quickstart.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Objective:\n",
    "1. setup langchain, langsmith\n",
    "2. create simple application with langchain\n",
    "    - use prompt templates, models, output parsers\n",
    "    - use lcel\n",
    "3. start tracing with langsmith\n",
    "\n",
    "\n",
    "Basics:\n",
    "- Primitives like llms, parsers, prompts, retrievers, agents -> implement LangChain Runnable Interface"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### simple chain"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Object \"install\" is unknown, try \"ip help\".\n"
     ]
    }
   ],
   "source": [
    "! pip install langchain --quiet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "\n",
    "with open(\"/home/ubuntu/config.json\") as file:\n",
    "    config = json.load(file)\n",
    "\n",
    "os.environ[\"LANGCHAIN_TRACING_V2\"]=\"true\"\n",
    "os.environ[\"LANGCHAIN_ENDPOINT\"]=\"https://api.smith.langchain.com\"\n",
    "os.environ[\"LANGCHAIN_API_KEY\"]=config[\"langsmith_api_key\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.prompts import (\n",
    "    PromptTemplate, ChatPromptTemplate \n",
    ")\n",
    "\n",
    "prompt = ChatPromptTemplate.from_messages([\n",
    "    (\"system\", \"You are a world class technical documentation writer\"),\n",
    "    (\"user\", \"{input}\")\n",
    "])\n",
    "\n",
    "# prompt.messages[0].pretty_print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import boto3\n",
    "from langchain_community.chat_models import BedrockChat\n",
    "\n",
    "llm = BedrockChat(\n",
    "    model_id=\"anthropic.claude-3-sonnet-20240229-v1:0\",\n",
    "    # model_id=\"anthropic.claude-3-haiku-20240307-v1:0\",\n",
    "    # model_id=\"anthropic.claude-instant-v1\",\n",
    "    # model_id=\"anthropic.claude-v2:1\",\n",
    "\n",
    "    client=boto3.client(\"bedrock-runtime\"),\n",
    "    model_kwargs={\n",
    "        \"max_tokens\": 128,\n",
    "        # \"max_tokens_to_sample\": 128,\n",
    "        \"temperature\": 0.5,\n",
    "    }\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "chain = prompt | llm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AIMessage(content=\"Langsmith can be a valuable tool for testing software applications and systems. Here are some ways Langsmith can assist with testing:\\n\\n1. **Test Case Generation**: Langsmith can be used to generate test cases automatically based on the requirements or specifications of the software. It can analyze the input data, expected outputs, and edge cases to create comprehensive test scenarios.\\n\\n2. **Natural Language Test Scripting**: With Langsmith's natural language processing capabilities, testers can write test scripts using plain English or other natural languages. This can make test script creation more accessible and easier to understand, especially for non-technical\")"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chain.invoke({\"input\": \"How can Langsmith help with testing ?\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.output_parsers import StrOutputParser\n",
    "\n",
    "chain = prompt | llm | StrOutputParser()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Langsmith can be a valuable tool for testing, particularly when it comes to testing software systems that involve natural language processing (NLP) or generation. Here are some ways Langsmith can assist with testing:\n",
      "\n",
      "1. **Test Data Generation**: Langsmith can generate large volumes of realistic and varied test data, including natural language text, dialogues, and scenarios. This can be useful for testing the robustness and edge cases of NLP systems, chatbots, or language models.\n",
      "\n",
      "2. **Test Case Generation**: Langsmith can help generate test cases for NLP systems by creating a diverse set of\n"
     ]
    }
   ],
   "source": [
    "print(chain.invoke({\"input\": \"How can Langsmith help with testing ?\"}))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### add context -> retrieval_chain"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thought: add context using a retriever (which contains text scraped from web)\n",
    "\n",
    "Steps:\n",
    "1. create documents (scrape from web)\n",
    "2. index documents\n",
    "3. create a retriever\n",
    "4. write q&a prompt\n",
    "5. create retrieval chain\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "! pip install beautifulsoup4 --quiet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.document_loaders import WebBaseLoader\n",
    "loader = WebBaseLoader(\"https://docs.smith.langchain.com/user_guide\")\n",
    "\n",
    "docs = loader.load()\n",
    "\n",
    "# print(docs[0].page_content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "! pip install cohere --quiet\n",
    "! pip install faiss-cpu --quiet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ[\"COHERE_API_KEY\"] = config[\"cohere_api_key\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_text_splitters import RecursiveCharacterTextSplitter\n",
    "from langchain_community.embeddings import CohereEmbeddings\n",
    "\n",
    "text_splitter = RecursiveCharacterTextSplitter()\n",
    "documents = text_splitter.split_documents(docs)\n",
    "\n",
    "embeddings = CohereEmbeddings()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.vectorstores import FAISS\n",
    "\n",
    "vectorstore = FAISS.from_documents(documents, embeddings)\n",
    "retriever = vectorstore.as_retriever()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "prompt = ChatPromptTemplate.from_template(\"\"\"Answer the following question based only on the provided context:\n",
    "\n",
    "<context>\n",
    "{context}\n",
    "</context>\n",
    "\n",
    "Question: {input}\"\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.chains.combine_documents import create_stuff_documents_chain\n",
    "from langchain.chains import create_retrieval_chain\n",
    "\n",
    "document_chain = create_stuff_documents_chain(llm, prompt)\n",
    "\n",
    "retrieval_chain = create_retrieval_chain(retriever, document_chain)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyError",
     "evalue": "\"Input to ChatPromptTemplate is missing variables {'chat_history'}.  Expected: ['chat_history', 'context', 'input'] Received: ['input', 'context']\"",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[31], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m response \u001b[38;5;241m=\u001b[39m \u001b[43mretrieval_chain\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minvoke\u001b[49m\u001b[43m(\u001b[49m\u001b[43m{\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43minput\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m:\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mhow can langsmith help with testing?\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m}\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:4069\u001b[0m, in \u001b[0;36mRunnableBindingBase.invoke\u001b[0;34m(self, input, config, **kwargs)\u001b[0m\n\u001b[1;32m   4063\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21minvoke\u001b[39m(\n\u001b[1;32m   4064\u001b[0m     \u001b[38;5;28mself\u001b[39m,\n\u001b[1;32m   4065\u001b[0m     \u001b[38;5;28minput\u001b[39m: Input,\n\u001b[1;32m   4066\u001b[0m     config: Optional[RunnableConfig] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m   4067\u001b[0m     \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs: Optional[Any],\n\u001b[1;32m   4068\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Output:\n\u001b[0;32m-> 4069\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbound\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minvoke\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   4070\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m   4071\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_merge_configs\u001b[49m\u001b[43m(\u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   4072\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43m{\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m}\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   4073\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:2075\u001b[0m, in \u001b[0;36mRunnableSequence.invoke\u001b[0;34m(self, input, config)\u001b[0m\n\u001b[1;32m   2073\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   2074\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m i, step \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msteps):\n\u001b[0;32m-> 2075\u001b[0m         \u001b[38;5;28minput\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[43mstep\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minvoke\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   2076\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2077\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;66;43;03m# mark each step as a child run\u001b[39;49;00m\n\u001b[1;32m   2078\u001b[0m \u001b[43m            \u001b[49m\u001b[43mpatch_config\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   2079\u001b[0m \u001b[43m                \u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcallbacks\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrun_manager\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_child\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43mf\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mseq:step:\u001b[39;49m\u001b[38;5;132;43;01m{\u001b[39;49;00m\u001b[43mi\u001b[49m\u001b[38;5;241;43m+\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[38;5;132;43;01m}\u001b[39;49;00m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m   