I would like to know that instead of using the file id in the api get request can I use the filename as (xxxx.xls)
You can specify the import data source or export name in the URL path instead of the numeric ID, although you may need to be careful with certain characters that must be encoded (the tool you use to access the API may handle this).
Use application/octet-stream for the Accept header.
If the file is large then use the /2/0 API or the chunks endpoints instead.
It looks like you're missing the path component "/files", which should follow the model ID. Test using a numeric identifier first, then substitute the name for the numeric identifier.
Now you've got /1/4 as a version instead of /1/3 or /2/0. That's not a valid API version, hence the not found error.