How can I read the text files faster and process the data using tokenizer?

Question

I am reading a text file with more than 64k lines of integer data placed in the form of columns. What I was doing earlier is reading the file line by line using wxTextInputStream (as it was wxWidgets application) and next from the read line , I was reading each integer using wxStringTokenizer tkz(line, separator).But this method is taking 5 min for reading the file and storing data into vectors. Can you suggest and better way to reduce the time of reading the file line by line and better usage of tokenizer.

 wxFileInputStream input(filepath);
 wxTextInputStream textInput(input, wxT("\x09"), wxConvUTF8);

 while (input.IsOk() && !input.Eof())
{
    line = textInput.ReadLine(); // read a text line
    wxStringTokenizer  tkz(line, separator);
    while (tkz.HasMoreTokens())
    {
        if (lineCount < 2)
        {
            data_set.headers.push_back(string(tkz.GetNextToken().mb_str()));
            columns++;
        }
        else if (lineCount > 2) {

            tkz.GetNextToken().ToDouble(&ver);
            versionNumbers.push_back(ver);
            dataPoints++;
        }
        else tkz.GetNextToken();
    }
    lineCount++;
}