feat: Add streaming support for QueryEngine (and unify streaming inte…

…rface with ChatEngine) (run-llama#393)

.changeset/spicy-rice-tan.md

@@ -0,0 +1,5 @@
+---
+"llamaindex": patch
+---
+
+Add streaming support for QueryEngine (and unify streaming interface with ChatEngine)

.changeset/tidy-mayflies-provide.md

@@ -0,0 +1,5 @@
+---
+"llamaindex": patch
+---
+
+Breaking: Use parameter object for query and chat methods of ChatEngine and QueryEngine

apps/docs/docs/modules/high_level/chat_engine.md

@@ -11,7 +11,16 @@ const retriever = index.asRetriever();
 const chatEngine = new ContextChatEngine({ retriever });
 
 // start chatting
-const response = await chatEngine.chat(query);
+const response = await chatEngine.chat({ message: query });
+```
+
+The `chat` function also supports streaming, just add `stream: true` as an option:
+
+```typescript
+const stream = await chatEngine.chat({ message: query, stream: true });
+for await (const chunk of stream) {
+  process.stdout.write(chunk.response);
+}
 ```
 
 ## Api References

apps/docs/docs/modules/high_level/query_engine.md

@@ -8,7 +8,16 @@ A query engine wraps a `Retriever` and a `ResponseSynthesizer` into a pipeline,
 
