A bug for show_result_pyplot(): 'EncoderDecoder' object has no attribute 'dataset_meta'

[VoyagerXvoyagerx]

mmsegmentation/mmseg/apis/inference.py

Line 197 in 7927591

classes=model.dataset_meta['classes'],

When running the https:/open-mmlab/mmsegmentation/blob/dev-1.x/demo/MMSegmentation_Tutorial.ipynb , an attribute error came.

When I referred to the doc and https://mmsegmentation.readthedocs.io/zh_CN/dev-1.x/_modules/mmseg/models/segmentors/encoder_decoder.html#EncoderDecoder (souce code) , I did not find the attribute'dataset_meta'.

1. What command or script did you run?
   https:/open-mmlab/mmsegmentation/blob/dev-1.x/demo/MMSegmentation_Tutorial.ipynb

2. Did you make any modifications on the code or config? Did you understand what you have modified?
   no modifications.

3. What dataset did you use?
   iccv09Data

Environment
sys.platform: linux
Python: 3.8.10 (default, Jun 4 2021, 15:09:15) [GCC 7.5.0]
CUDA available: True
numpy_random_seed: 2147483648
GPU 0: NVIDIA GeForce RTX 2080 Ti
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 11.3, V11.3.109
GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0
PyTorch: 1.11.0+cu113
PyTorch compiling details: PyTorch built with:

• GCC 7.3
• C++ Version: 201402
• Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.5.2 (Git Hash a9302535553c73243c632ad3c4c80beec3d19a1e)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• LAPACK is enabled (usually provided by MKL)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 11.3
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
• CuDNN 8.2
• Magma 2.5.2
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.3, CUDNN_VERSION=8.2.0, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.11.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=OFF, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF,

TorchVision: 0.12.0+cu113
OpenCV: 4.6.0
MMEngine: 0.3.1
MMSegmentation: 1.0.0rc1+7927591

Error traceback

/root/miniconda3/lib/python3.8/site-packages/mmengine/visualization/visualizer.py:170: UserWarning: `Visualizer` backend is not initialized because save_dir is None.
  warnings.warn('`Visualizer` backend is not initialized '
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In [32], line 9
      7 result = inference_model(model, img)
      8 plt.figure(figsize=(8, 6))
----> 9 vis_result = show_result_pyplot(model, img, result, palette)
     10 plt.imshow(mmcv.bgr2rgb(vis_result))

File ~/mmsegmentation/mmseg/apis/inference.py:197, in show_result_pyplot(model, img, result, opacity, title, draw_gt, draw_pred, wait_time, show, save_dir, out_file)
    191 # init visualizer
    192 visualizer = SegLocalVisualizer(
    193     vis_backends=[dict(type='LocalVisBackend')],
    194     save_dir=save_dir,
    195     alpha=opacity)
    196 visualizer.dataset_meta = dict(
--> 197     classes=model.dataset_meta['classes'],
    198     palette=model.dataset_meta['palette'])
    199 visualizer.add_datasample(
    200     name=title,
    201     image=image,
   (...)
    206     out_file=out_file,
    207     show=show)
    208 vis_img = visualizer.get_image()

File ~/miniconda3/lib/python3.8/site-packages/torch/nn/modules/module.py:1185, in Module.__getattr__(self, name)
   1183     if name in modules:
   1184         return modules[name]
-> 1185 raise AttributeError("'{}' object has no attribute '{}'".format(
   1186     type(self).__name__, name))

AttributeError: 'EncoderDecoder' object has no attribute 'dataset_meta'
<Figure size 800x600 with 0 Axes>

[sanbuphy]

you can refer to this :https:/open-mmlab/mmengine/blob/main/mmengine/runner/runner.py#L1939

[VoyagerXvoyagerx]

Thanks a lot! I read the code you referred to and added cfg.train_dataloader.dataset.metainfo = metainfo in the cfg, but the same error occurred.
In runner.py #L2098 , train_dataloader.dataset.metainfo is assigned to checkpoint.meta.dataset_meta , but model.dataset_meta is not assigned, while in show_result_pyplot()， inference.py#L148 I could not find chekpoint loaded to get dataset_meta. That maybe why my modification did not work.
fix
assign metainfo (a dict with two keys) to model.dataset_meta before excuting show_result_pyplot() .

result = inference_model(model, img)
plt.figure(figsize=(8, 6))
model.dataset_meta = dict(classes = classes, palette = palette)
vis_result = show_result_pyplot(model, img, result, save_dir='result_demo.jpg')
plt.imshow(mmcv.bgr2rgb(vis_result))

[sanbuphy]

Thanks a lot! I read the code you referred to and added cfg.train_dataloader.dataset.metainfo = metainfo in the cfg, but the same error occurred. In runner.py #L2098 , train_dataloader.dataset.metainfo is assigned to checkpoint.meta.dataset_meta , but model.dataset_meta is not assigned, while in show_result_pyplot()， inference.py#L148 I could not find chekpoint loaded to get dataset_meta. That maybe why my modification did not work. fix assign metainfo (a dict with two keys) to model.dataset_meta before excuting show_result_pyplot() .

result = inference_model(model, img)
plt.figure(figsize=(8, 6))
model.dataset_meta = dict(classes = classes, palette = palette)
vis_result = show_result_pyplot(model, img, result, save_dir='result_demo.jpg')
plt.imshow(mmcv.bgr2rgb(vis_result))

The genuine model.dataset_meta just like :
{'CLASSES':('xxx','xxxx'),'PALETTE':[xxxxxx]}
maybe you can change dict(classes = classes, palette = palette) into dict('CLASSES' = classes, 'PALETTE' = palette)
and the palette can be set to None.

In addition, you can make sure the dataset_meta is in your config.
in config they named 'metainfo' , there are four occurrences in the complete configuration：‘train_dataloader/val/test’ and Outermost layer metainfo.

[VoyagerXvoyagerx]

#2326
Thanks for your suggestion! I found dataset_meta saved in the checkpoint. So we should init model from the checkpoint before visualization. I have made a pull request to fix the bug in this tutorial.

[xiexinch]

#2326

Closing the issue, as there is no activity for a while.
We hope your issue has been resolved.
If not, please feel free to open a new one.