question

KenichiSegawa-8553 avatar image
0 Votes"
KenichiSegawa-8553 asked KenichiSegawa-8553 commented

How to solve train error from running model fit in ES_RNN model?

I cloned Github notebooks(https://github.com/Azure/DeepLearningForTimeSeriesForecasting) provided by Microsoft and was trying to run 4_ES_RNN.ipynb.

However, I am getting the following error for train model fit. Below is the code and error that I received.

Code:

model.fit(train_inputs['X'],
train_inputs['target'],
batch_size=BATCH_SIZE,
shuffle=False,
epochs=EPOCHS,
#validation_data=(valid_inputs['X'], valid_inputs['target']),
validation_split=0.1,
callbacks=[earlystop],
verbose=1)

Error:

InvalidArgumentError: Expected size[0] in [0, 21], but got 29
[[{<!-- -->{node es_1/Slice_13}} = Slice[Index=DT_INT32, T=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_input_1_0_0, es_1/Slice_13/begin, es_1/Slice_13/size)]]

Any recommendation would be greatly appreciated, thanks!







azure-machine-learning
· 3
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

Hello,

Thanks for reaching out to us. I am trying to reproduce your issue. May I know where are you running your .ipynb file?


Regards,
Yutong

0 Votes 0 ·

HI Yutong,

I am running this notebook in AzureML. I cloned it using the provided HTTPS from the Github link.

0 Votes 0 ·

Hi Yutong,

This notebook (4_ES_RNN.ipynb) runs without any issues with energy data that is hourly provided in the original notebook.

I provided my own dataset which is in daily data so in the TimeSeriesTensor function in utils.py from the common folder, I modified from freq='H' to freq='D' and that's how I received the error when I run model fit code.

If there are any suggestions that would be helpful. I am happy to provide more details if needed.

Thanks in advance.

0 Votes 0 ·

0 Answers