Add `num_processes` to `simulation_info.json`

[robgjansen]

During the parsing phase, tornettools will store some useful context info into simulation_info.json, for example:

{
  "net_scale": 0.05,
  "num_circuits_ten_minutes": 74448,
  "num_dir_authorities": 3,
  "num_emulated_users": 39597,
  "num_public_relays": 6910,
  "num_sampled_relays": 346,
  "num_tgen_markov_clients": 396,
  "num_tgen_perf_clients": 40,
  "num_tgen_servers": 40,
  "tornettools_generate_seed": 1191921776
}

We should add a num_processes field to this json file so we can quickly find how many Linux processes will be used during the sim.