GitBook: [master] 4 pages modified

ml-studio-app · Feb 3, 2020 · ac7a9b0 · ac7a9b0
1 parent 24af72d
commit ac7a9b0
Show file tree

Hide file tree

Showing 4 changed files with 96 additions and 10 deletions.
diff --git a/README.md b/README.md
@@ -9,14 +9,15 @@ description: Free Data Scientists to focus on hard and creative ML tasks
 ## Key features
 
 * **Try locally** before commiting to anything.
-* **Open:** your deployment specs and models are stored in clean format.
+* **Open:** your deployment specs and models are stored in open format.
+* **Reproducibility**: reproducing previous results easily. 
 * **Auto Deployments:** deploy your models in minutes.
 * **Autoscaling:** automatically scales APIs to handle production workloads.
+* **Zero DevOps**: focus on what matters, you w'not have to worry about devops.
 * **CPU / GPU instances:** ML Studio just works on your infrastructure.
 * **Preemptible \(Spot\) instances:** use preemptible VMs for cost savings.
-* **Zero DevOps**: focus on what matters, you w'not have to worry about devops.
 
-## Can I become who I want to be?
+## What is ML Studio?
 
 **ML Studio** is a cloud-native app build on-top of Kubernetes. This means it is cloud-agnostic and can be deployed on its own or to existing clusters. 
 
@@ -28,7 +29,15 @@ If you are already using Git then you are going to fit right in because every ch
 
 All that is open because we do not want to lock you in, we believe that after you try ML Studio you will be hooked. And if you ever feel want to leave you know that all your work and deployments configuration are yours and that you will be taking them with you.
 
-
-
-
+## Roadmap
+
+* **Testing**: add A/B, canaray and shadow testing support.
+* **1-click app installs**: apps like AirFlow, Meltano, Seldon, ...
+* **Drift alerts**: data drift detection alerts.
+* **Pipelines:** simple pipelines tools
+* HTTPS deployments.
+* Better user calibration.
+* Experiments tracking.
+* Multiplce workspaces.
+* [Request a feature](http://bit.ly/ml-studio-feature-suggestions).
 
diff --git a/SUMMARY.md b/SUMMARY.md
@@ -10,11 +10,11 @@
 * [Istio](install/istio.md)
 * [ML Studio](install/ml-studio.md)
 * [Usage](usage.md)
+* [Cleanup](cleanup.md)
 
 ## Deploying a model
 
 * [Untitled](deploying-a-model/untitled.md)
-* [Cleanup](cleanup.md)
 * [Demo Workspace](http://examples.mlstudioapp.com)
 * [Feature Requests](http://bit.ly/ml-studio-feature-suggestions)
 * [Survey](http://bit.ly/ml-survey-2020)

diff --git a/install/create-a-cluster.md b/install/create-a-cluster.md
@@ -4,6 +4,67 @@ description: Skip this step if you already have a cluster
 
 # Create a cluster
 
+Spinning up a cluster in easy, whether you want to try ML Studio locally or use it in the cloud.
+
+## Cloud
+
+#### Create a GKE k8s cluster
+
+You need to have Google Cloud SDK for this step, follow the [instructions to install it here](https://cloud.google.com/sdk/docs/downloads-interactive#mac).
+
+```text
+export cluster_name=mlstudio-cluster
+export cluster_zone=us-central1-a
+
+gcloud container clusters create $cluster_name \
+    --machine-type=n1-standard-4 \
+    --num-nodes 1 \
+    --enable-autoscaling --min-nodes 0 --max-nodes 6 \
+    --zone $cluster_zone
+```
+
+For cost savings you can also append `--preemptible` to the previous command. They offer the same machine types and options as regular compute instances and last for up to 24 hours.
+
+You can also create a GPU accelerated cluster by appending `--accelerator type=nvidia-tesla-t4,count=1` to the previous command. And then creating a `DaemonSet` to instal Nvidia drivers.
+
+```text
+kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/master/nvidia-driver-installer/cos/daemonset-preloaded.yaml
+```
+
+So the final commands to create a preemptible GPU accelerated cluster would look like:
+
+```text
+export cluster_name=mlstudio-cluster
+export cluster_zone=us-central1-a
+
+gcloud container clusters create $cluster_name \
+     --machine-type=n1-standard-4 \
+     --accelerator type=nvidia-tesla-t4,count=1 \
+     --num-nodes 1 \
+     --enable-autoscaling --min-nodes 0 --max-nodes 6 \
+     --zone $cluster_zone
+     --preemptible
+
+# Install NVIDIA GPU device drivers
+kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/master/nvidia-driver-installer/cos/daemonset-preloaded.yaml
+```
+
+## Local
+
+#### Create a [Minikube](https://minikube.sigs.k8s.io) k8s cluster
+
+```text
+minikube start --cpus 5 --memory 10096
+```
+
+#### Docker Desktop
+
+Download Docker Desktop application and follow the instructions on enabling Kuberentes.
+
+
+
+
+
 
 
 ```text
@@ -54,3 +115,7 @@ gcloud container clusters get-credentials $cluster_name --zone $cluster_zone
 ```
 ```
 
+
+
+
+
diff --git a/prerequisites.md b/prerequisites.md
@@ -1,6 +1,18 @@
 # Prerequisites
 
-1. A cluster, local or remote.
-2. HELM \(ML Studio has been tested with HELM 2.x\)
-3. Istio
+1. A **Kubernetes cluster**, local or remote.
+2. **HELM** \(ML Studio has been tested with HELM 2.x\)
+3. **Istio**
+
+## To try ML Studio locally
+
+To try ML Studio locally
+
+You need a local Kubernetes cluster with at least the following specs.
+
+* 5 CPUs
+* 8 Gb of RAM
+* 15 Gb of fee desk space
+
+