Extended refactoring of the benchmark pipeline

[SimonGrouard]

Refactoring of the benchmarking pipeline!

To run the benchmark, one should:

• Create their own config.yaml config file within the benchmark_configs folder.
• Every possible config parameter is explained within the @dataclass of the RunBenchmarkConfig inside run_benchmark_config_dataclass.py in which value compatibilities of the different provided arguments are checked.
• To run a benchmark with the created config, run inside a terminal python /home/owkin/deepdeconv-fork/run_benchmark.py --config /home/owkin/deepdeconv-fork/benchmark_configs.config.yaml.
• If inside the config, save: True, then every output of the experiment will be saved inside /home/owkin/project/run_benchmark_experiments/{your_experiment_name}. Otherwise, only final plots will be saved there.

[khalilouardini]

Thank you @SimonGrouard the benchmark is great, that's definitely going to help scale the benchmarking!