Quantize Pointrend

[adamantboy]

I compressed a model of pointrend based on mmdet-2.2.1，the training looks normal, but a error occurs when i convert it to onnx using functions of pytorch2onnx.py in mmdet-2.3.1 (commit id:6495391) . It seems like _bbox_forward() and _mask_forward() have some problems, do you know how to fix it? the error info is as follows:

File "/home/mechmind/projects/mech_learning/mmdet/models/detectors/base.py", line 180, in forward
return self.forward_test(img, img_metas, **kwargs)
File "/home/mechmind/projects/mech_learning/mmdet/models/detectors/base.py", line 138, in forward_test
return self.forward_dummy(imgs[0])
File "/home/mechmind/projects/mech_learning/mmdet/models/detectors/two_stage.py", line 101, in forward_dummy
roi_outs = self.roi_head.forward_dummy(x, proposals)
File "/home/mechmind/projects/mech_learning/mmdet/models/roi_heads/standard_roi_head.py", line 60, in forward_dummy
bbox_results = self._bbox_forward(x, rois)
File "/home/mechmind/projects/mech_learning/mmdet/models/roi_heads/standard_roi_head.py", line 139, in _bbox_forward
x[:self.bbox_roi_extractor.num_inputs], rois)
File "/home/mechmind/projects/nncf_pytorch/nncf/dynamic_graph/wrappers.py", line 83, in wrapped
retval = module_call(self, *args, **kwargs)
File "/home/mechmind/miniconda3/envs/nncf/lib/python3.6/site-packages/torch/nn/modules/module.py", line 722, in call_impl
result = self.forward(*input, **kwargs)
File "/home/mechmind/projects/mech_learning/mmdet/core/fp16/decorators.py", line 131, in new_func
return old_func(*args, **kwargs)
File "/home/mechmind/projects/mech_learning/mmdet/models/roi_heads/roi_extractors/single_level_roi_extractor.py", line 73, in forward
roi_feats_t = self.roi_layers[i](feats[i], rois)
File "/home/mechmind/projects/nncf_pytorch/nncf/dynamic_graph/wrappers.py", line 83, in wrapped
retval = module_call(self, *args, **kwargs)
File "/home/mechmind/miniconda3/envs/nncf/lib/python3.6/site-packages/torch/nn/modules/module.py", line 722, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/mechmind/projects/mech_learning/mmdet/ops/roi_align/roi_align.py", line 144, in forward
self.sample_num, self.aligned)
File "/home/mechmind/projects/mech_learning/mmdet/ops/roi_align/roi_align.py", line 30, in forward
aligned)
RuntimeError: roi_width >= 0 && roi_height >= 0 INTERNAL ASSERT FAILED at "/home/mechmind/projects/mech_learning/mmdet/ops/roi_align/src/cpu/roi_align_v2.cpp":134, please report a bug to PyTorch. ROIs in ROIAlign cannot have non-negative size!

[adamantboy]

@aleksei-kashapov, @AlexKoff88

[adamantboy]

@vanyalzr Could you please give me some suggestions? Thanks.

[adamantboy]

I compress a pointrend model based on mmdet-2.3.0(commit id: 7ab8f44 ) with nncf patch, the same error occurs, could you please give me a direction? @aleksei-kashapov

[lzrvch]

Hi @adamantboy,

Unfortunately the ONNX export in the original mmdetection repo is still limited in terms of the scope of exportable models. You can try to export the model to ONNX using our mmdetection fork.
The export script is located in tools/export.py and the corresponding config for INT8-quantized PointRend model is in configs/nncf_compression/point_rend/point_rend_r50_caffe_fpn_mstrain_1x_coco_int8.py. You can specify your compressed .pth checkpoint to loaded in the config script and export it to ONNX using the export.py script:

python tools/export.py configs/nncf_compression/point_rend/point_rend_r50_caffe_fpn_mstrain_1x_coco_int8.py /path/to/your/checkpoint/ ./ onnx.

cc @aleksei-kashapov