| COMPMID-2893: Bare-metal build failures |
|
More...
|
about 7 years ago
|
| COMPMID-2892: [CL] use int accumulator for quantized MxN DWC |
|
More...
|
about 7 years ago
|
| COMPMID-2305: NEDepthwiseConvolution 3x3: support for QUANT8_PER_CHANNEL_SYMM |
|
More...
|
about 7 years ago
|
| COMPMID-2452: Add mnist example |
|
More...
|
about 7 years ago
|
| COMPMID-2138: Create safe math integer functions |
|
More...
|
about 7 years ago
|
| COMPMID-2106: Extend get_input_accessor() to run input images through QASYMM8 networks |
|
More...
|
about 7 years ago
|
| COMPMID-2853: VGG16 regression for fp32 |
|
More...
|
about 7 years ago
|
| COMPMID-2672: add comments for GEMM helper macros |
|
More...
|
about 7 years ago
|
| COMPMID-2563 Change how the best overall GEMM configuration is selected |
|
More...
|
about 7 years ago
|
| COMPMID-2309 : CLConvolutionLayer: support QUANT8_SYMM_PER_CHANNEL filters |
|
More...
|
about 7 years ago
|
| COMPMID-2862: remove forward declaration of ReductionOperation |
|
More...
|
about 7 years ago
|
| COMPMID-2672: add comments for OpenCL helper macros |
|
More...
|
about 7 years ago
|
| COMPMID-2700: Use NEON wrapper on SoftmaxLayer |
|
More...
|
about 7 years ago
|
| COMPMID-2266: [CL] add support for Log Softmax |
|
More...
|
about 7 years ago
|
| COMPMID-2608: Enable quantization with multiplier greater than 1 on NEON |
|
More...
|
about 7 years ago
|
| COMPMID-1889: Fuse bias addition and output stage in CLFCLayer. |
|
More...
|
about 7 years ago
|
| COMPMID-1671: Allow fp mixed precision in CLFCLayer. |
|
More...
|
about 7 years ago
|
| COMPMID-2549: Add GLES Runtime Context interfaces. |
|
More...
|
about 7 years ago
|
| COMPMID-1730: Document best practices |
|
More...
|
about 7 years ago
|
| MLCE-146: VTS failures A55 |
|
More...
|
about 7 years ago
|
| COMPMID-2588: Optimize the output detection kernel required by MobileNet-SSD (~27% improvement) |
|
More...
|
about 7 years ago
|
| COMPMID-2481: Add InputQuantizationInfo!=OutputQuantization Info tests for DeconvolutionLayer |
|
More...
|
about 7 years ago
|
| COMPMID-2452: (3RDPARTY_UPDATE) Internal benchmark fixes |
|
More...
|
about 7 years ago
|
| COMPMID-2609: Enable quantization with multiplier greater than 1 on OpenCL |
|
More...
|
about 7 years ago
|
| COMPMID-2452: Fix CLDepthwiseConvolutionLayerNativeKernel |
|
More...
|
about 7 years ago
|
| COMPMID-1816: Use parallel reduction on 0 axis in CL ARG_MIN/ARG_MAX |
|
More...
|
about 7 years ago
|
| COMPMID-2576: Fuse activation in Winograd output transform. |
|
More...
|
about 7 years ago
|
| COMPMID-2486: Fix bare arm64-v8.2-a builds |
|
More...
|
about 7 years ago
|
| COMPMID-2452: Update documentation for Weights Manager |
|
More...
|
about 7 years ago
|
| COMPMID-2486: Fix bare metal builds |
|
More...
|
about 7 years ago
|
