时间戳

首页 > 解决方案 > RuntimeError：无法从输入流中解析字典

tensorflow - RuntimeError：无法从输入流中解析字典

问题描述

我对cntk完全陌生。我最近使用 GeForce 940MX GPU 在我的电脑（Windows 10，i5-7200U CPU）上安装了 cntk 2.7（GPU 版本）。我正在尝试设置 cntk 和以下链接中提供的更快的 rcnn 对象检测示例 https://docs.microsoft.com/en-us/cognitive-toolkit/object-detection-using-faster-r-cnn

我正在尝试运行玩具示例。

在 Examples/Image/Detection/FastRCNN 文件夹中运行 install_data_and_model.py 后

我运行 run_faster_rcnn.py Examples/Image/Detection/FasterRCNN 文件夹

我收到以下错误：

Selected GPU[0] GeForce 940MX as the process wide default device.

About to throw exception 'Failed to parse Dictionary from the input stream.'
Traceback (most recent call last):
File "run_faster_rcnn.py", line 34, in 
trained_model = train_faster_rcnn(cfg)

File "C:\Users\HP-PC\Anaconda3\Lib\site-packages\cntk\Examples\Image\Detection\FasterRCNN\FasterRCNN_train.py", line 291, in train_faster_rcnn
eval_model = train_faster_rcnn_e2e(cfg)

File "C:\Users\HP-PC\Anaconda3\Lib\site-packages\cntk\Examples\Image\Detection\FasterRCNN\FasterRCNN_train.py", line 314, in train_faster_rcnn_e2e
loss, pred_error = create_faster_rcnn_model(image_input, roi_input, dims_node, cfg)

File "C:\Users\HP-PC\Anaconda3\Lib\site-packages\cntk\Examples\Image\Detection\FasterRCNN\FasterRCNN_train.py", line 177, in create_faster_rcnn_model
base_model = load_model(cfg['BASE_MODEL_PATH'])

File "C:\Users\HP-PC\Anaconda3\lib\site-packages\cntk\internal\swig_helper.py", line 69, in wrapper
result = f(*args, **kwds)

File "C:\Users\HP-PC\Anaconda3\lib\site-packages\cntk\ops\functions.py", line 1721, in load_model
return Function.load(model, device, format)

File "C:\Users\HP-PC\Anaconda3\lib\site-packages\cntk\internal\swig_helper.py", line 69, in wrapper
result = f(*args, **kwds)

File "C:\Users\HP-PC\Anaconda3\lib\site-packages\cntk\ops\functions.py", line 1635, in load
return cntk_py.Function.load(str(model), device, format.value)

RuntimeError: Failed to parse Dictionary from the input stream.


[CALL STACK]
> CNTK::Internal:: UseSparseGradientAggregationInDataParallelSGD

- CNTK::operator>>

- CNTK::Function:: Load

- PyInit__cntk_py (x2)

- PyCFunction_Call

- PyEval_GetFuncDesc

- PyEval_EvalFrameEx (x2)

- PyFunction_SetAnnotations

- PyObject_Call

- PyEval_GetFuncDesc

- PyEval_EvalFrameEx (x2)

- PyEval_GetFuncDesc (x2)

有人可以帮我解决问题的全部内容吗？

标签： tensorflowpytorchobject-detectioncntkfaster-rcnn

解决方案

保存最后一个模型时关闭cntk时总是发生此错误，因此模型文件中断

推荐阅读