Mike Wang

[simple] add duckduckgo to load_tools.py

2

- duckduckgo was introduced in https://github.com/hwchase17/langchain/pull/3206/, but it’s not enabled in load_tools.py yet. - this PR simply enables it in load_tools.py with a straightforward agent call (as described in get_started...

I found most other search classes are using -Run postfix except DuckDuckGo. It might be good to keep them consistent. (screenshot below) - Updated all places (using big search and...

- added unittest for schema.py covering utility functions and token counting. - fixed a nit. based on huggingface doc, the tokenizer model is gpt-2. [link](https://huggingface.co/transformers/v4.8.2/_modules/transformers/models/gpt2/tokenization_gpt2_fast.html) - make lint && make...

It turns out arxiv is not hooked up with load_tools yet. As titled, bridged them and covered with integration tests. Added a specific test to cover all wrapper params.

[structure][agent class] Move Overlapping Methods to a Shared Base Class

4

- `BaseSingleActionAgent` and `BaseMultiActionAgent` are pretty much the same. In the[ original PR](https://github.com/hwchase17/langchain/pull/2362) where MultiActionAgent was introduced, @hwchase17 shared the worry that MultiActionAgent is so under-explored that it would potentially...

- ActionAgent has a property called, `allowed_tools`, which is declared as `List`. It stores all provided tools which is available to use during agent action. - This collection shouldn’t allow...

[agent] Add Spark Agent

1

- added support for spark through pyspark library. - added jupyter notebook as example.

[RFC] Question of Tooling Consistency Check Between Agent<>AgentExecutor

3

This is a follow up of https://github.com/hwchase17/langchain/pull/3840 Question: Why do we allow `None` `allowed_tools` in `Agent` when there are tools provided in `AgentExecutor`? AgentExecutor is able to fetch the `allowed_tools`...

[test] Add integration_test for PandasAgent

4

- confirm creation - confirm functionality with a simple dimension check. The test now is calling OpenAI API directly, but learning from @vowelparrot that we’re caching the requests, so that...

[agent] Add Support to Spark Connect

1

Spark Connect allows computation delegation to clusters though DataFrame API. It has different data type. Trying to re-use as much as possible here. (spark connect: https://spark.apache.org/docs/latest/spark-connect-overview.html) - example notebook added