mmrazor

mmrazor copied to clipboard

[Bug] AutoSlim For faster_rcnn_r50_fpn Error

1

### Describe the bug error log: ``` Traceback (most recent call last): File "/root/project/mmrazor/mmrazor/models/algorithms/autoslim.py", line 69, in _init_pruner pseudo_pruner.prepare_from_supernet(pseudo_architecture) File "/root/project/mmrazor/mmrazor/models/pruners/structure_pruning.py", line 152, in prepare_from_supernet pseudo_img = supernet.forward_dummy(pseudo_img) File "/root/project/mmrazor/mmrazor/models/architectures/base.py",...

tanghy2016

distill result

7

在修改了配置文件训练自己的数据后，我得到了teacher的权重文件，运行的代码为： ```python python tools/mmcls/train_mmcls.py configs/distill/rkd/rkd_neck_resnet34_resnet18_8xb32_in1k.py \ --work-dir /data/dataset_81/jd_result/distill_mmcls \ --cfg-options algorithm.distiller.teacher.init_cfg.type=Pretrained ``` 生成的结果为：  但是出现了一个问题： 生成的经过distill 的权重文件的大小是167MB，但是teacher的权重文件只有163MB，没有起到模型瘦身的效果，这是为什么呢

JennieDD

[Bug] Fix a bug in base distiller and add yolox distill example

1

## Motivation As the title ## BC-breaking (Optional) Does the modification introduce changes that break the backward compatibility of the downstream repositories? If so, please describe how it breaks the...

pppppM

NAS on semantic segmentation

1

Hi, I am wondering if its possible to currently run a NAS algorithm (e.g., SPOS) on a semantic segmentation model from mmsg (e.g., Deeplab). I see that for SPOS, the...

Hrayo712

mmrazor output file writing

2

mmdetection can write output to json file. but, mmrazor/tools/mmdet/test_mmdet.py cannot write output ex ) python tools/mmdet/test_mmdet.py ./configs/distill/cwd/cwd_cls_head_deformable_detr_rx101_r50.py ./deformable_detr_rx101_r50/epoch_211.pth --eval bbox --eval--options="jsonfile_prefix=./output_path" how can I write output file? the config file...

kdh-awraw1019

How could I use MMRazor to train YoloX with KD ?

5

I need the MMRazor API to the yolox to make KD, and the config file of YOLOX too.

LiuXuan66

Distillation in fp16 mode

1

when student and teacher network traing in fp16, if align_module is none, training is work well. but if i set align_module, the align_module's weight is fp32, but th student's outputs...

Veiiling

[Bug]distll for training point_rend

5

### Describe the bug A clear and concise description of what the bug is. \[here\]Traceback (most recent call last): File "tools/mmdet/train_mmdet.py", line 210, in main() File "tools/mmdet/train_mmdet.py", line 206, in...

xuhao-anhe

[Bug]Distll for point_rend

3

### Describe the bug A clear and concise description of what the bug is. \[here\]I try to use CWD for point_ Rend for distillation.It has the following bugs：Traceback (most recent...

xuhao-anhe

find_make_group_parser for depthwise

3

### Describe the bug def find_make_group_parser(self, node_name, name2module): """Find the corresponding make_group_parser according to the ``node_name``""" if 'concat' in node_name and node_name not in name2module: return MAKE_GROUP_PARSER_DICT['concat'] elif 'chunk' in...

NickYangMin