2080\u001b[0m \u001b[43m            \u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2081\u001b[0m \u001b[43m        \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   2082\u001b[0m \u001b[38;5;66;03m# finish the root run\u001b[39;00m\n\u001b[1;32m   2083\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/passthrough.py:419\u001b[0m, in \u001b[0;36mRunnableAssign.invoke\u001b[0;34m(self, input, config, **kwargs)\u001b[0m\n\u001b[1;32m    413\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21minvoke\u001b[39m(\n\u001b[1;32m    414\u001b[0m     \u001b[38;5;28mself\u001b[39m,\n\u001b[1;32m    415\u001b[0m     \u001b[38;5;28minput\u001b[39m: Dict[\u001b[38;5;28mstr\u001b[39m, Any],\n\u001b[1;32m    416\u001b[0m     config: Optional[RunnableConfig] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m    417\u001b[0m     \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs: Any,\n\u001b[1;32m    418\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Dict[\u001b[38;5;28mstr\u001b[39m, Any]:\n\u001b[0;32m--> 419\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_with_config\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_invoke\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:1262\u001b[0m, in \u001b[0;36mRunnable._call_with_config\u001b[0;34m(self, func, input, config, run_type, **kwargs)\u001b[0m\n\u001b[1;32m   1258\u001b[0m     context \u001b[38;5;241m=\u001b[39m copy_context()\n\u001b[1;32m   1259\u001b[0m     context\u001b[38;5;241m.\u001b[39mrun(var_child_runnable_config\u001b[38;5;241m.\u001b[39mset, child_config)\n\u001b[1;32m   1260\u001b[0m     output \u001b[38;5;241m=\u001b[39m cast(\n\u001b[1;32m   1261\u001b[0m         Output,\n\u001b[0;32m-> 1262\u001b[0m         \u001b[43mcontext\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   1263\u001b[0m \u001b[43m            \u001b[49m\u001b[43mcall_func_with_variable_args\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1264\u001b[0m \u001b[43m            \u001b[49m\u001b[43mfunc\u001b[49m\u001b[43m,\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# type: ignore[arg-type]\u001b[39;49;00m\n\u001b[1;32m   1265\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# type: ignore[arg-type]\u001b[39;49;00m\n\u001b[1;32m   1266\u001b[0m \u001b[43m            \u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1267\u001b[0m \u001b[43m            \u001b[49m\u001b[43mrun_manager\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1268\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1269\u001b[0m \u001b[43m        \u001b[49m\u001b[43m)\u001b[49m,\n\u001b[1;32m   1270\u001b[0m     )\n\u001b[1;32m   1271\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[1;32m   1272\u001b[0m     run_manager\u001b[38;5;241m.\u001b[39mon_chain_error(e)\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/config.py:326\u001b[0m, in \u001b[0;36mcall_func_with_variable_args\u001b[0;34m(func, input, config, run_manager, **kwargs)\u001b[0m\n\u001b[1;32m    324\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m run_manager \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m accepts_run_manager(func):\n\u001b[1;32m    325\u001b[0m     kwargs[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrun_manager\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;241m=\u001b[39m run_manager\n\u001b[0;32m--> 326\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/passthrough.py:406\u001b[0m, in \u001b[0;36mRunnableAssign._invoke\u001b[0;34m(self, input, run_manager, config, **kwargs)\u001b[0m\n\u001b[1;32m    393\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_invoke\u001b[39m(\n\u001b[1;32m    394\u001b[0m     \u001b[38;5;28mself\u001b[39m,\n\u001b[1;32m    395\u001b[0m     \u001b[38;5;28minput\u001b[39m: Dict[\u001b[38;5;28mstr\u001b[39m, Any],\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    398\u001b[0m     \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs: Any,\n\u001b[1;32m    399\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Dict[\u001b[38;5;28mstr\u001b[39m, Any]:\n\u001b[1;32m    400\u001b[0m     \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(\n\u001b[1;32m    401\u001b[0m         \u001b[38;5;28minput\u001b[39m, \u001b[38;5;28mdict\u001b[39m\n\u001b[1;32m    402\u001b[0m     ), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mThe input to RunnablePassthrough.assign() must be a dict.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    404\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m {\n\u001b[1;32m    405\u001b[0m         \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39m\u001b[38;5;28minput\u001b[39m,\n\u001b[0;32m--> 406\u001b[0m         \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mmapper\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minvoke\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    407\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    408\u001b[0m \u001b[43m            \u001b[49m\u001b[43mpatch_config\u001b[49m\u001b[43m(\u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcallbacks\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrun_manager\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_child\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    409\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    410\u001b[0m \u001b[43m        \u001b[49m\u001b[43m)\u001b[49m,\n\u001b[1;32m    411\u001b[0m     }\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:2712\u001b[0m, in \u001b[0;36mRunnableParallel.invoke\u001b[0;34m(self, input, config)\u001b[0m\n\u001b[1;32m   2699\u001b[0m     \u001b[38;5;28;01mwith\u001b[39;00m get_executor_for_config(config) \u001b[38;5;28;01mas\u001b[39;00m executor:\n\u001b[1;32m   2700\u001b[0m         futures \u001b[38;5;241m=\u001b[39m [\n\u001b[1;32m   2701\u001b[0m             executor\u001b[38;5;241m.\u001b[39msubmit(\n\u001b[1;32m   2702\u001b[0m                 step\u001b[38;5;241m.\u001b[39minvoke,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   2710\u001b[0m             \u001b[38;5;28;01mfor\u001b[39;00m key, step \u001b[38;5;129;01min\u001b[39;00m steps\u001b[38;5;241m.\u001b[39mitems()\n\u001b[1;32m   2711\u001b[0m         ]\n\u001b[0;32m-> 2712\u001b[0m         output \u001b[38;5;241m=\u001b[39m {key: future\u001b[38;5;241m.\u001b[39mresult() \u001b[38;5;28;01mfor\u001b[39;00m key, future \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mzip\u001b[39m(steps, futures)}\n\u001b[1;32m   2713\u001b[0m \u001b[38;5;66;03m# finish the root run\u001b[39;00m\n\u001b[1;32m   2714\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:2712\u001b[0m, in \u001b[0;36m<dictcomp>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m   2699\u001b[0m     \u001b[38;5;28;01mwith\u001b[39;00m get_executor_for_config(config) \u001b[38;5;28;01mas\u001b[39;00m executor:\n\u001b[1;32m   2700\u001b[0m         futures \u001b[38;5;241m=\u001b[39m [\n\u001b[1;32m   2701\u001b[0m             executor\u001b[38;5;241m.\u001b[39msubmit(\n\u001b[1;32m   2702\u001b[0m                 step\u001b[38;5;241m.\u001b[39minvoke,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   2710\u001b[0m             \u001b[38;5;28;01mfor\u001b[39;00m key, step \u001b[38;5;129;01min\u001b[39;00m steps\u001b[38;5;241m.\u001b[39mitems()\n\u001b[1;32m   2711\u001b[0m         ]\n\u001b[0;32m-> 2712\u001b[0m         output \u001b[38;5;241m=\u001b[39m {key: \u001b[43mfuture\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mresult\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m \u001b[38;5;28;01mfor\u001b[39;00m key, future \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mzip\u001b[39m(steps, futures)}\n\u001b[1;32m   2713\u001b[0m \u001b[38;5;66;03m# finish the root run\u001b[39;00m\n\u001b[1;32m   2714\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/concurrent/futures/_base.py:458\u001b[0m, in \u001b[0;36mFuture.result\u001b[0;34m(self, timeout)\u001b[0m\n\u001b[1;32m    456\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m CancelledError()\n\u001b[1;32m    457\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_state \u001b[38;5;241m==\u001b[39m FINISHED:\n\u001b[0;32m--> 458\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m__get_result\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    459\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    460\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mTimeoutError\u001b[39;00m()\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/concurrent/futures/_base.py:403\u001b[0m, in \u001b[0;36mFuture.__get_result\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    401\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_exception:\n\u001b[1;32m    402\u001b[0m     \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m--> 403\u001b[0m         \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_exception\n\u001b[1;32m    404\u001b[0m     \u001b[38;5;28;01mfinally\u001b[39;00m:\n\u001b[1;32m    405\u001b[0m         \u001b[38;5;66;03m# Break a reference cycle with the exception in self._exception\u001b[39;00m\n\u001b[1;32m    406\u001b[0m         \u001b[38;5;28mself\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/concurrent/futures/thread.py:58\u001b[0m, in \u001b[0;36m_WorkItem.run\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     55\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m\n\u001b[1;32m     57\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m---> 58\u001b[0m     result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     59\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m exc:\n\u001b[1;32m     60\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mfuture\u001b[38;5;241m.