· gpt4all langchain til generative-ai

GPT4All/LangChain: Model.__init__() got an unexpected keyword argument 'ggml_model' (type=type_error)

I’m starting to realise that things move insanely fast in the world of LLMs (Large Language Models) and you will run into issues because you aren’t using the latest version of libraries. I say this because I’ve been following Sami Maameri’s blog post which explains how to run an LLM on your own machine and ran into an error, which we’ll explore in this blog post.

Sami’s post is based around a library called GPT4All, but he also uses LangChain to glue things together. I first installed the following libraries:

pip install gpt4all langchain pyllamacpp

And then launched a Python REPL, into which I pasted the following code:

from langchain.llms import GPT4All
llm = GPT4All(model='/Users/markhneedham/Library/Application Support/nomic.ai/GPT4All/ggml-gpt4all-j-v1.3-groovy.bin')

The model path that I’ve used is the location where the GPT4All app downloaded the model, but you can choose to put it elsewhere. When that second line of code ran, I got the following exception:

Exception ignored in: <function Model.__del__ at 0x10ab2b7e0>
Traceback (most recent call last):
  File "/Users/markhneedham/projects/docs-bot/env/lib/python3.11/site-packages/pyllamacpp/model.py", line 402, in __del__
    if self._ctx:
       ^^^^^^^^^
AttributeError: 'Model' object has no attribute '_ctx'
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__
pydantic.error_wrappers.ValidationError: 1 validation error for GPT4All
__root__
  Model.__init__() got an unexpected keyword argument 'ggml_model' (type=type_error)

After a bit of exploration, I came across this GitHub thread, which suggests that the GPT4All code had changed and LangChain was passing in the wrong parameters. I also learnt that it’s been fixed for about a month, so my version of LangChain was clearly lagging. Let’s check my version:

pip freeze | grep langchain
Output
langchain==0.0.142

From searching the ttps://pypi.org/project/langchain/#history[release history of LangChain^], I discovered that I installed it on 17th April, since when there have been more than 60 releases! I updated by running pip install langchain -U and now I’m up to date (at least for now!):

Output
langchain==0.0.208
langchainplus-sdk==0.0.16

I’m also able to run the example from Sami’s blog post:

llm("""
You are a friendly chatbot assistant that responds in a conversational
manner to users questions. Keep the answers short, unless specifically
asked by the user to elaborate on something.

Question: Where is Paris?

Answer:""")
Output
' The capital city of France and Europe located at 47°23′N 8°54″W'
  • LinkedIn
  • Tumblr
  • Reddit
  • Google+
  • Pinterest
  • Pocket