Merge pull request 'Readme with instruction for Gcloud' (#1) from impl/sky/cloud-deploy_preps into main

Reviewed-on: https://git.myzel.io/PyMC/PyMC-Server/pulls/1

Author: chris-aeviator

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "nutpie"]
+	path = nutpie
+	url = https://github.com/pymc-devs/nutpie

README.md

@@ -0,0 +1,48 @@
+# Quickstart
+
+## Installation
+
+Install this library from PiPy `pip3 install pymc-server`
+
+### Setting up a cloud: Google Cloud
+
+1. Run `pymcs check` to see if you already have credentials setup. If you see a green checkmark for GCP (Google Cloud Platform), skip to the next section
+2. Install the google cloud SDK and authenticate.
+  ```bash
+  conda install -c conda-forge google-cloud-sdk
+  gcloud init
+
+  # Run this if you don't have a credentials file.
+  # This will generate ~/.config/gcloud/application_default_credentials.json.
+  gcloud auth application-default login
+  ```
+  > Tip:
+  > If you are using multiple GCP projects, list all the projects by gcloud projects list and activate one by gcloud config set project <PROJECT_ID> (see GCP docs).
+
+3. Follow the link to Google and authorize the Google Cloud SDK. This will create and persist credentials on your local computer. Running pymc-server commands via Google Cloud will introduce cost to your Google Cloud bill according to the VMs provisioned by your configuration (see next section).
+
+  ![Gcloud SDK Auth](./assets/gcloud_auth.png)
+
+  Follow through the rest of the GCP configuration in your terminal.
+
+  > Tip:
+  > You can see a global status of active VMs by running `pymcs status`
+
+
+
+### Status of your deployments
+
+Running `pymcs status` all your deployments are checked and displayed in a table.
+
+```bash
+❯ pymcs status
+Clusters
+NAME  LAUNCHED      RESOURCES                 STATUS  AUTOSTOP  COMMAND
+tc    5 months ago  1x Kubernetes(2CPU--2GB)  UP      -         pymcs launch -c tc hello_sk...
+
+Managed jobs
+No in-progress managed jobs. (See: pymcs jobs -h)
+
+Services
+No live services. (See: pymcs serve -h)
+```

assets/gcloud_auth.png

@@ -3,3 +3,5 @@ nutpie
 lifetimes
 numba
 ray[default]
+skypilot
+skypilot[gcp]

examples/other_nuts_samplers.ipynb

@@ -0,0 +1,56 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+
+# You can run this code outside of the Ray cluster!
+import ray
+
+# Starting the Ray client. This connects to a remote Ray cluster.
+#  ray.init(_node_ip_address='192.168.88.197')
+ray.init('ray://192.168.88.197:10002')
+
+
+# Normal Ray code follows
+@ray.remote
+def do_work(x):
+    return x ** x
+
+
+@ray.remote
+def fit_model():
+    import arviz as az
+    import matplotlib.pyplot as plt
+    from lifetimes.datasets import load_cdnow_summary
+
+    from pymc_marketing import clv
+
+    az.style.use("arviz-darkgrid")
+    plt.rcParams["figure.figsize"] = [12, 7]
+    plt.rcParams["figure.dpi"] = 100
+    plt.rcParams["figure.facecolor"] = "white"
+
+    df = (
+        load_cdnow_summary(index_col=[0])
+        .reset_index()
+        .rename(columns={"ID": "customer_id"})
+    )
+    sampler_kwargs = {
+        "draws": 2_000,
+        "target_accept": 0.9,
+        "chains": 5,
+        "random_seed": 42,
+    }
+
+    model = clv.BetaGeoModel(data=df)
+    idata_nutpie = model.fit(nuts_sampler="nutpie", **sampler_kwargs)
+    return idata_nutpie
+
+object_ref = fit_model.remote()
+x = ray.get(object_ref)
+print(x)
+
+"""
+object_refs = do_work.remote(2)
+x = ray.get(object_refs)
+"""
+
+