\u001b[39mset_exception(exc)\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:4069\u001b[0m, in \u001b[0;36mRunnableBindingBase.invoke\u001b[0;34m(self, input, config, **kwargs)\u001b[0m\n\u001b[1;32m   4063\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21minvoke\u001b[39m(\n\u001b[1;32m   4064\u001b[0m     \u001b[38;5;28mself\u001b[39m,\n\u001b[1;32m   4065\u001b[0m     \u001b[38;5;28minput\u001b[39m: Input,\n\u001b[1;32m   4066\u001b[0m     config: Optional[RunnableConfig] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m   4067\u001b[0m     \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs: Optional[Any],\n\u001b[1;32m   4068\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Output:\n\u001b[0;32m-> 4069\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbound\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minvoke\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   4070\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m   4071\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_merge_configs\u001b[49m\u001b[43m(\u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   4072\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43m{\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m}\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   4073\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:2075\u001b[0m, in \u001b[0;36mRunnableSequence.invoke\u001b[0;34m(self, input, config)\u001b[0m\n\u001b[1;32m   2073\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m   2074\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m i, step \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msteps):\n\u001b[0;32m-> 2075\u001b[0m         \u001b[38;5;28minput\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[43mstep\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minvoke\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   2076\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2077\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;66;43;03m# mark each step as a child run\u001b[39;49;00m\n\u001b[1;32m   2078\u001b[0m \u001b[43m            \u001b[49m\u001b[43mpatch_config\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   2079\u001b[0m \u001b[43m                \u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcallbacks\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrun_manager\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_child\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43mf\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mseq:step:\u001b[39;49m\u001b[38;5;132;43;01m{\u001b[39;49;00m\u001b[43mi\u001b[49m\u001b[38;5;241;43m+\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[38;5;132;43;01m}\u001b[39;49;00m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m   2080\u001b[0m \u001b[43m            \u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2081\u001b[0m \u001b[43m        \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   2082\u001b[0m \u001b[38;5;66;03m# finish the root run\u001b[39;00m\n\u001b[1;32m   2083\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/prompts/base.py:113\u001b[0m, in \u001b[0;36mBasePromptTemplate.invoke\u001b[0;34m(self, input, config)\u001b[0m\n\u001b[1;32m    111\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtags:\n\u001b[1;32m    112\u001b[0m     config[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtags\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mextend(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtags)\n\u001b[0;32m--> 113\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_with_config\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    114\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_format_prompt_with_error_handling\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    115\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    116\u001b[0m \u001b[43m    \u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    117\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrun_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mprompt\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    118\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/base.py:1262\u001b[0m, in \u001b[0;36mRunnable._call_with_config\u001b[0;34m(self, func, input, config, run_type, **kwargs)\u001b[0m\n\u001b[1;32m   1258\u001b[0m     context \u001b[38;5;241m=\u001b[39m copy_context()\n\u001b[1;32m   1259\u001b[0m     context\u001b[38;5;241m.\u001b[39mrun(var_child_runnable_config\u001b[38;5;241m.\u001b[39mset, child_config)\n\u001b[1;32m   1260\u001b[0m     output \u001b[38;5;241m=\u001b[39m cast(\n\u001b[1;32m   1261\u001b[0m         Output,\n\u001b[0;32m-> 1262\u001b[0m         \u001b[43mcontext\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   1263\u001b[0m \u001b[43m            \u001b[49m\u001b[43mcall_func_with_variable_args\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1264\u001b[0m \u001b[43m            \u001b[49m\u001b[43mfunc\u001b[49m\u001b[43m,\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# type: ignore[arg-type]\u001b[39;49;00m\n\u001b[1;32m   1265\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# type: ignore[arg-type]\u001b[39;49;00m\n\u001b[1;32m   1266\u001b[0m \u001b[43m            \u001b[49m\u001b[43mconfig\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1267\u001b[0m \u001b[43m            \u001b[49m\u001b[43mrun_manager\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1268\u001b[0m \u001b[43m            \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1269\u001b[0m \u001b[43m        \u001b[49m\u001b[43m)\u001b[49m,\n\u001b[1;32m   1270\u001b[0m     )\n\u001b[1;32m   1271\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mBaseException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[1;32m   1272\u001b[0m     run_manager\u001b[38;5;241m.\u001b[39mon_chain_error(e)\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/runnables/config.py:326\u001b[0m, in \u001b[0;36mcall_func_with_variable_args\u001b[0;34m(func, input, config, run_manager, **kwargs)\u001b[0m\n\u001b[1;32m    324\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m run_manager \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m accepts_run_manager(func):\n\u001b[1;32m    325\u001b[0m     kwargs[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrun_manager\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;241m=\u001b[39m run_manager\n\u001b[0;32m--> 326\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/opt/conda/envs/pytorch/lib/python3.10/site-packages/langchain_core/prompts/base.py:98\u001b[0m, in \u001b[0;36mBasePromptTemplate._format_prompt_with_error_handling\u001b[0;34m(self, inner_input)\u001b[0m\n\u001b[1;32m     96\u001b[0m missing \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mset\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39minput_variables)\u001b[38;5;241m.\u001b[39mdifference(inner_input)\n\u001b[1;32m     97\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m missing:\n\u001b[0;32m---> 98\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mKeyError\u001b[39;00m(\n\u001b[1;32m     99\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mInput to \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__class__\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m is missing variables \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mmissing\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m. \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    100\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m Expected: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39minput_variables\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    101\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m Received: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mlist\u001b[39m(inner_input\u001b[38;5;241m.\u001b[39mkeys())\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    102\u001b[0m     )\n\u001b[1;32m    103\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mformat_prompt(\u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39minner_input)\n",
      "\u001b[0;31mKeyError\u001b[0m: \"Input to ChatPromptTemplate is missing variables {'chat_history'}.  Expected: ['chat_history', 'context', 'input'] Received: ['input', 'context']\""
     ]
    }
   ],
   "source": [
    "response = retrieval_chain.invoke({\"input\": \"how can langsmith help with testing?\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'input': 'how can langsmith help with testing?',\n",
