feat: use llamacloud for chat

.changeset/tough-pugs-destroy.md

@@ -0,0 +1,5 @@
+---
+"create-llama": patch
+---
+
+use llamacloud for chat

helpers/python.ts

@@ -118,6 +118,12 @@ const getAdditionalDependencies = (
             version: "^2.9.9",
           });
           break;
+        case "llamacloud":
+          dependencies.push({
+            name: "llama-index-indices-managed-llama-cloud",
+            version: "^0.2.1",
+          });
+          break;
       }
     }
   }

helpers/types.ts

@@ -26,7 +26,8 @@ export type TemplateVectorDB =
   | "milvus"
   | "astra"
   | "qdrant"
-  | "chroma";
+  | "chroma"
+  | "llamacloud";
 export type TemplatePostInstallAction =
   | "none"
   | "VSCode"
@@ -36,7 +37,7 @@ export type TemplateDataSource = {
   type: TemplateDataSourceType;
   config: TemplateDataSourceConfig;
 };
-export type TemplateDataSourceType = "file" | "web" | "db";
+export type TemplateDataSourceType = "file" | "web" | "db" | "llamacloud";
 export type TemplateObservability = "none" | "opentelemetry";
 // Config for both file and folder
 export type FileSourceConfig = {

questions.ts

@@ -123,7 +123,13 @@ export const getDataSourceChoices = (
   framework: TemplateFramework,
   selectedDataSource: TemplateDataSource[],
 ) => {
+  // If LlamaCloud is already selected, don't show any other options
+  if (selectedDataSource.find((s) => s.type === "llamacloud")) {
+    return [];
+  }
+
   const choices = [];
+
   if (selectedDataSource.length > 0) {
     choices.push({
       title: "No",
@@ -171,6 +177,13 @@ export const getDataSourceChoices = (
       value: "db",
     });
   }
+
+  if (!selectedDataSource.length) {
+    choices.push({
+      title: "Use LlamaCloud",
+      value: "llamacloud",
+    });
+  }
   return choices;
 };
 
@@ -484,17 +497,19 @@ export const askQuestions = async (
       // continue asking user for data sources if none are initially provided
       while (true) {
         const firstQuestion = program.dataSources.length === 0;
+        const choices = getDataSourceChoices(
+          program.framework,
+          program.dataSources,
+        );
+        if (choices.length === 0) break;
         const { selectedSource } = await prompts(
           {
             type: "select",
             name: "selectedSource",
             message: firstQuestion
               ? "Which data source would you like to use?"
               : "Would you like to add another data source?",
-            choices: getDataSourceChoices(
-              program.framework,
-              program.dataSources,
-            ),
+            choices,
             initial: firstQuestion ? 1 : 0,
           },
           questionHandlers,
@@ -591,6 +606,13 @@ export const askQuestions = async (
               config: await prompts(dbPrompts, questionHandlers),
             });
           }
+          case "llamacloud": {
+            program.dataSources.push({
+              type: "llamacloud",
+              config: {},
+            });
+            break;
+          }
         }
       }
     }
@@ -635,7 +657,16 @@ export const askQuestions = async (
     }
   }
 
-  if (program.dataSources.length > 0 && !program.vectorDb) {
+  const isUsingLlamaCloud = program.dataSources.some(
+    (ds) => ds.type === "llamacloud",
+  );
+
+  if (isUsingLlamaCloud) {
+    // If using LlamaCloud, don't ask for vector database and use `llamacloud` folder for vector database
+    const vectorDb = "llamacloud";
+    program.vectorDb = vectorDb;
+    preferences.vectorDb = vectorDb;
+  } else if (program.dataSources.length > 0 && !program.vectorDb) {
     if (ciInfo.isCI) {
       program.vectorDb = getPrefOrDefault("vectorDb");
     } else {

templates/components/engines/typescript/agent/chat.ts

@@ -2,7 +2,6 @@ import { BaseToolWithCall, OpenAIAgent, QueryEngineTool } from "llamaindex";
 import fs from "node:fs/promises";
 import path from "node:path";
 import { getDataSource } from "./index";
-import { STORAGE_CACHE_DIR } from "./shared";
 import { createTools } from "./tools";
 
 export async function createChatEngine() {
@@ -17,7 +16,7 @@ export async function createChatEngine() {
         queryEngine: index.asQueryEngine(),
         metadata: {
           name: "data_query_engine",
-          description: `A query engine for documents in storage folder: ${STORAGE_CACHE_DIR}`,
+          description: `A query engine for documents from your data source.`,
         },
       }),
     );

templates/components/vectordbs/python/llamacloud/generate.py

@@ -0,0 +1,45 @@
+from dotenv import load_dotenv
+
+load_dotenv()
+
+import os
+import logging
+from app.settings import init_settings
+from app.engine.loaders import get_documents
+from llama_index.indices.managed.llama_cloud import LlamaCloudIndex
+
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger()
+
+
+def generate_datasource():
+    init_settings()
+    logger.info("Generate index for the provided data")
+
+    name = os.getenv("LLAMA_CLOUD_NAME")
+    project_name = os.getenv("LLAMA_CLOUD_PROJECT_NAME")
+    api_key = os.getenv("LLAMA_CLOUD_API_KEY")
+    base_url = os.getenv("LLAMA_CLOUD_BASE_URL")
+
+    if not name or not project_name or not api_key:
+        raise ValueError(
+            "Please set LLAMA_CLOUD_NAME, LLAMA_CLOUD_PROJECT_NAME and LLAMA_CLOUD_API_KEY"
+            " to your environment variables or config them in .env file"
+        )
+
+    documents = get_documents()
+
+    LlamaCloudIndex.from_documents(
+        documents,
+        name,
+        project_name=project_name,
+        api_key=api_key,
+        base_url=base_url,
+    )
+
+    logger.info("Finished generating the index")
+
+
+if __name__ == "__main__":
+    generate_datasource()

