I am trying to use data i have in CSV files for training. To use them I'm following these instructions: https://www.tensorflow.org/api_guides/python/reading_data#QueueRunner
since my data are no integers but decimals i have changed
record_defaults = [[1], [1], [1], [1], [1]]
to record_defaults = [[1.], [1.], [1.], [1.], [1.]]
so it accepts floats.
I still get following error though:
tensorflow.python.framework.errors_impl.InvalidArgumentError: Field 0 in record 0 is not a valid float: 15.5
[[Node: DecodeCSV = DecodeCSV[OUT_TYPE=[DT_FLOAT, ... , DT_FLOAT], field_delim=";", na_value="", select_cols=[], use_quote_delim=true, _device="/job:localhost/replica:0/task:0/device:CPU:0"](ReaderReadV2:1, DecodeCSV/record_defaults_0, ... , DecodeCSV/record_defaults_0)]]
Which doesn't make sense to me, 15.5 is in fact the first cell from my CSV file and if it isn't a float, what else?
any help on how to eat with this would be greatly appreciated