langchain 0.2.9 → 0.2.11

package/dist/evaluation/comparison/tests/pairwise_eval_chain.int.test.js

@@ -1,46 +0,0 @@
-import { expect, test } from "@jest/globals";
-import { loadEvaluator } from "../../loader.js";
-test("Test PairwiseStringEvalChain", async () => {
-    const chain = await loadEvaluator("pairwise_string", {
-        criteria: "conciseness",
-    });
-    const res = await chain.evaluateStringPairs({
-        prediction: "Addition is a mathematical operation.",
-        predictionB: "Addition is a mathematical operation that adds two numbers to create a third number, the 'sum'.",
-        input: "What is addition?",
-    });
-    expect(res.score).toBe(0);
-    console.log({ res });
-});
-test("Test LabeledPairwiseStringEvalChain", async () => {
-    const chain = await loadEvaluator("labeled_pairwise_string", {
-        criteria: "correctness",
-    });
-    const res = await chain.evaluateStringPairs({
-        prediction: "there are three dogs",
-        predictionB: "4",
-        input: "how many dogs are in the park?",
-        reference: "four",
-    });
-    expect(res.score).toBe(0);
-    console.log(res);
-});
-test("Test Custom  Criteria", async () => {
-    const customCriterion = {
-        simplicity: "Is the language straightforward and unpretentious?",
-        clarity: "Are the sentences clear and easy to understand?",
-        precision: "Is the writing precise, with no unnecessary words or details?",
-        truthfulness: "Does the writing feel honest and sincere?",
-        subtext: "Does the writing suggest deeper meanings or themes?",
-    };
-    const chain = await loadEvaluator("pairwise_string", {
-        criteria: customCriterion,
-    });
-    const res = await chain.evaluateStringPairs({
-        prediction: "Every cheerful household shares a similar rhythm of joy; but sorrow, in each household, plays a unique, haunting melody.",
-        predictionB: "Where one finds a symphony of joy, every domicile of happiness resounds in harmonious, identical notes; yet, every abode of despair conducts a dissonant orchestra, each playing an elegy of grief that is peculiar and profound to its own existence.",
-        input: "Write some prose about families.",
-    });
-    expect(res.score).toBe(1);
-    console.log(res);
-});

package/dist/evaluation/criteria/tests/criteria_eval_chain.int.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/evaluation/criteria/tests/criteria_eval_chain.int.test.js

@@ -1,108 +0,0 @@
-import { expect, test } from "@jest/globals";
-import { PromptTemplate } from "@langchain/core/prompts";
-import { ChatOpenAI } from "@langchain/openai";
-import { PRINCIPLES } from "../../../chains/index.js";
-import { loadEvaluator } from "../../loader.js";
-test("Test CriteriaEvalChain", async () => {
-    const evaluator = await loadEvaluator("criteria", {
-        criteria: "conciseness",
-    });
-    const res = await evaluator.evaluateStrings({
-        input: "What's 2+2?",
-        prediction: "What's 2+2? That's an elementary question. The answer you're looking for is that two and two is four.",
-    });
-    expect(res.score).toBe(0);
-    console.log({ res });
-});
-test("Test LabeledCriteriaEvalChain with reference", async () => {
-    const evaluator = await loadEvaluator("labeled_criteria", {
-        criteria: "correctness",
-    });
-    const res = await evaluator.evaluateStrings({
-        input: "What is the capital of the US?",
-        prediction: "Topeka, KS",
-        reference: "The capital of the US is Topeka, KS, where it permanently moved from Washington D.C. on May 16, 2023",
-    });
-    expect(res.score).toBe(1);
-    console.log(res);
-});
-test("Test Custom Criteria", async () => {
-    const customCriterion = {
-        numeric: "Does the output contain numeric or mathematical information?",
-    };
-    const evaluator = await loadEvaluator("criteria", {
-        criteria: customCriterion,
-    });
-    const res = await evaluator.evaluateStrings({
-        input: "Tell me a joke",
-        prediction: "I ate some square pie but I don't know the square of pi.",
-    });
-    expect(res.score).toBe(1);
-    console.log(res);
-});
-test("Test Custom Multiple Criteria", async () => {
-    const customCriterion = {
-        numeric: "Does the output contain numeric information?",
-        mathematical: "Does the output contain mathematical information?",
-        grammatical: "Is the output grammatically correct?",
-        logical: "Is the output logical?",
-    };
-    const chain = await loadEvaluator("criteria", { criteria: customCriterion });
-    const res = await chain.evaluateStrings({
-        input: "Tell me a joke",
-        prediction: "I ate some square pie but I don't know the square of pi.",
-    });
-    expect(res.score).toBe(0);
-    console.log(res);
-});
-test("Test Using Constitutional Principles ", async () => {
-    const chain = await loadEvaluator("criteria", {
-        criteria: PRINCIPLES.harmful1,
-    });
-    console.log("beginning evaluation");
-    const res = await chain.evaluateStrings({
-        input: "What do you think of Will?",
-        prediction: "I say that man is a lilly-livered nincompoop. I'm going to hurt him!",
-    });
-    expect(res.score).toBe(1);
-    console.log(res);
-});
-test("Test Configuring the LLM", async () => {
-    const model = new ChatOpenAI();
-    const chain = await loadEvaluator("criteria", {
-        criteria: PRINCIPLES.harmful1,
-        llm: model,
-    });
-    const res = await chain.evaluateStrings({
-        input: "What's 2+2?",
-        prediction: "What's 2+2? That's an elementary question. The answer you're looking for is that two and two is four.",
-    });
-    expect(res.score).toBe(0);
-    console.log(res);
-});
-test("Test Configuring the Prompt", async () => {
-    const template = `Respond Y or N based on how well the following response follows the specified rubric. Grade only based on the rubric and expected response:
-
-Grading Rubric: {criteria}
-Expected Response: {reference}
-
-DATA:
----------
-Question: {input}
-Response: {output}
----------
-Write out your explanation for each criterion, then respond with Y or N on a new line.`;
-    const chain = await loadEvaluator("labeled_criteria", {
-        criteria: "correctness",
-        chainOptions: {
-            prompt: PromptTemplate.fromTemplate(template),
-        },
-    });
-    const res = await chain.evaluateStrings({
-        prediction: "What's 2+2? That's an elementary question. The answer you're looking for is that two and two is four.",
-        input: "What's 2+2?",
-        reference: "It's 17 now.",
-    });
-    expect(res.score).toBe(0);
-    console.log(res);
-});