       " 'context': [Document(page_content='Skip to main content🦜️🛠️ LangSmith DocsLangChain Python DocsLangChain JS/TS DocsLangSmith API DocsSearchGo to AppLangSmithUser GuideSetupPricing (Coming Soon)Self-HostingTracingEvaluationMonitoringPrompt HubProxyUser GuideOn this pageLangSmith User GuideLangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.Prototyping\\u200bPrototyping LLM applications often involves quick experimentation between prompts, model types, retrieval strategy and other parameters.\\nThe ability to rapidly understand how the model is performing — and debug where it is failing — is incredibly important for this phase.Debugging\\u200bWhen developing new LLM applications, we suggest having LangSmith tracing enabled by default.\\nOftentimes, it isn’t necessary to look at every single trace. However, when things go wrong (an unexpected end result, infinite agent loop, slower than expected execution, higher than expected token usage), it’s extremely helpful to debug by looking through the application traces. LangSmith gives clear visibility and debugging information at each step of an LLM sequence, making it much easier to identify and root-cause issues.\\nWe provide native rendering of chat messages, functions, and retrieve documents.Initial Test Set\\u200bWhile many developers still ship an initial version of their application based on “vibe checks”, we’ve seen an increasing number of engineering teams start to adopt a more test driven approach. LangSmith allows developers to create datasets, which are collections of inputs and reference outputs, and use these to run tests on their LLM applications.\\nThese test cases can be uploaded in bulk, created on the fly, or exported from application traces. LangSmith also makes it easy to run custom evaluations (both LLM and heuristic based) to score test results.Comparison View\\u200bWhen prototyping different versions of your applications and making changes, it’s important to see whether or not you’ve regressed with respect to your initial test cases.\\nOftentimes, changes in the prompt, retrieval strategy, or model choice can have huge implications in responses produced by your application.\\nIn order to get a sense for which variant is performing better, it’s useful to be able to view results for different configurations on the same datapoints side-by-side. We’ve invested heavily in a user-friendly comparison view for test runs to track and diagnose regressions in test scores across multiple revisions of your application.Playground\\u200bLangSmith provides a playground environment for rapid iteration and experimentation.\\nThis allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.', metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'}),\n",
       "  Document(page_content=\"This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\\nEvery playground run is logged in the system and can be used to create test cases or compare with other runs.Beta Testing\\u200bBeta testing allows developers to collect more data on how their LLM applications are performing in real-world scenarios. In this phase, it’s important to develop an understanding for the types of inputs the app is performing well or poorly on and how exactly it’s breaking down in those cases. Both feedback collection and run annotation are critical for this workflow. This will help in curation of test cases that can help track regressions/improvements and development of automatic evaluations.Collecting Feedback\\u200bWhen launching your application to an initial set of users, it’s important to gather human feedback on the responses it’s producing. This helps draw attention to the most interesting runs and highlight edge cases that are causing problematic responses. LangSmith allows you to attach feedback scores to logged traces (oftentimes, this is hooked up to a feedback button in your app), then filter on traces that have a specific feedback tag and score. A common workflow is to filter on traces that receive a poor user feedback score, then drill down into problematic points using the detailed trace view.Annotating Traces\\u200bLangSmith also supports sending runs to annotation queues, which allow annotators to closely inspect interesting traces and annotate them with respect to different criteria. Annotators can be PMs, engineers, or even subject matter experts. This allows users to catch regressions across important evaluation criteria.Adding Runs to a Dataset\\u200bAs your application progresses through the beta testing phase, it's essential to continue collecting data to refine and improve its performance. LangSmith enables you to add runs as examples to datasets (from both the project page and within an annotation queue), expanding your test coverage on real-world scenarios. This is a key benefit in having your logging system and your evaluation/testing system in the same platform.Production\\u200bClosely inspecting key data points, growing benchmarking datasets, annotating traces, and drilling down into important data in trace view are workflows you’ll also want to do once your app hits production. However, especially at the production stage, it’s crucial to get a high-level overview of application performance with respect to latency, cost, and feedback scores. This ensures that it's delivering desirable results at scale.Monitoring and A/B Testing\\u200bLangSmith provides monitoring charts that allow you to track key metrics over time. You can expand to view metrics for a given period and drill down into a specific data point to get a trace table for that time period — this is especially handy for debugging production issues.LangSmith also allows for tag and metadata grouping, which allows users to mark different versions of their applications with different identifiers and view how they are performing side-by-side within each chart. This is helpful for A/B testing changes in prompt, model, or retrieval strategy.Was this page helpful?PreviousLangSmithNextSetupPrototypingBeta TestingProductionCommunityDiscordTwitterGitHubDocs CodeLangSmith SDKPythonJS/TSMoreHomepageBlogCopyright © 2024 LangChain, Inc.\", metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'}),\n",
       "  Document(page_content='LangSmith User Guide | 🦜️🛠️ LangSmith', metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'})],\n",
       " 'answer': 'According to the context provided, LangSmith can help with testing LLM applications in the following ways:\\n\\n1. Initial Test Set: LangSmith allows developers to create datasets which are collections of inputs and reference outputs. These datasets can be used to run tests on their LLM applications. The test cases can be uploaded in bulk, created on the fly, or exported from application traces.\\n\\n2. Comparison View: When making changes to the application, LangSmith provides a user-friendly comparison view that allows side-by-side comparison of test results across multiple revisions of the application. This'}"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "response"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "According to the context provided, LangSmith can help with testing LLM applications in the following ways:\n",
      "\n",
      "1. Initial Test Set: LangSmith allows developers to create datasets which are collections of inputs and reference outputs. These datasets can be used to run tests on their LLM applications. The test cases can be uploaded in bulk, created on the fly, or exported from application traces.\n",
      "\n",
      "2. Comparison View: When making changes to the application, LangSmith provides a user-friendly comparison view that allows side-by-side comparison of test results across multiple revisions of the application. This\n"
     ]
    }
   ],
   "source": [
    "print(response[\"answer\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### conversational retrieval chain"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.chains import create_history_aware_retriever\n",
    "from langchain_core.prompts import MessagesPlaceholder, ChatPromptTemplate\n",
    "\n",
    "\n",
    "prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        MessagesPlaceholder(variable_name=\"chat_history\"),\n",
    "        (\n",
    "            \"user\", \n",
    "            \"{input}\"\n",
    "            \"Given the above conversation, generate a search query to look up in order to get information relevant to the conversation\"\n",
    "        )\n",
    "    ]\n",
    ")\n",
    "\n",
    "# prompt.messages[2].pretty_print()\n",
    "\n",
    "retriever_chain = create_history_aware_retriever(llm, retriever, prompt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(page_content='Skip to main content🦜️🛠️ LangSmith DocsLangChain Python DocsLangChain JS/TS DocsLangSmith API DocsSearchGo to AppLangSmithUser GuideSetupPricing (Coming Soon)Self-HostingTracingEvaluationMonitoringPrompt HubProxyUser GuideOn this pageLangSmith User GuideLangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.Prototyping\\u200bPrototyping LLM applications often involves quick experimentation between prompts, model types, retrieval strategy and other parameters.\\nThe ability to rapidly understand how the model is performing — and debug where it is failing — is incredibly important for this phase.Debugging\\u200bWhen developing new LLM applications, we suggest having LangSmith tracing enabled by default.\\nOftentimes, it isn’t necessary to look at every single trace. However, when things go wrong (an unexpected end result, infinite agent loop, slower than expected execution, higher than expected token usage), it’s extremely helpful to debug by looking through the application traces. LangSmith gives clear visibility and debugging information at each step of an LLM sequence, making it much easier to identify and root-cause issues.\\nWe provide native rendering of chat messages, functions, and retrieve documents.Initial Test Set\\u200bWhile many developers still ship an initial version of their application based on “vibe checks”, we’ve seen an increasing number of engineering teams start to adopt a more test driven approach. LangSmith allows developers to create datasets, which are collections of inputs and reference outputs, and use these to run tests on their LLM applications.\\nThese test cases can be uploaded in bulk, created on the fly, or exported from application traces. LangSmith also makes it easy to run custom evaluations (both LLM and heuristic based) to score test results.Comparison View\\u200bWhen prototyping different versions of your applications and making changes, it’s important to see whether or not you’ve regressed with respect to your initial test cases.\\nOftentimes, changes in the prompt, retrieval strategy, or model choice can have huge implications in responses produced by your application.\\nIn order to get a sense for which variant is performing better, it’s useful to be able to view results for different configurations on the same datapoints side-by-side. We’ve invested heavily in a user-friendly comparison view for test runs to track and diagnose regressions in test scores across multiple revisions of your application.Playground\\u200bLangSmith provides a playground environment for rapid iteration and experimentation.\\nThis allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.', metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'}),\n",
