Add option to specify number of job runs on API

[fridex]

As we will use Amun for benchmarks where we will run jobs multiple times, it would be nice to adjust Amun API to specify number of jobs that should be run on user request.

An example - benchmark this TensorFlow stack on the given hardware 3 times. The API will respect this and schedule 3 jobs for this purpose.

There needs to be adjusted:

• Introduce a parameter job_count on user API request.
• Adjust swagger/openAPI specification to accept parameter job_count
• The result of the /inspect endpoint will return a list of job ids that are scheduled.
• Runned jobs will have set label to inspection_id so that these jobs can easily queried

[ace2107]

#55