WEBDIENSTE/langgraph
1
0
Fork 0
mirror of https://github.com/langchain-ai/langgraph.git synced 2026-09-09 19:27:54 +02:00
Code Issues Packages Projects Releases Wiki Activity
Files
0687378c620148e9443edad4f0e644f1b9d05abb
langgraph/docs/docs/tutorials/usaco/usaco.ipynb
T
Dionysis GlytsosandGitHub a9d4e0da29 docs: fix typos (#4992) 
Fix typos
2025-06-08 13:46:41 +00:00

2089 lines
590 KiB
Plaintext
Raw Blame History

{
"cells": [
{
"attachments": {
"a53accbd-8074-456e-8268-547edff14571.png": {
"image/png": "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"
}
},
"cell_type": "markdown",
"id": "69ad077f-af4e-49b8-b549-fb3112299aa8",
"metadata": {},
"source": [
"# Competitive Programming\n",
"\n",
"In this tutorial, you will build a computing olympiad agent that leverages three complementary techniques to boost performance: **reflection**, **retrieval**, and **human-in-the-loop** collaboration. These techniques and data are all adapted from the paper \"Can Language Models Solve Olympiad Programming?\" by Quan Shi, Michael Tang, Karthik Narasimhan, and Shunyu Yao. You can check out their paper at the following link:\n",
"\n",
"[![arXiv](http://img.shields.io/badge/cs.CL-arXiv%3A2404.10952v1-B31B1B.svg)](https://arxiv.org/abs/2404.10952v1)\n",
"\n",
"You will construct an agentic graph capable of answering programming questions of increasing difficulty.\n",
"\n",
"1. **Reflection**: In part 1, you will create a zero-shot tool calling agent and prompt it to reflect on the test case results to correct its initial errors. This is similar to the agent the paper reported as having a pass rate of 12.38 on the USACO benchmark.\n",
"2. **Retrieval**: In Part 2, you will implement an initial retrieval step as \"episodic memory\" for the agent that retrieves high-quality few-shot examples from our corpora of programming problems to help solve the **bronze** level question. This agent is similar to the one the paper benchmarked at 20.2.\n",
"3. **Human-in-the-loop**: In part 3, you will use `interrupt_after` to let the user copilot the agent to a better answer. The benchmark performance then is constrained only by the competitiveness of the human it is paired with.\n",
"\n",
"Your final agent graph will be structured like the diagram below:\n",
"\n",
"<img src=\"./img/diagram.png\" src=\"../img/diagram.png\" >\n",
"\n",
"Parts 1 and 2 are analogous to the systems benchmarked in the paper as having a pass rate of 12.38 and 20.2 respectively.\n",
"\n",
"![Benchmark system results](attachment:a53accbd-8074-456e-8268-547edff14571.png)\n",
"\n",
"\n",
"While LLMs are not yet capable of autonomously solving all these problems, we can design the system that far surpasses the capabilities of a basic ReAct agent at answering these questions. \n",
"\n",
"Before diving in, let's set up our machine. This will involve installing dependencies, fetching the dataset, and defining a utility function.\n",
"\n",
"## Setup\n",
"\n",
"For this tutorial, we will need to install some dependencies, fetch the Olympiad dataset, and define a utility function to help run the candidate solutions to see if they pass the test cases.\n",
"\n",
"First, let's install the required packages and set our API keys"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "c686827a-8078-4fd4-af7a-638ca1362796",
"metadata": {},
"outputs": [],
"source": [
"%%capture --no-stderr\n",
"%pip install -U langgraph langsmith langchain_anthropic datasets langchain langchainhub"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "e2e542bb-a99e-44d3-8ebb-6a952dcbf2bf",
"metadata": {},
"outputs": [],
"source": [
"import getpass\n",
"import os\n",
"\n",
"\n",
"def _get_env(var: str):\n",
" if not os.environ.get(var):\n",
" os.environ[var] = getpass.getpass(f\"{var}: \")\n",
"\n",
"\n",
"_get_env(\"ANTHROPIC_API_KEY\")"
]
},
{
"cell_type": "markdown",
"id": "10284e28",
"metadata": {},
"source": [
"<div class=\"admonition tip\">\n",
" <p class=\"admonition-title\">Set up <a href=\"https://smith.langchain.com\">LangSmith</a> for LangGraph development</p>\n",
" <p style=\"padding-top: 5px;\">\n",
" Sign up for LangSmith to quickly spot issues and improve the performance of your LangGraph projects. LangSmith lets you use trace data to debug, test, and monitor your LLM apps built with LangGraph — read more about how to get started <a href=\"https://docs.smith.langchain.com\">here</a>. \n",
" </p>\n",
"</div>"
]
},
{
"cell_type": "markdown",
"id": "c86ebbeb-c070-45d4-99ef-de33b53d447d",
"metadata": {},
"source": [
"#### Data\n",
"\n",
"Fetch the USACO benchmark data using the util below:"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "f7a0c7bd-512d-4e5b-ab43-1bc3b8c97fd4",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import zipfile\n",
"\n",
"import datasets\n",
"import requests\n",
"\n",
"usaco_url = \"https://storage.googleapis.com/benchmarks-artifacts/usaco/usaco_sampled_with_tests.zip\"\n",
"zip_path = \"usaco.zip\"\n",
"extract_path = \"usaco_datasets\"\n",
"\n",
"response = requests.get(usaco_url)\n",
"with open(zip_path, \"wb\") as file:\n",
" file.write(response.content)\n",
"\n",
"with zipfile.ZipFile(zip_path, \"r\") as zip_ref:\n",
" zip_ref.extractall(extract_path)\n",
"\n",
"os.remove(zip_path)\n",
"\n",
"ds = datasets.load_from_disk(os.path.join(extract_path, \"usaco_v3_sampled_with_tests\"))"
]
},
{
"cell_type": "markdown",
"id": "2e4ad035-2c1a-4e31-93b4-58793d219bc9",
"metadata": {},
"source": [
"#### Test Evaluation Utils\n",
"\n",
"We also need a way to evaluate our generated code. We will use this unsafe code execution program to run the generated code against our test cases.\n",
"**Note:** The code below runs arbitrary code on your local machine! Proceed with caution."
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "54f9d037-121e-412f-857a-3e0ccc73892e",
"metadata": {},
"outputs": [],
"source": [
"import multiprocessing\n",
"import queue\n",
"import subprocess\n",
"import sys\n",
"import time\n",
"import traceback\n",
"\n",
"multiprocessing.set_start_method(\"fork\", force=True)\n",
"# WARNING\n",
"# This program exists to execute untrusted model-generated code. Although\n",
"# it is highly unlikely that model-generated code will do something overtly\n",
"# malicious in response to this test suite, model-generated code may act\n",
"# destructively due to a lack of model capability or alignment.\n",
"# Users are strongly encouraged to sandbox this evaluation suite so that it\n",
"# does not perform destructive actions on their host or network.\n",
"# Proceed at your own risk:\n",
"\n",
"\n",
"def exec_program(q, program, input_data, expected_output, timeout):\n",
" try:\n",
" start_time = time.time()\n",
" process = subprocess.Popen(\n",
" [sys.executable, \"-c\", program],\n",
" stdin=subprocess.PIPE,\n",
" stdout=subprocess.PIPE,\n",
" stderr=subprocess.PIPE,\n",
" text=True,\n",
" )\n",
" stdout, stderr = process.communicate(input=input_data, timeout=timeout)\n",
" if time.time() - start_time > timeout:\n",
" raise TimeoutError(\"Execution timed out.\")\n",
" if process.returncode != 0:\n",
" q.put(f\"failed: {stderr}\")\n",
" else:\n",
" if stdout.strip() == expected_output.strip():\n",
" q.put(\"passed\")\n",
" else:\n",
" q.put(f\"wrong answer. Expected '{expected_output}', got '{stdout}'\")\n",
" except subprocess.TimeoutExpired:\n",
" process.kill()\n",
" q.put(\"timed out\")\n",
" except Exception:\n",
" q.put(f\"failed: {traceback.format_exc()}\")\n",
"\n",
"\n",
"def check_correctness(\n",
" program: str, input_data: str, expected_output: str, timeout: float\n",
") -> str:\n",
" q = multiprocessing.Queue()\n",
" process = multiprocessing.Process(\n",
" target=exec_program, args=(q, program, input_data, expected_output, timeout)\n",
" )\n",
" process.start()\n",
" process.join(timeout=timeout + 1)\n",
" if process.is_alive():\n",
" process.terminate()\n",
" process.join()\n",
" result = \"timed out\"\n",
" else:\n",
" try:\n",
" result = q.get_nowait()\n",
" except queue.Empty:\n",
" result = \"no result returned\"\n",
" return result"
]
},
{
"cell_type": "markdown",
"id": "1e799866-6334-4c3a-8d03-7b7b1cf730ab",
"metadata": {},
"source": [
"Let's check an example program and output to see how it works:"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "411cfc6a-d430-4642-8f48-2d6335430dd9",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Example 1: passed\n",
"Example 2: wrong answer. Expected 'hi there', got 'goodbye\n",
"'\n"
]
}
],
"source": [
"program_code = \"print('hello, world!')\"\n",
"input_data = \"\"\n",
"expected_output = \"hello, world!\"\n",
"timeout = 2\n",
"\n",
"test_result = check_correctness(program_code, input_data, expected_output, timeout)\n",
"print(\"Example 1: \", test_result)\n",
"test_result = check_correctness(\"print('goodbye')\", input_data, \"hi there\", timeout)\n",
"print(\"Example 2: \", test_result)"
]
},
{
"cell_type": "markdown",
"id": "d4ebf535-d86c-440e-9820-ae9b872db477",
"metadata": {},
"source": [
"## Part 1: Zero-Shot with Reflection\n",
"\n",
"In our first section, we will build a simple zero-shot tool-calling agent to try to solve these problems. We will incorporate a simple form of [reflection](https://www.youtube.com/watch?v=v5ymBTXNqtk) directly in the agent's tool calling schema by adding a \"reasoning\" field. Furthermore, Claude was trained to \"reason\" with freeform text prior to invoking any tools. Together, this should induce reflective \"chain-of-thought\" prompting.\n",
"\n",
"_Note: this diverges somewhat from the paper's implementation, which uses an explicit reflection step with a variation of the [Reflexion](../../reflexion/reflexion) prompt._\n",
"\n",
"By the end of this section, we will have built a reflective zero-shot programming agent that looks like the section marked \"Part 1\" in the system diagram below:\n",
"\n",
"<img src=\"./img/diagram-part-1.png\" src=\"../img/diagram-part-1.png\" >"
]
},
{
"cell_type": "markdown",
"id": "00f91dac-d13b-4221-be1b-9254ca849c8d",
"metadata": {},
"source": [
"### State\n",
"\n",
"LangGraph's main primitive is the `StateGraph`, which you use to define an agent as a controllable state machine. The graph has `node`'s (python functions) that perform the work, and `edge`s that define how to route between the nodes.\n",
"The `State` defines the interface between each node and carries all the information your agent needs.\n",
"\n",
"Below, define a `State` for our programming olympiad agent. The `messages` will track the sequence of submissions (and test case feedback) as chat history. The `status` field will flip from `in_progress` to `success` if the submission passes all test cases.\n",
"The other fields (test_cases, runtime_limit) are used by the `evaluation` node to test the agent's submissions. These values are not seen by the agent itself."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "f43d68d9-10be-4544-879a-88a33db18bea",
"metadata": {},
"outputs": [],
"source": [
"from typing import Annotated\n",
"\n",
"from typing_extensions import TypedDict\n",
"\n",
"from langgraph.graph.message import AnyMessage, add_messages\n",
"\n",
"\n",
"class TestCase(TypedDict):\n",
" inputs: str\n",
" outputs: str\n",
"\n",
"\n",
"class State(TypedDict):\n",
" # Append-only chat memory so the agent can try to recover from initial mistakes.\n",
" messages: Annotated[list[AnyMessage], add_messages]\n",
" # From the dataset. These are used for testing.\n",
" test_cases: list[TestCase]\n",
" runtime_limit: int\n",
" status: str"
]
},
{
"cell_type": "markdown",
"id": "64921a60-411a-4a9b-aaf0-8476d02d8a3a",
"metadata": {},
"source": [
"Now, convert the dataset into inputs our graph will accept."