 ```typescript
 const queryEngine = index.asQueryEngine();
-const response = await queryEngine.query("query string");
+const response = await queryEngine.query({ query: "query string" });
+```
+
+The `query` function also supports streaming, just add `stream: true` as an option:
+
+```typescript
+const stream = await queryEngine.query({ query: "query string", stream: true });
+for await (const chunk of stream) {
+  process.stdout.write(chunk.response);
+}
 ```
 
 ## Sub Question Query Engine

examples/astradb/query.ts

@@ -18,7 +18,9 @@ async function main() {
 
     const queryEngine = await index.asQueryEngine({ retriever });
 
-    const results = await queryEngine.query("What is the best reviewed movie?");
+    const results = await queryEngine.query({
+      query: "What is the best reviewed movie?",
+    });
 
     console.log(results.response);
   } catch (e) {

examples/chatEngine.ts

@@ -25,8 +25,11 @@ async function main() {
 
   while (true) {
     const query = await rl.question("Query: ");
-    const response = await chatEngine.chat(query);
-    console.log(response.toString());
+    const stream = await chatEngine.chat({ message: query, stream: true });
+    console.log();
+    for await (const chunk of stream) {
+      process.stdout.write(chunk.response);
+    }
   }
 }
 

examples/chromadb/test.ts

@@ -28,9 +28,9 @@ async function main() {
 
     console.log("Querying index");
     const queryEngine = index.asQueryEngine();
-    const response = await queryEngine.query(
-      "Tell me about Godfrey Cheshire's rating of La Sapienza.",
-    );
+    const response = await queryEngine.query({
+      query: "Tell me about Godfrey Cheshire's rating of La Sapienza.",
+    });
     console.log(response.toString());
   } catch (e) {
     console.error(e);

examples/huggingface.ts

@@ -32,12 +32,15 @@ async function main() {
 
   // Query the index
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(
-    "What did the author do in college?",
-  );
+  const stream = await queryEngine.query({
+    query: "What did the author do in college?",
+    stream: true,
+  });
 
   // Output response
-  console.log(response.toString());
+  for await (const chunk of stream) {
+    process.stdout.write(chunk.response);
+  }
 }
 
 main().catch(console.error);

examples/keywordIndex.ts

@@ -20,9 +20,9 @@ async function main() {
         mode,
       }),
     });
-    const response = await queryEngine.query(
-      "What did the author do growing up?",
-    );
+    const response = await queryEngine.query({
+      query: "What did the author do growing up?",
+    });
     console.log(response.toString());
   });
 }

examples/markdown.ts

@@ -11,7 +11,9 @@ async function main() {
   // Query the index
   const queryEngine = index.asQueryEngine();
 
-  const response = await queryEngine.query("What does the example code do?");
+  const response = await queryEngine.query({
+    query: "What does the example code do?",
+  });
 
   // Output response
   console.log(response.toString());

examples/mistral.ts

@@ -27,7 +27,7 @@ async function rag(llm: LLM, embedModel: BaseEmbedding, query: string) {
 
   // Query the index
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(query);
+  const response = await queryEngine.query({ query });
   return response.response;
 }
 

examples/mongo.ts

@@ -54,7 +54,7 @@ async function main() {
       break;
     }
 
-    const response = await queryEngine.query(query);
+    const response = await queryEngine.query({ query });
 
     // Output response
     console.log(response.toString());

examples/mongodb/3_query.ts

@@ -24,9 +24,9 @@ async function query() {
 
   const retriever = index.asRetriever({ similarityTopK: 20 });
   const queryEngine = index.asQueryEngine({ retriever });
-  const result = await queryEngine.query(
-    "What does the author think of web frameworks?",
-  );
+  const result = await queryEngine.query({
+    query: "What does the author think of web frameworks?",
+  });
   console.log(result.response);
   await client.close();
 }

examples/multimodal/rag.ts

@@ -46,9 +46,9 @@ async function main() {
     responseSynthesizer: new MultiModalResponseSynthesizer({ serviceContext }),
     retriever: index.asRetriever({ similarityTopK: 3, imageSimilarityTopK: 1 }),
   });
-  const result = await queryEngine.query(
-    "Tell me more about Vincent van Gogh's famous paintings",
-  );
+  const result = await queryEngine.query({
+    query: "Tell me more about Vincent van Gogh's famous paintings",
+  });
   console.log(result.response, "\n");
   images.forEach((image) =>
     console.log(`Image retrieved and used in inference: ${image.toString()}`),

examples/pg-vector-store/query.ts

@@ -31,7 +31,7 @@ async function main() {
       }
 
       try {
-        const answer = await queryEngine.query(question);
+        const answer = await queryEngine.query({ query: question });
         console.log(answer.response);
       } catch (error) {
         console.error("Error:", error);

examples/pinecone-vector-store/query.ts

@@ -29,7 +29,7 @@ async function main() {
       }
 
       try {
-        const answer = await queryEngine.query(question);
+        const answer = await queryEngine.query({ query: question });
         console.log(answer.response);
       } catch (error) {
         console.error("Error:", error);

examples/readers/load-assemblyai.ts

@@ -48,7 +48,7 @@ program
         break;
       }
 
-      const response = await queryEngine.query(query);
+      const response = await queryEngine.query({ query });
 
       console.log(response.toString());
     }

examples/readers/load-csv.ts

@@ -38,9 +38,9 @@ Given the CSV file, generate me Typescript code to answer the question: ${query}
