Deepak S V
Deepak S V
The following improvements are made to the Python REPL Tool: 1) Improved the sanitization of query (using regex), by removing python command (since gpt-3.5-turbo sometimes assumes python console as a...
Finally Solved the ```ValueError(f"Could not parse LLM output: `{llm_output}`")``` error, whenever llm (especially gpt-3.5-turbo) does not follow the format of MRKL Agent, while returning "Action:" & "Action Input:". **Note:** If...
This is a highly optimized update to the pull request https://github.com/hwchase17/langchain/pull/3269 Summary: 1) Added ability to MRKL agent to self solve the ValueError(f"Could not parse LLM output: `{llm_output}`") error, whenever...
Update to pull request https://github.com/hwchase17/langchain/pull/3215 Summary: 1) Improved the sanitization of query (using regex), by removing python command (since gpt-3.5-turbo sometimes assumes python console as a terminal, and runs python...
@vowelparrot Added regex expressions (superset to the previously present non regex expressions) to also solve the case where unnecessary characters are present after the triple backticks but before the {,...
Fixed the issue of blank Thoughts being printed in verbose when `handle_parsing_errors=True`, as below: Before Fix: ``` Observation: There are 38175 accounts available in the dataframe. Thought: Observation: Invalid or...
The code works only if the Global Frame id of the costmap is given as "map" and if any other global frame id is provided, there is a mismatch in...