templates/components/vectordbs/python/llamacloud/index.py

@@ -0,0 +1,32 @@
+from dotenv import load_dotenv
+
+load_dotenv()
+
+import logging
+import os
+from llama_index.indices.managed.llama_cloud import LlamaCloudIndex
+
+
+logger = logging.getLogger("uvicorn")
+
+
+def get_index():
+    name = os.getenv("LLAMA_CLOUD_NAME")
+    project_name = os.getenv("LLAMA_CLOUD_PROJECT_NAME")
+    api_key = os.getenv("LLAMA_CLOUD_API_KEY")
+    base_url = os.getenv("LLAMA_CLOUD_BASE_URL")
+
+    if not name or not project_name or not api_key:
+        raise ValueError(
+            "Please set LLAMA_CLOUD_NAME, LLAMA_CLOUD_PROJECT_NAME and LLAMA_CLOUD_API_KEY"
+            " to your environment variables or config them in .env file"
+        )
+
+    index = LlamaCloudIndex(
+        name,
+        project_name=project_name,
+        api_key=api_key,
+        base_url=base_url,
+    )
+
+    return index

templates/components/vectordbs/typescript/llamacloud/generate.ts

@@ -0,0 +1,26 @@
+import * as dotenv from "dotenv";
+import { LlamaCloudIndex } from "llamaindex";
+import { getDocuments } from "./loader";
+import { initSettings } from "./settings";
+import { checkRequiredEnvVars } from "./shared";
+
+dotenv.config();
+
+async function loadAndIndex() {
+  const documents = await getDocuments();
+  await LlamaCloudIndex.fromDocuments({
+    documents,
+    name: process.env.LLAMA_CLOUD_NAME!,
+    projectName: process.env.LLAMA_CLOUD_PROJECT_NAME!,
+    apiKey: process.env.LLAMA_CLOUD_API_KEY,
+    baseUrl: process.env.LLAMA_CLOUD_BASE_URL,
+  });
+  console.log(`Successfully created embeddings!`);
+}
+
+(async () => {
+  checkRequiredEnvVars();
+  initSettings();
+  await loadAndIndex();
+  console.log("Finished generating storage.");
+})();

templates/components/vectordbs/typescript/llamacloud/index.ts

@@ -0,0 +1,13 @@
+import { LlamaCloudIndex } from "llamaindex/cloud/LlamaCloudIndex";
+import { checkRequiredEnvVars } from "./shared";
+
+export async function getDataSource() {
+  checkRequiredEnvVars();
+  const index = new LlamaCloudIndex({
+    name: process.env.LLAMA_CLOUD_NAME!,
+    projectName: process.env.LLAMA_CLOUD_PROJECT_NAME!,
+    apiKey: process.env.LLAMA_CLOUD_API_KEY,
+    baseUrl: process.env.LLAMA_CLOUD_BASE_URL,
+  });
+  return index;
+}

templates/components/vectordbs/typescript/llamacloud/shared.ts

@@ -0,0 +1,22 @@
+const REQUIRED_ENV_VARS = [
+  "LLAMA_CLOUD_NAME",
+  "LLAMA_CLOUD_PROJECT_NAME",
+  "LLAMA_CLOUD_API_KEY",
+];
+
+export function checkRequiredEnvVars() {
+  const missingEnvVars = REQUIRED_ENV_VARS.filter((envVar) => {
+    return !process.env[envVar];
+  });
+
+  if (missingEnvVars.length > 0) {
+    console.log(
+      `The following environment variables are required but missing: ${missingEnvVars.join(
+        ", ",
+      )}`,
+    );
+    throw new Error(
+      `Missing environment variables: ${missingEnvVars.join(", ")}`,
+    );
+  }
+}

templates/types/streaming/express/package.json

@@ -20,7 +20,7 @@
     "dotenv": "^16.3.1",
     "duck-duck-scrape": "^2.2.5",
     "express": "^4.18.2",
-    "llamaindex": "0.4.3",
+    "llamaindex": "0.4.6",
     "pdf2json": "3.0.5",
     "ajv": "^8.12.0",
     "@e2b/code-interpreter": "^0.0.5",

templates/types/streaming/fastapi/pyproject.toml

@@ -14,8 +14,8 @@ fastapi = "^0.109.1"
 uvicorn = { extras = ["standard"], version = "^0.23.2" }
 python-dotenv = "^1.0.0"
 aiostream = "^0.5.2"
-llama-index = "0.10.41"
-llama-index-core = "0.10.41"
+llama-index = "0.10.50"
+llama-index-core = "0.10.50"
 cachetools = "^5.3.3"
 
 [build-system]

templates/types/streaming/nextjs/package.json

@@ -24,7 +24,7 @@
     "duck-duck-scrape": "^2.2.5",
     "formdata-node": "^6.0.3",
     "got": "^14.4.1",
-    "llamaindex": "0.4.3",
+    "llamaindex": "0.4.6",
     "lucide-react": "^0.294.0",
     "next": "^14.2.4",
     "pdf2json": "3.0.5",