-
Notifications
You must be signed in to change notification settings - Fork 173
Open
Description
I'm using LMStudio with Llama hoping to use custom langchain functions with it. I can see that the tools are loaded:
TOOLS: [
%LangChain.Function{
name: "validate_promql",
description: "Validate the syntax of a PromQL query. Return 'ok' or an error message.",
display_text: nil,
strict: false,
function: #Function<0.126923400/2 in Intel.PromQL.Model.validate_promql_tool/0>,
async: true,
parameters_schema: %{
type: "object",
required: ["query"],
properties: %{
query: %{type: "string", description: "The PromQL query to validate"}
}
},
parameters: []
}
]
and that the model attempts to call the function, but it's trying it via python?
SINGLE MESSAGE RESPONSE: %LangChain.Message{
content: "<|python_tag|>{\"type\": \"function\", \"function\": {\"name\": \"validate_promql\", \"parameters\": {\"query\": \"rate(last(accepted_metrics) by item)\"}}",
processed_content: nil,
index: 0,
status: :complete,
role: :assistant,
name: nil,
tool_calls: [],
tool_results: nil,
metadata: nil
}
Is there something custom i need to do in LMStudio or langchain to make this work?
Metadata
Metadata
Assignees
Labels
No labels