[BMs] Make it easy to reproduce multisim benchmarking plots

[graceful-coder]

Background / motivation

Plots improve the understanding of ML model benchmarks for Predictoor. Alternative tools like Codepen.io are difficult to integrate into pdr-backend, while Plotly is an easy addition to pdr-backend with the option to plot live multisim data.

TODOs

• [ ] Implement Python Plotly to generate plots for benchmarks of different ML models.

[graceful-coder]

• Created a new benchmarks dir under the pdr-backend/pdr_backend dir
• Initially committed three plot files, one for each model.
• Reduced DRY violations by consolidating into one simpler file and deleted the three previous files.

[trentmc]

done