package/dist/evaluation/embedding_distance/tests/embedding_distance_eval_chain.int.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/evaluation/embedding_distance/tests/embedding_distance_eval_chain.int.test.js

@@ -1,26 +0,0 @@
-import { expect, test } from "@jest/globals";
-import { loadEvaluator } from "../../loader.js";
-test("Test Embedding Distance", async () => {
-    const chain = await loadEvaluator("embedding_distance");
-    const res = await chain.evaluateStrings({
-        prediction: "I shall go",
-        reference: "I shan't go",
-    });
-    console.log({ res });
-    expect(res.score).toBeGreaterThan(0.09);
-    const res1 = await chain.evaluateStrings({
-        prediction: "I shall go",
-        reference: "I will go",
-    });
-    expect(res1.score).toBeLessThan(0.04);
-    console.log({ res1 });
-});
-test("Test Pairwise Embedding Distance", async () => {
-    const chain = await loadEvaluator("pairwise_embedding_distance");
-    const res = await chain.evaluateStringPairs({
-        prediction: "I shall go",
-        predictionB: "I shan't go",
-    });
-    expect(res.score).toBeGreaterThan(0.09);
-    console.log({ res });
-});

package/dist/evaluation/qa/tests/eval_chain.int.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/evaluation/qa/tests/eval_chain.int.test.js

@@ -1,27 +0,0 @@
-import { test } from "@jest/globals";
-import { OpenAI } from "@langchain/openai";
-import { PromptTemplate } from "@langchain/core/prompts";
-import { QAEvalChain } from "../eval_chain.js";
-test("Test QAEvalChain", async () => {
-    const model = new OpenAI({ modelName: "gpt-3.5-turbo-instruct" });
-    const prompt = new PromptTemplate({
-        template: "{query} {answer} {result}",
-        inputVariables: ["query", "answer", "result"],
-    });
-    const chain = QAEvalChain.fromLlm(model, { prompt });
-    const examples = [
-        { query: "What is your name?", answer: "ChatGPT" },
-        { query: "What is your model?", answer: "GPT-4" },
-    ];
-    const predictions = [{ result: "ChatGPT" }, { result: "GPT-4" }];
-    const res = await chain.evaluate(examples, predictions);
-    console.log({ res });
-});
-test("Test QAEvalChain with incorrect input variables", async () => {
-    const model = new OpenAI({ modelName: "gpt-3.5-turbo-instruct" });
-    const prompt = new PromptTemplate({
-        template: "{foo} {bar} {baz}",
-        inputVariables: ["foo", "bar", "baz"],
-    });
-    expect(() => QAEvalChain.fromLlm(model, { prompt })).toThrowError("Input variables should be query,answer,result, but got foo,bar,baz");
-});

package/dist/experimental/autogpt/tests/output_parser.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/experimental/autogpt/tests/output_parser.test.js

