Pengguang He comments

Results 9 comments of


                                            Pengguang He

ImportError: cannot import name 'Di1stributedDataParallel' from 'apex.parallel'

maybe it's beacause the platform, we can directly annotate the line "from apex.parallel import Di1stributedDataParallel as DDP"

How to get the student model checkpoint after finishing knowledge distillation?

> > Would you like to make a PR? @maxrumi > > ok, I will make a PR. Hi, how to extract the student chpt of mmseg model, we need...

When I ran the script pruning/get_channel_units.py, I got this error: TypeError: expected Tensor as element 0 in argument 0, but got tuple.

> Please use "Fxtracer" to replace "BackwardTracer". We cannot maintain BackwardTracer anymore. Thanks for your reply, do you mean "MMFxTracer" or "FxTracer"? I tried to modify the "BackwardTracer" to "MMFxTracer"...

When I ran the script pruning/get_channel_units.py, I got this error: TypeError: expected Tensor as element 0 in argument 0, but got tuple.

> Please use "Fxtracer" to replace "BackwardTracer". We cannot maintain BackwardTracer anymore. https://github.com/open-mmlab/mmrazor/tree/main/configs/pruning/mmseg/dcff I refer to this recommendation command. ![image](https://github.com/open-mmlab/mmrazor/assets/73171209/45d12629-e3c4-439c-8662-9a43956e5451)

When I ran the script pruning/get_channel_units.py, I got this error: TypeError: expected Tensor as element 0 in argument 0, but got tuple.

> Please use "Fxtracer" to replace "BackwardTracer". We cannot maintain BackwardTracer anymore. **Is it possible that your documentation does not match the repository code? As someone with some coding experience,...

运行mmdetection image_demo.py时报错 unknown argument {'texts', 'custom_entities'} for `preprocess`, `forward`, `visualize` and `postprocess`

> @HePengguang Have you updated your code and configurations to the latest 3.1.0 ? thanks, it works. The cause of the error was that there were both 3.0.0 and 3.0.1...

ValueError: need at least one array to concatenate when training with custom .tif dataset

I met the same issue, did you solve this problem?

Quantization for MMDet and MMSeg

> still no doc has been updated? could anyone help for doing quantization among mmseg via mmrazor? > > sos Did you know how to quantize mmseg in mmrazor now?...

What should I do to prune my own segmentation model with my own datasets?

But > Please refer to [link](https://github.com/open-mmlab/mmrazor/blob/main/configs/pruning/base/group_fisher/README.md), where we provide several configs to prune classification and detection models It is different with mmcls or mmdet. And I cannot use this command...