- 環境構築中に起こったバグをまとめたもの
- Collect some issues when I try to export onnx file.
- エラー内容
> mim download mmaction2 --config tsn_imagenet-pretrained-r50_8xb32-1x1x3-100e_kinetics400-rgb --dest .
Traceback (most recent call last):
File "<frozen importlib._bootstrap>", line 602, in _exec
File "<frozen zipimport>", line 259, in load_module
File "/usr/share/python-wheels/pkg_resources-0.0.0-py2.py3-none-any.whl/pkg_resources/__init__.py", line 57, in <module>
ImportError: cannot import name 'six' from 'pkg_resources.extern' (/usr/local/lib/python3.8/dist-packages/pkg_resources/extern/__init__.py)
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/local/bin/mim", line 8, in <module>
sys.exit(cli())
File "/usr/local/lib/python3.8/dist-packages/click/core.py", line 1157, in __call__
return self.main(*args, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/click/core.py", line 1078, in main
rv = self.invoke(ctx)
File "/usr/local/lib/python3.8/dist-packages/click/core.py", line 1688, in invoke
return _process_result(sub_ctx.command.invoke(sub_ctx))
File "/usr/local/lib/python3.8/dist-packages/click/core.py", line 1434, in invoke
return ctx.invoke(self.callback, **ctx.params)
File "/usr/local/lib/python3.8/dist-packages/click/core.py", line 783, in invoke
return __callback(*args, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/mim/commands/download.py", line 70, in cli
download(package, configs, dest_root, check_certificate, dataset)
File "/usr/local/lib/python3.8/dist-packages/mim/commands/download.py", line 107, in download
return _download_configs(package, configs, dest_root,
File "/usr/local/lib/python3.8/dist-packages/mim/commands/download.py", line 126, in _download_configs
if not is_installed(package):
File "/usr/local/lib/python3.8/dist-packages/mim/utils/utils.py", line 59, in is_installed
importlib.reload(pkg_resources)
File "/usr/lib/python3.8/importlib/__init__.py", line 169, in reload
_bootstrap._exec(spec, module)
File "<frozen importlib._bootstrap>", line 608, in _exec
KeyError: 'pkg_resources'
- mmcvのバージョンが高すぎると起きるバグ？
pip install mmcv==1.7.1を実行し，mmcvのバージョンを下げてから再度mim download ~~を実行．mim download ~~が終わったあとは，再度pip install mmcv==2.0.1を実行してバージョンを戻す．
- deploy時に，
--device cudaを指定するとどこかにcpuを使っていますよというエラーが出る- モデルの中身などどこかにcpuを使っているところがある模様，修正できませんでした．
- CPUを使ってdeployしてください
> python3 include/mmaction2/tools/deployment/export_onnx_stdet.py ./configs/mm_action/detection/video_mae/vit-base-p16_videomae-k400-pre_8xb8-16x4x1-20e-adamw_ava-kinetics-rgb.py ./configs/mm_action/detection/video_mae/vit-base-p16_videomae-k400-pre_8xb8-16x4x1-20e-adamw_ava-kinetics-rgb_20230314-3dafab75.pth --num_frames 16 --output_file videomae.onnx --device cuda:0
cuda:0
Loads checkpoint by local backend from path: ./configs/mm_action/detection/video_mae/vit-base-p16_videomae-k400-pre_8xb8-16x4x1-20e-adamw_ava-kinetics-rgb_20230314-3dafab75.pth
Model output shape: torch.Size([3, 81])
cuda:0
cuda:0
/usr/local/lib/python3.8/dist-packages/mmcv/cnn/bricks/wrappers.py:65: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 4)):
/usr/local/lib/python3.8/dist-packages/mmaction/models/backbones/vit_mae.py:362: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if (h, w) != self.grid_size:
/usr/local/lib/python3.8/dist-packages/mmcv/cnn/bricks/wrappers.py:167: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 5)):
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:78: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
assert rois.size(1) == 5, 'RoI must be (idx, x1, y1, x2, y2)!'