       " Document(page_content='LangSmith User Guide | 🦜️🛠️ LangSmith', metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'}),\n",
       " Document(page_content=\"This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\\nEvery playground run is logged in the system and can be used to create test cases or compare with other runs.Beta Testing\\u200bBeta testing allows developers to collect more data on how their LLM applications are performing in real-world scenarios. In this phase, it’s important to develop an understanding for the types of inputs the app is performing well or poorly on and how exactly it’s breaking down in those cases. Both feedback collection and run annotation are critical for this workflow. This will help in curation of test cases that can help track regressions/improvements and development of automatic evaluations.Collecting Feedback\\u200bWhen launching your application to an initial set of users, it’s important to gather human feedback on the responses it’s producing. This helps draw attention to the most interesting runs and highlight edge cases that are causing problematic responses. LangSmith allows you to attach feedback scores to logged traces (oftentimes, this is hooked up to a feedback button in your app), then filter on traces that have a specific feedback tag and score. A common workflow is to filter on traces that receive a poor user feedback score, then drill down into problematic points using the detailed trace view.Annotating Traces\\u200bLangSmith also supports sending runs to annotation queues, which allow annotators to closely inspect interesting traces and annotate them with respect to different criteria. Annotators can be PMs, engineers, or even subject matter experts. This allows users to catch regressions across important evaluation criteria.Adding Runs to a Dataset\\u200bAs your application progresses through the beta testing phase, it's essential to continue collecting data to refine and improve its performance. LangSmith enables you to add runs as examples to datasets (from both the project page and within an annotation queue), expanding your test coverage on real-world scenarios. This is a key benefit in having your logging system and your evaluation/testing system in the same platform.Production\\u200bClosely inspecting key data points, growing benchmarking datasets, annotating traces, and drilling down into important data in trace view are workflows you’ll also want to do once your app hits production. However, especially at the production stage, it’s crucial to get a high-level overview of application performance with respect to latency, cost, and feedback scores. This ensures that it's delivering desirable results at scale.Monitoring and A/B Testing\\u200bLangSmith provides monitoring charts that allow you to track key metrics over time. You can expand to view metrics for a given period and drill down into a specific data point to get a trace table for that time period — this is especially handy for debugging production issues.LangSmith also allows for tag and metadata grouping, which allows users to mark different versions of their applications with different identifiers and view how they are performing side-by-side within each chart. This is helpful for A/B testing changes in prompt, model, or retrieval strategy.Was this page helpful?PreviousLangSmithNextSetupPrototypingBeta TestingProductionCommunityDiscordTwitterGitHubDocs CodeLangSmith SDKPythonJS/TSMoreHomepageBlogCopyright © 2024 LangChain, Inc.\", metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'})]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from langchain_core.messages import HumanMessage, AIMessage\n",
    "\n",
    "chat_history = [\n",
    "    HumanMessage(\n",
    "        # name=\"H\", \n",
    "        content=\"Can LangSmith help test my LLM applications?\"\n",
    "    ), \n",
    "    AIMessage(\n",
    "        # name=\"A\", \n",
    "        content=\"Yes!\"\n",
    "    )    \n",
    "]\n",
    "retriever_chain.invoke({\n",
    "    \"chat_history\": chat_history,\n",
    "    \"input\": \"Tell me how\"\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "prompt = ChatPromptTemplate.from_messages([\n",
    "    (\"system\", \"Answer the user's questions based on the below context:\\n\\n{context}\"),\n",
    "    MessagesPlaceholder(variable_name=\"chat_history\"),\n",
    "    (\"user\", \"{input}\"),\n",
    "])\n",
    "document_chain = create_stuff_documents_chain(llm, prompt)\n",
    "retrieval_chain = create_retrieval_chain(retriever_chain, document_chain)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'chat_history': [HumanMessage(content='Can LangSmith help test my LLM applications?'),\n",
       "  AIMessage(content='Yes!')],\n",
       " 'input': 'Tell me how',\n",
       " 'context': [Document(page_content='Skip to main content🦜️🛠️ LangSmith DocsLangChain Python DocsLangChain JS/TS DocsLangSmith API DocsSearchGo to AppLangSmithUser GuideSetupPricing (Coming Soon)Self-HostingTracingEvaluationMonitoringPrompt HubProxyUser GuideOn this pageLangSmith User GuideLangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.Prototyping\\u200bPrototyping LLM applications often involves quick experimentation between prompts, model types, retrieval strategy and other parameters.\\nThe ability to rapidly understand how the model is performing — and debug where it is failing — is incredibly important for this phase.Debugging\\u200bWhen developing new LLM applications, we suggest having LangSmith tracing enabled by default.\\nOftentimes, it isn’t necessary to look at every single trace. However, when things go wrong (an unexpected end result, infinite agent loop, slower than expected execution, higher than expected token usage), it’s extremely helpful to debug by looking through the application traces. LangSmith gives clear visibility and debugging information at each step of an LLM sequence, making it much easier to identify and root-cause issues.\\nWe provide native rendering of chat messages, functions, and retrieve documents.Initial Test Set\\u200bWhile many developers still ship an initial version of their application based on “vibe checks”, we’ve seen an increasing number of engineering teams start to adopt a more test driven approach. LangSmith allows developers to create datasets, which are collections of inputs and reference outputs, and use these to run tests on their LLM applications.\\nThese test cases can be uploaded in bulk, created on the fly, or exported from application traces. LangSmith also makes it easy to run custom evaluations (both LLM and heuristic based) to score test results.Comparison View\\u200bWhen prototyping different versions of your applications and making changes, it’s important to see whether or not you’ve regressed with respect to your initial test cases.\\nOftentimes, changes in the prompt, retrieval strategy, or model choice can have huge implications in responses produced by your application.\\nIn order to get a sense for which variant is performing better, it’s useful to be able to view results for different configurations on the same datapoints side-by-side. We’ve invested heavily in a user-friendly comparison view for test runs to track and diagnose regressions in test scores across multiple revisions of your application.Playground\\u200bLangSmith provides a playground environment for rapid iteration and experimentation.\\nThis allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.', metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'}),\n",
       "  Document(page_content='LangSmith User Guide | 🦜️🛠️ LangSmith', metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'}),\n",
