open-mmlab/mmrazor

Issues

• [Bug] use mmrazor/main/tools/visualizations/feature_diff_visualization.py error: inference_mot() missing 2 required positional arguments: 'frame_id' and 'video_len'

  #624 opened a year ago by zhanghay
  4

• AssertionError: "neck" is not in the model.

  #654 opened 3 months ago by XiaoSiJi-GCY
  3

• Are there any tutorials that are friendly to newcomers?

  #616 opened a year ago by pepsicoke7
  1

• [Bug] Custom Distillation MMSeg CWD loss nan problem

  #638 opened 9 months ago by Priyanshu88
  1

• 我在用mmrazor通过yolov5-x蒸馏yolov5-s时候遇到了问题

  #644 opened 4 months ago by KungFuPandaPro
  2

• torch.fx.proxy.TraceError: class `MMArchitectureQuant`

  #621 opened a year ago by Levi-zhan
  20

• Recorder for gradients of certian modules

  #653 opened 4 months ago by lanqz7766
  0

• TypeError: 'ABCMeta' Object is not subscriptable when configuring mmseg model for PTQ and QAT

  #652 opened 4 months ago by ktantoulas
  4

• Problematic difference between FakeQuantize for activations and weights

  #651 opened 5 months ago by Veccoy
  1

• [Bug] (suggested fix) `mmrazor.models.algorithms.quantization.mm_architecture.MMArchitectureQuant.sync_qparams()` fails if there are modules present in other modes but not in forward `mode='tensor'`

  #634 opened 9 months ago by elisa-aleman
  4

• [Bug] (suggested temporary fix) Pytorch >= 2 causes mmrazor.engine to fail

  #632 opened 10 months ago by elisa-aleman
  4

• Missing keys after RTMDET knowledge distillation

  #650 opened 6 months ago by Ramzes30765
  1

• [Docs] A100算力加持！书生大模型实战营第3期全面升级，趣味闯关模式等你开启

  #649 opened 6 months ago by boshallen
  0

• MAP is stucked at 0 for Mobilenet V2 SSD QAT without pretrained model [Bug]

  #648 opened 6 months ago by HeleMartin
  0

• No Sign of activation quantization with QAT

  #647 opened 7 months ago by Veccoy
  1

• [Bug] (suggested fix) `mmrazor.models.algorithms.mm_architecture.MMArchitectureQuant.get_deploy_model()` fails if `predict` mode lacks nodes from the `model.quantizer.tracer.skipped_methods` configuration, but the architecture `quantizer.prepare(fp32_model)` has these nodes.

  #642 opened 9 months ago by elisa-aleman
  4

• Quantizing rtmdet to int8 fails

  #612 opened a year ago by AbhiM98
  2

• ImportError: `torch>=1.13` is not installed properly, plz check.

  #615 opened a year ago by FPPMXB
  1

• Is this a dead project ?

  #643 opened 9 months ago by taofuyu
  1

• why the accuracy of distilled student is lower than model of mmpretrain

  #607 opened a year ago by sofpya
  2

• How to load the training weights of the student model before distillation？

  #613 opened a year ago by Ye1133Zi
  1

• [Bug] (suggested fix) `nn.Parameter` are not added to root after being traced in `mmrazor.models.task_utils.tracer.fx.custom_tracer.build_graphmodule()`

  #633 opened 9 months ago by elisa-aleman
  2

• Regarding tables and accuracy

  #641 opened 9 months ago by PriyanshuPandeyNA
  0

• When I use methodoutputs to access the results of assigner, I only obtain one sample

  #639 opened 9 months ago by pkyzh2006
  0

• [Bug] `mmrazor.engine.runner.quantization_loops.QATValLoop` calls `after_val_epoch` hook twice with different keys, causing `mmengine.hooks.checkpoint_hook._save_best_checkpoint()` to fail with `KeyError` for the `save_best` config

  #637 opened 9 months ago by elisa-aleman
  0

• cannot use recorder to obtain panoptic_head info from mask2former

  #636 opened 9 months ago by pkyzh2006
  0

• I want to obtain the current epoch value and associate it with the custom distillation loss

  #635 opened 9 months ago by pkyzh2006
  0

• Try to reproduce CWD in VOC data set

  #631 opened 10 months ago by nameLiMing
  0

• [Bug] TypeError: 'NoneType' object is not iterable

  #630 opened 10 months ago by Priyanshu88
  0

• TypeError: __init__() takes 1 positional argument but 2 were given

  #608 opened a year ago by kpyon-w
  2

• How to get started??

  #629 opened 10 months ago by milinbhade1214
  0

• I can't reproduce dfad results

  #628 opened a year ago by c122-ode
  0

• Was support for GN ever added to mmrazor?

  #627 opened a year ago by qraleq
  0

• Error: assert len(module.recorded_input) > 0

  #626 opened a year ago by yourbikun
  0

• How to use the mmrazor tool to lightweight the end2end.onnx model generated by mmdeploy

  #625 opened a year ago by xqc-qc
  0

• generate a yolov5_prune_config file

  #623 opened a year ago by mayilin0714
  0

• KD connector

  #620 opened a year ago by FPPMXB
  0

• The student model maintains an mAP of 0 throughout the distiller training process.

  #619 opened a year ago by Rax-Lie
  0

• compile mmrazor gets a wrong results with pip install -v -e .

  #618 opened a year ago by forever10086
  0

• Does mmrazor now support mmroate?

  #614 opened a year ago by LakeOcean
  0

• [Bug] class `CustomTracer` in mmrazor/models/task_modules/tracer/fx/custom_tracer.py: __init__() takes 1 positional argument but 3 were given

  #606 opened a year ago by RX0-Zed
  1

• RTM Head

  #600 opened a year ago by grpinto
  4

• TypeError: 'Proxy' object cannot be interpreted as an integer

  #609 opened a year ago by chunxt
  1

• [Bug]mmcls needs mmcv-full, while mmrazor now need mmcv>=2.0.0

  #602 opened a year ago by CaMi1le
  6

• [Feature] merge DFND algorithm to mmrazor

  #605 opened a year ago by Cbtor
  0

• Why student's performance started at 0 during distillation?

  #604 opened a year ago by MR-hyj
  0

• Distillation failed

  #603 opened a year ago by yaoyuan10475
  0

• I am running the group fisher prunning on colab and I am geeting an error

  #601 opened a year ago by grpinto
  0

• [Bug]

  #599 opened a year ago by grpinto
  2

• [Bug]

  #598 opened a year ago by grpinto
  1