   const queryEngine = index.asQueryEngine({ responseSynthesizer });
 
   // Query the index
-  const response = await queryEngine.query(
-    "What is the correlation between survival and age?",
-  );
+  const response = await queryEngine.query({
+    query: "What is the correlation between survival and age?",
+  });
 
   // Output response
   console.log(response.toString());

examples/readers/load-docx.ts

@@ -15,7 +15,7 @@ async function main() {
 
   // Test query
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(SAMPLE_QUERY);
+  const response = await queryEngine.query({ query: SAMPLE_QUERY });
   console.log(`Test query > ${SAMPLE_QUERY}:\n`, response.toString());
 }
 

examples/readers/load-html.ts

@@ -10,9 +10,9 @@ async function main() {
 
   // Query the index
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(
-    "What were the notable changes in 18.1?",
-  );
+  const response = await queryEngine.query({
+    query: "What were the notable changes in 18.1?",
+  });
 
   // Output response
   console.log(response.toString());

examples/readers/load-md.ts

@@ -15,7 +15,7 @@ async function main() {
 
   // Test query
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(SAMPLE_QUERY);
+  const response = await queryEngine.query({ query: SAMPLE_QUERY });
   console.log(`Test query > ${SAMPLE_QUERY}:\n`, response.toString());
 }
 

examples/readers/load-notion.ts

@@ -79,7 +79,7 @@ program
         break;
       }
 
-      const response = await queryEngine.query(query);
+      const response = await queryEngine.query({ query });
 
       // Output response
       console.log(response.toString());

examples/readers/load-pdf.ts

@@ -10,7 +10,9 @@ async function main() {
 
   // Query the index
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query("What mistakes did they make?");
+  const response = await queryEngine.query({
+    query: "What mistakes did they make?",
+  });
 
   // Output response
   console.log(response.toString());

examples/sentenceWindow.ts

@@ -31,9 +31,9 @@ async function main() {
   const queryEngine = index.asQueryEngine({
     nodePostprocessors: [new MetadataReplacementPostProcessor("window")],
   });
-  const response = await queryEngine.query(
-    "What did the author do in college?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do in college?",
+  });
 
   // Output response
   console.log(response.toString());

examples/storageContext.ts

@@ -20,9 +20,9 @@ async function main() {
 
   // Query the index
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(
-    "What did the author do in college?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do in college?",
+  });
 
   // Output response
   console.log(response.toString());
@@ -35,9 +35,9 @@ async function main() {
     storageContext: secondStorageContext,
   });
   const loadedQueryEngine = loadedIndex.asQueryEngine();
-  const loadedResponse = await loadedQueryEngine.query(
-    "What did the author do growing up?",
-  );
+  const loadedResponse = await loadedQueryEngine.query({
+    query: "What did the author do growing up?",
+  });
   console.log(loadedResponse.toString());
 }
 

examples/subquestion.ts

@@ -18,9 +18,9 @@ import essay from "./essay";
     ],
   });
 
-  const response = await queryEngine.query(
-    "How was Paul Grahams life different before and after YC?",
-  );
+  const response = await queryEngine.query({
+    query: "How was Paul Grahams life different before and after YC?",
+  });
 
   console.log(response.toString());
 })();

examples/summaryIndex.ts

@@ -20,9 +20,9 @@ async function main() {
   const queryEngine = index.asQueryEngine({
     retriever: index.asRetriever({ mode: SummaryRetrieverMode.LLM }),
   });
-  const response = await queryEngine.query(
-    "What did the author do growing up?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do growing up?",
+  });
   console.log(response.toString());
 }
 

examples/together-ai/vector-index.ts

@@ -29,9 +29,9 @@ async function main() {
 
   const queryEngine = index.asQueryEngine();
 
-  const response = await queryEngine.query(
-    "What did the author do in college?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do in college?",
+  });
 
   console.log(response.toString());
 }

examples/vectorIndex.ts

@@ -16,9 +16,9 @@ async function main() {
 
   // Query the index
   const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query(
-    "What did the author do in college?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do in college?",
+  });
 
   // Output response
   console.log(response.toString());

examples/vectorIndexAnthropic.ts

@@ -35,9 +35,9 @@ async function main() {
 
   // Query the index
   const queryEngine = index.asQueryEngine({ responseSynthesizer });
-  const response = await queryEngine.query(
-    "What did the author do in college?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do in college?",
+  });
 
   // Output response
   console.log(response.toString());

examples/vectorIndexCustomize.ts

@@ -33,9 +33,9 @@ async function main() {
     [nodePostprocessor],
   );
 
-  const response = await queryEngine.query(
-    "What did the author do growing up?",
-  );
+  const response = await queryEngine.query({
+    query: "What did the author do growing up?",
+  });
   console.log(response.response);
 }
 

examples/vectorIndexFromVectorStore.ts

@@ -189,7 +189,9 @@ async function main() {
     },
   });
 
-  const response = await queryEngine.query("How many results do you have?");
+  const response = await queryEngine.query({
+    query: "How many results do you have?",
+  });
 
   console.log(response.toString());
 }