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "6d56776f-993b-4ca7-89ef-21dec01dc9d3",
"metadata": {},
"outputs": [],
"source": [
"input_states = [\n",
" {\n",
" \"messages\": [(\"user\", row[\"description\"])],\n",
" \"test_cases\": row[\"test_cases\"],\n",
" \"runtime_limit\": row[\"runtime_limit\"],\n",
" \"status\": \"in_progress\",\n",
" \"problem_level\": row[\"problem_level\"],\n",
" }\n",
" for row in ds\n",
"]"
]
},
{
"cell_type": "markdown",
"id": "9e7883ee-b6b3-4d89-b5a5-2e139fe361c9",
"metadata": {},
"source": [
"#### Node 1: Solver\n",
"\n",
"Create a `solver` node that prompts an LLM \"agent\" to use a [writePython tool](https://python.langchain.com/docs/integrations/chat/anthropic/#integration-details) to generate the submitted code."
]
},
{
"cell_type": "markdown",
"id": "7c1f057e-b2cb-4085-9f4b-54ef005df0c6",
"metadata": {},
"source": [
"<div class=\"admonition note\">\n",
" <p class=\"admonition-title\">Using Pydantic with LangChain</p>\n",
" <p>\n",
" This notebook uses Pydantic v2 <code>BaseModel</code>, which requires <code>langchain-core >= 0.3</code>. Using <code>langchain-core < 0.3</code> will result in errors due to mixing of Pydantic v1 and v2 <code>BaseModels</code>.\n",
" </p>\n",
"</div>"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "7b9e7742-16a3-4ad2-bc63-5f9cd4fd734b",
"metadata": {},
"outputs": [],
"source": [
"from langchain_core.language_models import BaseChatModel\n",
"from langchain_core.prompts import ChatPromptTemplate\n",
"\n",
"from pydantic import BaseModel, Field\n",
"\n",
"\n",
"class writePython(BaseModel):\n",
" \"\"\"Write python code that resolves the problem.\"\"\"\n",
"\n",
" reasoning: str = Field(..., description=\"Conceptual solution.\")\n",
" pseudocode: str = Field(..., description=\"Detailed English pseudocode.\")\n",
" code: str = Field(..., description=\"Valid Python 3 solution to the problem\")\n",
"\n",
"\n",
"class Solver:\n",
" def __init__(self, llm: BaseChatModel, prompt: ChatPromptTemplate):\n",
" self.runnable = prompt | llm.bind_tools([writePython])\n",
"\n",
" def __call__(self, state: State) -> dict:\n",
" # Our agent only can see the \"messages\" and will ignore the test info\n",
" return {\"messages\": [self.runnable.invoke({\"messages\": state[\"messages\"]})]}"
]
},
{
"cell_type": "markdown",
"id": "c22eaa6d-36ae-4526-9b80-49d615fc055c",
"metadata": {},
"source": [
"Now, create the solver below. We'll use Claude Opus"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "6cc472f1-b9b3-4f81-a797-c64704bb07d5",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"***********************************Prompt***********************************\n",
"================================\u001b[1m System Message \u001b[0m================================\n",
"\n",
"You are a world-class competitive programmer.\n",
"Please reply with a Python 3 solution to the problem below. \n",
"First, reason through the problem and conceptualize a solution.\n",
"Then write detailed pseudocode to uncover any potential logical errors or omissions.\n",
"Finally output the working Python code for your solution, ensuring to fix any errors uncovered while writing pseudocode.\n",
"\n",
"No outside libraries are allowed.\u001b[33;1m\u001b[1;3m{examples}\u001b[0m\n",
"\n",
"=============================\u001b[1m Messages Placeholder \u001b[0m=============================\n",
"\n",
"\u001b[33;1m\u001b[1;3m{messages}\u001b[0m\n"
]
}
],
"source": [
"from langchain import hub\n",
"from langchain_anthropic import ChatAnthropic\n",
"\n",
"# For this section, we are testing zero-shot performance and won't have\n",
"# any examples. Partial them out to pre-fill the template.\n",
"prompt = hub.pull(\"wfh/usaco-draft-solver\").partial(examples=\"\")\n",
"print(\"*\" * 35 + \"Prompt\" + \"*\" * 35)\n",
"prompt.pretty_print()\n",
"\n",
"# Use Haiku if you want to save $$ while (almost) never correctly answering the question\n",
"# llm = ChatAnthropic(model=\"claude-3-haiku-20240307\")\n",
"llm = ChatAnthropic(model=\"claude-3-opus-20240229\")\n",
"\n",
"solver = Solver(llm, prompt)"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "5d9560ba-900a-43d1-ad38-f132fd660337",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"********************************** Example **********************************\n",
"==================================\u001b[1m Ai Message \u001b[0m==================================\n",
"\n",
"[{'text': \"<thinking>\\nTo address this problem, we need to use the writePython function, which requires the following parameters:\\n- reasoning: a conceptual solution to the problem\\n- pseudocode: detailed pseudocode for the solution\\n- code: working Python code implementing the solution\\n\\nThe key aspects to address in the solution are:\\n1. We have an infinite stream, so we can't store all elements. Need an online algorithm.\\n2. Need to ensure each element has an equal probability of being in the final sample.\\n\\nI believe I have enough information to provide values for all the required parameters.\\n</thinking>\", 'type': 'text'}, {'id': 'toolu_01UqpLYyueky5GtYMidS9oLF', 'input': {'reasoning': 'To get a perfectly random sample of size k from an infinite stream:\\n\\n1. Store the first k elements in an array (reservoir). \\n2. For each ith element after the kth element (i > k):\\n - Generate a random integer j between 0 and i (inclusive)\\n - If j < k, replace the jth element of the reservoir with the ith element\\n3. At the end, the reservoir contains the random sample.\\n\\nThis works because for any element, when we process the nth element, the probability that it is in the reservoir is:\\n- k/n when n <= k (first k elements always selected)\\n- k/n * k/(n-1) * k/(n-2) * ... * k/(k+1) = k/n when n > k\\n\\nSo any element has k/n probability of being in final reservoir, giving a perfectly random sample.', 'pseudocode': '```\\nfunction selectKItems(stream, k):\\n reservoir = [0..k-1] # store first k elements\\n\\n i = k\\n while stream has next item:\\n item = stream.next()\\n j = random(0, i) # generate random index between 0 and i\\n if j < k:\\n reservoir[j] = item # replace element at random index with new item\\n i += 1\\n\\n return reservoir\\n```', 'code': 'import random\\n\\ndef reservoir_sampling(stream, k):\\n reservoir = []\\n \\n # Store first k elements in reservoir\\n for i in range(k):\\n reservoir.append(next(stream))\\n\\n i = k\\n for item in stream:\\n # Generate random index between 0 and i\\n j = random.randint(0, i) \\n \\n # Replace element at random index with new item\\n if j < k:\\n reservoir[j] = item\\n i += 1\\n\\n return reservoir'}, 'name': 'writePython', 'type': 'tool_use'}]\n"
]
}
],
"source": [
"print(\"*\" * 34 + \" Example \" + \"*\" * 34)\n",
"result = solver(\n",
" {\n",
" \"messages\": [\n",
" (\n",
" \"user\",\n",
" \"How do I get a perfectly random sample from an infinite stream\",\n",
" )\n",
" ]\n",
" }\n",
")\n",
"result[\"messages\"][0].pretty_print()\n",
"# Could expand to include (1)\n",
"# 1. Restate the problem in plain English\n",
"# 2. Closely following the explanation, restate and explain the solution in plain English\n",
"# 3. Write a pseudocode solution\n",
"# 4. Output the final Python solution with your solution steps in comments."
]
},
{
"cell_type": "markdown",
"id": "58c5fa7a-a2eb-4954-be34-4194a58f00d3",
"metadata": {},
"source": [
"#### Node 2: Evaluate\n",
"\n",
"Now define the \"`evaluate`\" node. This node takes the `solver`'s submitted code and executes it against the `test_cases` in our `State`.\n",
"This uses the unsafe `check_correctness` utility we defined in the setup above."