       "  Document(page_content=\"This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\\nEvery playground run is logged in the system and can be used to create test cases or compare with other runs.Beta Testing\\u200bBeta testing allows developers to collect more data on how their LLM applications are performing in real-world scenarios. In this phase, it’s important to develop an understanding for the types of inputs the app is performing well or poorly on and how exactly it’s breaking down in those cases. Both feedback collection and run annotation are critical for this workflow. This will help in curation of test cases that can help track regressions/improvements and development of automatic evaluations.Collecting Feedback\\u200bWhen launching your application to an initial set of users, it’s important to gather human feedback on the responses it’s producing. This helps draw attention to the most interesting runs and highlight edge cases that are causing problematic responses. LangSmith allows you to attach feedback scores to logged traces (oftentimes, this is hooked up to a feedback button in your app), then filter on traces that have a specific feedback tag and score. A common workflow is to filter on traces that receive a poor user feedback score, then drill down into problematic points using the detailed trace view.Annotating Traces\\u200bLangSmith also supports sending runs to annotation queues, which allow annotators to closely inspect interesting traces and annotate them with respect to different criteria. Annotators can be PMs, engineers, or even subject matter experts. This allows users to catch regressions across important evaluation criteria.Adding Runs to a Dataset\\u200bAs your application progresses through the beta testing phase, it's essential to continue collecting data to refine and improve its performance. LangSmith enables you to add runs as examples to datasets (from both the project page and within an annotation queue), expanding your test coverage on real-world scenarios. This is a key benefit in having your logging system and your evaluation/testing system in the same platform.Production\\u200bClosely inspecting key data points, growing benchmarking datasets, annotating traces, and drilling down into important data in trace view are workflows you’ll also want to do once your app hits production. However, especially at the production stage, it’s crucial to get a high-level overview of application performance with respect to latency, cost, and feedback scores. This ensures that it's delivering desirable results at scale.Monitoring and A/B Testing\\u200bLangSmith provides monitoring charts that allow you to track key metrics over time. You can expand to view metrics for a given period and drill down into a specific data point to get a trace table for that time period — this is especially handy for debugging production issues.LangSmith also allows for tag and metadata grouping, which allows users to mark different versions of their applications with different identifiers and view how they are performing side-by-side within each chart. This is helpful for A/B testing changes in prompt, model, or retrieval strategy.Was this page helpful?PreviousLangSmithNextSetupPrototypingBeta TestingProductionCommunityDiscordTwitterGitHubDocs CodeLangSmith SDKPythonJS/TSMoreHomepageBlogCopyright © 2024 LangChain, Inc.\", metadata={'source': 'https://docs.smith.langchain.com/user_guide', 'title': 'LangSmith User Guide | 🦜️🛠️ LangSmith', 'description': 'LangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.', 'language': 'en'})],\n",
       " 'answer': 'Yes, LangSmith provides several powerful features to help test and evaluate LLM applications:\\n\\n1. Initial Test Set\\nLangSmith allows you to create datasets which are collections of inputs and reference outputs. You can use these datasets to run tests on your LLM application and evaluate its performance against the expected outputs. Test cases can be uploaded in bulk, created on the fly, or exported from application traces.\\n\\n2. Comparison View\\nWhen making changes to your application (prompts, retrieval strategy, model choice etc.), LangSmith provides a user-friendly comparison view that lets you see the'}"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chat_history = [HumanMessage(content=\"Can LangSmith help test my LLM applications?\"), AIMessage(content=\"Yes!\")]\n",
    "retrieval_chain.invoke({\n",
    "    \"chat_history\": chat_history,\n",
    "    \"input\": \"Tell me how\"\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### agents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m\u001b[33mWARNING: Ignoring invalid distribution -ackaging (/opt/conda/envs/pytorch/lib/python3.10/site-packages)\u001b[0m\u001b[33m\n",
      "\u001b[0m\u001b[31mERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.\n",
      "botocore 1.34.30 requires urllib3<2.1,>=1.25.4; python_version >= \"3.10\", but you have urllib3 2.2.1 which is incompatible.\n",
      "sagemaker 2.205.0 requires urllib3<1.27, but you have urllib3 2.2.1 which is incompatible.\u001b[0m\u001b[31m\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "! pip install langchainhub --quiet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.tools.retriever import create_retriever_tool\n",
    "\n",
    "retriever_tool = create_retriever_tool(\n",
    "    retriever,\n",
    "    \"langsmith_search\",\n",
    "    \"Search for information about LangSmith. For any questions about LangSmith, you must use this tool!\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.tools.tavily_search import TavilySearchResults\n",
    "os.environ['TAVILY_API_KEY'] = config[\"tavily_api_key\"]\n",
    "search = TavilySearchResults(max_results=1)\n",
    "\n",
    "# name: str = \"tavily_search_results_json\"\n",
    "# description: str = (\n",
    "#     \"A search engine optimized for comprehensive, accurate, and trusted results. \"\n",
    "#     \"Useful for when you need to answer questions about current events. \"\n",
    "#     \"Input should be a search query.\"\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Tool(name='langsmith_search', description='Search for information about LangSmith. For any questions about LangSmith, you must use this tool!', args_schema=<class 'langchain.tools.retriever.RetrieverInput'>, func=functools.partial(<function _get_relevant_documents at 0x7f3f282a9480>, retriever=VectorStoreRetriever(tags=['FAISS', 'CohereEmbeddings'], vectorstore=<langchain_community.vectorstores.faiss.FAISS object at 0x7f3f2a94ac80>), document_prompt=PromptTemplate(input_variables=['page_content'], template='{page_content}'), document_separator='\\n\\n'), coroutine=functools.partial(<function _aget_relevant_documents at 0x7f3f282a9510>, retriever=VectorStoreRetriever(tags=['FAISS', 'CohereEmbeddings'], vectorstore=<langchain_community.vectorstores.faiss.FAISS object at 0x7f3f2a94ac80>), document_prompt=PromptTemplate(input_variables=['page_content'], template='{page_content}'), document_separator='\\n\\n')),\n",
       " TavilySearchResults(max_results=1)]"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tools = [retriever_tool, search]\n",
    "\n",
    "tools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from langchain import hub\n",
    "\n",
    "# # prompt = hub.pull(\"hwchase17/openai-functions-agent\")\n",
    "# # prompt = hub.pull(\"hwchase17/react\")\n",
    "# prompt = hub.pull(\"hwchase17/react-chat\")\n",
    "# prompt.pretty_print()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "prompt_template = \"\"\"Assistant is a large language model trained by Anthropic.\n",
    "\n",
    "Assistant is designed to be able to assist with a wide range of tasks, from answering simple questions to providing in-depth explanations and discussions on a wide range of topics. As a language model, Assistant is able to generate human-like text based on the input it receives, allowing it to engage in natural-sounding conversations and provide responses that are coherent and relevant to the topic at hand.\n",
    "\n",
    "Assistant is constantly learning and improving, and its capabilities are constantly evolving. It is able to process and understand large amounts of text, and can use this knowledge to provide accurate and informative responses to a wide range of questions. Additionally, Assistant is able to generate its own text based on the input it receives, allowing it to engage in discussions and provide explanations and descriptions on a wide range of topics.\n",
    "\n",
    "Overall, Assistant is a powerful tool that can help with a wide range of tasks and provide valuable insights and information on a wide range of topics. Whether you need help with a specific question or just want to have a conversation about a particular topic, Assistant is here to assist.\n",
    "\n",
    "TOOLS:\n",
    "------\n",
    "\n",
    "Assistant has access to the following tools:\n",
    "\n",
    "{tools}\n",
    "\n",
    "To use a tool, please use the following format:\n",
    "\n",
    "```\n",
    "Thought: Do I need to use a tool? Yes\n",
    "Action: the action to take, should be one of [{tool_names}]\n",
    "Action Input: the input to the action\n",
    "Observation: the result of the action\n",
    "```\n",
    "\n",
    "When you have a response to say to the Human, or if you do not need to use a tool, you MUST use the format:\n",
    "\n",
    "```\n",
    "Thought: Do I need to use a tool? No\n",
    "Final Answer: [your response here]\n",
    "```\n",
    "\n",
    "Begin!\n",
    "\n",
    "Previous conversation history:\n",
    "{chat_history}\n",
    "\n",
    "New input: {input}\n",
    "{agent_scratchpad}\n",
    "\n",
    "\"\"\"\n",
    "\n",
    "\n",
    "prompt = PromptTemplate(\n",
    "    input_variables=[\n",
    "        \"tools\", \"tool_names\", \"chat_history\", \"input\", \"agent_scratchpad\"\n",
    "    ],\n",
    "    template=prompt_template\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "import boto3\n",
    "from langchain_community.chat_models import BedrockChat\n",
    "\n",
    "from langchain.agents import create_react_agent\n",
    "from langchain.agents import AgentExecutor\n",
    "\n",
    "\n",
    "# prompt = hub.pull(\"hwchase17/openai-functions-agent\")\n",
    "# prompt.pretty_print()\n",
    "llm = BedrockChat(\n",
    "    # model_id=\"anthropic.claude-3-sonnet-20240229-v1:0\",\n",
    "    model_id=\"anthropic.claude-3-haiku-20240307-v1:0\",\n",
    "    # model_id=\"anthropic.claude-instant-v1\",\n",
    "    # model_id=\"anthropic.claude-v2:1\",\n",
    "    client=boto3.client(\"bedrock-runtime\"),\n",
    "    model_kwargs={\n",
    "        \"max_tokens\": 128,\n",
    "        # \"max_tokens_to_sample\": 128,\n",
    "        \"temperature\": 0.0,\n",
    "    }\n",
    ")\n",
    "agent = create_react_agent(llm, tools, prompt)\n",
