getting ValueError: chunk_overlap_ratio must be a float between 0. and 1. error

[vsneelesh]

Am getting an the following error
ValueError Traceback (most recent call last)
in <cell line: 1>()
----> 1 construct_index('/content/meditations.mb.txt')

1 frames
/usr/local/lib/python3.10/dist-packages/llama_index/indices/prompt_helper.py in init(self, context_window, num_output, chunk_overlap_ratio, chunk_size_limit, tokenizer, separator, max_input_size, embedding_limit, max_chunk_overlap)
70 self.chunk_overlap_ratio = chunk_overlap_ratio
71 if self.chunk_overlap_ratio > 1.0 or self.chunk_overlap_ratio < 0.0:
---> 72 raise ValueError("chunk_overlap_ratio must be a float between 0. and 1.")
73 self.chunk_size_limit = chunk_size_limit
74

ValueError: chunk_overlap_ratio must be a float between 0. and 1.

for the code

!pip install langchain
!pip install llama_index
from llama_index import SimpleDirectoryReader, GPTListIndex, GPTVectorStoreIndex ,LLMPredictor, PromptHelper
from langchain import OpenAI
import sys
import os
def ask_bot(question):
storage_context=StorageContext.from_defaults(persist_dir='input_index')
index=load_index_from_storage(storage_context)
query_engine=index.as_query_engine()
response=query_engine.query(question)
return response
!wget http://classics.mit.edu/Antoninus/meditations.mb.txt
construct_index('/content/meditations.mb.txt')

The above code should actually create an json file from which I can access the data but its not being created .

[xincxiong]

anyone already solved this issue?

[AICardiologist]

This is how I fixed this: set chunk_overlap_ratio= 0.1, and delete: max_chunk_overlap option. For example
...
max_input_size = 4096

num_output = 4096    

prompt_helper = PromptHelper(max_input_size, num_output, chunk_overlap_ratio= 0.1, chunk_size_limit=1024)

...
This fixed my problem

[wnxub]

With your solution, I got this error: PromptHelper.init() got multiple values for argument 'chunk_overlap_ratio'

[kenil-samyak]

same error I have. It's not working. Trying since 2 days

[wnxub]

You need to check out the updated llama-index's methods with its params