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:31: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
g.op('Constant', value_t=torch.tensor([0], dtype=torch.long)))
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:26: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
return g.op('Gather', self, index, axis_i=dim)
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:32: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
batch_indices = g.op('Squeeze', batch_indices, axes_i=[1])
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:33: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
batch_indices = g.op(
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:38: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
g.op(
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:44: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
aligned_offset = g.op(
/usr/local/lib/python3.8/dist-packages/torch/onnx/symbolic_opset9.py:364: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
return g.op("Sub", self, other)
/usr/local/lib/python3.8/dist-packages/mmcv/ops/roi_align.py:50: FutureWarning: 'torch.onnx._patch_torch._graph_op' is deprecated in version 1.13 and will be removed in version 1.14. Please note 'g.op()' is to be removed from torch.Graph. Please open a GitHub issue if you need this functionality..
return g.op(
/usr/local/lib/python3.8/dist-packages/torch/onnx/utils.py:687: UserWarning: Constant folding in symbolic shape inference fails: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument tensors in method wrapper_cat) (Triggered internally at ../torch/csrc/jit/passes/onnx/shape_type_inference.cpp:413.)
_C._jit_pass_onnx_graph_shape_type_inference(
Traceback (most recent call last):
File "include/mmaction2/tools/deployment/export_onnx_stdet.py", line 208, in <module>
main()
File "include/mmaction2/tools/deployment/export_onnx_stdet.py", line 167, in main
torch.onnx.export(
File "/usr/local/lib/python3.8/dist-packages/torch/onnx/utils.py", line 504, in export
_export(
File "/usr/local/lib/python3.8/dist-packages/torch/onnx/utils.py", line 1529, in _export
graph, params_dict, torch_out = _model_to_graph(
File "/usr/local/lib/python3.8/dist-packages/torch/onnx/utils.py", line 1172, in _model_to_graph
params_dict = _C._jit_pass_onnx_constant_fold(
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument tensors in method wrapper_cat)pip install onnxruntime-gpuを行ったあと，cpuモードでonnxファイルを出力した場合に起こるエラー- onnxファイルは問題なく生成できているので，気にしなくても良い
- onnxファイルにエクスポートしたあと，評価のために
inferenceが走ることになっているが，そこで出ているエラーである．
- 実行時に，
['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider']をつける必要があるsession = onnxruntime.InferenceSession(args.onnx_file, providers=['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider'])
- you add
['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider']when your inference timesession = onnxruntime.InferenceSession(args.onnx_file, providers=['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider'])
Successfully export the onnx file to videomae.onnx
Traceback (most recent call last):
File "include/mmaction2/tools/deployment/export_onnx_stdet.py", line 207, in <module>
main()
File "include/mmaction2/tools/deployment/export_onnx_stdet.py", line 194, in main
session = onnxruntime.InferenceSession(args.output_file)
File "/usr/local/lib/python3.8/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 347, in __init__
self._create_inference_session(providers, provider_options, disabled_optimizers)
File "/usr/local/lib/python3.8/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 375, in _create_inference_session
raise ValueError(
ValueError: This ORT build has ['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider'] enabled. Since ORT 1.9, you are required to explicitly set the providers parameter when instantiating InferenceSession. For example, onnxruntime.InferenceSession(..., providers=['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider'], ...)- moviepyのバージョンが古いと起きるバグ？
- 現在は，これで解決可能
pip install moviepy --upgrade
Performing SpatioTemporal Action Detection for each clip
[>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ] 27/28, 13.7 task/s, elapsed: 2s, ETA: 0sPerforming visualization
Moviepy - Building video include/mmaction2/demo/demo_det_onnx.mp4.