    "agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[32;1m\u001b[1;3mThought: I should use the langsmith_search tool to find information about how LangSmith can help with testing.\n",
      "\n",
      "Action: langsmith_search\n",
      "Action Input: langsmith testing\n",
      "\u001b[0m\u001b[36;1m\u001b[1;3mSkip to main content🦜️🛠️ LangSmith DocsLangChain Python DocsLangChain JS/TS DocsLangSmith API DocsSearchGo to AppLangSmithUser GuideSetupPricing (Coming Soon)Self-HostingTracingEvaluationMonitoringPrompt HubProxyUser GuideOn this pageLangSmith User GuideLangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.Prototyping​Prototyping LLM applications often involves quick experimentation between prompts, model types, retrieval strategy and other parameters.\n",
      "The ability to rapidly understand how the model is performing — and debug where it is failing — is incredibly important for this phase.Debugging​When developing new LLM applications, we suggest having LangSmith tracing enabled by default.\n",
      "Oftentimes, it isn’t necessary to look at every single trace. However, when things go wrong (an unexpected end result, infinite agent loop, slower than expected execution, higher than expected token usage), it’s extremely helpful to debug by looking through the application traces. LangSmith gives clear visibility and debugging information at each step of an LLM sequence, making it much easier to identify and root-cause issues.\n",
      "We provide native rendering of chat messages, functions, and retrieve documents.Initial Test Set​While many developers still ship an initial version of their application based on “vibe checks”, we’ve seen an increasing number of engineering teams start to adopt a more test driven approach. LangSmith allows developers to create datasets, which are collections of inputs and reference outputs, and use these to run tests on their LLM applications.\n",
      "These test cases can be uploaded in bulk, created on the fly, or exported from application traces. LangSmith also makes it easy to run custom evaluations (both LLM and heuristic based) to score test results.Comparison View​When prototyping different versions of your applications and making changes, it’s important to see whether or not you’ve regressed with respect to your initial test cases.\n",
      "Oftentimes, changes in the prompt, retrieval strategy, or model choice can have huge implications in responses produced by your application.\n",
      "In order to get a sense for which variant is performing better, it’s useful to be able to view results for different configurations on the same datapoints side-by-side. We’ve invested heavily in a user-friendly comparison view for test runs to track and diagnose regressions in test scores across multiple revisions of your application.Playground​LangSmith provides a playground environment for rapid iteration and experimentation.\n",
      "This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\n",
      "\n",
      "LangSmith User Guide | 🦜️🛠️ LangSmith\n",
      "\n",
      "This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\n",
      "Every playground run is logged in the system and can be used to create test cases or compare with other runs.Beta Testing​Beta testing allows developers to collect more data on how their LLM applications are performing in real-world scenarios. In this phase, it’s important to develop an understanding for the types of inputs the app is performing well or poorly on and how exactly it’s breaking down in those cases. Both feedback collection and run annotation are critical for this workflow. This will help in curation of test cases that can help track regressions/improvements and development of automatic evaluations.Collecting Feedback​When launching your application to an initial set of users, it’s important to gather human feedback on the responses it’s producing. This helps draw attention to the most interesting runs and highlight edge cases that are causing problematic responses. LangSmith allows you to attach feedback scores to logged traces (oftentimes, this is hooked up to a feedback button in your app), then filter on traces that have a specific feedback tag and score. A common workflow is to filter on traces that receive a poor user feedback score, then drill down into problematic points using the detailed trace view.Annotating Traces​LangSmith also supports sending runs to annotation queues, which allow annotators to closely inspect interesting traces and annotate them with respect to different criteria. Annotators can be PMs, engineers, or even subject matter experts. This allows users to catch regressions across important evaluation criteria.Adding Runs to a Dataset​As your application progresses through the beta testing phase, it's essential to continue collecting data to refine and improve its performance. LangSmith enables you to add runs as examples to datasets (from both the project page and within an annotation queue), expanding your test coverage on real-world scenarios. This is a key benefit in having your logging system and your evaluation/testing system in the same platform.Production​Closely inspecting key data points, growing benchmarking datasets, annotating traces, and drilling down into important data in trace view are workflows you’ll also want to do once your app hits production. However, especially at the production stage, it’s crucial to get a high-level overview of application performance with respect to latency, cost, and feedback scores. This ensures that it's delivering desirable results at scale.Monitoring and A/B Testing​LangSmith provides monitoring charts that allow you to track key metrics over time. You can expand to view metrics for a given period and drill down into a specific data point to get a trace table for that time period — this is especially handy for debugging production issues.LangSmith also allows for tag and metadata grouping, which allows users to mark different versions of their applications with different identifiers and view how they are performing side-by-side within each chart. This is helpful for A/B testing changes in prompt, model, or retrieval strategy.Was this page helpful?PreviousLangSmithNextSetupPrototypingBeta TestingProductionCommunityDiscordTwitterGitHubDocs CodeLangSmith SDKPythonJS/TSMoreHomepageBlogCopyright © 2024 LangChain, Inc.\u001b[0m\u001b[32;1m\u001b[1;3mBased on the information from the LangSmith user guide, here is how LangSmith can help with testing LLM applications:\n",
      "\n",
      "Thought: Do I need to use a tool? No\n",
      "Final Answer: LangSmith provides several key features to support testing of LLM applications:\n",
      "\n",
      "1. Debugging: LangSmith provides detailed tracing and visibility into each step of an LLM sequence, making it easier to identify and debug issues.\n",
      "\n",
      "2. Initial Test Sets: LangSmith allows developers to create datasets of inputs and reference outputs to run tests on their LLM applications\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'input': 'how can langsmith help with testing?',\n",
       " 'chat_history': [],\n",
       " 'output': 'LangSmith provides several key features to support testing of LLM applications:\\n\\n1. Debugging: LangSmith provides detailed tracing and visibility into each step of an LLM sequence, making it easier to identify and debug issues.\\n\\n2. Initial Test Sets: LangSmith allows developers to create datasets of inputs and reference outputs to run tests on their LLM applications'}"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent_executor.invoke({\n",
    "    \"input\": \"how can langsmith help with testing?\",\n",
    "    \"chat_history\": []\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[32;1m\u001b[1;3mThought: Do I need to use a tool? Yes\n",
      "Action: tavily_search_results_json\n",
      "Action Input: weather in SF\u001b[0m\u001b[33;1m\u001b[1;3m[{'url': 'https://forecast.weather.gov/zipcity.php?inputstring=San francisco,CA', 'content': 'NOAA National Weather Service National Weather Service. Toggle navigation. HOME; FORECAST . Local; Graphical; Aviation; Marine; Rivers and Lakes; Hurricanes; Severe Weather; Fire Weather; ... San Francisco CA 37.77°N 122.41°W (Elev. 131 ft) Last Update: 8:40 pm PDT Mar 14, 2024. Forecast Valid: 12am PDT Mar 15, 2024-6pm PDT Mar 20, 2024 .'}]\u001b[0m\u001b[32;1m\u001b[1;3mThought: Do I need to use a tool? Yes\n",
      "Action: tavily_search_results_json\n",
      "Action Input: weather in SF\u001b[0m\u001b[33;1m\u001b[1;3m[{'url': 'https://weather.com/weather/tenday/l/San Francisco CA USCA0987:1:US', 'content': \"Comfy & Cozy\\nThat's Not What Was Expected\\nOutside\\n'No-Name Storms' In Florida\\nGifts From On High\\nWhat To Do For Wheezing\\nSurviving The Season\\nStay Safe\\nAir Quality Index\\nAir quality is considered satisfactory, and air pollution poses little or no risk.\\n Health & Activities\\nSeasonal Allergies and Pollen Count Forecast\\nNo pollen detected in your area\\nCold & Flu Forecast\\nFlu risk is low in your area\\nWe recognize our responsibility to use data and technology for good. recents\\nSpecialty Forecasts\\n10 Day Weather-San Francisco, CA\\nToday\\nMon 18 | Day\\nConsiderable cloudiness. Tue 19\\nTue 19 | Day\\nLight rain early...then remaining cloudy with showers in the afternoon. Wed 27\\nWed 27 | Day\\nOvercast with rain showers at times.\"}]\u001b[0m\u001b[32;1m\u001b[1;3mThought: Do I need to use a tool? No\n",
      "Final Answer: According to the weather information I found, the current weather in San Francisco, CA is:\n",
      "\n",
