updated cluser compute, environment and workloads

README.md

@@ -83,7 +83,7 @@ We'll start by setting up our cluster with the environment and compute configura
   - Project: `madewithml`
   - Cluster environment name: `madewithml-cluster-env`
   # Toggle `Select from saved configurations`
-  - Compute config: `madewithml-cluster-compute`
+  - Compute config: `madewithml-cluster-compute-g5.4xlarge`
   ```
 
   > Alternatively, we can use the [CLI](https://docs.anyscale.com/reference/anyscale-cli) to create the workspace via `anyscale workspace create ...`
@@ -423,7 +423,7 @@ anyscale cluster-env build deploy/cluster_env.yaml --name $CLUSTER_ENV_NAME
 The compute configuration determines **what** resources our workloads will be executes on. We've already created this [compute configuration](./deploy/cluster_compute.yaml) for us but this is how we can create it ourselves.
 
 ```bash
-export CLUSTER_COMPUTE_NAME="madewithml-cluster-compute"
+export CLUSTER_COMPUTE_NAME="madewithml-cluster-compute-g5.4xlarge"
 anyscale cluster-compute create deploy/cluster_compute.yaml --name $CLUSTER_COMPUTE_NAME
 ```
 

deploy/cluster_compute.yaml

@@ -1,12 +1,12 @@
-cloud: madewithml-us-east-2
-region: us-east2
+cloud: education-us-west-2
+region: us-west-2
 head_node_type:
   name: head_node_type
-  instance_type: m5.2xlarge  # 8 CPU, 0 GPU, 32 GB RAM
+  instance_type: g5.4xlarge
 worker_node_types:
 - name: gpu_worker
-  instance_type: g4dn.xlarge  # 4 CPU, 1 GPU, 16 GB RAM
-  min_workers: 0
+  instance_type: g5.4xlarge
+  min_workers: 1
   max_workers: 1
   use_spot: False
 aws:

deploy/cluster_env.yaml

@@ -1,4 +1,4 @@
-base_image: anyscale/ray:2.6.0-py310-cu118
+base_image: anyscale/ray:2.7.0optimized-py310-cu118
 env_vars: {}
 debian_packages:
   - curl

deploy/jobs/workloads.sh

@@ -1,6 +1,5 @@
 #!/bin/bash
 export PYTHONPATH=$PYTHONPATH:$PWD
-export RAY_AIR_REENABLE_DEPRECATED_SYNC_TO_HEAD_NODE=1
 mkdir results
 
 # Test data