Caffe Framework for DesignWare EV Processors, T-2022.09

This is release T-2022.09 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit T-2022.09, from Synopsys.

New Features in this Release

• Updated docker validation script for regression check.

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, U-2023.03-RC1

This is release U-2023.03-RC1 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit U-2023.03, from Synopsys.

New Features in this Release

• Updated building scripts and packages versions to support python 3.8 environment for both Linux and Windows platforms.

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, T-2022.06

This is release T-2022.06 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit T-2022.06, from Synopsys.

New Features in this Release

• Updated APIs in code to support protobuf version >=3.19.1
• Refined Quantize shift and scale computation method for pre-quantize models support

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, T-2022.06-RC1

This is release T-2022.06-RC1 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit T-2022.06, from Synopsys.

New Features in this Release

• Updated APIs in code to support protobuf version >=3.19.1

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, S-2021.12

This is release S-2021.12 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit S-2021.12, from Synopsys.

New Features in this Release:

• Ev support enhancement
  • Refined SpatialBatchingPooling and MatMul layer implementation
  • Added round_mode (derived from origin bvlc Caffe) in Pooling Parameter
• Custom Quantized model support
  • Added per_channel_scale_output feature for Conv layer
  • Fixed depthwise Conv quantize rounding computation issue
• Added support for sparse submanifold implementation in Conv layer
• Uploaded CentOS build script with full environment setup; Updated Ubuntu build script with newer packages dependency support
• Small updates in python scripts to avoid deprecated API warnings

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, T-2022.03-RC1

This is release T-2022.03-RC1 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit T-2022.03, from Synopsys.

New Features in this Release

• EV converter and mapping support enhancement
  • Refined DepthToSpace layer implementation, added onnx style data format support
  • Fixed SpaceToBatchND layer data initialization issue for multiple instances inference
• Added automatic docker build regression check

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, S-2021.12-RC1

This is release S-2021.12-RC1 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit S-2021.12, from Synopsys.

New Features in this Release:

• Ev support enhancement: Refined SpatialBatchingPooling and MatMul layer implementation
• Custom Quantized model support: added per_channel_scale_output feature for Conv layer
• Added support for sparse submanifold implementation in Conv layer
• Uploaded CentOS build script with full environment setup; Updated Ubuntu build script with newer packages dependency support
• Small updates in python scripts to avoid deprecated API warnings

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, S-2021.09

This is release S-2021.09 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit S-2021.09, from Synopsys.

New Features in this Release:

• evconvert enhancement: Added 2 new layers spatial_batching_pooling_layer and tensor2box_layer, and refined existing layers to better support TensorFlow/Keras/ONNX/Pytorch/Caffe2 models conversion
• Custom Quantized model support: Introduced EV style quantization and rounding mechanism for key layers
• Updated python dependencies and scripts with newer packages

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, S-2021.06

This is release S-2021.06 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit S-2021.06, from Synopsys.

New Features in this Release:

• evconvert enhancement
  • Refined Mul layer to better support TensorFlow/Keras/ONNX/Pytorch models conversion
  • Updated DetectionOutput layer to support tflite style processing
• Custom Quantize support: Implement quantized features for key layers to support quantization processing in cpu-mode for tflite/onnx/caffe2 converted models
  • Details could be found at QUANTIZED_OP.md

Full details see: FEATURES.md

Caffe Framework for DesignWare EV Processors, R-2021.03-1

This is release R-2021.03-1 of Synopsys-caffe, a modified version of the popular Caffe Deep Learning framework adapted for use with DesignWare EV Family of Processors.

Synopsys-caffe is meant to be used together with synopsys-caffe-models and the MetaWare EV Development Toolkit R-2021.03-1, from Synopsys.

New Features in this Release:

• evconvert enhancement: Added 1 new C++ layer to better support TensorFlow/Keras/ONNX/Pytorch models conversion
• Updated RNNV2 layer to better integrate with EV mapping tools and support LSTM models
• Custom Quantize support: Added scale and zero_point features for key layers to support quantization processing in cpu-mode

Full details see: FEATURES.md