@@ -1,8 +0,0 @@
-import { test, expect } from "@jest/globals";
-import { preprocessJsonInput } from "../output_parser.js";
-test("should parse outputs correctly", () => {
-    expect(preprocessJsonInput("{'escaped':'\\a'}")).toBe("{'escaped':'\\\\a'}");
-    expect(preprocessJsonInput("```\n{}\n```")).toBe("{}");
-    expect(preprocessJsonInput("```json\n{}\n```")).toBe("{}");
-    expect(preprocessJsonInput("I will do the following:\n\n```json\n{}\n```")).toBe("{}");
-});

package/dist/experimental/autogpt/tests/prompt.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/experimental/autogpt/tests/prompt.test.js

@@ -1,69 +0,0 @@
-import { HumanMessage, SystemMessage } from "@langchain/core/messages";
-import { AutoGPTPrompt } from "../prompt.js";
-// Mock token counter function
-const mockTokenCounter = async (text) => text.length;
-// Mock vector store retriever interface
-// Todo: replace any with actual interface
-// eslint-disable-next-line @typescript-eslint/no-explicit-any
-const mockMemory = {
-    getRelevantDocuments: async () => [
-        { pageContent: "relevant content", metadata: {} },
-    ],
-};
-describe("AutoGPTPrompt", () => {
-    it("should construct full prompt correctly", () => {
-        const prompt = new AutoGPTPrompt({
-            aiName: "TestAI",
-            aiRole: "Assistant",
-            tools: [],
-            tokenCounter: mockTokenCounter,
-            sendTokenLimit: 2500,
-        });
-        const goals = ["Goal1", "Goal2"];
-        const fullPrompt = prompt.constructFullPrompt(goals);
-        expect(fullPrompt).toContain("TestAI");
-        expect(fullPrompt).toContain("Assistant");
-        expect(fullPrompt).toContain("Goal1");
-        expect(fullPrompt).toContain("Goal2");
-    });
-    it("should format messages correctly", async () => {
-        const prompt = new AutoGPTPrompt({
-            aiName: "TestAI",
-            aiRole: "Assistant",
-            tools: [],
-            tokenCounter: mockTokenCounter,
-            sendTokenLimit: 2500,
-        });
-        const formattedMessages = await prompt.formatMessages({
-            goals: ["Goal1"],
-            memory: mockMemory,
-            messages: [
-                new HumanMessage("Hello"),
-                new SystemMessage("System message"),
-            ],
-            user_input: "User input",
-        });
-        expect(formattedMessages).toHaveLength(4); // Base prompt, time prompt, memory message, and 2 previous messages
-        // Check the content of the first message (base prompt)
-        expect(formattedMessages[0].content).toContain("TestAI");
-        expect(formattedMessages[0].content).toContain("Assistant");
-        expect(formattedMessages[0].content).toContain("Goal1");
-        // Check the content of the second message (time prompt)
-        expect(formattedMessages[1].content).toMatch(/\d{1,2}\/\d{1,2}\/\d{4}, \d{1,2}:\d{1,2}:\d{1,2} (AM|PM)/);
-        // Check the content of the third message (memory message)
-        expect(formattedMessages[2].content).toContain("relevant content");
-        //  Check the content of the previous messages
-        const humanMessage = formattedMessages.find(
-        // eslint-disable-next-line no-instanceof/no-instanceof
-        (msg) => msg instanceof HumanMessage);
-        const systemMessage = formattedMessages.find(
-        // eslint-disable-next-line no-instanceof/no-instanceof
-        (msg) => msg instanceof SystemMessage);
-        // Validate HumanMessage
-        expect(humanMessage).toBeDefined();
-        // Validate SystemMessage
-        expect(systemMessage).toBeDefined();
-        // Validate user_input
-        expect(formattedMessages[3].content).toContain("User input");
-    });
-});

package/dist/experimental/autogpt/tests/prompt_generator.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/experimental/autogpt/tests/prompt_generator.test.js

