From 1c02c8d9606bcb45340cb300e4619606774d308c Mon Sep 17 00:00:00 2001
From: Cheng <git@zcbenz.com>
Date: Sun, 15 Sep 2024 20:07:45 +0900
Subject: [PATCH] Update README with screencast

---
 README.md  | 14 ++++++++++++--
 src/cli.ts |  2 +-
 2 files changed, 13 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index eeb0aac..910b3cd 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,12 @@
 # Semantic Image Search CLI (sisi)
 
-:construction:
-
 CLI tool for semantic image search, locally without Internet.
 
 Powered by [node-mlx](https://github.com/frost-beta/node-mlx), a machine
 learning framework for Node.js.
 
+https://github.com/user-attachments/assets/66e6e437-c27b-48cf-80cc-a5a0c8c0bdfb
+
 ## Supported platforms
 
 GPU support:
@@ -87,6 +87,16 @@ It works with local files too:
 sisi search file:///Users/Your/Pictures/cat.jpg
 ```
 
+## Under the hood
+
+The index is built by computing the embeddings of images using the [CLIP
+model](https://github.com/openai/CLIP), and then stored in a binary JSON file.
+
+Searching the images is computing cosine similarities between the query string
+and the indexed embeddings. There is no database involved here, everytime you do
+a search the computation is done for all the embeddings stored, which is very
+fast even when you have tens of thousands of pictures.
+
 ## License
 
 MIT
diff --git a/src/cli.ts b/src/cli.ts
index bbeabe8..0fce6ae 100644
--- a/src/cli.ts
+++ b/src/cli.ts
@@ -51,7 +51,7 @@ export class SearchCommand extends Command {
 
   query = Option.String();
   target = Option.String('--in', {description: 'The directory where images are searched.'});
-  max = Option.String('--max', 20, {description: 'The maximum number of results to return.'});
+  max = Option.String('--max', '20', {description: 'The maximum number of results to return.'});
   print = Option.Boolean('--print', {description: 'Print the results to stdout.'});
 
   async execute() {