You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I want to create a new binary which takes as input the mcts json path and a question and tries to answer the question based on the content of the mcts json.
The mcts json can be deserialized into a search tree minimal or something struct and then we can get the action nodes, look at the tool input and tool output and create a trajectory for what the agent did.
Once we have that we can pass it to a llm to run inference on it and answer any user question, I think this should all be part of the tool_use_agent.rs and the binary should talk to it via the session (so uses the same flow) there is a bit of unknown here so I would like to see your approach to it.
use cargo check to make sure the code works
The text was updated successfully, but these errors were encountered:
I want to create a new binary which takes as input the mcts json path and a question and tries to answer the question based on the content of the mcts json.
The mcts json can be deserialized into a search tree minimal or something struct and then we can get the action nodes, look at the tool input and tool output and create a trajectory for what the agent did.
Once we have that we can pass it to a llm to run inference on it and answer any user question, I think this should all be part of the tool_use_agent.rs and the binary should talk to it via the session (so uses the same flow) there is a bit of unknown here so I would like to see your approach to it.
use
cargo check
to make sure the code worksThe text was updated successfully, but these errors were encountered: