feat: use latest llamaindex version (#76)

app/api/fetch/embeddings.ts

@@ -7,27 +7,27 @@ import {
  Document,
  MetadataMode,
  SentenceSplitter,
+ SimpleNodeParser,
  VectorStoreIndex,
- getNodesFromDocument,
  serviceContextFromDefaults,
 } from "llamaindex";
 
 export default async function splitAndEmbed(
  document: string,
 ): Promise<Embedding[]> {
- const nodes = getNodesFromDocument(
- new Document({ text: document }),
- new SentenceSplitter({
+ const nodeParser = new SimpleNodeParser({
+ textSplitter: new SentenceSplitter({
  chunkSize: DATASOURCES_CHUNK_SIZE,
  chunkOverlap: DATASOURCES_CHUNK_OVERLAP,
  }),
- );
-
- const nodesWithEmbeddings = await VectorStoreIndex.getNodeEmbeddingResults(
- nodes,
- serviceContextFromDefaults(),
- true,
- );
+ });
+ const nodes = nodeParser.getNodesFromDocuments([
+ new Document({ text: document }),
+ ]);
+ const index = await VectorStoreIndex.fromDocuments(nodes, {
+ serviceContext: serviceContextFromDefaults(),
+ });
+ const nodesWithEmbeddings = await index.getNodeEmbeddingResults(nodes);
 
  return nodesWithEmbeddings.map((nodeWithEmbedding) => ({
  text: nodeWithEmbedding.getContent(MetadataMode.NONE),

app/api/llm/route.ts

@@ -1,16 +1,17 @@
 import {
  ChatHistory,
  ChatMessage,
- DefaultContextGenerator,
- HistoryChatEngine,
+ ContextChatEngine,
  IndexDict,
  OpenAI,
  ServiceContext,
+ SimpleChatEngine,
  SimpleChatHistory,
  SummaryChatHistory,
  TextNode,
  VectorStoreIndex,
  serviceContextFromDefaults,
+ Response,
 } from "llamaindex";
 import { NextRequest, NextResponse } from "next/server";
 import { LLMConfig, MessageContent } from "@/app/client/platforms/llm";
@@ -27,7 +28,6 @@ async function createChatEngine(
  datasource?: string,
  embeddings?: Embedding[],
 ) {
- let contextGenerator;
  if (datasource || embeddings) {
  let index;
  if (embeddings) {
@@ -39,12 +39,14 @@ async function createChatEngine(
  const retriever = index!.asRetriever();
  retriever.similarityTopK = 5;
 
- contextGenerator = new DefaultContextGenerator({ retriever });
+ return new ContextChatEngine({
+ chatModel: serviceContext.llm,
+ retriever,
+ });
  }
 
- return new HistoryChatEngine({
+ return new SimpleChatEngine({
  llm: serviceContext.llm,
- contextGenerator,
  });
 }
 
@@ -75,9 +77,10 @@ async function createIndex(
 }
 
 function createReadableStream(
- stream: AsyncGenerator<string, void, unknown>,
+ stream: AsyncIterable<Response>,
  chatHistory: ChatHistory,
 ) {
+ const it = stream[Symbol.asyncIterator]();
  let responseStream = new TransformStream();
  const writer = responseStream.writable.getWriter();
  let aborted = false;
@@ -88,10 +91,12 @@ function createReadableStream(
  const encoder = new TextEncoder();
  const onNext = async () => {
  try {
- const { value, done } = await stream.next();
+ const { value, done } = await it.next();
  if (aborted) return;
  if (!done) {
- writer.write(encoder.encode(`data: ${JSON.stringify(value)}\n\n`));
+ writer.write(
+ encoder.encode(`data: ${JSON.stringify(value.response)}\n\n`),
+ );
  onNext();
  } else {
  writer.write(
@@ -168,7 +173,11 @@ export async function POST(request: NextRequest) {
  ? new SummaryChatHistory({ llm, messages })
  : new SimpleChatHistory({ messages });
 
- const stream = await chatEngine.chat(message, chatHistory, true);
+ const stream = await chatEngine.chat({
+ message,
+ chatHistory,
+ stream: true,
+ });
  const readableStream = createReadableStream(stream, chatHistory);
 
  return new NextResponse(readableStream, {

package.json

@@ -38,7 +38,7 @@
  "dotenv": "^16.3.1",
  "emoji-picker-react": "^4.4.12",
  "encoding": "^0.1.13",
- "llamaindex": "0.0.0-20231110031459",
+ "llamaindex": "0.1.10",
  "lucide-react": "^0.277.0",
  "mermaid": "^10.3.1",
  "nanoid": "^5.0.2",