fmt, add env vars

templates/rag-chroma-multi-modal-multi-vector/README.md

@@ -1,30 +1,30 @@
 
-# rag-chroma-multi-modal
+# rag-chroma-multi-modal-multi-vector
 
 Presentations (slide decks, etc) contain visual content that challenges conventional RAG.
 
 Multi-modal LLMs unlock new ways to build apps over visual content like presentations.
 
 This template performs multi-modal RAG using Chroma with the multi-vector retriever (see [blog](https://blog.langchain.dev/multi-modal-rag-template/)):
 
-* Extract the slides as images
-* Use GPT-4V to summarize each image
-* Embed the image summaries with a link to the original images
-* Retrieve relevant image based on similarity between the image summary and the user input
+* Extracts the slides as images
+* Uses GPT-4V to summarize each image
+* Embeds the image summaries with a link to the original images
+* Retrieves relevant image based on similarity between the image summary and the user input
 * Finally pass those images to GPT-4V for answer synthesis
 
 ## Storage
 
-We will use Upstash to store the images.
+We will use Upstash to store the images, which offers Redis with a REST API.
 
 Simply login [here](https://upstash.com/) and create a database.
 
-This will give you: 
+This will give you a REST API with:
 
 * UPSTASH_URL
 * UPSTASH_TOKEN
 
-Set these in chain.py (***TODO: Update this? Env var?***)
+Set `UPSTASH_URL` and `UPSTASH_TOKEN` as environment variables to access your database.
 
 We will use Chroma to store and index the image summaries, which will be created locally in the template directory.
 
@@ -47,6 +47,8 @@ The app will retrieve images using multi-modal embeddings, and pass them to GPT-
 
 Set the `OPENAI_API_KEY` environment variable to access the OpenAI GPT-4V.
 
+Set `UPSTASH_URL` and `UPSTASH_TOKEN` as environment variables to access your database.
+
 ## Usage
 
 To use this package, you should first have the LangChain CLI installed:

templates/rag-chroma-multi-modal-multi-vector/ingest.py

@@ -1,5 +1,6 @@
 import base64
 import io
+import os
 import uuid
 from io import BytesIO
 from pathlib import Path
@@ -64,8 +65,8 @@ def generate_img_summaries(img_base64_list):
         try:
             image_summaries.append(image_summarize(base64_image, prompt))
             processed_images.append(base64_image)
-        except:
-            print(f"BadRequestError with image {i+1}")
+        except Exception as e:
+            print(f"Error with image {i+1}: {e}")
 
     return image_summaries, processed_images
 
@@ -136,8 +137,8 @@ def create_multi_vector_retriever(vectorstore, image_summaries, images):
     """
 
     # Initialize the storage layer for images
-    UPSTASH_URL = "https://usw1-bright-beagle-34178.upstash.io"
-    UPSTASH_TOKEN = "AYWCACQgNzk3OTJjZTItMGIxNy00MTEzLWIyZTAtZWI0ZmI1ZGY0NjFhNGRhMGZjNDE4YjgxNGE4MTkzOWYxMzllM2MzZThlOGY="
+    UPSTASH_URL = os.getenv("UPSTASH_URL")
+    UPSTASH_TOKEN = os.getenv("UPSTASH_TOKEN")
     store = UpstashRedisByteStore(url=UPSTASH_URL, token=UPSTASH_TOKEN)
     id_key = "doc_id"