Error: "Chain run errored with error: this.llm.pipe is not a function"

[ehiaig]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

chatbot.js:


export async function run() {
  const { LlamaCpp } = await import('@langchain/community/llms/llama_cpp')
  const model = new LlamaCpp({ modelPath: modelPath })
  const memory = new ConversationSummaryMemory({
    memoryKey: 'chat_history',
    llm: model,
    inputKey: 'question'
  })
  console.log("our memory", memory)
  const template = `You are a helpful assistant. Answer all questions to the best of your ability. If you do not know the answer to a question, truthfully say you do not know.

  Current conversation:
  {chat_history}
  Human: {question}
  AI:`
  const prompt = PromptTemplate.fromTemplate(template)
  const chain = new LLMChain({ llm: model, prompt: prompt, memory: memory, verbose: true })

  async function callChain(input) {
    const result = await chain.call({ question: input })
    console.log("our res", res)
    console.log({ 'res and memory': res, memory: await memory.loadMemoryVariables({}) })
    return res
  }
  return callChain
}

Description

I'm using an internal_K.gguf base model(not a chat model) in my electronjs app. When the line const result = await chain.call({ question: input }) runs, I get this:

[chain/start] [1:chain:LLMChain] Entering Chain run with input: {
  "question": "who is elon musk",
  "chat_history": ""
}
[chain/error] [1:chain:LLMChain] [1ms] Chain run errored with error: "this.llm.pipe is not a function\n\nTypeError: this.llm.pipe is not a function\n    at LLMChain._call 

Without using chain, I'm able to send and receive answers to the model without issues but I'm using chain for history/memory purpose and I can't seem to understand why this is happening. Would really pepreciate any help I can get.

System Info

my system info:

[email protected] | MIT | deps: 17 | versions: 261
Typescript bindings for langchain
https://github.com/langchain-ai/langchainjs/tree/main/langchain/

keywords: llm, ai, gpt3, chain, prompt, prompt engineering, chatgpt, machine learning, ml, openai, embeddings, vectorstores

dist
.tarball: https://registry.npmjs.org/langchain/-/langchain-0.1.30.tgz
.shasum: e1adb3f1849fcd5c596c668300afd5dc8cb37a97
.integrity: sha512-5h/vNMmutQ98tbB0sPDlAileZVca6A2McFgGa3+D56Dm8mSSCzTQL2DngPA6h09DlKDpSr7+6PdFw5Hoj0ZDSw==
.unpackedSize: 4.1 MB

dependencies:
@anthropic-ai/sdk: ^0.9.1 js-yaml: ^4.1.0 p-retry: 4
@langchain/community: ~0.0.41 jsonpointer: ^5.0.1 uuid: ^9.0.0
@langchain/core: ~0.1.44 langchainhub: ~0.0.8 yaml: ^2.2.1
@langchain/openai: ~0.0.19 langsmith: ~0.1.7 zod-to-json-schema: ^3.22.3
binary-extensions: ^2.2.0 ml-distance: ^4.0.0 zod: ^3.22.4
js-tiktoken: ^1.0.7 openapi-types: ^12.1.3

maintainers:

• nfcampos [email protected]
• jacoblee93 [email protected]
• davidduong [email protected]
• hwchase17 [email protected]
• basproul [email protected]

dist-tags:
latest: 0.1.30 next: 0.1.29-rc.0

published 5 days ago by jacoblee93 [email protected]

[dosubot[bot]]

Hey @ehiaig, great to see you tackling new challenges! Hope this one hasn't been too much of a head-scratcher. Let's dive in and untangle it together 🚀

Thank you for providing the detailed information. The error message you're encountering, "this.llm.pipe is not a function", suggests that the pipe method is being called on the llm object, which does not have such a method. This typically happens when the object does not match the expected interface or class instance that would normally provide the pipe method.

In your case, you're using the LlamaCpp class as the llm object. However, the LlamaCpp class might not have the pipe method, which is expected by the LLMChain class. This could be the reason why you're encountering the error.

To resolve this issue, you might need to ensure that the llm object you're using has the pipe method. If the LlamaCpp class does not have this method, you might need to use a different class that has it. You can check the LangChain JS documentation or the source code to find out which classes have the pipe method.

If you're sure that the LlamaCpp class should have the pipe method, then the issue might be with the version of LangChain JS you're using. You mentioned that you're using version 0.1.30. There might have been changes in the later versions that added the pipe method to the LlamaCpp class. In this case, updating to the latest version might resolve the issue.

If none of these solutions work, it would be helpful if you could provide more information about the LlamaCpp class and how you're using it. This would help in understanding the issue better and finding a suitable solution.

I hope this helps! If you have any more questions or need further clarification, feel free to ask.

Sources

• examples/src/chains/llm_chain_stream.ts
• langchain-core/src/language_models/llms.ts

About Dosu

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[ehiaig]

Can you point me to where/how to confirm that the llm object has the pipe method?
I have tested with using a mistral-7b base model as the modelPath value and it's same issue. Does this also mean that that mistral doesn't have the pipe method?