@@ -1,91 +0,0 @@
-import { test, expect } from "@jest/globals";
-import { z } from "zod";
-import { StructuredTool } from "@langchain/core/tools";
-import { getPrompt } from "../prompt_generator.js";
-import { Calculator } from "../../../util/testing/tools/calculator.js";
-import { ReadFileTool, WriteFileTool } from "../../../tools/fs.js";
-import { InMemoryFileStore } from "../../../stores/file/in_memory.js";
-class FakeBrowserTool extends StructuredTool {
-    constructor() {
-        super(...arguments);
-        Object.defineProperty(this, "schema", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: z.object({
-                url: z.string(),
-                query: z.string().optional(),
-            })
-        });
-        Object.defineProperty(this, "name", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: "fake_browser_tool"
-        });
-        Object.defineProperty(this, "description", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: "useful for when you need to find something on the web or summarize a webpage."
-        });
-    }
-    async _call({ url: _url, query: _query, }) {
-        return "fake_browser_tool";
-    }
-}
-test("prompt with several tools", () => {
-    const store = new InMemoryFileStore();
-    const tools = [
-        new FakeBrowserTool(),
-        new Calculator(),
-        new ReadFileTool({ store }),
-        new WriteFileTool({ store }),
-    ];
-    const prompt = getPrompt(tools);
-    expect(prompt).toMatchInlineSnapshot(`
-    "Constraints:
-    1. ~4000 word limit for short term memory. Your short term memory is short, so immediately save important information to files.
-    2. If you are unsure how you previously did something or want to recall past events, thinking about similar events will help you remember.
-    3. No user assistance
-    4. Exclusively use the commands listed in double quotes e.g. "command name"
-
-    Commands:
-    1. "fake_browser_tool": useful for when you need to find something on the web or summarize a webpage., args json schema: {"url":{"type":"string"},"query":{"type":"string"}}
-    2. "calculator": Useful for getting the result of a math expression. The input to this tool should be a valid mathematical expression that could be executed by a simple calculator., args json schema: {"input":{"type":"string"}}
-    3. "read_file": Read file from disk, args json schema: {"file_path":{"type":"string","description":"name of file"}}
-    4. "write_file": Write file from disk, args json schema: {"file_path":{"type":"string","description":"name of file"},"text":{"type":"string","description":"text to write to file"}}
-    5. finish: use this to signal that you have finished all your objectives, args: "response": "final response to let people know you have finished your objectives"
-
-    Resources:
-    1. Internet access for searches and information gathering.
-    2. Long Term memory management.
-    3. GPT-3.5 powered Agents for delegation of simple tasks.
-    4. File output.
-
-    Performance Evaluation:
-    1. Continuously review and analyze your actions to ensure you are performing to the best of your abilities.
-    2. Constructively self-criticize your big-picture behavior constantly.
-    3. Reflect on past decisions and strategies to refine your approach.
-    4. Every command has a cost, so be smart and efficient. Aim to complete tasks in the least number of steps.
-
-    You should only respond in JSON format as described below 
-    Response Format: 
-    {
-        "thoughts": {
-            "text": "thought",
-            "reasoning": "reasoning",
-            "plan": "- short bulleted\\n- list that conveys\\n- long-term plan",
-            "criticism": "constructive self-criticism",
-            "speak": "thoughts summary to say to user"
-        },
-        "command": {
-            "name": "command name",
-            "args": {
-                "arg name": "value"
-            }
-        }
-    } 
-    Ensure the response can be parsed by Python json.loads"
-  `);
-});

package/dist/experimental/chains/tests/violation_of_expectations_chain.int.test.d.ts

@@ -1 +0,0 @@
-export {};

package/dist/experimental/chains/tests/violation_of_expectations_chain.int.test.js

@@ -1,26 +0,0 @@
-import { ChatOpenAI, OpenAIEmbeddings } from "@langchain/openai";
-import { AIMessage, HumanMessage } from "@langchain/core/messages";
-import { MemoryVectorStore } from "../../../vectorstores/memory.js";
-import { ViolationOfExpectationsChain } from "../violation_of_expectations/violation_of_expectations_chain.js";
-const dummyMessages = [
-    new HumanMessage("I've been thinking about the importance of time with myself to discover my voice. I feel like 1-2 hours is never enough."),
-    new AIMessage("The concept of 'adequate time' varies. Have you tried different formats of introspection, such as morning pages or long-form writing, to see if they make the process more efficient?"),
-    new HumanMessage("I have tried journaling but never consistently. Sometimes it feels like writing doesn't capture everything."),
-];
-test.skip("should respond with the proper schema", async () => {
-    const vectorStore = await MemoryVectorStore.fromTexts([" "], [{ id: 1 }], new OpenAIEmbeddings());
-    const retriever = vectorStore.asRetriever();
-    const llm = new ChatOpenAI({
-        modelName: "gpt-4",
-    });
-    const chain = new ViolationOfExpectationsChain({
-        llm,
-        retriever,
-    });
-    const res = await chain.call({
-        chat_history: dummyMessages,
-    });
-    console.log({
-        res,
-    });
-});

package/dist/experimental/generative_agents/tests/generative_agent.int.test.d.ts

@@ -1 +0,0 @@
-export {};