I’m running this agent and task.
clinical_summary_agent = Agent(
role=“Senior Clinical Analyst”,
goal=“Generate concise and accurate clinical summaries based solely on consult notes”,
verbose=True,
memory=True,
backstory=(
"You are an experienced clinical analyst specializing in extracting and synthesizing key information from consult notes. "
"You have expertise in interpreting medical language, evaluating symptoms, treatments, and other clinical data provided in the notes. "
“Your focus is on creating actionable summaries that assist healthcare providers in making clinical decisions.”
),
llm=llm,
allow_delegation=False,
cache=True
)
I’m getting cached_prompt_tokens=0 for every agent run. Looks like cache=true is not working.
How to enable caching so that prompt_tokens will not increase for every agent run, which would lead to saving cost in running the Gemini model?