在下载所有数据并安装依赖项后，我正在训练一个模型（RNN）以使用Theano，numpy，h5py，scipy，mnist和Microsoft Coco数据集从自然语言生成图像斑点，我尝试使用以下命令开始训练：
python alignDraw.py models / mnist-captions.json
但由于代码尝试访问此路径中的文件，因此出现IO错误：
/ais/gobi3/u/nitish/mnist/mnist.h5
Code: (https://github.com/mansimov/text2image)
我发现gobi是一个与我假设的kaggle类似的网站。
我尝试过的事情：
- 重复下载mnist数据
- 使用pbd逐行调试直到602行
- 我确认已安装依赖项依赖项
- 仍在检查skipthough，无法确定是否正确设置
if "data" in model:
train_data_key = model["data"]["train_data"]["key"]
train_data_file = model["data"]["train_data"]["file"]
train_labels_key = model["data"]["train_labels"]["key"]
train_labels_file = model["data"]["train_labels"]["file"]
val_data_key = model["data"]["validation_data"]["key"]
val_data_file = model["data"]["validation_data"]["file"]
val_labels_key = model["data"]["validation_labels"]["key"]
val_labels_file = model["data"]["validation_labels"]
else:
train_file = "/ais/gobi3/u/nitish/mnist/mnist.h5"
train_key = "train"
val_file = "/ais/gobi3/u/nitish/mnist/mnist.h5"
val_key = "validation"
我期望开始培训并举起我的GPU，但我得到了以下信息：
警告（theano.tensor.blas）：对BLAS函数使用基于NumPy C-API的实现。
追溯（最近一次通话）：
File "alignDraw.py", line 607, in <module>
train_data =np.copy(h5py.File(train_data_file,'r'[train_data_key])
File "/home/Myusername/.local/lib/python2.7/site-packages/h5py/_hl/files.py", line 408, in __init__ swmr=swmr)
File "/home/Myusername/.local/lib/python2.7/site-packages/h5py/_hl/files.py", line 173, in make_fid fid = h5f.open(name, flags, fapl=fapl)
File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
File "h5py/h5f.pyx", line 88, in h5py.h5f.open
IOError: Unable to open file (unable to open file: name = '/ais/gobi3/u/nitish/mnist/mnist.h5', errno = 2, error message = 'No such file or directory', flags = 0, o_flags = 0)
谢谢您阅读此篇。
道歉的穷人仍然学习xD