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "1785015b-24f8-415f-b950-e229b5137887",
"metadata": {},
"outputs": [],
"source": [
"from langchain_core.messages import AIMessage, HumanMessage, ToolMessage\n",
"\n",
"\n",
"# This is the node we will add to the graph.\n",
"# Most tool-calling APIs require that the `ToolMessage` contain the ID\n",
"# of the\n",
"def format_tool_message(response: str, ai_message: AIMessage):\n",
" return ToolMessage(\n",
" content=response + \"\\nMake all fixes using the writePython tool.\",\n",
" tool_call_id=ai_message.tool_calls[0][\"id\"],\n",
" )\n",
"\n",
"\n",
"def evaluate(state: State):\n",
" test_cases = state[\"test_cases\"]\n",
" ai_message: AIMessage = state[\"messages\"][-1]\n",
" if not ai_message.tool_calls:\n",
" return {\n",
" \"messages\": [\n",
" HumanMessage(\n",
" content=\"No code submitted. Please try again using the correct python code.\"\n",
" )\n",
" ]\n",
" }\n",
" try:\n",
" code = ai_message.tool_calls[0][\"args\"][\"code\"]\n",
" except Exception as e:\n",
" return {\"messages\": [format_tool_message(repr(e), ai_message)]}\n",
" num_test_cases = len(test_cases)\n",
" succeeded = 0\n",
" test_results = []\n",
" # TODO: Multiprocess\n",
" for test_case in test_cases:\n",
" input_data = test_case[\"inputs\"]\n",
" expected_output = test_case[\"outputs\"]\n",
" test_result = check_correctness(code, input_data, expected_output, timeout)\n",
" test_results.append(test_result)\n",
" if test_result == \"passed\":\n",
" succeeded += 1\n",
" pass_rate = succeeded / num_test_cases if num_test_cases else \"N/A\"\n",
" if pass_rate == 1:\n",
" return {\"status\": \"success\"}\n",
"\n",
" responses = \"\\n\".join(\n",
" [f\"<test id={i}>\\n{r}\\n</test>\" for i, r in enumerate(test_results)]\n",
" )\n",
" response = f\"Incorrect submission. Please respond with updated code.\\nPass rate: {succeeded}/{num_test_cases}\\nResults:\\n{responses}\"\n",
" formatted_message = format_tool_message(response, ai_message)\n",
" return {\"messages\": [formatted_message]}"
]
},
{
"cell_type": "markdown",
"id": "76f2a46d-c3e2-49d4-b44b-cadb6fd4e5a0",
"metadata": {},
"source": [
"#### Create Graph\n",
"\n",
"Now, put it all together! Once you've defined each node, defining the connectivity / state transitions is fairly easy.\n",
"\n",
"Our Zero-shot graph defines a loop. If we visualize the data flow, we want the logic to:\n",
"1. First go to the `solver`, which attempts a first solution.\n",
"2. Next go to the `evaluate` node, which tests the solution.\n",
"3. If the solution passes, end, otherwise, return to the `solver` to try again.\n",
"\n",
"In LangGraph, we use `conditional_edges` to define state transitions that contain conditional logic.\n",
"Below, define the graph, adding a `control_edge` to handle step (3) above."
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "caf1560e-1517-4229-8a43-186816da6a3a",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import END, StateGraph, START\n",
"\n",
"builder = StateGraph(State)\n",
"builder.add_node(\"solver\", solver)\n",
"builder.add_edge(START, \"solver\")\n",
"builder.add_node(\"evaluate\", evaluate)\n",
"builder.add_edge(\"solver\", \"evaluate\")\n",
"\n",
"\n",
"def control_edge(state: State):\n",
" if state.get(\"status\") == \"success\":\n",
" return END\n",
" return \"solver\"\n",
"\n",
"\n",
"builder.add_conditional_edges(\"evaluate\", control_edge, {END: END, \"solver\": \"solver\"})\n",
"graph = builder.compile()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "7275a2c3-1818-4d14-a7a5-97bc56243a9b",
"metadata": {},
"outputs": [
{
"data": {
"image/jpeg": "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",
"text/plain": [
"<IPython.core.display.Image object>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from IPython.display import Image, display\n",
"\n",
"try:\n",
" display(Image(graph.get_graph().draw_mermaid_png()))\n",
"except Exception:\n",
" # This requires some extra dependencies and is optional\n",
" pass"
]
},
{
"cell_type": "markdown",
"id": "fc4b1bb4-262f-41c3-822d-e015daf0744a",
"metadata": {},
"source": [
"Now that we've created our graph, let's see the type of question it will have to solve."
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "308d6e18-69ad-4e8d-9c5f-06111b0806ee",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Farmer John has $N$ ($1 \\leq N \\leq 2 \\cdot 10^5$) farms, numbered from $1$ to\n",
"$N$. It is known that FJ closes farm $i$ at time $c_i$. Bessie wakes up at time\n",
"$S$, and wants to maximize the productivity of her day by visiting as many farms\n",
"as possible before they close. She plans to visit farm $i$ on time $t_i + S$.\n",
"Bessie must arrive at a farm strictly before Farmer John closes it to actually visit it.\n",
"\n",
"Bessie has $Q$ $(1 \\leq Q \\leq 2 \\cdot 10^5)$ queries. For each query, she gives\n",
"you two integers $S$ and $V$. For each query, output whether Bessie can visit at\n",
"least $V$ farms if she wakes up at time $S$.\n",
"\n",
"INPUT FORMAT (input arrives from the terminal / stdin):\n",
"The first line consists of $N$ and $Q$.\n",
"\n",
"The second line consists of $c_1, c_2, c_3 \\dots c_N$ ($1 \\leq c_i \\leq 10^6$).\n",
"\n",
"The third line consists of $t_1, t_2, t_3 \\dots t_N$ ($1 \\leq t_i \\leq 10^6$).\n",
"\n",
"The next $Q$ lines each consist of two integers $V$ ($1 \\leq V \\leq N$) and $S$\n",
"($1 \\leq S \\leq 10^6$).\n",
"\n",
"OUTPUT FORMAT (print output to the terminal / stdout):\n",
"For each of the $Q$ queries, output YES or NO on a new line.\n",
"\n",
"SAMPLE INPUT:\n",
"5 5\n",
"3 5 7 9 12\n",
"4 2 3 3 8\n",
"1 5\n",
"1 6\n",
"3 3\n",
"4 2\n",
"5 1\n",
"SAMPLE OUTPUT: \n",
"YES\n",
"NO\n",
"YES\n",
"YES\n",
"NO\n",
"\n",
"For the first query, Bessie will visit the farms at time $t = [9, 7, 8, 8, 13]$,\n",
"so she will only get to visit farm $4$ on time before FJ closes the farm.\n",
"\n",
"For the second query, Bessie will not be able to visit any of the farms on time.\n",
"\n",
"For the third query, Bessie will visit farms $3, 4, 5$ on time.\n",
"\n",
"For the fourth and fifth queries, Bessie will be able to visit all but the first\n",
"farm on time.\n",
"\n",
"SCORING:\n",
"Inputs 2-4: $N,Q\\le 10^3$Inputs 5-9: $c_i, t_i \\le 20$Inputs 10-17: No additional constraints.\n",
"\n",
"\n",
"Problem credits: Chongtian Ma\n",
"\n"
]
}
],
"source": [
"input_state = input_states[0].copy()\n",
"# We will reduce the test cases to speed this notebook up\n",
"input_state[\"test_cases\"] = input_state[\"test_cases\"][:3]\n",
"print(input_state[\"messages\"][0][1])"
]
},
{
"cell_type": "markdown",
"id": "3c444be0-4c7d-49cc-9aaa-2b99dabe3922",
"metadata": {},
"source": [
"Pretty difficult! Let's run our simple \"zero-shot\" agent below to see how it fares. **It most likely will not be able to solve this question** (unless you are using a more powerful model than what I had available at the time of writing this tutorial (2024/04/20).\n",
"We will trace the trajectory to LangSmith to review the series of submissions. To reduce the packet size, we will use \"`hide_inputs`\" and filter out the test_cases. All this is optional but useful for development. \n",
"\n",
"**Note:** We _expect_ a **GraphRecursionError** here from it not being able to answer it correctly in the allocated number of steps."
]
},
{
"cell_type": "code",
"execution_count": 25,
"id": "2ebe0da0-8f27-4805-829c-54bc1dc29ba4",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Assistant: [{'text': '<thinking>\\nThe key steps to solve this\n",
"Assistant: KeyError('code')\\nMake all fixes using the writePy\n",
"Assistant: [{'id': 'toolu_01KimhKt8aqQjGZJmrHVnAtE', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01CMZTqAd7BZQ2nSgtk9djRW', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01Kbaq9gX4BnHvps6TMfVGHL', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01MiSnpiGK5Yy4Cpp6GGbjmT', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01GWuvJezXLMVurUBG84odDP', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01W8DGmhcpFVctySmx58scf9', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_018bhYtCKDK6S4MHiAxUZCrb', 'input':\n",
"Assistant: KeyError('code')\\nMake all fixes using the writePy\n",
"Assistant: [{'id': 'toolu_01LCwaCjX9uZBV3jt9eAkmAa', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01WqJvdE2WDeTZXoKp2V7PWb', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_01DGevkunt9zWx7SVDCHdBuv', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'id': 'toolu_013comYKVxNSzTM4ZbH3L3FP', 'input':\n",
"Assistant: Incorrect submission. Please respond with updated \n"
]
},
{
"ename": "GraphRecursionError",
"evalue": "Recursion limit of 25 reachedwithout hitting a stop condition. You can increase the limit by setting the `recursion_limit` config key.",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mGraphRecursionError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[0;32mIn[25], line 17\u001b[0m\n\u001b[1;32m 15\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m tracing_v2_enabled(client\u001b[38;5;241m=\u001b[39mclient):\n\u001b[1;32m 16\u001b[0m events \u001b[38;5;241m=\u001b[39m graph\u001b[38;5;241m.\u001b[39mstream(input_state)\n\u001b[0;32m---> 17\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mevent\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mevents\u001b[49m\u001b[43m:\u001b[49m\n\u001b[1;32m 18\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mvalue\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mevent\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mvalues\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m:\u001b[49m\n\u001b[1;32m 19\u001b[0m \u001b[43m \u001b[49m\u001b[43mmessages\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m \u001b[49m\u001b[43mvalue\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mmessages\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n",
"File \u001b[0;32m~/.pyenv/versions/3.11.2/lib/python3.11/site-packages/langgraph/pregel/__init__.py:645\u001b[0m, in \u001b[0;36mPregel.stream\u001b[0;34m(self, input, config, stream_mode, output_keys, input_keys, interrupt_before_nodes, interrupt_after_nodes, debug)\u001b[0m\n\u001b[1;32m 643\u001b[0m \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[1;32m 644\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m step \u001b[38;5;241m==\u001b[39m config[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrecursion_limit\u001b[39m\u001b[38;5;124m\"\u001b[39m]:\n\u001b[0;32m--> 645\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m GraphRecursionError(\n\u001b[1;32m 646\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mRecursion limit of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mconfig[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mrecursion_limit\u001b[39m\u001b[38;5;124m'\u001b[39m]\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m reached\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 647\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mwithout hitting a stop condition. You can increase the \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 648\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlimit by setting the `recursion_limit` config key.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 649\u001b[0m )\n\u001b[1;32m 651\u001b[0m \u001b[38;5;66;03m# before execution, check if we should interrupt\u001b[39;00m\n\u001b[1;32m 652\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m _should_interrupt(\n\u001b[1;32m 653\u001b[0m checkpoint,\n\u001b[1;32m 654\u001b[0m interrupt_before_nodes,\n\u001b[1;32m 655\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mstream_channels_list,\n\u001b[1;32m 656\u001b[0m next_tasks,\n\u001b[1;32m 657\u001b[0m ):\n",
"\u001b[0;31mGraphRecursionError\u001b[0m: Recursion limit of 25 reachedwithout hitting a stop condition. You can increase the limit by setting the `recursion_limit` config key."