Moviepy - Writing video include/mmaction2/demo/demo_det_onnx.mp4
Traceback (most recent call last):
File "./include/mmaction2/demo/demo_spatiotemporal_det_onnx.py", line 358, in <module>
main()
File "./include/mmaction2/demo/demo_spatiotemporal_det_onnx.py", line 352, in main
vid.write_videofile(args.out_filename)
File "/home/yuga/.local/lib/python3.8/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/usr/local/lib/python3.8/dist-packages/moviepy/decorators.py", line 54, in requires_duration
return f(clip, *a, **k)
File "/home/yuga/.local/lib/python3.8/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/usr/local/lib/python3.8/dist-packages/moviepy/decorators.py", line 135, in use_clip_fps_by_default
return f(clip, *new_a, **new_kw)
File "/home/yuga/.local/lib/python3.8/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/usr/local/lib/python3.8/dist-packages/moviepy/decorators.py", line 22, in convert_masks_to_RGB
return f(clip, *a, **k)
File "/usr/local/lib/python3.8/dist-packages/moviepy/video/VideoClip.py", line 300, in write_videofile
ffmpeg_write_video(self, filename, fps, codec,
File "/usr/local/lib/python3.8/dist-packages/moviepy/video/io/ffmpeg_writer.py", line 213, in ffmpeg_write_video
with FFMPEG_VideoWriter(filename, clip.size, fps, codec = codec,
File "/usr/local/lib/python3.8/dist-packages/moviepy/video/io/ffmpeg_writer.py", line 88, in __init__
'-r', '%.02f' % fps,
TypeError: must be real number, not NoneType-
解決策: pyopensslのバージョンが古い
-
以下のコマンドで解決
pip install pip --upgrade pip install pyopenssl --upgrade
-
エラー内容
Error in sys.excepthook:
Traceback (most recent call last):
File "/usr/lib/python3/dist-packages/apport_python_hook.py", line 72, in apport_excepthook
from apport.fileutils import likely_packaged, get_recent_crashes
File "/usr/lib/python3/dist-packages/apport/__init__.py", line 5, in <module>
from apport.report import Report
File "/usr/lib/python3/dist-packages/apport/report.py", line 32, in <module>
import apport.fileutils
File "/usr/lib/python3/dist-packages/apport/fileutils.py", line 12, in <module>
import os, glob, subprocess, os.path, time, pwd, sys, requests_unixsocket
File "/usr/lib/python3/dist-packages/requests_unixsocket/__init__.py", line 1, in <module>
import requests
File "/usr/lib/python3/dist-packages/requests/__init__.py", line 95, in <module>
from urllib3.contrib import pyopenssl
File "/usr/lib/python3/dist-packages/urllib3/contrib/pyopenssl.py", line 46, in <module>
import OpenSSL.SSL
File "/usr/lib/python3/dist-packages/OpenSSL/__init__.py", line 8, in <module>
from OpenSSL import crypto, SSL
File "/usr/lib/python3/dist-packages/OpenSSL/crypto.py", line 1553, in <module>
class X509StoreFlags(object):
File "/usr/lib/python3/dist-packages/OpenSSL/crypto.py", line 1573, in X509StoreFlags
CB_ISSUER_CHECK = _lib.X509_V_FLAG_CB_ISSUER_CHECK
AttributeError: module 'lib' has no attribute 'X509_V_FLAG_CB_ISSUER_CHECK'
- pklファイルをダウンロードしてくる必要がある
/usr/local/lib/python3.8/dist-packages/mmdet/apis/inference.py:90: UserWarning: dataset_meta or class names are not saved in the checkpoint's meta data, use COCO classes by default.
warnings.warn(
/usr/lib/python3/dist-packages/apport/report.py:13: DeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
import fnmatch, glob, traceback, errno, sys, atexit, imp, stat
Traceback (most recent call last):
File "demo/webcam_demo_spatiotemporal_det.py", line 865, in <module>
main(parse_args())
File "demo/webcam_demo_spatiotemporal_det.py", line 793, in main
stdet_predictor = StdetPredictor(
File "demo/webcam_demo_spatiotemporal_det.py", line 298, in __init__
model = init_detector(config, checkpoint, device=device)
File "/usr/local/lib/python3.8/dist-packages/mmdet/apis/inference.py", line 102, in init_detector
metainfo = DATASETS.build(test_dataset_cfg).metainfo
File "/usr/local/lib/python3.8/dist-packages/mmengine/registry/registry.py", line 570, in build
return self.build_func(cfg, *args, **kwargs, registry=self)
File "/usr/local/lib/python3.8/dist-packages/mmengine/registry/build_functions.py", line 121, in build_from_cfg
obj = obj_cls(**args) # type: ignore
File "/usr/local/lib/python3.8/dist-packages/mmaction/datasets/ava_dataset.py", line 158, in __init__
self.proposals = load(self.proposal_file)
File "/usr/local/lib/python3.8/dist-packages/mmengine/fileio/io.py", line 855, in load
with BytesIO(file_backend.get(file)) as f:
File "/usr/local/lib/python3.8/dist-packages/mmengine/fileio/backends/local_backend.py", line 33, in get
with open(filepath, 'rb') as f:
FileNotFoundError: [Errno 2] No such file or directory: 'data/ava/annotations/ava_dense_proposals_val.FAIR.recall_93.9.pkl'