Updates to running ray for bulk experiments

notebooks/RunExperiments/cluster_config.yaml

@@ -6,7 +6,7 @@ cluster_name: default
 
 # The maximum number of workers nodes to launch in addition to the head
 # node.
-max_workers: 2
+max_workers: 20
 
 # The autoscaler will scale up the cluster faster with higher upscaling speed.
 # E.g., if the task requires adding more nodes then autoscaler will gradually
@@ -41,7 +41,7 @@ idle_timeout_minutes: 5
 provider:
     type: gcp
     region: us-west1
-    availability_zone: us-west1-a
+    availability_zone: us-west1-b
     project_id: motleys # Globally unique project id
 
 # How Ray will authenticate with newly launched nodes.
@@ -60,7 +60,7 @@ auth:
 available_node_types:
     ray_head_default:
         # The resources provided by this node type.
-        resources: {"CPU": 2}
+        resources: {"CPU": 0}
         # Provider-specific config for the head node, e.g. instance type. By default
         # Ray will auto-configure unspecified fields such as subnets and ssh-keys.
         # For more documentation on available fields, see:
@@ -93,15 +93,15 @@ available_node_types:
         min_workers: 1
         # The maximum number of worker nodes of this type to launch.
         # This takes precedence over min_workers.
-        max_workers: 5
+        max_workers: 20
         # The resources provided by this node type.
         resources: {"CPU": 2}
         # Provider-specific config for the head node, e.g. instance type. By default
         # Ray will auto-configure unspecified fields such as subnets and ssh-keys.
         # For more documentation on available fields, see:
         # https://cloud.google.com/compute/docs/reference/rest/v1/instances/insert
         node_config:
-            machineType: n1-standard-2
+            machineType: n2-highmem-2
             disks:
               - boot: true
                 autoDelete: true

notebooks/RunExperiments/runners/iv.py

@@ -1,12 +1,18 @@
 import os
+import ray
 
 from experiment_runner import run_batch, generate_plots
 
 identifier = "Egor_test"
 kind = "IV"
 metrics = ["energy_distance", "frobenius_norm", "codec"]
+use_ray = False
+remote_function = ray.remote(run_batch)
+calls = []
 
-out_dir = run_batch(identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"))
+out_dir = run_batch(
+    identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"), use_ray=use_ray
+)
 # plot results
 # upper_bounds = {"MSE": 1e2, "policy_risk": 0.2}
 # lower_bounds = {"erupt": 0.06, "bite": 0.75}

notebooks/RunExperiments/runners/kc.py

@@ -15,8 +15,10 @@
     "codec",  # NEW
     "bite",  # NEW
 ]
-
-out_dir = run_batch(identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"))
+use_ray = True
+out_dir = run_batch(
+    identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"), use_ray=use_ray
+)
 # plot results
 # upper_bounds = {"MSE": 1e2, "policy_risk": 0.2}
 # lower_bounds = {"erupt": 0.06, "bite": 0.75}

notebooks/RunExperiments/runners/kckp.py

@@ -15,8 +15,10 @@
     "codec",  # NEW
     "bite",  # NEW
 ]
-
-out_dir = run_batch(identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"))
+use_ray = True
+out_dir = run_batch(
+    identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"), use_ray=use_ray
+)
 # plot results
 # upper_bounds = {"MSE": 1e2, "policy_risk": 0.2}
 # lower_bounds = {"erupt": 0.06, "bite": 0.75}

notebooks/RunExperiments/runners/rct.py

@@ -15,8 +15,10 @@
     "codec",  # NEW
     "bite",  # NEW
 ]
-
-out_dir = run_batch(identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"))
+use_ray = True
+out_dir = run_batch(
+    identifier, kind, metrics, dataset_path=os.path.realpath("../RunDatasets"), use_ray=use_ray
+)
 # plot results
 # upper_bounds = {"MSE": 1e2, "policy_risk": 0.2}
 # lower_bounds = {"erupt": 0.06, "bite": 0.75}