]
}
],
"source": [
"from langchain_core.tracers.context import tracing_v2_enabled\n",
"from langsmith import Client\n",
"\n",
"\n",
"# We don't need to include all the test cases in our traces.\n",
"def _hide_test_cases(inputs):\n",
" copied = inputs.copy()\n",
" # These are tens of MB in size. No need to send them up\n",
" copied[\"test_cases\"] = \"...\"\n",
" return copied\n",
"\n",
"\n",
"client = Client(hide_inputs=_hide_test_cases, hide_outputs=_hide_test_cases)\n",
"with tracing_v2_enabled(client=client):\n",
" events = graph.stream(input_state)\n",
" for event in events:\n",
" for value in event.values():\n",
" messages = value.get(\"messages\")\n",
" if messages:\n",
" if isinstance(messages, list):\n",
" messages = value[\"messages\"][-1]\n",
" print(\n",
" \"Assistant:\",\n",
" str(messages.content).replace(\"\\n\", \"\\\\n\")[:50],\n",
" )"
]
},
{
"cell_type": "markdown",
"id": "2c496d9a-a95b-4cab-86fa-daafc7ddd065",
"metadata": {},
"source": [
"It wasn't able to solve it in time **but that's OK**! If it were easy, this paper would be a lot shorter :)\n",
"\n",
"You can view the [agent's full LangSmith trace](https://smith.langchain.com/public/61c84ad0-51db-40f1-b50d-6983d9481ca1/r) at the provided link.\n",
"\n",
"In the next section we will add an improvement the paper terms \"episodic memory\", which in this case is really few-shot retrieval."
]
},
{
"attachments": {},
"cell_type": "markdown",
"id": "090a1639-09c5-4682-813f-71100f236eb3",
"metadata": {},
"source": [
"## Part 2: Few-shot Retrieval\n",
"\n",
"Even with reflective tool calling, our baseline agent from part 1 struggled with this difficult task. One way to \"teach\" an LLM how to better perform a task is through demonstrations, also known as \"few-shot examples.\"\n",
"\n",
"What the authors of the USACO paper call \"episodic memory\" **is really just few-shot prompting over similar examples.**\n",
"\n",
"Each examples in this case is a different problems + solution within the dataset. The term \"episodic memory\" makes sense if you pretend your agent has already \"solved\" these problems and is recalling its solutions to them.\n",
"\n",
"This section adds the \"Episodic Memory\" components from \"Part 2\" in the diagram below.\n",
"\n",
"<img src=\"./img/diagram-part-2.png\" src=\"../img/diagram-part-2.png\" >\n",
"\n",
"Note that this memory step is performed **one time**, **before** the logic of our zero-shot loop from part 1. The steps are as follows:\n",
"\n",
"1. Prompt the LLM to generate a candidate solution.\n",
"2. Use the text of the candidate solution to retrieve the N most similar (problem, solution) pairs.\n",
"3. Format this result in the Zero-shot agent's prompt.\n",
"\n",
"Below, let's implement our episodic memory as a retriever. We will follow the paper's retriever selection and use [BM25](https://en.wikipedia.org/wiki/Okapi_BM25)."
]
},
{
"cell_type": "code",
"execution_count": 26,
"id": "d612dd8d-31af-426c-944b-203acd55ace0",
"metadata": {},
"outputs": [],
"source": [
"%%capture --no-stderr\n",
"%pip install --upgrade --quiet rank_bm25"
]
},
{
"cell_type": "markdown",
"id": "32e0c485-8eba-41e6-ae29-64bcea98d19e",
"metadata": {},
"source": [
"#### State\n",
"\n",
"The state is mostly recycled from part 1. Add additional \"candidate\" and \"examples\" fields to store the information for the memory steps."
]
},
{
"cell_type": "code",
"execution_count": 27,
"id": "16937fef-58b9-4ab2-bbfc-5237aad235ec",
"metadata": {},
"outputs": [],
"source": [
"from typing import Annotated\n",
"\n",
"from typing_extensions import TypedDict\n",
"\n",
"from langgraph.graph.message import AnyMessage, add_messages\n",
"\n",
"\n",
"class TestCase(TypedDict):\n",
" inputs: str\n",
" outputs: str\n",
"\n",
"\n",
"class State(TypedDict):\n",
" # NEW! Candidate for retrieval + formatted fetched examples as \"memory\"\n",
" candidate: AIMessage\n",
" examples: str\n",
" # Repeated from Part 1\n",
" messages: Annotated[list[AnyMessage], add_messages]\n",
" test_cases: list[TestCase]\n",
" runtime_limit: int\n",
" status: str"
]
},
{
"cell_type": "markdown",
"id": "61fb9104-e73f-47db-9af5-84a375a8d323",
"metadata": {},
"source": [
"#### Nodes 1 and 3: Draft & Solver\n",
"\n",
"Let's create our \"agent\". We will modify the `Solver` from Part 1 to reuse it for for the agent node and for the candidate program generation node (\"draft\")."
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "25f947a7-15bb-4119-a47e-b5c33ca0a249",
"metadata": {},
"outputs": [],
"source": [
"from langchain import hub\n",
"from langchain_anthropic import ChatAnthropic\n",
"\n",
"\n",
"class Solver:\n",
" def __init__(self, llm: BaseChatModel, prompt: ChatPromptTemplate):\n",
" self.runnable = prompt | llm.bind_tools([writePython])\n",
"\n",
" def __call__(self, state: State) -> dict:\n",
" # Our agent only can see the \"messages\" and will ignore the test info\n",
" inputs = {\"messages\": state[\"messages\"]}\n",
" has_examples = bool(state.get(\"examples\"))\n",
" output_key = \"candidate\" # Used in the draft node\n",
" if has_examples:\n",
" output_key = \"messages\"\n",
" # Used in the solve node\n",
" inputs[\"examples\"] = state[\"examples\"]\n",
" response = self.runnable.invoke(inputs)\n",
" if not response.content:\n",
" return {\n",
" output_key: AIMessage(\n",
" content=\"I'll need to think about this step by step.\"\n",
" )\n",
" }\n",
" return {output_key: response}\n",
"\n",
"\n",
"prompt = hub.pull(\"wfh/usaco-draft-solver\")\n",
"llm = ChatAnthropic(model=\"claude-3-opus-20240229\")\n",
"\n",
"draft_solver = Solver(llm, prompt.partial(examples=\"\"))\n",
"solver = Solver(llm, prompt)"
]
},
{
"cell_type": "markdown",
"id": "273a526b-3b52-4630-a58e-0317a0034609",
"metadata": {},
"source": [
"#### Node 2: Retrieve\n",
"\n",
"The retrieve node takes a candidate solution (made by the 'solver' node), uses _this_ to search for similar examples, then formats those in the message."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e5e0aa40-79a4-4071-9ad2-9aa2f36599ce",
"metadata": {},
"outputs": [],
"source": [
"# We will test our agent on index 0 (the same as above).\n",
"# Later, we will test on index 2 (the first 'silver difficulty' question)\n",
"test_indices = [0, 2]\n",
"train_ds = [row for i, row in enumerate(ds) if i not in test_indices]\n",
"test_ds = [row for i, row in enumerate(ds) if i in test_indices]"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "96a1ff96-7556-4959-9f54-1ade3bd1c01a",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.retrievers import BM25Retriever\n",
"\n",
"\n",
"def format_example(row):\n",
" question = row[\"description\"]\n",
" answer = row[\"solution\"]\n",
" return f\"\"\"<problem>\n",
"{question}\n",
"</problem>\n",
"<solution>\n",
"{answer}\n",
"</solution>\"\"\"\n",
"\n",
"\n",
"# Skip our 'test examples' to avoid cheating\n",
"# This is \"simulating\" having seen other in-context examples\n",
"retriever = BM25Retriever.from_texts([format_example(row) for row in train_ds])"
]
},
{
"cell_type": "markdown",
"id": "6ad20d3a-291c-41a5-a343-bd14284f76f2",
"metadata": {},
"source": [
"Now define the node. Any node can optionally accept a second `config` positional argument. This contains `configurable` params you can adjust when invoking the graph. For instance, we can\n",
"adjust the top `k` examples to retrieve for our agent."
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "af42962d-c06e-4b6e-96df-72ad48f17617",
"metadata": {},
"outputs": [],
"source": [
"from langchain_core.runnables import RunnableConfig\n",
"\n",
"\n",
"def retrieve_examples(state: State, config: RunnableConfig):\n",
" top_k = config[\"configurable\"].get(\"k\") or 2\n",
" ai_message: AIMessage = state[\"candidate\"]\n",
" if not ai_message.tool_calls:\n",
" # We err here. To make more robust, you could loop back\n",
" raise ValueError(\"Draft agent did not produce a valid code block\")\n",
" code = ai_message.tool_calls[0][\"args\"][\"code\"]\n",
" examples_str = \"\\n\".join(\n",
" [doc.page_content for doc in retriever.invoke(code)[:top_k]]\n",
" )\n",
" examples_str = f\"\"\"\n",
"You previously solved the following problems in this competition:\n",
"<Examples>\n",
"{examples_str}\n",
"<Examples>\n",
"Approach this new question with similar sophistication.\"\"\"\n",
" return {\"examples\": examples_str}"
]
},
{
"cell_type": "markdown",
"id": "df5cbe72-fca0-4974-a769-284267d3df91",
"metadata": {},
"source": [
"#### Graph\n",
"\n",
"Now let's put it all together. The graph is slightly more complicated than in part 1, since we have to add the initial \"draft\" and \"retrieve\" nodes to our agent loop."
