Releases: renesas-rz/meta-renesas-ai
Renesas RZ/G AI BSP v4.0.0
Main features of this release:
- Upgrade base Renesas RZ/G1 BSP to tag certified-linux-v2.1.7
- Upgrade base Renesas RZ/G2 BSP to tag BSP-1.0.6-update1
- Upgrade TensorFlow and TensorFlow Lite support to v2.3.1
- Upgrade Google Coral TPU support to the frogfish release (including moving to the new libedgetpu repository)
- Added 'htop' package support
- Added support for all Coral TPU image classification models (including EfficientNet (S, M, L), MobileNet (v1, v2, v3, iNet insects, iNat plants, iNat birds), Inception (v1, v2, v3, v4)) to meta-benchmark
- Added support to build 'wic' images
- Added a simple script to automate the building process
Supported AI tools:
- Arm Neural Network SDK v19.08.01
- Caffe2 v0.8.1
- Google Coral USB TPU (frogfish)
- ONNX Runtime v1.1.2
- OpenCV v4.1.1
- PyTorch v1.5.1
- TensorFlow v2.3.1
- TensorFlow Lite v2.3.1
Supported Platforms:
- Renesas RZ/G2H hihope-rzg2h
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2N hihope-rzg2n
- Renesas RZ/G2E ek874
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1M iwg20m-g1m
- Renesas RZ/G1E iwg22m
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.7.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.6-update1.
Renesas RZ/G AI BSP v3.5.0
Main features of this release:
- Upgrade base Renesas RZ/G1 BSP to tag certified-linux-v2.1.7
- Upgrade base Renesas RZ/G2 BSP to tag BSP-1.0.6-update1
- Upgrade TensorFlow and TensorFlow Lite support to v2.3.1
- Upgrade Google Coral TPU support to the frogfish release (including moving to the new libedgetpu repository)
- Added 'htop' package support
- Added support for all Coral TPU image classification models (including EfficientNet (S, M, L), MobileNet (v1, v2, v3, iNet insects, iNat plants, iNat birds), Inception (v1, v2, v3, v4)) to meta-benchmark
- Added support to build 'wic' images
- Added a simple script to automate the building process
Supported AI tools:
- Arm Neural Network SDK v19.08.01
- Caffe2 v0.8.1
- Google Coral USB TPU (diploria2)
- ONNX Runtime v1.1.2
- OpenCV v4.1.1
- PyTorch v1.5.1
- TensorFlow v2.3.1
- TensorFlow Lite v2.3.1
Supported Platforms:
- Renesas RZ/G2H hihope-rzg2h
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2N hihope-rzg2n
- Renesas RZ/G2E ek874
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1M iwg20m-g1m
- Renesas RZ/G1E iwg22m
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.7.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.6-update1.
Renesas RZ/G AI BSP v3.4.0
Main features of this release:
- Add support for PyTorch v1.5.0
- Add support for RZ/G2N hihope-rzg2n reference platform
- Upgraded base Renesas RZ/G1 BSP to tag certified-linux-v2.1.5
- Upgraded base Renesas RZ/G2 BSP to tag BSP-1.0.3
- Added support for the TensorFlow Lite MobileNet v3 pre-trained models to meta-benchmark
- Added ability to select the Google Coral USB accelerator operating frequency
Supported AI tools:
- Arm Neural Network SDK v19.08.01
- Caffe2 v0.8.1
- Google Coral USB TPU (diploria2)
- ONNX Runtime v1.1.2
- OpenCV v4.1.1
- PyTorch v1.5.0
- TensorFlow v2.0.0
- TensorFlow Lite v2.0.0
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2N hihope-rzg2n
- Renesas RZ/G2E ek874
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.5.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.3.
Renesas RZ/G AI BSP v3.3.0
Main features of this release:
- Upgraded ONX Runtime support to v1.1.2
Supported AI tools:
- Arm Neural Network SDK v19.08.01
- Caffe2 v0.8.1
- Google Coral USB TPU (diploria2)
- ONNX Runtime v1.1.2
- OpenCV v4.1.1
- TensorFlow v2.0.0
- TensorFlow Lite v2.0.0
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2E ek874
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.1.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.1-update1.
Renesas RZ/G AI BSP v3.2.0
Main features of this release:
- Upgraded ArmNN support to v19.08.01
- Upgraded TensorFlow support to v2.0.0
- Upgraded TensorFlow Lite support to v2.0.0
- Added support for the Google Coral Edge TPU USB device
Supported AI tools:
- Arm Neural Network SDK v19.08.01
- Caffe2 v0.8.1
- Google Coral USB TPU (diploria2)
- ONNX Runtime v0.4.0
- OpenCV v4.1.1
- TensorFlow v2.0.0
- TensorFlow Lite v2.0.0
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2E ek874
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.1.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.1-update1.
Renesas RZ/G AI BSP v3.1.0
This release adds support for OpenCV v4.1.1.
Supported AI tools:
- Arm Neural Network SDK v19.02
- Caffe2 v0.8.1
- ONNX Runtime v0.4.0
- OpenCV v4.1.1
- TensorFlow v1.10.0
- TensorFlow Lite v1.10.0
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2E ek874
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.1.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.1-update1.
Renesas RZ/G AI BSP v3.0.0
This release adds support for the RZ/G2M and RZ/G2E reference platforms.
RZ/G2M platform details: https://mp.renesas.com/en-eu/rzg/marketplace/board/RZGB000012.html
RZ/G2E platform details: https://mp.renesas.com/en-eu/rzg/marketplace/board/RZGB000011.html
Also added to this release is meta-benchmark. This layer will install a variety of models for each supported framework along with some simple test applications. See meta-benchmark/README.md for more details.
Supported AI tools:
- TensorFlow v1.10.0
- TensorFlow Lite v1.10.0
- Caffe2 v0.8.1
- Arm Neural Network SDK v19.02
- ONNX Runtime v0.4.0
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
- Renesas RZ/G2M hihope-rzg2m
- Renesas RZ/G2E ek874
For RZ/G1 platforms use meta-renesas tag certified-linux-v2.1.1.
For RZ/G2 platforms use meta-rzg2 tag BSP-1.0.1-update1.
Renesas RZ/G AI BSP v2.3.0
The release adds support for the ONNX Runtime v0.4.0 from Microsoft: https://github.com/microsoft/onnxruntime.
Supported AI tools:
- TensorFlow v1.10.0
- TensorFlow Lite v1.10.0
- Caffe2 v0.8.1
- Arm Neural Network SDK v19.02
- ONNX Runtime v0.4.0
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
Based on meta-renesas certified-linux-v2.1.1.
Renesas RZ/G AI BSP v2.2.0
The release upgrades Arm NN SDK support to v19.02 and rebases to the latest Renesas RZ/G BSP: certified-linux-v2.1.0.
Supported AI tools:
- TensorFlow v1.10.0
- TensorFlow Lite v1.10.0
- Caffe2 v0.8.1
- Arm Neural Network SDK v19.02
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
Based on meta-renesas certified-linux-v2.1.1.
Renesas RZ/G AI BSP v2.1.0
The release adds support for the Arm NN SDK.
Supported AI tools:
- TensorFlow v1.10.0
- TensorFlow Lite v1.10.0
- Caffe2 v0.8.1
- Arm Neural Network SDK v18.11
Supported Platforms:
- Renesas RZ/G1M iwg20m
- Renesas RZ/G1H iwg21m
- Renesas RZ/G1E iwg22m
Based on meta-renesas certified-linux-v2.1.0.