Speech to text for speech watermark

Sir, there is problem that

when i am save my speech into text file value of it is in minus and too small. some of the text value is written below:

 0.0000000e+000
 -3.0517578e-005
  0.0000000e+000
   3.0517578e-005
  3.0517578e-005
  0.0000000e+000
  0.0000000e+000
  0.0000000e+000
 -3.0517578e-005
  3.0517578e-005
-2.1362305e-004
 -1.8310547e-004
 -1.8310547e-004
 -2.1362305e-004
 -2.7465820e-004
 -2.7465820e-004
 -3.0517578e-004
 -3.3569336e-004
 -3.0517578e-004
 -2.4414063e-004
 -2.4414063e-004
 -2.1362305e-004
 -2.4414063e-004
 -3.0517578e-004
 -3.9672852e-004
 -4.2724609e-004
 -3.9672852e-004
 -3.6621094e-004
 -3.0517578e-004
 -2.7465820e-004
 -2.7465820e-004
 -3.0517578e-004
 -3.0517578e-004
 -3.6621094e-004
 -3.9672852e-004
 -3.6621094e-004
 -3.6621094e-004
 -3.3569336e-004
 -3.3569336e-004
Now problem is that how can i read that text value in verilog because that value is in floting type. and verilog can't read it.

I am thinking on take unsigned integer for that value but when i am trying same because of small value it shows output as 0, so any other option available for the same.

My project is on speech watermarking on FPGA. So if you have any idea suggest me

[Walter Roberson]

How are you doing the writing at the moment?

Does verilog only accept unsigned integers?