identify if fileId before sending message

src/components/ChatWindow.jsx

@@ -10,6 +10,7 @@ function ChatWindow({
   maxTokens,
 }) {
   const { loadingStatus, send, isGenerating, setOnMessage } = useLLM();
+  const [fileId, setFileId] = useState();
   const [userInput, setUserInput] = useState("");  
 
   const handleChange = (event) => {
@@ -23,6 +24,10 @@ function ChatWindow({
       return;
     }
     
+    if (fileId) {
+      console.log('we have a fileId, so this should be contextual chat')
+    }
+
     send(userInput, maxTokens, stopStrings);
     setUserInput("");
   }, [
@@ -48,6 +53,16 @@ function ChatWindow({
     };
   }, [handleSubmit]);
 
+  const loadFile = async () => {
+    console.log('test');
+  }
+
+  useEffect(() => {
+    console.log('have a fileId, look it up and summarize after uploading for demo');
+    loadFile();
+
+  }, [fileId])
+
   return (
     <div className="window sm:w-[500px] w-full">
       
@@ -98,7 +113,7 @@ function ChatWindow({
                       height="40"
                     />
                   </button>
-                  <FileLoader />
+                  <FileLoader setFileId={setFileId} />
                   </div>
                   
                   <div
@@ -117,6 +132,7 @@ function ChatWindow({
             </div>
           )}
           {!isReady && <Loader />}
+          {fileId && <div>loaded {fileId}</div>}
         </div>
       </div>
     </div>

src/components/FileLoader.jsx

@@ -26,7 +26,7 @@ export default class Pdf {
   }
 }
 
-export const FileLoader = () => {
+export const FileLoader = ({ setFileId }) => {
   const [files, setFiles] = useState();
   const [uploadStatus, setUploadStatus] = useState("Embed");
   const handleEmbed = (files) => {
@@ -58,14 +58,13 @@ export const FileLoader = () => {
               body: JSON.stringify({ text }),
             });
             const docChat = await response.json();
-            console.log('handleDocs-getModels: ', docChat);
             const id = await db.docs.add({
               fileName: file.name || '',
               fileSourceData: text,
               model: docChat,
             });
             setUploadStatus("Embedding Completed");
-            console.log('handleDocs-setToDb', id);
+            setFileId(id)
           }
         }}
       >

src/embed/hf.ts

@@ -5,7 +5,7 @@ import { pipeline } from "@xenova/transformers";
    model?: string;
  }
 
- export class XenovaTransformersEmbeddings
+ export default class XenovaTransformersEmbeddings
    extends Embeddings
    implements XenovaTransformersEmbeddingsParams
  {
@@ -22,10 +22,11 @@ import { pipeline } from "@xenova/transformers";
      if (!this.client) {
        this.client = await pipeline("embeddings", this.model);
      }
+     console.log(this.client, texts);
 
      return this.caller.call(async () => {
        return await Promise.all(
-         texts.map(async (t) => (await this.client(t)).data)
+         texts.map(async (t) => (await this.client(t, { pooling: 'mean', normalize: true })).data)
        );
      });
    }

src/pages/api/docChat.ts

@@ -3,7 +3,7 @@ import type { NextApiRequest, NextApiResponse } from 'next';
 import { storesDir } from '@/utils/file-handler';
 
 import { makeChain } from '@/utils/make-chain';
-import { XenovaTransformersEmbeddings } from '../../embed/hf';
+import XenovaTransformersEmbeddings from '../../embed/hf';
 import { HNSWLib } from 'langchain/vectorstores/hnswlib';
 
 let vectorStore: HNSWLib;

src/pages/api/docHandle.ts

@@ -1,5 +1,4 @@
 import type { NextApiRequest, NextApiResponse } from 'next';
-
 import {
   readHNSWLibModelFromLocal,
   storesDir,
@@ -8,16 +7,12 @@ import {
 import fs from 'fs-extra';
 import { RecursiveCharacterTextSplitter } from 'langchain/text_splitter';
 import { HNSWLib } from 'langchain/vectorstores/hnswlib';
-import { XenovaTransformersEmbeddings } from '../../embed/hf'
+import XenovaTransformersEmbeddings from '../../embed/hf'
 
 async function handleDocs(text: string) {
   const textSplitter = new RecursiveCharacterTextSplitter({ chunkSize: 1000 });
   const docs = await textSplitter.createDocuments([text]);
-  console.log(docs);
-
   const vectorStore = await HNSWLib.fromDocuments(docs, new XenovaTransformersEmbeddings());
-  console.log(vectorStore);
-
   return vectorStore;
 }
 
@@ -32,6 +27,8 @@ export default async function handler(
     return res.status(400).json({ message: 'No question in the request' });
   }
 
+  /*
+  use dexie instead to get contents 
   const exists = await fs.exists(storesDir);
   console.log(exists);
 
@@ -42,6 +39,7 @@ export default async function handler(
       ...model,
     });
   }
+  */
 
   const vectorStore = await handleDocs(text);
   const model = await vectorStoreToHNSWLibModel(vectorStore);

src/utils/file-handler.ts

@@ -1,4 +1,5 @@
-import { XenovaTransformersEmbeddings } from '@/embed/hf';
+import type XenovaTransformersEmbeddings from '@/embed/hf';
+// import { HuggingFaceInferenceEmbeddings } from 'langchain/embeddings/hf';
 import fs from 'fs-extra';
 import {
   HNSWLib,