What's the pricing model for dynamic worker loader
is it counted as independent worker request with subrequest and cpu costs or is it charged only as part of compute on the worker calling it ?
I'm asking this because if we use it as interpreter for LLM generated code it'll be huge cost if it costs a seperate worker request per LLM call
I'm asking this because if we use it as interpreter for LLM generated code it'll be huge cost if it costs a seperate worker request per LLM call
