From 188d39f61eef1ceb550ef2593c6dbe7681869a46 Mon Sep 17 00:00:00 2001
From: Florents Tselai <florents.tselai@gmail.com>
Date: Sun, 30 Jun 2024 21:42:38 +0300
Subject: [PATCH] 16 binary embedding for blob types eg using llm clip (#21)

* Test binary with hazo

* Add image embedding example
---
 README.md            | 17 +++++++++++++++++
 setup.py             |  2 +-
 tests/test_tsellm.py | 17 +++++++++++------
 3 files changed, 29 insertions(+), 7 deletions(-)

diff --git a/README.md b/README.md
index cdfe13c..c9a993e 100644
--- a/README.md
+++ b/README.md
@@ -16,6 +16,8 @@ pip install tsellm
 Behind the scenes, **tsellm** is based on the beautiful [llm](https://llm.datasette.io) library,
 so you can use any of its plugins:
 
+## Generative
+
 For example, to access `gpt4all` models
 
 ```shell
@@ -33,6 +35,21 @@ llm sentence-transformers register all-MiniLM-L12-v2
 tsellm :memory: "select embed('Hello', 'sentence-transformers/all-MiniLM-L12-v2')"
 ```
 
+### Embeddings for binary (`BLOB`) columns
+
+```shell
+wget https://tselai.com/img/flo.jpg
+sqlite3 images.db <<EOF
+CREATE TABLE images(name TEXT, type TEXT, img BLOB);
+INSERT INTO images(name,type,img) VALUES('flo','jpg',readfile('flo.jpg'));
+EOF
+```
+
+```shell
+llm install llm-clip
+tsellm images.db "select embed(img, 'clip') from images"
+```
+
 ## Examples
 
 Things get more interesting if you
diff --git a/setup.py b/setup.py
index ce554a2..4d813e8 100644
--- a/setup.py
+++ b/setup.py
@@ -40,7 +40,7 @@ def get_long_description():
             "ruff",
             "sqlite_utils",
             "llm-markov",
-            "llm-embed-hazo",
+            "llm-embed-hazo==0.2.1",
         ]
     },
     python_requires=">=3.11",
diff --git a/tests/test_tsellm.py b/tests/test_tsellm.py
index cdaa6f2..51a1f51 100644
--- a/tests/test_tsellm.py
+++ b/tests/test_tsellm.py
@@ -95,13 +95,18 @@ def test_embed_hazo(self):
             out,
         )
 
+    def test_embed_hazo_binary(self):
+        self.assertTrue(llm.get_embedding_model("hazo").supports_binary)
+        self.expect_success(":memory:", "select embed(randomblob(16), 'hazo')")
+
+
     def test_embed_default_hazo(self):
-        self.assertEqual(llm_cli.get_default_embedding_model(), "hazo")
-        out = self.expect_success(":memory:", "select embed('hello world')")
-        self.assertEqual(
-            "('[5.0, 5.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]',)\n",
-            out,
-        )
+            self.assertEqual(llm_cli.get_default_embedding_model(), "hazo")
+            out = self.expect_success(":memory:", "select embed('hello world')")
+            self.assertEqual(
+                "('[5.0, 5.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]',)\n",
+                out,
+            )
 
 
 if __name__ == "__main__":