process_future: Process prompts in parallel chunks with error handling and...

In hellmer: Batch Processing for Chat Models

Process prompts in parallel chunks with error handling and state management

Description

Process prompts in parallel chunks with error handling and state management

Usage

process_future(
  chat_obj,
  prompts,
  type_spec,
  judgements,
  state_path,
  workers,
  chunk_size,
  plan,
  max_chunk_attempts,
  max_retries,
  initial_delay,
  max_delay,
  backoff_factor,
  beep,
  progress,
  echo,
  ...
)

Arguments

chat_obj	Chat model object for API calls
prompts	Vector or list of prompts to process
type_spec	Optional type specification for structured data extraction
judgements	Number of judgements for structured data extraction resulting in refined data
state_path	Path to save intermediate state
workers	Number of parallel workers
chunk_size	Number of prompts to process in parallel at a time
plan	Parallel backend
max_chunk_attempts	Maximum retries per failed chunk
max_retries	Maximum retries per prompt
initial_delay	Initial delay before first retry
max_delay	Maximum delay between retries
backoff_factor	Delay multiplier after each retry
beep	Play sound on completion/error
progress	Whether to show progress bars

Value

Batch results object

hellmer documentation built on June 8, 2025, 11:09 a.m.