The log file you shared is a bit odd, because it doesn't seem to contain the command line that you used (this would typically appear in the first few lines of the file). Did you remove it? Can you send the command line that you used?
My first suggestion would be to use param-medic to make sure you are using the right parameters for your data. But the out-of-range peaks are very strange, so I wonder if something else is going on. You might also consider using --spectrum-parser mstoolkit and see if that makes a difference.