updated quantization modules to support mmdetection, using Hardtanh for fixed range...
authorManu Mathew <a0393608@ti.com>
Tue, 16 Jun 2020 08:02:49 +0000 (13:32 +0530)
committerManu Mathew <a0393608@ti.com>
Tue, 16 Jun 2020 08:06:14 +0000 (13:36 +0530)
commitecc9d7c2b74a728b12bacb0604c9ad94de27a394
tree9a4f1fb00a6878d56807d73246ea2c8338a02de7
parentfaca665ae4c35d90c613fd5a69172bb10f3cbf22
updated quantization modules to support mmdetection, using Hardtanh for fixed range activation, other minor fixes, updated docs
20 files changed:
docs/Quantization.md
modules/pytorch_jacinto_ai/engine/train_classification.py
modules/pytorch_jacinto_ai/engine/train_pixel2pixel.py
modules/pytorch_jacinto_ai/vision/models/pixel2pixel/pixel2pixelnet_utils.py
modules/pytorch_jacinto_ai/xnn/__init__.py
modules/pytorch_jacinto_ai/xnn/layers/activation.py
modules/pytorch_jacinto_ai/xnn/layers/common_blocks.py
modules/pytorch_jacinto_ai/xnn/layers/conv_blocks.py
modules/pytorch_jacinto_ai/xnn/onnx/__init__.py [new file with mode: 0644]
modules/pytorch_jacinto_ai/xnn/optim/lr_scheduler.py
modules/pytorch_jacinto_ai/xnn/quantize/hooked_module.py
modules/pytorch_jacinto_ai/xnn/quantize/quant_base_module.py
modules/pytorch_jacinto_ai/xnn/quantize/quant_calib_module.py
modules/pytorch_jacinto_ai/xnn/quantize/quant_graph_module.py
modules/pytorch_jacinto_ai/xnn/quantize/quant_test_module.py
modules/pytorch_jacinto_ai/xnn/quantize/quant_train_module.py
modules/pytorch_jacinto_ai/xnn/utils/load_weights.py
run_quantization.sh
scripts/train_classification_main.py
scripts/train_segmentation_main.py