]
},
{
"cell_type": "code",
"execution_count": 17,
"id": "e6e73e85-1232-4848-beba-3139ac7d0a64",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.checkpoint.memory import MemorySaver\n",
"from langgraph.graph import END, StateGraph, START\n",
"\n",
"builder = StateGraph(State)\n",
"builder.add_node(\"draft\", draft_solver)\n",
"builder.add_edge(START, \"draft\")\n",
"builder.add_node(\"retrieve\", retrieve_examples)\n",
"builder.add_node(\"solve\", solver)\n",
"builder.add_node(\"evaluate\", evaluate)\n",
"# Add connectivity\n",
"builder.add_edge(\"draft\", \"retrieve\")\n",
"builder.add_edge(\"retrieve\", \"solve\")\n",
"builder.add_edge(\"solve\", \"evaluate\")\n",
"\n",
"\n",
"def control_edge(state: State):\n",
" if state.get(\"status\") == \"success\":\n",
" return END\n",
" return \"solve\"\n",
"\n",
"\n",
"builder.add_conditional_edges(\"evaluate\", control_edge, {END: END, \"solve\": \"solve\"})\n",
"\n",
"\n",
"checkpointer = MemorySaver()\n",
"graph = builder.compile(checkpointer=checkpointer)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "57acf78d-5e68-46dd-adae-2259e5c5d3f1",
"metadata": {},
"outputs": [
{
"data": {
"image/jpeg": "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",
"text/plain": [
"<IPython.core.display.Image object>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from IPython.display import Image, display\n",
"\n",
"try:\n",
" display(Image(graph.get_graph().draw_mermaid_png()))\n",
"except Exception:\n",
" # This requires some extra dependencies and is optional\n",
" pass"
]
},
{
"cell_type": "markdown",
"id": "3bb633e0-0d02-4050-96a8-18265594385b",
"metadata": {},
"source": [
"Let's try again on this problem:"
]
},
{
"cell_type": "code",
"execution_count": 34,
"id": "dc6f0455-a6e9-46a0-8f8b-03de00fb1f8d",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'text': \"<thinking>\\nThis problem essentially asks to find the number of farms Bessie can visit before they close at each query. The key insights are:\\n\\n1. Bessie's arrival time at each farm is S +\n",
"Retrieved examples:\n",
"\n",
" \n",
"You previously solved the following problems in this competition:\n",
"<Examples>\n",
"<problem>\n",
"\n",
"Farmer John...\n",
"Assistant: [{'text': \"<thinking>\\nThe key information given i\n"
]
}
],
"source": [
"config = {\"configurable\": {\"thread_id\": \"question-recall\", \"k\": 3}}\n",
"with tracing_v2_enabled(client=client):\n",
" events = graph.stream(input_state, config)\n",
" for event in events:\n",
" for value in event.values():\n",
" messages = value.get(\"messages\")\n",
" if messages:\n",
" if isinstance(messages, list):\n",
" messages = value[\"messages\"][-1]\n",
" print(\n",
" \"Assistant:\",\n",
" str(messages.content).replace(\"\\n\", \"\\\\n\")[:50],\n",
" )\n",
" elif value.get(\"examples\"):\n",
" print(\"Retrieved examples:\\n\\n\", value[\"examples\"][:100] + \"...\")\n",
" elif value.get(\"candidate\"):\n",
" print(str(value[\"candidate\"].content)[:200])"
]
},
{
"cell_type": "markdown",
"id": "1389ffb0-839e-4646-be68-dbc79279c8ae",
"metadata": {},
"source": [
"**No recursion error!** You can view the [full LangSmith trace](https://smith.langchain.com/public/1f1c4db3-b53c-49bf-a287-a2b51c081156/r/31f90ddd-8ae9-4b23-a2b5-b0c0d67c5cc3) of the graph's execution at the provided link to confirm the results. You can also check the graph state to confirm that it passed all test cases successfully:"
]
},
{
"cell_type": "code",
"execution_count": 35,
"id": "d55a1d95-9fa4-45b2-9f5e-7ca02fb8415b",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"'success'"
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"checkpoint = graph.get_state(config)\n",
"checkpoint.values[\"status\"]"
]
},
{
"cell_type": "markdown",
"id": "50348c66-4d38-44d0-bb20-ae1a14517f78",
"metadata": {},
"source": [
"**Congrats!** You added \"episodic memory\" to your agent to fetch few-shot examples and solve this bronze level programming olympiad question!\n",
"\n",
"Our agent is still limited, however. Let's test it out on a more challenging 🪙🏆silver✨ level question:"
]
},
{
"cell_type": "code",
"execution_count": 36,
"id": "1ef0b06f-c448-49a5-8f1f-f7041a5d6b87",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"'silver'"
]
},
"execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"silver_row = test_ds[1]\n",
"silver_row[\"problem_level\"]"
]
},
{
"cell_type": "code",
"execution_count": 37,
"id": "2ae55ddc-e629-4be7-badd-756608eec50e",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'text': \"<thinking>\\nThe relevant tool for this problem is writePython. It requires the following parameters:\\n- reasoning: To solve this problem, we need to simulate the cruise by following the seq\n",
"Retrieved examples:\n",
"\n",
" \n",
"You previously solved the following problems in this competition:\n",
"<Examples>\n",
"<problem>\n",
"\n",
"Farmer John...\n",
"Assistant: [{'text': \"<thinking>\\nTo solve this problem, we n\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nAfter reviewing the failed \n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nAfter reviewing the latest \n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nOops, looks like I made a s\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nHmm, some of the test cases\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': '<thinking>\\nOops, looks like I accident\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nLooks like the code is now \n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': '<thinking>\\nOops, looks like I accident\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nHmm, the optimization to si\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nOops, I did it again - acci\n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Assistant: [{'text': \"<thinking>\\nHmm, the latest code is sti\n",
"Assistant: Incorrect submission. Please respond with updated \n"
]
},
{
"ename": "GraphRecursionError",
"evalue": "Recursion limit of 25 reachedwithout hitting a stop condition. You can increase the limit by setting the `recursion_limit` config key.",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mGraphRecursionError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[0;32mIn[37], line 12\u001b[0m\n\u001b[1;32m 10\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m tracing_v2_enabled(client\u001b[38;5;241m=\u001b[39mclient):\n\u001b[1;32m 11\u001b[0m events \u001b[38;5;241m=\u001b[39m graph\u001b[38;5;241m.\u001b[39mstream(silver_input, config)\n\u001b[0;32m---> 12\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mevent\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mevents\u001b[49m\u001b[43m:\u001b[49m\n\u001b[1;32m 13\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mvalue\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mevent\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mvalues\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m:\u001b[49m\n\u001b[1;32m 14\u001b[0m \u001b[43m \u001b[49m\u001b[43mmessages\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m \u001b[49m\u001b[43mvalue\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mmessages\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n",
"File \u001b[0;32m~/.pyenv/versions/3.11.2/lib/python3.11/site-packages/langgraph/pregel/__init__.py:645\u001b[0m, in \u001b[0;36mPregel.stream\u001b[0;34m(self, input, config, stream_mode, output_keys, input_keys, interrupt_before_nodes, interrupt_after_nodes, debug)\u001b[0m\n\u001b[1;32m 643\u001b[0m \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[1;32m 644\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m step \u001b[38;5;241m==\u001b[39m config[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrecursion_limit\u001b[39m\u001b[38;5;124m\"\u001b[39m]:\n\u001b[0;32m--> 645\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m GraphRecursionError(\n\u001b[1;32m 646\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mRecursion limit of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mconfig[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mrecursion_limit\u001b[39m\u001b[38;5;124m'\u001b[39m]\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m reached\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 647\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mwithout hitting a stop condition. You can increase the \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 648\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlimit by setting the `recursion_limit` config key.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 649\u001b[0m )\n\u001b[1;32m 651\u001b[0m \u001b[38;5;66;03m# before execution, check if we should interrupt\u001b[39;00m\n\u001b[1;32m 652\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m _should_interrupt(\n\u001b[1;32m 653\u001b[0m checkpoint,\n\u001b[1;32m 654\u001b[0m interrupt_before_nodes,\n\u001b[1;32m 655\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mstream_channels_list,\n\u001b[1;32m 656\u001b[0m next_tasks,\n\u001b[1;32m 657\u001b[0m ):\n",
"\u001b[0;31mGraphRecursionError\u001b[0m: Recursion limit of 25 reachedwithout hitting a stop condition. You can increase the limit by setting the `recursion_limit` config key."
]
}
],
"source": [
"silver_input = {\n",
" \"messages\": [(\"user\", silver_row[\"description\"])],\n",
" \"test_cases\": silver_row[\"test_cases\"],\n",
" \"runtime_limit\": silver_row[\"runtime_limit\"],\n",
" \"status\": \"in_progress\",\n",
"}\n",
"\n",
"\n",
"config = {\"configurable\": {\"thread_id\": \"silver-question-1\", \"k\": 2}}\n",
"with tracing_v2_enabled(client=client):\n",
" events = graph.stream(silver_input, config)\n",
" for event in events:\n",
" for value in event.values():\n",
" messages = value.get(\"messages\")\n",
" if messages:\n",
" if isinstance(messages, list):\n",
" messages = value[\"messages\"][-1]\n",
" print(\n",
" \"Assistant:\",\n",
" str(messages.content).replace(\"\\n\", \"\\\\n\")[:50],\n",
" )\n",
" elif value.get(\"examples\"):\n",
" print(\"Retrieved examples:\\n\\n\", value[\"examples\"][:100] + \"...\")\n",
" elif value.get(\"candidate\"):\n",
" print(str(value[\"candidate\"].content)[:200])"
]
},
{
"cell_type": "markdown",
"id": "7b4c9217-38de-4f48-9070-2092fd9ecc15",
"metadata": {},
"source": [
"**Still too hard!** AGI not achieved yet. To investigate our agent's trajectory in detail, check out the [full LangSmith trace](https://smith.langchain.com/public/13018b44-0c4f-4f1a-9e6d-dea1f3fd4705/r).\n",
"\n",
"Our agent isn't good enough to be autonomous. The great thing about LangGraph is you don't have to decide between \"autonomous agent\" and \"simple DAG\": you can inject control and user-interfaces wherever it can usefully benefit your application."
