Upload 4 files

Dockerfile

@@ -0,0 +1,20 @@
+# read the doc: https://huggingface.co/docs/hub/spaces-sdks-docker
+# you will also find guides on how best to write your Dockerfile
+
+FROM python:3.9
+
+WORKDIR /code
+
+COPY ./requirements.txt /code/requirements.txt
+
+RUN python3 -m pip install --no-cache-dir --upgrade pip
+RUN python3 -m pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+COPY . .
+
+CMD ["panel", "serve", "/code/app.ipynb", "--address", "0.0.0.0", "--port", "7860", "--allow-websocket-origin", "kaansahin-communicatelocalpdf.hf.space",  "--allow-websocket-origin", "0.0.0.0:7860"]
+
+
+

app.ipynb

@@ -0,0 +1,202 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#!pip install langchain openai chromadb tiktoken pypdf panel"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os \n",
+    "from langchain.chains import RetrievalQA\n",
+    "from langchain.llms import OpenAI\n",
+    "from langchain.document_loaders import TextLoader\n",
+    "from langchain.document_loaders import PyPDFLoader\n",
+    "from langchain.indexes import VectorstoreIndexCreator\n",
+    "from langchain.text_splitter import CharacterTextSplitter\n",
+    "from langchain.embeddings import OpenAIEmbeddings\n",
+    "from langchain.vectorstores import Chroma\n",
+    "import panel as pn\n",
+    "import tempfile"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "pn.extension('texteditor', template=\"bootstrap\", sizing_mode='stretch_width')\n",
+    "pn.state.template.param.update(\n",
+    "    main_max_width=\"690px\",\n",
+    "    header_background=\"#F08080\",\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "file_input = pn.widgets.FileInput(width=300)\n",
+    "\n",
+    "openaikey = pn.widgets.PasswordInput(\n",
+    "    value=\"\", placeholder=\"Enter your OpenAI API Key here...\", width=300\n",
+    ")\n",
+    "prompt = pn.widgets.TextEditor(\n",
+    "    value=\"\", placeholder=\"Enter your questions here...\", height=160, toolbar=False\n",
+    ")\n",
+    "run_button = pn.widgets.Button(name=\"Run!\")\n",
+    "\n",
+    "select_k = pn.widgets.IntSlider(\n",
+    "    name=\"Number of relevant chunks\", start=1, end=5, step=1, value=2\n",
+    ")\n",
+    "select_chain_type = pn.widgets.RadioButtonGroup(\n",
+    "    name='Chain type', \n",
+    "    options=['stuff', 'map_reduce', \"refine\", \"map_rerank\"]\n",
+    ")\n",
+    "\n",
+    "widgets = pn.Row(\n",
+    "    pn.Column(prompt, run_button, margin=5),\n",
+    "    pn.Card(\n",
+    "        \"Chain type:\",\n",
+    "        pn.Column(select_chain_type, select_k),\n",
+    "        title=\"Advanced settings\", margin=10\n",
+    "    ), width=600\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def qa(file, query, chain_type, k):\n",
+    "    # load document\n",
+    "    loader = PyPDFLoader(file)\n",
+    "    documents = loader.load()\n",
+    "    # split the documents into chunks\n",
+    "    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)\n",
+    "    texts = text_splitter.split_documents(documents)\n",
+    "    # select which embeddings we want to use\n",
+    "    embeddings = OpenAIEmbeddings()\n",
+    "    # create the vectorestore to use as the index\n",
+    "    db = Chroma.from_documents(texts, embeddings)\n",
+    "    # expose this index in a retriever interface\n",
+    "    retriever = db.as_retriever(search_type=\"similarity\", search_kwargs={\"k\": k})\n",
+    "    # create a chain to answer questions \n",
+    "    qa = RetrievalQA.from_chain_type(\n",
+    "        llm=OpenAI(), chain_type=chain_type, retriever=retriever, return_source_documents=True)\n",
+    "    result = qa({\"query\": query})\n",
+    "    print(result['result'])\n",
+    "    return result"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# result = qa(\"example.pdf\", \"what is the total number of AI publications?\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "convos = []  # store all panel objects in a list\n",
+    "\n",
+    "def qa_result(_):\n",
+    "    os.environ[\"OPENAI_API_KEY\"] = openaikey.value\n",
+    "    \n",
+    "    # save pdf file to a temp file \n",
+    "    if file_input.value is not None:\n",
+    "        file_input.save(\"/.cache/temp.pdf\")\n",
+    "    \n",
+    "        prompt_text = prompt.value\n",
+    "        if prompt_text:\n",
+    "            result = qa(file=\"/.cache/temp.pdf\", query=prompt_text, chain_type=select_chain_type.value, k=select_k.value)\n",
+    "            convos.extend([\n",
+    "                pn.Row(\n",
+    "                    pn.panel(\"\\U0001F60A\", width=10),\n",
+    "                    prompt_text,\n",
+    "                    width=600\n",
+    "                ),\n",
+    "                pn.Row(\n",
+    "                    pn.panel(\"\\U0001F916\", width=10),\n",
+    "                    pn.Column(\n",
+    "                        result[\"result\"],\n",
+    "                        \"Relevant source text:\",\n",
+    "                        pn.pane.Markdown('\\n--------------------------------------------------------------------\\n'.join(doc.page_content for doc in result[\"source_documents\"]))\n",
+    "                    )\n",
+    "                )\n",
+    "            ])\n",
+    "            #return convos\n",
+    "    return pn.Column(*convos, margin=15, width=575, min_height=400)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "qa_interactive = pn.panel(\n",
+    "    pn.bind(qa_result, run_button),\n",
+    "    loading_indicator=True,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "output = pn.WidgetBox('*Output will show up here:*', qa_interactive, width=630, scroll=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# layout\n",
+    "pn.Column(\n",
+    "    pn.pane.Markdown(\"\"\"\n",
+    "    ## \\U0001F60A! Question Answering with your PDF file\n",
+    "    \n",
+    "    1) Upload a PDF. 2) Enter OpenAI API key. This costs $. Set up billing at [OpenAI](https://platform.openai.com/account). 3) Type a question and click \"Run\".\n",
+    "    \n",
+    "    \"\"\"),\n",
+    "    pn.Row(file_input,openaikey),\n",
+    "    output,\n",
+    "    widgets\n",
+    "\n",
+    ").servable()"
+   ]
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  },
+  "orig_nbformat": 4
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

requirements.txt

@@ -0,0 +1,7 @@
+langchain
+openai
+chromadb
+pypdf
+tiktoken
+panel
+notebook