Merge pull request #12 from StacklokLabs/rename

Rename to GoRag

Author: lukehinds

.golangci.yml

@@ -11,7 +11,7 @@ linters-settings:
     sections:
       - standard
       - default
-      - prefix(github.com/stackloklabs/gollm)
+      - prefix(github.com/stackloklabs/gorag)
   revive:
     # see https://github.com/mgechev/revive#available-rules for details.
     ignore-generated-header: true

Makefile

@@ -1,7 +1,7 @@
 # Makefile for Go Project
 
 # Project variables
-PROJECT_NAME := gollm
+PROJECT_NAME := gorag
 BINARY_NAME := $(PROJECT_NAME)
 BUILD_DIR := build
 GO_FILES := $(shell find . -name '*.go' -not -path "./vendor/*")

README.md

@@ -1,28 +1,27 @@
-# Gollm: Go Interface for LLM development with RAG 📜
+# GoRag: Go Interface for LLM  / RAG development via Multiple Vector Database types 📜
 
-[![Go Report Card](https://goreportcard.com/badge/github.com/stackloklabs/gollm)](https://goreportcard.com/report/github.com/stackloklabs/gollm)
-[![License](https://img.shields.io/github/license/stackloklabs/gollm)](LICENSE)
+[![Go Report Card](https://goreportcard.com/badge/github.com/stackloklabs/gorag)](https://goreportcard.com/report/github.com/stackloklabs/gorag)
+[![License](https://img.shields.io/github/license/stackloklabs/gorag)](LICENSE)
 
-Gollm is a Go library that provides an easy interface to interact with Large 
+GoRag is a Go library that provides an easy interface to interact with Large 
 Language Model backends including [Ollama](https://ollama.com) and [OpenAI](https://openai.com), along with an embeddings interface for RAG (currently with Postgres pgvector).
 
 
 ## 🌟 Features
 
 - **Interact with Ollama & OpenAI:** Generate responses from multiple AI backends.
 - **RAG / Embeddings Generation:** Generate text embeddings store / load to a vector database for RAG.
-
+- **Multiple Vector Database Support:** Currently Postgres with pgvector is supported, along with qdrant (others to follow, open an issue if you want to see something included).
 ---
 
 ## 🚀 Getting Started
 
-
 ### 1. Installation
 
-gollm needs to be installed as a dependency in your project. You can do it by importing it in your codebase:
+gorag needs to be installed as a dependency in your project. You can do it by importing it in your codebase:
 
 ```go
-import "github.com/stackloklabs/gollm"
+import "github.com/stackloklabs/gorag"
 ```
 
 Then make sure that you have Go installed, and run:
@@ -31,7 +30,6 @@ Then make sure that you have Go installed, and run:
 go mod tidy
 ```
 
-
 ##  2. Setting Up Ollama
 
 You'll need to have an Ollama server running and accessible.
@@ -54,17 +52,17 @@ You'll need an OpenAI API key to use the OpenAI backend.
 ## 4. Configuration
 
 Currently Postgres is supported, and the database should be created before
-running the application, with the schena provided in `db/init.sql`
+running the application, with the schema provided in `db/init.sql`
 
-Should you wish, the docker-compose will automate the setup of the database.
+Should you prefer, the docker-compose will automate the setup of the database.
 
 # 🛠️ Usage
 
 Best bet is to see `/examples/*` for reference, this explains how to use
-the library with examples for generation, embeddings and implementing RAG.
+the library with examples for generation, embeddings and implementing RAG for pgvector or qdrant.
 
 There are currently two backend systems supported, Ollama and OpenAI, with
-the ability to generate embeddings for RAG.
+the ability to generate embeddings for RAG on both.
 
 ## Ollama
 
@@ -139,10 +137,50 @@ if err != nil {
     log.Fatalf("Error generating embedding: %v", err)
 }
 log.Println("Embedding generated")
+
+// Embed the query using the specified embedding backend
+queryEmbedding, err := embeddingBackend.Embed(ctx, query)
+if err != nil {
+    log.Fatalf("Error generating query embedding: %v", err)
+}
+log.Println("Vector embeddings generated")
+
+// Retrieve relevant documents for the query embedding
+retrievedDocs, err := vectorDB.QueryRelevantDocuments(ctx, queryEmbedding, "ollama")
+if err != nil {
+    log.Fatalf("Error retrieving relevant documents: %v", err)
+}
+
+// Log the retrieved documents to see if they include the inserted content
+for _, doc := range retrievedDocs {
+    log.Printf("Retrieved Document: %v", doc)
+}
+
+// Augment the query with retrieved context
+augmentedQuery := db.CombineQueryWithContext(query, retrievedDocs)
+
+prompt := backend.NewPrompt().
+    AddMessage("system", "You are an AI assistant. Use the provided context to answer the user's question as accurately as possible.").
+    AddMessage("user", augmentedQuery).
+    SetParameters(backend.Parameters{
+        MaxTokens:   150, // Supported by LLaMa
+        Temperature: 0.7, // Supported by LLaMa
+        TopP:        0.9, // Supported by LLaMa
+    })
 ```
 