]
},
{
"attachments": {},
"cell_type": "markdown",
"id": "59177162-3fb8-4307-8e7f-ea67e785d4cf",
"metadata": {},
"source": [
"## Part 3: Human-in-the-loop\n",
"\n",
"Our retrieval-enhanced agent was able to solve the `bronze`-level question but still failed for those with the more challenging **silver** difficulty. \n",
"\n",
"Recall that the paper presented 3 complementary techniques that improved performance:\n",
"\n",
"1. Reflection: explicitly prompting the LLM to \"reflect\" on its mistakes can help it\n",
"2. Few-shot prompting: retrieving relevant, high-quality examples as \"memory\"\n",
"3. **Human-in-the-loop collaboration:** without giving the correct answer, the human is allowed to help the agent reflect on its approach and point it in a better direction.\n",
"\n",
"\n",
"In this section, we will add the \"human\" node (marked as \"part 3\" in the diagram below), completing our agent graph:\n",
"\n",
"<img src=\"./img/diagram.png\" src=\"../img/diagram.png\" >\n",
"\n",
"From an ML perspective, this is a bit of a [clever hans](https://en.wikipedia.org/wiki/Clever_Hans), but from the application designer's perspective, where the primary goal is to achieve a higher combined success rate, letting the human interject with thoughts and insights is only natural. \n",
"\n",
"In either case, adding a human check to a LangGraph instance requires no extra lines of code. Let's do so by instructing the graph to `interrupt_after` the \"`evaluate`\" node to give the user a chance to modify the trajectory.\n",
"\n",
"Start assembling your graph below. The following section is identical to our application in part 2:"
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "3c6456ba-363c-4133-8631-6dabb042b6ce",
"metadata": {},
"outputs": [],
"source": [
"# This is all the same as before\n",
"from langgraph.checkpoint.memory import MemorySaver\n",
"from langgraph.graph import END, StateGraph, START\n",
"\n",
"builder = StateGraph(State)\n",
"prompt = hub.pull(\"wfh/usaco-draft-solver\")\n",
"llm = ChatAnthropic(model=\"claude-3-opus-20240229\", max_tokens_to_sample=4000)\n",
"\n",
"draft_solver = Solver(llm, prompt.partial(examples=\"\"))\n",
"builder.add_node(\"draft\", draft_solver)\n",
"builder.add_edge(START, \"draft\")\n",
"builder.add_node(\"retrieve\", retrieve_examples)\n",
"solver = Solver(llm, prompt)\n",
"builder.add_node(\"solve\", solver)\n",
"builder.add_node(\"evaluate\", evaluate)\n",
"builder.add_edge(\"draft\", \"retrieve\")\n",
"builder.add_edge(\"retrieve\", \"solve\")\n",
"builder.add_edge(\"solve\", \"evaluate\")\n",
"\n",
"\n",
"def control_edge(state: State):\n",
" if state.get(\"status\") == \"success\":\n",
" return END\n",
" return \"solve\"\n",
"\n",
"\n",
"builder.add_conditional_edges(\"evaluate\", control_edge, {END: END, \"solve\": \"solve\"})\n",
"checkpointer = MemorySaver()"
]
},
{
"cell_type": "markdown",
"id": "3d837103-ede4-4515-a0cd-0b3319382013",
"metadata": {},
"source": [
"Now finish by compiling the graph. Set`interrupt_after=[\"evaluate\"]` to instruct the agent to wait for human input before continuing execution."
]
},
{
"cell_type": "code",
"execution_count": 20,
"id": "461c13ba-01cc-44e1-b837-6a64d03069d9",
"metadata": {},
"outputs": [],
"source": [
"graph = builder.compile(\n",
" checkpointer=checkpointer,\n",
" # New: this tells the graph to break any time it goes to the \"human\" node\n",
" interrupt_after=[\"evaluate\"],\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 28,
"id": "040b2100-5b2e-40c9-b6af-a1b680e16ee3",
"metadata": {},
"outputs": [
{
"data": {
"image/jpeg": "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",
"text/plain": [
"<IPython.core.display.Image object>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from IPython.display import Image, display\n",
"\n",
"try:\n",
" display(Image(graph.get_graph().draw_mermaid_png()))\n",
"except Exception:\n",
" # This requires some extra dependencies and is optional\n",
" pass"
]
},
{
"cell_type": "markdown",
"id": "372462bf-c74e-4c18-bbee-c21c8132b2e5",
"metadata": {},
"source": [
"As you can see in the graph above, the structure is the same as Part 2, except that we've inserted a \"`human`\" breakpoint between the \"`evaluate`\" and \"`solve`\" nodes.\n",
"\n",
"Let's try this question again!"
]
},
{
"cell_type": "code",
"execution_count": 41,
"id": "5f9ad0d0-cdaf-4ba2-9527-b5b9739e7b67",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'text': \"<thinking>\\nTo solve this problem, we need to:\\n1. Read in the input data - number of ports N, length of direction sequence M, number of repetitions K, the port connections, and the directi\n",
"Retrieved examples:\n",
"\n",
" \n",
"You previously solved the following problems in this competition:\n",
"<Examples>\n",
"<problem>\n",
"Farmer John ...\n",
"Assistant: [{'text': '<thinking>\\nTo determine where Bessie e\n",
"Assistant: Incorrect submission. Please respond with updated \n"
]
}
],
"source": [
"config = {\"configurable\": {\"thread_id\": \"silver-hl-1\", \"k\": 2}}\n",
"with tracing_v2_enabled(client=client):\n",
" events = graph.stream(silver_input, config)\n",
" for event in events:\n",
" for value in event.values():\n",
" messages = value.get(\"messages\")\n",
" if messages:\n",
" if isinstance(messages, list):\n",
" messages = value[\"messages\"][-1]\n",
" print(\n",
" \"Assistant:\",\n",
" str(messages.content).replace(\"\\n\", \"\\\\n\")[:50],\n",
" )\n",
" elif value.get(\"examples\"):\n",
" print(\"Retrieved examples:\\n\\n\", value[\"examples\"][:100] + \"...\")\n",
" elif value.get(\"candidate\"):\n",
" print(str(value[\"candidate\"].content)[:200])"
]
},
{
"cell_type": "markdown",
"id": "947e3089-6ae5-40e4-b993-7ef7e8630a12",
"metadata": {},
"source": [
"**⏰Time to weigh in⏰:** our model failed in its first attempt, so we have the opportunity to give it some advice.\n",
"\n",
"Recall the original question:"
]
},
{
"cell_type": "code",
"execution_count": 42,
"id": "4fcdf8c9-6a5a-4463-90ae-eec89a57cd05",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Problem 3: Luxury River Cruise [Josh Alman and Nathan Pinsker, 2013]\n",
"\n",
"Farmer John is taking Bessie and the cows on a cruise! They are sailing on a \n",
"network of rivers with N ports (1 <= N <= 1,000) labeled 1..N, and Bessie \n",
"starts at port 1. Each port has exactly two rivers leading out of it which \n",
"lead directly to other ports, and rivers can only be sailed one way.\n",
"\n",
"At each port, the tour guides choose either the \"left\" river or the \"right\" \n",
"river to sail down next, but they keep repeating the same choices over and \n",
"over. More specifically, the tour guides have chosen a short sequence of M \n",
"directions (1 <= M <= 500), each either \"left\" or \"right\", and have\n",
"repeated it K times (1 <= K <= 1,000,000,000). Bessie thinks she is going\n",
"in circles -- help her figure out where she ends up!\n",
"\n",
"PROBLEM NAME: cruise\n",
"\n",
"INPUT FORMAT:\n",
"\n",
"* Line 1: Three space-separated integers N, M, and K.\n",
"\n",
"* Lines 2..N+1: Line i+1 has two space-separated integers,\n",
" representing the number of the ports that port i's left and\n",
" right rivers lead to, respectively.\n",
"\n",
"* Line N+2: M space-separated characters, either 'L' or 'R'. 'L'\n",
" represents a choice of 'left' and 'R' represents a choice of\n",
" 'right'.\n",
"\n",
"SAMPLE INPUT:\n",
"\n",
"4 3 3\n",
"2 4\n",
"3 1\n",
"4 2\n",
"1 3\n",
"L L R\n",
"\n",
"INPUT DETAILS:\n",
"\n",
"The port numbers are arranged clockwise in a circle, with 'L' being a \n",
"clockwise rotation and 'R' being a counterclockwise rotation. The sequence \n",
"taken is LLRLLRLLR.\n",
"\n",
"OUTPUT FORMAT:\n",
"\n",
"* Line 1: A single integer giving the number of the port where\n",
" Bessie's cruise ends.\n",
"\n",
"SAMPLE OUTPUT:\n",
"\n",
"4\n",
"\n",
"OUTPUT DETAILS:\n",
"\n",
"After the first iteration of the sequence of directions, Bessie is at port\n",
"2 (1 -> 2 -> 3 -> 2); after the second, she is at port 3 (2 -> 3 -> 4 ->\n",
"3), and at the end she is at port 4 (3 -> 4 -> 1 -> 4).\n",
"\n"
]
}
],
"source": [
"snapshot = graph.get_state(config)\n",
"print(snapshot.values[\"messages\"][0].content)"
]
},
{
"cell_type": "markdown",
"id": "232ed165-e2cf-495f-a4bb-a0f3f87a719c",
"metadata": {},
"source": [
"And then review the agent's current submission:"
]
},
{
"cell_type": "code",
"execution_count": 43,
"id": "2049980a-0a0c-4135-98c7-d4de1af4757b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"<thinking>\n",
"To determine where Bessie ends up, we need to:\n",
"1. Simulate the cruise by following the sequence of left/right directions\n",
"2. Repeat this sequence K times to find the final destination port\n",
"\n",
"The problem provides:\n",
"- The number of ports N\n",
"- The connections between ports (left and right rivers for each port)\n",
"- The sequence of M directions (L or R) to follow\n",
"- The number of times K to repeat the sequence\n",
"\n",
"With this information, we have everything needed to simulate the cruise and find the ending port. The key steps will be:\n",
"1. Read in the input data to initialize the river connections and direction sequence \n",
"2. Iterate K times:\n",
" - For each direction in the M-length sequence:\n",
" - Move to the next port based on the current port and direction \n",
"3. Output the final port number after K iterations\n",
"\n",
"The solution will require loops to repeat the sequence K times and follow the M directions. Since K can be up to 1 billion, simulating all K iterations directly would be too slow. Instead, we can find a pattern in how the port changes after each M-length sequence, and then \"fast-forward\" by calculating which port we reach after K repetitions of the pattern.\n",
"</thinking>\n",
"\n",
"\n",
"Code:\n",
"\n",
"\n",
"N, M, K = map(int, input().split())\n",
"\n",
"ports = []\n",
"for _ in range(N):\n",
" left, right = map(int, input().split())\n",
" ports.append((left, right))\n",
"\n",
"directions = input().split()\n",
"\n",
"cur = 1\n",
"pattern = []\n",
"seen = set() \n",
"steps = 0\n",
"\n",
"while cur not in seen:\n",
" seen.add(cur)\n",
" for d in directions:\n",
" steps += 1\n",
" if d == 'L': \n",
" cur = ports[cur-1][0]\n",
" else:\n",
" cur = ports[cur-1][1]\n",
" pattern.append((cur, steps))\n",
"\n",
"K %= steps\n",
"for port, step in pattern:\n",
" if step > K:\n",
" cur = port\n",
" break\n",
" K -= step\n",
" \n",
"print(cur)\n"
]
}
],
"source": [
"snapshot = graph.get_state(config)\n",
"print(snapshot.values[\"messages\"][-2].content[0][\"text\"])\n",
"print(\"\\n\\nCode:\\n\\n\")\n",
"print(snapshot.values[\"messages\"][-2].tool_calls[0][\"args\"][\"code\"])"
]
},
{
"cell_type": "code",
"execution_count": 44,
"id": "69fd7efa-acbc-4450-927e-fb2b760066b2",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Incorrect submission. Please respond with updated code.\n",
"Pass rate: 4/10\n",
"Results:\n",
"<test id=0>\n",
"wrong answer. Expected '4\n",
"', got '3\n",
"'\n",
"</test>\n",
"<test id=1>\n",
"wrong answer. Expected '50\n",
"', got '2\n",
"'\n",
"</test>\n",
"<t\n"
]
}
],
"source": [
"print(snapshot.values[\"messages\"][-1].content[:200])"
]
},
{
"cell_type": "markdown",
"id": "0afb3b34-8c67-4ceb-95ff-4ac65053c8a4",
"metadata": {},
"source": [
"The agent failed. It's on the right track but clearly doesn't handle all the edge cases.\n",
"\n",
"The agent needs to remember that simulation should include the cycle + whatever steps led up to the example. It could use the \"tortoise and hare\" algo for cycle detection, use the simulated path and break if and when a repeat is detected, and then \n",
"\n",
"Let's let the agent know this by **updating the graph state**."
