Ajinkya Indulkar

@AlgoHunt can you please add a requirements.txt file for easy use of this repository?

Can we publish this repository on GitHub Pages for ease of accessibility? The process to do so seems quite simple to execute and would be a great addition to this...

## Description Closes #333. As a trial user, I am unable to index my data source due to rate limits imposed by OpenAI on trial accounts. As I'm not planning...

bug: `return_intermediate_steps=True` doesn't work for `"zero-shot-react-description"` agent

3

## Description I was going through the Intermediate Steps documentation: https://langchain.readthedocs.io/en/latest/modules/agents/examples/intermediate_steps.html. My use case is to use this feature for conversation agent. Sample code: ``` from langchain.agents import initialize_agent, load_tools...

Fix hallucination problem with `SQLDatabaseChain`

16

## Description Based on my experiments with `SQLDatabaseChain` class, I have noticed that the LLM tends to hallucinate `Answer` if `SQLResult` field results as `[]`. As I primarily used OpenAI...

## Description ### Motivation Using configuration files are great for fast prototyping but given the current state of the framework, it is limited to pre-defined agents, tools and other components....

[Feat] `langchain.register`: dynamically register custom tools and agents

6

## Description Closes #1360. ### ✨ Highlights Introducing `langchain.register` module: collection of decorator functions to dynamically register new tools and agents. This module will allow users to register custom agents...

Update optional dependencies

1

## Description This PR updates the list of dependencies and removes the ones which are optional to run `gpt_index` with minimal features. Related #524. ### Changelog - 🔧 updated setup...

Reduce package size with `options.extras_require` feature of setuptools

5

At the moment, the package is quite heavy due to certain dependencies like `transformers`. This makes it difficult to deploy on serverless solutions like AWS Lambda. Is it possible to...

I was going through [Vectorstore Agent](https://langchain.readthedocs.io/en/latest/modules/agents/agent_toolkits/vectorstore.html?highlight=vectorstore%20agent#vectorstore-agent) tutorial and I am facing issues with the `VectorStoreQAWithSourcesTool`. Looking closely at the code https://github.com/hwchase17/langchain/blob/3c2468452284ee37b8a88a20b864255fa4385b65/langchain/chains/qa_with_sources/base.py#L119-L120 it appears, the parsing rule might be too strict...