cleanup instant vector search recipe

hypermodeinc · Dec 9, 2024 · 0eed6bf · 0eed6bf
1 parent 740639c
commit 0eed6bf
Show file tree

Hide file tree

Showing 3 changed files with 30 additions and 14 deletions.
diff --git a/README.md b/README.md
@@ -4,7 +4,7 @@ Code recipes for cooking with [Modus](https://github.com/hypermodeinc/modus), th
 
 ## 🚀 Getting started with Modus
 
-For more infromation on getting started with Modus, check out [the docs](https://docs.hypermode.com/modus/overview).
+For more information on getting started with Modus, check out [the docs](https://docs.hypermode.com/modus/overview).
 
 ## 📺 Video walkthroughs
 
@@ -16,5 +16,6 @@ The following recipes have associated recorded content:
 | [modus-getting-started](modus-getting-started/) | [Getting Started With Modus video](https://www.youtube.com/watch?v=3CcJTXTmz88)                                                                    |
 | [modushack-data-models](modushack-data-models/) | [ModusHack: Working With Data & AI Models livestream](https://www.youtube.com/watch?v=gB-v7YWwkCw&list=PLzOEKEHv-5e3zgRGzDysyUm8KQklHQQgi&index=3) |
 | [modus-press](modus-press/)                     | Coming soon                                                                                                                                        |
-| [dgraph-101](dgraph-101/)      | [Working with Dgraph in Modus](https://youtu.be/Z2fB-nBf4Wo)  |
-| [function-calling](function-calling/)                     | [Use LLM function calling aka tools in Modus.](https://youtu.be/afFk7JzSIm0)                                                                                                                                       |
+| [dgraph-101](dgraph-101/)                       | [Working with Dgraph in Modus](https://youtu.be/Z2fB-nBf4Wo)                                                                                       |
+| [function-calling](function-calling/)           | [Use LLM function calling aka tools in Modus.](https://youtu.be/afFk7JzSIm0)                                                                       |
+| [instant-vector-search](instant-vector-search)  | [Instant vector search with Modus & Hypermode](https://www.youtube.com/watch?v=4H_xPTUbwL8)                                                        |
diff --git a/instant-vector-search/README.md b/instant-vector-search/README.md
@@ -1,32 +1,32 @@
 # instant-vector-search
 
-A simplified template demonstrating how to build instant vector search using Hypermode and Modus. This project showcases how to configure collections, embed data, and deploy APIs to perform real-time vector search with minimal setup.
+A simplified example demonstrating how to build instant vector search using Hypermode and Modus. This project showcases how to configure collections, embed data, and deploy APIs to perform real-time vector search with minimal setup.
 
 ## Guide
 
-Check out the guide on how to create a project using this template:
-[Instant Vector Search Guide](TODO)
+Check out the blog post on how to create a project using this example:
+[How to Build an Instant Vector Search App using Hypermode & Modus](https://hypermode.com/blog/instant-vector-search-guide)
 
 ## Semantic Search in Action
 
-This template demonstrates how instant vector search can enhance your application by generating embeddings from user queries and performing vector searches in less than 200ms.
+This example demonstrates how instant vector search can enhance your application by generating embeddings from user queries and performing vector searches in less than 200ms.
 
 ## How to Use the Template
 
 Fork or clone this repository to get started:
 
-```
-git clone https://github.com/hypermodeinc/instant-vector-search.git
-cd instant-vector-search
+```bash
+git clone https://github.com/hypermodeinc/modus-recipes.git
+cd modus-recipes/instant-vector-search
 ```
 
 Install Modus CLI:
 
-```
+```bash
 npm install -g @hypermode/modus-cli
 ```
 
-## Initialize and deploy your app:
+## Initialize and deploy your app
 
 You can deploy using either the Hypermode Console or Hyp CLI.
 
@@ -37,9 +37,15 @@ Option 1: Hypermode Console
 
 Option 2: Hyp CLI
 
-- Run the following command to import and deploy your app:
+- Install the Hyp CLI
 
+```bash
+npm install -g @hypermode/hyp-cli
 ```
+
+- Run the following command to import and deploy your app:
+
+```bash
 hyp init
 ```
 
@@ -57,7 +63,7 @@ Once your data is in place, the embeddings will be generated automatically, and
 
 You can test the API via the Query page in the Hypermode Console. Run a sample vector search query:
 
-```
+```GraphQL
 query {
   search(query: "your search term here")
 }

diff --git a/instant-vector-search/assembly/index.ts b/instant-vector-search/assembly/index.ts
@@ -6,6 +6,9 @@ const textsCollection = "texts";
 const searchMethod = "searchMethod1";
 const embeddingModelName = "minilm";
 
+/**
+ * Add text(s) to the collection
+ */
 export function upsertTexts(ids: string[], texts: string[]): string[] {
   const errors: string[] = [];
 
@@ -24,6 +27,9 @@ export function upsertTexts(ids: string[], texts: string[]): string[] {
   return ids;
 }
 
+/**
+ * Perform a vector search using an embedding of the input string
+ */
 export function search(query: string): string[] {
   const searchResults = collections.search(
     textsCollection,
@@ -48,6 +54,9 @@ export function search(query: string): string[] {
   return searchTexts;
 }
 
+/**
+ * Embed the input text(s) using the miniLM embedding model
+ */
 export function miniLMEmbed(texts: string[]): f32[][] {
   const model = models.getModel<EmbeddingsModel>(embeddingModelName);
   const input = model.createInput(texts);