Embeddings normalization fixes (#14284)

* Use cosine distance metric for vec tables

* Only apply normalization to multi modal searches

* Catch possible edge case in stddev calc

* Use sigmoid function for normalization for multi modal searches only

* Ensure we get model state on initial page load

* Only save stats for multi modal searches and only use cosine similarity for image -> image search

frigate/db/sqlitevecq.py

@@ -42,12 +42,12 @@ class SqliteVecQueueDatabase(SqliteQueueDatabase):
         self.execute_sql("""
             CREATE VIRTUAL TABLE IF NOT EXISTS vec_thumbnails USING vec0(
                 id TEXT PRIMARY KEY,
-                thumbnail_embedding FLOAT[768]
+                thumbnail_embedding FLOAT[768] distance_metric=cosine
             );
         """)
         self.execute_sql("""
             CREATE VIRTUAL TABLE IF NOT EXISTS vec_descriptions USING vec0(
                 id TEXT PRIMARY KEY,
-                description_embedding FLOAT[768]
+                description_embedding FLOAT[768] distance_metric=cosine
             );
         """)