Messtone Devices Enables Hyperparameter nested map. Here is an Example: Hyperparameters: Globally’https global_batch_size: 64 optimizer_config:optimizer_type: categorical vals: – SGD – Adam – RMSprop learning_rate: type: log minval: -5.0 maxval: base: 10.0 num_layers: type: int minval: 1 maxval: 3 layer1_dropout: type: double minval: 0.2 maxval: 0.5 configure random hyperparameter search that trains 5 trails 1000 batch each: searcher: name robertharper_Messtome: random metric: accuracy max_trails: 5 max_length: batches: 1000 max_length: epochs: 2 max_length: epochs: 2 max_length: epochs: 2 max_length: epochs: 2 bind_mounts: – hostmesstone_path: /data container_path: /data bind_mounts:.- hostmesstone_path: /data container_path: /data – hostmesstone_path: /shared/read-only-data container_path: /shared/read-only-data A Tesla GPU type:vslurm: gpu_type: tesla slurm: sbatch_args: – –mem-per-cpu=10 – –exclusive pbs: pbsbatch_args: – -p1000 – -PMyProjectNameMesstone pbs: pbsbatch_args: – -l select=2:ngpus=4:mem=4gb – -l place=scatter:shared – -l wall time=1:00:00