]
},
{
"cell_type": "code",
"execution_count": 45,
"id": "b10fcbc9-6dd4-41ad-98f7-1cf1685035e6",
"metadata": {},
"outputs": [],
"source": [
"updated_config = graph.update_state(\n",
" config,\n",
" values={\n",
" \"messages\": [\n",
" (\n",
" \"user\",\n",
" \"\"\"Consider breaking down the algorithm into separate parts: reading inputs, detecting cycles using the tortoise and hare algorithm, and determining Bessie's final position by skipping ahead K steps.\n",
"\n",
"Read the inputs into three arrays:\n",
"- Two arrays L and R for the ports (adjust for 0-based indexing)\n",
"- A third array S for the direction sequence\n",
"\n",
"Optimize by multiplying K by M before the main loop to convert the number of repetitions into the total number of steps.\n",
"\n",
"Use the tortoise and hare algorithm to detect the cycle:\n",
"- Define a helper function get_next(v) that returns the next position and direction index\n",
"- Initialize two pointers s0 and s1 to (0, 0)\n",
"- In each iteration:\n",
" - Move s0 by 1 step and s1 by 2 steps using get_next()\n",
" - If s0 equals s1, decrement K by 1 and break out of the loop\n",
" - Otherwise, decrement K by 1\n",
"- After the loop, if K is not 0, there is a cycle\n",
"\n",
"To find the cycle length:\n",
"- Initialize a counter variable rho to 1\n",
"- Move s0 by 1 step using get_next()\n",
"- Enter a loop:\n",
" - Move s0 by 1 step using get_next()\n",
" - Increment rho\n",
" - If s0 equals s1, break out of the loop\n",
"\n",
"Skip ahead by reducing K modulo rho.\n",
"\n",
"Simulate the remaining steps:\n",
"- While K > 0, move s0 to the next position using get_next() and decrement K\n",
"\n",
"Print the final position (converted to 1-based indexing).\n",
"\n",
"Pay close attention to the initialization and movement of pointers during cycle detection and length calculation. Ensure that the logic is correct and handles all cases accurately.\"\"\",\n",
" )\n",
" ]\n",
" },\n",
")"
]
},
{
"cell_type": "markdown",
"id": "e0542e9b-0992-407e-967f-c43bf1a75cc6",
"metadata": {},
"source": [
"Now the graph's state contains our new message."
]
},
{
"cell_type": "code",
"execution_count": 46,
"id": "818e93f2-2204-4704-a832-f5c102d330f8",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"HumanMessage(content=\"Consider breaking down the algorithm into separate parts: reading inputs, detecting cycles using the tortoise and hare algorithm, and determining Bessie's final position by skipping ahead K steps.\\n\\nRead the inputs into three arrays:\\n- Two arrays L and R for the ports (adjust for 0-based indexing)\\n- A third array S for the direction sequence\\n\\nOptimize by multiplying K by M before the main loop to convert the number of repetitions into the total number of steps.\\n\\nUse the tortoise and hare algorithm to detect the cycle:\\n- Define a helper function get_next(v) that returns the next position and direction index\\n- Initialize two pointers s0 and s1 to (0, 0)\\n- In each iteration:\\n - Move s0 by 1 step and s1 by 2 steps using get_next()\\n - If s0 equals s1, decrement K by 1 and break out of the loop\\n - Otherwise, decrement K by 1\\n- After the loop, if K is not 0, there is a cycle\\n\\nTo find the cycle length:\\n- Initialize a counter variable rho to 1\\n- Move s0 by 1 step using get_next()\\n- Enter a loop:\\n - Move s0 by 1 step using get_next()\\n - Increment rho\\n - If s0 equals s1, break out of the loop\\n\\nSkip ahead by reducing K modulo rho.\\n\\nSimulate the remaining steps:\\n- While K > 0, move s0 to the next position using get_next() and decrement K\\n\\nPrint the final position (converted to 1-based indexing).\\n\\nPay close attention to the initialization and movement of pointers during cycle detection and length calculation. Ensure that the logic is correct and handles all cases accurately.\", id='98888982-a469-4c5a-ab65-743d2f2608dc')"
]
},
"execution_count": 46,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"graph.get_state(config).values[\"messages\"][-1]"
]
},
{
"cell_type": "markdown",
"id": "4eb67198-c84f-458b-8baf-783d7246dddc",
"metadata": {},
"source": [
"Let's let the agent try again. Call `stream` with `None` to just use the inputs loaded from the memory. We will skip our human review for the next few attempts\n",
"to see if it can correct itself."
]
},
{
"cell_type": "code",
"execution_count": 47,
"id": "d5d76d8f-e49b-46bf-a762-a6c6978ee96c",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Assistant: [{'text': '<thinking>\\nThank you for the detailed \n",
"Assistant: Incorrect submission. Please respond with updated \n",
"Continuing...\n"
]
}
],
"source": [
"num_trials = 1\n",
"with tracing_v2_enabled(client=client):\n",
" for _ in range(num_trials):\n",
" events = graph.stream(None, updated_config)\n",
" for event in events:\n",
" for value in event.values():\n",
" messages = value.get(\"messages\")\n",
" if messages:\n",
" if isinstance(messages, list):\n",
" messages = value[\"messages\"][-1]\n",
" print(\n",
" \"Assistant:\",\n",
" str(messages.content).replace(\"\\n\", \"\\\\n\")[:50],\n",
" )\n",
" elif value.get(\"examples\"):\n",
" print(\"Retrieved examples:\\n\\n\", value[\"examples\"][:100] + \"...\")\n",
" elif value.get(\"candidate\"):\n",
" print(str(value[\"candidate\"].content)[:200])\n",
" if graph.get_state(config).values[\"status\"] == \"success\":\n",
" break\n",
" print(\"Continuing...\")"
]
},
{
"cell_type": "code",
"execution_count": 48,
"id": "20ee7535-1bc8-4105-87c4-0e7a89a011ff",
"metadata": {},
"outputs": [],
"source": [
"most_recent_state = list(graph.get_state_history(config))[0]"
]
},
{
"cell_type": "markdown",
"id": "f9a492fd-03fc-4de5-8ba5-3ccdaeb1791c",
"metadata": {},
"source": [
"OK so the agent tried again. Check out the [LangSmith trace](https://smith.langchain.com/public/707be522-9eaf-4b6a-994e-1742f421a433/r/add3d8e7-85b1-40cf-bbd3-e78c50f835e8) from this step to see its update."