      "- Considerable cloudiness today, with light rain early on Tuesday followed by remaining cloudy conditions with showers in the afternoon. \n",
      "- Wednesday is expected to be overcast with rain showers at times.\n",
      "- The 10-day forecast shows a mix of cloudy, rainy, and cooler conditions in the San Francisco area over the next week.\n",
      "\n",
      "So it looks like the weather in San Francisco is expected to be rather unsettled\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'input': 'what is the weather in SF?',\n",
       " 'chat_history': [],\n",
       " 'output': 'According to the weather information I found, the current weather in San Francisco, CA is:\\n\\n- Considerable cloudiness today, with light rain early on Tuesday followed by remaining cloudy conditions with showers in the afternoon. \\n- Wednesday is expected to be overcast with rain showers at times.\\n- The 10-day forecast shows a mix of cloudy, rainy, and cooler conditions in the San Francisco area over the next week.\\n\\nSo it looks like the weather in San Francisco is expected to be rather unsettled'}"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent_executor.invoke({\n",
    "    \"input\": \"what is the weather in SF?\",\n",
    "    \"chat_history\": []\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[32;1m\u001b[1;3mThought: Do I need to use a tool? Yes\n",
      "Action: langsmith_search\n",
      "Action Input: LangSmith\u001b[0m\u001b[36;1m\u001b[1;3mLangSmith User Guide | 🦜️🛠️ LangSmith\n",
      "\n",
      "Skip to main content🦜️🛠️ LangSmith DocsLangChain Python DocsLangChain JS/TS DocsLangSmith API DocsSearchGo to AppLangSmithUser GuideSetupPricing (Coming Soon)Self-HostingTracingEvaluationMonitoringPrompt HubProxyUser GuideOn this pageLangSmith User GuideLangSmith is a platform for LLM application development, monitoring, and testing. In this guide, we’ll highlight the breadth of workflows LangSmith supports and how they fit into each stage of the application development lifecycle. We hope this will inform users how to best utilize this powerful platform or give them something to consider if they’re just starting their journey.Prototyping​Prototyping LLM applications often involves quick experimentation between prompts, model types, retrieval strategy and other parameters.\n",
      "The ability to rapidly understand how the model is performing — and debug where it is failing — is incredibly important for this phase.Debugging​When developing new LLM applications, we suggest having LangSmith tracing enabled by default.\n",
      "Oftentimes, it isn’t necessary to look at every single trace. However, when things go wrong (an unexpected end result, infinite agent loop, slower than expected execution, higher than expected token usage), it’s extremely helpful to debug by looking through the application traces. LangSmith gives clear visibility and debugging information at each step of an LLM sequence, making it much easier to identify and root-cause issues.\n",
      "We provide native rendering of chat messages, functions, and retrieve documents.Initial Test Set​While many developers still ship an initial version of their application based on “vibe checks”, we’ve seen an increasing number of engineering teams start to adopt a more test driven approach. LangSmith allows developers to create datasets, which are collections of inputs and reference outputs, and use these to run tests on their LLM applications.\n",
      "These test cases can be uploaded in bulk, created on the fly, or exported from application traces. LangSmith also makes it easy to run custom evaluations (both LLM and heuristic based) to score test results.Comparison View​When prototyping different versions of your applications and making changes, it’s important to see whether or not you’ve regressed with respect to your initial test cases.\n",
      "Oftentimes, changes in the prompt, retrieval strategy, or model choice can have huge implications in responses produced by your application.\n",
      "In order to get a sense for which variant is performing better, it’s useful to be able to view results for different configurations on the same datapoints side-by-side. We’ve invested heavily in a user-friendly comparison view for test runs to track and diagnose regressions in test scores across multiple revisions of your application.Playground​LangSmith provides a playground environment for rapid iteration and experimentation.\n",
      "This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\n",
      "\n",
      "This allows you to quickly test out different prompts and models. You can open the playground from any prompt or model run in your trace.\n",
      "Every playground run is logged in the system and can be used to create test cases or compare with other runs.Beta Testing​Beta testing allows developers to collect more data on how their LLM applications are performing in real-world scenarios. In this phase, it’s important to develop an understanding for the types of inputs the app is performing well or poorly on and how exactly it’s breaking down in those cases. Both feedback collection and run annotation are critical for this workflow. This will help in curation of test cases that can help track regressions/improvements and development of automatic evaluations.Collecting Feedback​When launching your application to an initial set of users, it’s important to gather human feedback on the responses it’s producing. This helps draw attention to the most interesting runs and highlight edge cases that are causing problematic responses. LangSmith allows you to attach feedback scores to logged traces (oftentimes, this is hooked up to a feedback button in your app), then filter on traces that have a specific feedback tag and score. A common workflow is to filter on traces that receive a poor user feedback score, then drill down into problematic points using the detailed trace view.Annotating Traces​LangSmith also supports sending runs to annotation queues, which allow annotators to closely inspect interesting traces and annotate them with respect to different criteria. Annotators can be PMs, engineers, or even subject matter experts. This allows users to catch regressions across important evaluation criteria.Adding Runs to a Dataset​As your application progresses through the beta testing phase, it's essential to continue collecting data to refine and improve its performance. LangSmith enables you to add runs as examples to datasets (from both the project page and within an annotation queue), expanding your test coverage on real-world scenarios. This is a key benefit in having your logging system and your evaluation/testing system in the same platform.Production​Closely inspecting key data points, growing benchmarking datasets, annotating traces, and drilling down into important data in trace view are workflows you’ll also want to do once your app hits production. However, especially at the production stage, it’s crucial to get a high-level overview of application performance with respect to latency, cost, and feedback scores. This ensures that it's delivering desirable results at scale.Monitoring and A/B Testing​LangSmith provides monitoring charts that allow you to track key metrics over time. You can expand to view metrics for a given period and drill down into a specific data point to get a trace table for that time period — this is especially handy for debugging production issues.LangSmith also allows for tag and metadata grouping, which allows users to mark different versions of their applications with different identifiers and view how they are performing side-by-side within each chart. This is helpful for A/B testing changes in prompt, model, or retrieval strategy.Was this page helpful?PreviousLangSmithNextSetupPrototypingBeta TestingProductionCommunityDiscordTwitterGitHubDocs CodeLangSmith SDKPythonJS/TSMoreHomepageBlogCopyright © 2024 LangChain, Inc.\u001b[0m\u001b[32;1m\u001b[1;3mOkay, let's summarize how LangSmith can help test your LLM applications:\n",
      "\n",
      "Thought: Do I need to use a tool? No\n",
      "Final Answer: LangSmith provides several key features to help test and debug LLM applications:\n",
      "\n",
      "1. Prototyping: LangSmith allows for rapid experimentation with different prompts, models, and retrieval strategies. It provides visibility into model performance to help identify issues quickly.\n",
      "\n",
      "2. Debugging: LangSmith provides detailed tracing of LLM sequences, making it easier to identify and root-cause issues in\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'chat_history': [HumanMessage(content='Can LangSmith help test my LLM applications?'),\n",
       "  AIMessage(content='Yes!')],\n",
       " 'input': 'Tell me how',\n",
       " 'output': 'LangSmith provides several key features to help test and debug LLM applications:\\n\\n1. Prototyping: LangSmith allows for rapid experimentation with different prompts, models, and retrieval strategies. It provides visibility into model performance to help identify issues quickly.\\n\\n2. Debugging: LangSmith provides detailed tracing of LLM sequences, making it easier to identify and root-cause issues in'}"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chat_history = [HumanMessage(content=\"Can LangSmith help test my LLM applications?\"), AIMessage(content=\"Yes!\")]\n",
    "agent_executor.invoke({\n",
    "    \"chat_history\": chat_history,\n",
    "    \"input\": \"Tell me how\"\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "pytorch",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}