Adding perf metrics to each node in both single and multi node scripts

[jfomhover]

This PR connects the dots between multiple precedent PRs:

• Add a perf metric collector to ensure number of measures stays under control #193
• Create performance reporting thread for single node components #189
• Remove singleton from MetricsLogger #190

It implements the performance metrics in single node and multi node scripts. Those metrics represent performance summary per node. When running on distributed, we use step to store metrics under one metric name for all nodes.

This is what those metrics look like in AzureML:

Here's an example of multinode perf (extract)

[github-actions]

Unit Test Results for Build

  1 files  ±0    1 suites  ±0   22s ⏱️ +7s
82 tests ±0  82 ✔️ ±0  0 💤 ±0  0 ❌ ±0 

Results for commit 5493c75. ± Comparison against base commit c7298fa.

♻️ This comment has been updated with latest results.

[github-actions]

Package	Line Rate	Branch Rate	Complexity
common	84%	0%	0
scripts	100%	0%	0
scripts.data_processing.generate_data	93%	0%	0
scripts.data_processing.lightgbm_data2bin	95%	0%	0
scripts.data_processing.partition_data	92%	0%	0
scripts.inferencing.custom_win_cli	65%	0%	0
scripts.inferencing.lightgbm_c_api	35%	0%	0
scripts.inferencing.lightgbm_python	95%	0%	0
scripts.inferencing.treelite_python	94%	0%	0
scripts.model_transformation.treelite_compile	92%	0%	0
scripts.sample	93%	0%	0
scripts.training.lightgbm_python	80%	0%	0
Summary	81% (1057 / 1303)	0% (0 / 0)	0