]
},
{
"cell_type": "code",
"execution_count": 49,
"id": "1f691881-f56e-4e2c-8b6a-5febb7ccf3ae",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'text': '<thinking>\\nThank you for the detailed algorithm breakdown! Let me go through each step to make sure I understand and have the necessary information to implement the solution.\\n\\nReading inputs:\\n- Read N, M, K and store in separate variables\\n- Create arrays L and R to store the left and right port connections (adjust for 0-based indexing)\\n- Create array S to store the M-length direction sequence \\n- Multiply K by M upfront to get the total number of steps\\n\\nDetecting cycles with tortoise and hare:\\n- Define get_next(v) to return the next position and direction index\\n - It will use the current position and direction to look up the next port in L/R\\n- Initialize two pointers s0 and s1 to (0, 0) \\n- Loop until s0 equals s1 or all K steps are taken:\\n - Move s0 by 1 step and s1 by 2 steps using get_next()\\n - Decrement K\\n- After the loop, check if K is 0 to determine if a cycle was found\\n\\nFinding cycle length:\\n- If a cycle was found, initialize rho to 1\\n- Move s0 by 1 step \\n- Loop until s0 equals s1 again:\\n - Move s0 by 1 step and increment rho\\n- rho will equal the cycle length\\n\\nSkipping ahead:\\n- Reduce K by taking it modulo rho\\n\\nSimulating remaining steps:\\n- While K is greater than 0:\\n - Move s0 using get_next()\\n - Decrement K\\n- s0 will hold the final position\\n\\nPrinting result:\\n- Add 1 to the final position to convert back to 1-based indexing before printing\\n\\nThe key aspects are:\\n- Handling the input format and 0-based indexing \\n- Defining get_next() to handle moving to the next port based on direction\\n- Correctly implementing the tortoise and hare cycle detection\\n- Finding the cycle length after detection\\n- Skipping ahead with modulo and simulating any remaining steps\\n- Adjusting the output back to 1-based indexing\\n\\nI believe I have all the necessary pieces to implement this solution now. Let me code it up using the writePython tool.\\n</thinking>', 'type': 'text'}, {'id': 'toolu_01EDrYeHJU7GxApRb1QfMA1b', 'input': {'reasoning': \"Here's the problem-solving approach:\\n\\n1. Read in the input data:\\n - N ports, M-length direction sequence, K repetitions\\n - L and R arrays for left/right port connections\\n - S array for direction sequence\\n - Multiply K by M to get total steps\\n\\n2. Define get_next(v) helper function:\\n - Takes current position and direction index\\n - Returns next position and incremented direction index\\n - Looks up next port in L/R arrays based on current direction\\n\\n3. Detect cycle using tortoise and hare algorithm:\\n - Initialize s0 and s1 pointers to (0, 0)\\n - Loop until match or all steps taken:\\n - Move s0 by 1 step, s1 by 2 steps\\n - Decrement K\\n - Check if K is 0 after loop\\n\\n4. If cycle found, find cycle length:\\n - Initialize rho to 1\\n - Move s0 by 1 step\\n - Loop until s0 equals s1 again:\\n - Move s0 and increment rho\\n - rho is the cycle length\\n\\n5. Skip ahead by K % rho steps\\n\\n6. Simulate remaining steps:\\n - While K > 0:\\n - Move s0 with get_next()\\n - Decrement K\\n \\n7. Print final position (+1 for 1-based indexing)\\n\\nKey points:\\n- Multiplying K*M avoids nested loop\\n- get_next() handles port transitions \\n- Tortoise and hare finds cycles\\n- Modulo skips ahead in cycle\\n- Adjust 0-based indexing for input/output\", 'pseudocode': \"1. Read input:\\n N, M, K = read_ints()\\n L = [0] * N\\n R = [0] * N\\n for i in 0..N-1:\\n L[i], R[i] = read_ints()\\n S = read_direction_sequence()\\n K *= M\\n\\n2. Define get_next(v):\\n def get_next(pos, dir_idx):\\n if S[dir_idx] == 'L':\\n next_pos = L[pos]\\n else:\\n next_pos = R[pos]\\n next_dir_idx = (dir_idx + 1) % M\\n return (next_pos, next_dir_idx)\\n\\n3. Find cycle:\\n s0 = (0, 0)\\n s1 = (0, 0) \\n while K:\\n s0 = get_next(s0[0], s0[1])\\n s1 = get_next(s1[0], get_next(s1[0], s1[1])[1])\\n K -= 1\\n if s0 == s1: break\\n if K != 0: no cycle, print s0[0] + 1\\n\\n4. Find cycle length:\\n rho = 1\\n s0 = get_next(s0[0], s0[1])\\n while s0 != s1:\\n s0 = get_next(s0[0], s0[1]) \\n rho += 1\\n\\n5. Skip steps:\\n K %= rho\\n\\n6. Remaining steps: \\n while K:\\n s0 = get_next(s0[0], s0[1])\\n K -= 1\\n \\n7. Print result:\\n print(s0[0] + 1)\", 'code': \"def read_ints():\\n return map(int, input().split())\\n\\nN, M, K = read_ints()\\n\\nL = [0] * N\\nR = [0] * N\\nfor i in range(N):\\n L[i], R[i] = read_ints()\\n L[i] -= 1\\n R[i] -= 1\\n\\nS = input().split()\\n\\nK *= M\\n\\ndef get_next(pos, dir_idx):\\n if S[dir_idx] == 'L':\\n next_pos = L[pos] \\n else:\\n next_pos = R[pos]\\n next_dir_idx = (dir_idx + 1) % M\\n return (next_pos, next_dir_idx)\\n\\ns0 = (0, 0) \\ns1 = (0, 0)\\n\\nwhile K:\\n if s0 == s1: break\\n \\n s0 = get_next(s0[0], s0[1])\\n s1 = get_next(s1[0], get_next(s1[0], s1[1])[1])\\n \\n K -= 1\\n \\nif K:\\n rho = 1\\n s0 = get_next(s0[0], s0[1])\\n while s0 != s1:\\n s0 = get_next(s0[0], s0[1])\\n rho += 1\\n \\n K %= rho\\n \\nwhile K: \\n s0 = get_next(s0[0], s0[1])\\n K -= 1\\n \\nprint(s0[0] + 1)\"}, 'name': 'writePython', 'type': 'tool_use'}]\n",
"\n",
"\n",
"Code:\n",
"\n",
"\n",
"def read_ints():\n",
" return map(int, input().split())\n",
"\n",
"N, M, K = read_ints()\n",
"\n",
"L = [0] * N\n",
"R = [0] * N\n",
"for i in range(N):\n",
" L[i], R[i] = read_ints()\n",
" L[i] -= 1\n",
" R[i] -= 1\n",
"\n",
"S = input().split()\n",
"\n",
"K *= M\n",
"\n",
"def get_next(pos, dir_idx):\n",
" if S[dir_idx] == 'L':\n",
" next_pos = L[pos] \n",
" else:\n",
" next_pos = R[pos]\n",
" next_dir_idx = (dir_idx + 1) % M\n",
" return (next_pos, next_dir_idx)\n",
"\n",
"s0 = (0, 0) \n",
"s1 = (0, 0)\n",
"\n",
"while K:\n",
" if s0 == s1: break\n",
" \n",
" s0 = get_next(s0[0], s0[1])\n",
" s1 = get_next(s1[0], get_next(s1[0], s1[1])[1])\n",
" \n",
" K -= 1\n",
" \n",
"if K:\n",
" rho = 1\n",
" s0 = get_next(s0[0], s0[1])\n",
" while s0 != s1:\n",
" s0 = get_next(s0[0], s0[1])\n",
" rho += 1\n",
" \n",
" K %= rho\n",
" \n",
"while K: \n",
" s0 = get_next(s0[0], s0[1])\n",
" K -= 1\n",
" \n",
"print(s0[0] + 1)\n"
]
}
],
"source": [
"snapshot = graph.get_state(most_recent_state.config)\n",
"ai_message = snapshot.values[\"messages\"][-2]\n",
"if ai_message.content:\n",
" print(ai_message.content)\n",
"print(\"\\n\\nCode:\\n\\n\")\n",
"print(ai_message.tool_calls[0][\"args\"][\"code\"] if ai_message.tool_calls else \"N/A\")"
]
},
{
"cell_type": "code",
"execution_count": 50,
"id": "b6fdebe7-557e-4751-97f3-15901c95600a",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Incorrect submission. Please respond with updated code.\n",
"Pass rate: 3/10\n",
"Results:\n",
"<test id=0>\n",
"passed\n",
"</test>\n",
"<test id=1>\n",
"timed out\n",
"</test>\n",
"<test id=2>\n",
"timed out\n",
"</test>\n",
"<test id=3>\n",
"timed out\n",
"</test>\n",
"<t\n"
]
}
],
"source": [
"print(snapshot.values[\"messages\"][-1].content[:200])"
]
},
{
"cell_type": "markdown",
"id": "ff2f9068-3c42-408c-a414-5914f923b78f",
"metadata": {},
"source": [
"Still getting most test cases wrong.\n",
"\n",
"Let's provide more feedback."
]
},
{
"cell_type": "code",
"execution_count": 53,
"id": "6eb46517-cdd1-4716-8a9e-df72cdd9ba67",
"metadata": {},
"outputs": [],
"source": [
"updated_config = graph.update_state(\n",
" updated_config,\n",
" values={\n",
" \"messages\": [\n",
" (\n",
" \"user\",\n",
" \"\"\"That's better, but you're still getting some errors. Let's double check some things:\n",
" \n",
"1. When calculating the cycle length, make sure the initialization and movement of the pointers is correct. Double-check the logic there and see if you can spot any discrepancies.\n",
"2. Check the condition for whether there's a cycle after the main loop to ensure it covers all cases, like if K becomes 0 in the last iteration.\n",
"\n",
"Think step by step through youur implementation and update using the writePython tool.\"\"\",\n",
" )\n",
" ]\n",
" },\n",
")"
]
},
{
"cell_type": "markdown",
"id": "cd222b55-9bc5-4db0-8d28-58f366826dd4",
"metadata": {},
"source": [
"Now that we've provided this feedback, let's give the agent a few attempts at solving it before we weigh in again."
]
},
{
"cell_type": "code",
"execution_count": 54,
"id": "ab89905b-ab75-4820-bde8-20ae01410aa7",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Assistant: [{'text': \"<thinking>\\nThe algorithm looks mostly \n"
]
}
],
"source": [
"num_trials = 2\n",
"with tracing_v2_enabled(client=client):\n",
" for _ in range(num_trials):\n",
" events = graph.stream(None, updated_config)\n",
" for event in events:\n",
" for value in event.values():\n",
" messages = value.get(\"messages\")\n",
" if messages:\n",
" if isinstance(messages, list):\n",
" messages = value[\"messages\"][-1]\n",
" print(\n",
" \"Assistant:\",\n",
" str(messages.content).replace(\"\\n\", \"\\\\n\")[:50],\n",
" )\n",
" elif value.get(\"examples\"):\n",
" print(\"Retrieved examples:\\n\\n\", value[\"examples\"][:100] + \"...\")\n",
" elif value.get(\"candidate\"):\n",
" print(str(value[\"candidate\"].content)[:200])\n",
" if graph.get_state(config).values[\"status\"] == \"success\":\n",
" break\n",
" print(\"Continuing...\")"
]
},
{
"cell_type": "markdown",
"id": "28446ca4-0063-4ac2-bd87-bb9319fef84c",
"metadata": {},
"source": [
"You can review [a LangSmith trace (link)](https://smith.langchain.com/public/d383e743-f8f1-4206-9dce-47627f152612/r/3f89582f-9107-461a-a34e-608d52641eeb) of the agent's response to your feedback at the provided link."
]
},
{
"cell_type": "code",
"execution_count": 55,
"id": "28afec75-d661-4b88-b295-edb886231693",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"success\n"
]
}
],
"source": [
"snapshot = graph.get_state(config)\n",
"print(snapshot.values[\"status\"])"
]
},
{
"cell_type": "markdown",
"id": "72327753-b8b6-4cbc-b5cc-b3cd90b7dd2e",
"metadata": {},
"source": [
"**Success!** - the LLM really wouldn't have been able to come to the correct answer without detailed human involvement."
]
},
{
"cell_type": "markdown",
"id": "d1e2d6e8-3daf-4cbe-acb9-387e3058e602",
"metadata": {},
"source": [
"## Conclusion\n",
"\n",
"Congrats on making it to the end! In this tutorial, you implemented an agent in LangGraph capable of solving challenging programming problems. You did so by leveraging a few common techniques to improve performance, including:\n",
"\n",
"1. **Reflection**: while we didn't implement an explicit reflection step, our prompt and tool invocation was designed to encourage critique of previous outputs. You added this in Part 1.\n",
"2. **Retrieval**: the \"episodic memory\" of the agent retrieves high-quality few-shot examples from our corpora of programming problems to help solve the **bronze** level question. In Part 2, you implemented a retrieval memory as an initial step.\n",
"3. **Human-in-the-loop**: LLM-powered agents are still too weak to answer all these questions autonomously, but at times, they can get most of the way there and land on the right answer with human feedback. In Part 3, you used `interrupt_after` on the `evaluate` node and then included your feedback by using `update_state` on the graph.\n",
"\n",
"\n",
"LLMs are not capable of solving all these problems autonomously, but through better prompting and clever engineering, you can create a system that is able to more reliably arrive at the proper solution."
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"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": "3.11.9"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Reference in New Issue View Git Blame Copy Permalink