 3.1.
* AUTO supports fall-back to CPU device in case of run-time failure of networks on accelerator devices.
* AUTO supports the option to disable CPU as the initial acceleration device to speed up first-inference latency.
* Implementing a parameter enabling network inference independently of IR precision. The default mode will be FP16
with FP32 as an option to increase accuracy. General hints (performance vs accuracy) to be provided.
* Adding support for new model use cases or optimizing the existing support (better accuracy or performance).
* Performance on dGPU optimized with Intel oneDNN v3.1, especially for transformer models.
* Dynamic shapes supported on iGPU and dGPU for Transformer(NLP) models. Not all dynamic models are supported but model coverage will be expanded in future releases.
* Performance improvement for Transformer models for NLP pipelines on CPU.
* Models:
* Migrated the CPU plugin to OneDNN 3.1.
* Enabled CPU fall-back for the AUTO plugin - in case of run-time failure of networks on accelerator devices, CPU is used.
* Now, AUTO supports the option to disable CPU as the initial acceleration device to speed up first-inference latency.
* Implemented ov::hint::inference_precision, which enables running network inference independently of the IR precision.
The default mode is FP16, it is possible to infer in FP32 to increase accuracy.
* Optimized performance on dGPU with Intel oneDNN v3.1, especially for transformer models.
* Enabled dynamic shapes on iGPU and dGPU for Transformer(NLP) models. Not all dynamic models are enabled but model coverage will be expanded in following releases.
* Improved performance for Transformer models for NLP pipelines on CPU.
* Extended support to the following models:
* Enabled MLPerf RNN-T model.
* Enabled Detectron2 MaskRCNN.
@@ -44,6 +44,11 @@ a general changelog for each version published under the current cycle.
* Optimized WeNet model.
Known issues:
* OpenVINO-dev wheel does not contain the benchmark_app package
.. dropdown:: OpenVINO Toolkit 2023.0.0.dev20230217
:animate: fade-in-slide-down