TooManyInvalidError: Maximum number of invalid lines reached!

[FirasNacef]

Hello,

I am trying to generate new lines for my dataframe which contains survey responses (So categorical variables that are not yet encoded). I ran the "synthetic_records" notebook hoping that it would create new responses that are similar but not identical to the ones already existing. When training, the RNN reached 91% accuracy, but when running the lines generating code the output kept giving me a list of lines in this format :

GenText(valid=False, text ="value1,value2,value3,etc...", explain='record not 6 parts', delimiter=',')

Here's the error message I got :

[johntmyers]

Hi @FirasNacef can you share more of the full traceback? Also this error usually means that the model cannot generate the expected number of columns, which could stem from a variety of issues like sparse columns. If you can share your dataset that would be helpful too.

[FirasNacef]

poll3.csv

You will find attached the dataset I used.

[zredlined]

Hey @FirasNacef - try running ACTGAN, it should work just fine for your dataset (the LSTM model you were using needed more examples to learn the structure and format).

https://github.com/gretelai/gretel-synthetics/blob/master/examples/ACTGAN_Demo.ipynb