-A database is also required, we have support for PostGres with pgvector. See `/examples/*` 
-for reference.
+Example output:
+
+```
+2024/10/28 15:08:25 Embedding backend LLM: mxbai-embed-large
+2024/10/28 15:08:25 Generation backend: llama3
+2024/10/28 15:08:25 Vector database initialized
+2024/10/28 15:08:26 Embedding generated
+2024/10/28 15:08:26 Vector Document generated
+2024/10/28 15:08:26 Vector embeddings generated
+2024/10/28 15:08:26 Retrieved Document: {doc-5630d3f2-bf61-4e13-8ec9-9e863bc1a962 map[content:Mickey mouse is a real human being]}
+2024/10/28 15:08:34 Retrieval-Augmented Generation influenced output from LLM model: In that case, since Mickey Mouse is a real human being (as per your context), I would say... **YES**, Mickey Mouse is indeed a human!
+```
 
 # 📝 Contributing
 

examples/ollama/README.md

@@ -1,6 +1,6 @@
 # Ollama Example Project
 
-This project demonstrates how to use GoLLM's OpenAI's API backend.
+This project demonstrates how to use gorag's OpenAI's API backend.
 
 This code demonstrates using Ollama embeddings and generation models, along with
 how RAG overides LLM knowledge by changing an established fact, already learned

examples/ollama/main.go

@@ -6,8 +6,8 @@ import (
 
 	"time"
 
-	"github.com/stackloklabs/gollm/pkg/backend"
-	"github.com/stackloklabs/gollm/pkg/db"
+	"github.com/stackloklabs/gorag/pkg/backend"
+	"github.com/stackloklabs/gorag/pkg/db"
 )
 
 var (
@@ -39,6 +39,9 @@ func main() {
 	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel()
 
+	// Close the connection when done
+	defer vectorDB.Close()
+
 	// We insert contextual information into the vector store so that the RAG system
 	// can use it to answer the query about the moon landing, effectively replacing 1969 with 2023
 	ragContent := "According to the Space Exploration Organization's official records, the moon landing occurred on July 20, 2023, during the Artemis Program. This mission marked the first successful crewed lunar landing since the Apollo program."

examples/openai/README.md

@@ -1,6 +1,6 @@
 # OpenAI Example Project
 
-This project demonstrates how to use GoLLM's OpenAI's API backend.
+This project demonstrates how to use gorag's OpenAI's API backend.
 
 This code demonstrates using OpenAI embeddings and generation models, along with
 how RAG overides LLM knowledge by changing an established fact, already learned

examples/openai/main.go

@@ -7,8 +7,8 @@ import (
 
 	"time"
 
-	"github.com/stackloklabs/gollm/pkg/backend"
-	"github.com/stackloklabs/gollm/pkg/db"
+	"github.com/stackloklabs/gorag/pkg/backend"
+	"github.com/stackloklabs/gorag/pkg/db"
 )
 
 var (
@@ -66,7 +66,6 @@ func main() {
 
 	// Insert the document into the vector store
 	err = vectorDB.InsertDocument(ctx, ragContent, embedding)
-
 	if err != nil {
 		log.Fatalf("Error inserting document: %v", err)
 	}
@@ -94,9 +93,6 @@ func main() {
 
 	// Augment the query with retrieved context
 	augmentedQuery := db.CombineQueryWithContext(query, retrievedDocs)
-	log.Printf("LLM Prompt: %s", query)
-
-	log.Printf("Augmented Query: %s", augmentedQuery)
 
 	prompt := backend.NewPrompt().
 		AddMessage("system", "You are an AI assistant. Use the provided context to answer the user's question as accurately as possible.").

examples/qdrant/main.go

@@ -3,11 +3,12 @@ package main
 import (
 	"context"
 	"fmt"
-	"github.com/google/uuid"
-	"github.com/stackloklabs/gollm/pkg/backend"
-	"github.com/stackloklabs/gollm/pkg/db"
 	"log"
 	"time"
+
+	"github.com/google/uuid"
+	"github.com/stackloklabs/gorag/pkg/backend"
+	"github.com/stackloklabs/gorag/pkg/db"
 )
 
 var (

go.mod

@@ -1,4 +1,4 @@
-module github.com/stackloklabs/gollm
+module github.com/stackloklabs/gorag
 
 go 1.22.2