#Commands to produce the plots in the paper
The commands below are based on the results generated in our experiments and is specific to the tag naming we used. Different tag names require adjustment of the commands
Fig2: python plots/plot_exp.py 'safa1' google_speech_resnet34 'fig1' 'Test' 0 1000
Fig3: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'plot_motive1' 'Test' 1 10
Fig4: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'plot_motive2' 'Test' 1 10
Selection Methods: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'avail_yogi' 'Test' 1 10
Staleness Aggregation: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'stale_yogi' 'Test' 1 10
Compare with Safa - Fig 10: python plots/plot_exp.py 'safa' google_speech_resnet34 'fig2' 'Test' 0 1000
Selection Adaptation (APT) - Fig 11: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'avail_selectadapt' 'Test' 0 50
Oort-Future: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'plot_scalesyspercent' 'Test' 1 10 1 oort
REFL-Future: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'plot_scalesyspercent' 'Test' 1 10 1 relay
Safa-large: python plots/plot_exp.py 'safa_large' google_speech_resnet34 'fig3' 'Test' 0 3000
stale update methods: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'avail_yogi_scale' 'Test' 1 10
avail_yogi_converge - Fig 9: python plots/plot_exp.py 'exp_type' google_speech_resnet34 'avail_yogi_converge' 'Test' 1 10
Reddit: python plots/plot_exp.py 'exp_type' reddit_albert-base-v2 'plot_avail' 'Test' 1 10
Stackoverflow: python plots/plot_exp.py 'exp_type' stackoverflow_albert-base-v2 'plot_avail' 'Test' 1 10
openimg: python plots/plot_exp.py 'exp_type' openimg_shufflenet 'plot_avail' 'Test' 1 10 None None 1.0
cifar10: python plots/plot_exp.py 'exp_type' cifar10_resnet18 'fedavg_avail' 'Test' 1 10