To be valid audio sample values, the data needs to be normalized to the range +/- 1.0.
Your largest absolute value is 203913, so you could divide each value by this number prior to exporting, or, because Audacity works in 32-bit float format and can handle values over 0 dB, you could import the data first and normalize after.
Because these are text (ASCII) values and not actual audio data (binary data), normal audio import will not work, and neither will “import RAW”. You need to read in the values as numeric text, which you can do with the “Sample Data Import” generator: http://manual.audacityteam.org/man/sample_data_import.html
If you want to import first and then normalize, use the “Normalize” effect after importing http://manual.audacityteam.org/man/normalize.html
