Skip to contents

Collect and resume live TypeSafe Jev decisions

Source: R/typesafe_live.R
submit_typesafe_pairs_live.Rd

Collect multiple binary decisions with the same scientific contract as typesafe_compare_pair_live(). Use submit_llm_pairs() for provider-neutral dispatch. These are live requests, not managed Batch jobs.

Usage

submit_typesafe_pairs_live(
  pairs,
  model = "jev-1.13.0",
  trait_name,
  trait_description,
  prompt_template = NULL,
  api_key = NULL,
  verbose = TRUE,
  status_every = 1,
  progress = TRUE,
  include_raw = FALSE,
  save_path = NULL,
  parallel = FALSE,
  workers = 1,
  execution_control = list(),
  ...
)

Arguments

pairs

Tibble or data frame with at least columns ID1, text1, ID2, text2. Typically created by make_pairs(), sample_pairs(), and randomize_pair_order().

model

Requested Jev identifier; defaults to versioned "jev-1.13.0".

trait_name

Trait name to pass through to the backend-specific comparison function (for example "Overall Quality").

trait_description

Full-text trait description passed to the backend.

prompt_template

Must be omitted or NULL; custom decision instructions are supplied with decision_template in ....

api_key

Optional API key for the selected backend. If NULL, the backend-specific helper will use its own default environment variable. For "ollama", this argument is ignored (no API key is required for local inference).

verbose

Logical; if TRUE, prints status, timing, and result summaries (for backends that support it).

status_every

Integer; print status and timing for every status_every-th pair. Defaults to 1 (every pair). Errors are always printed.

progress

Logical; if TRUE, shows a textual progress bar for backends that support it.

include_raw

Logical; if TRUE, each row of the returned tibble will include a raw_response list-column with the parsed JSON body from the backend (for backends that support this).

save_path

Character string; optional file path (e.g., "output.csv") to save results incrementally. If the file exists, the function reads it to identify and skip pairs that have already been processed (resume mode). Supported by all backends.

parallel

Logical; if TRUE, enables parallel processing using future.apply for generative backends, requiring future and future.apply. Decision concurrency uses optional promises/later.

workers

Integer; the number of parallel workers (threads) to use if parallel = TRUE. Defaults to 1.

execution_control

Named list of decision-only scheduling, retry, timeout and lock controls; defaults to list(). See decision_bulk for all defaults, persistent budgets, probability storage and recovery guarantees.

...

Only decision_template is supported. See typesafe_compare_pair_live() for the complete scientific/provider contract.

Value

The results, failed_pairs, and failed_attempts list documented in submit_llm_pairs().

Details

See decision_bulk for execution controls, persistent retry budgets, locking, probability storage, and interruption/recovery guarantees.

See also

submit_llm_pairs(), submit_openai_pairs_live(), decision_bulk

Other live backends: anthropic_compare_pair_live(), check_llm_api_keys(), gemini_compare_pair_live(), llm_compare_pair(), ollama_compare_pair_live(), openai_compare_pair_live(), submit_anthropic_pairs_live(), submit_gemini_pairs_live(), submit_llm_pairs(), submit_ollama_pairs_live(), submit_openai_pairs_live(), submit_together_pairs_live(), submit_vertex_pairs_live(), together_compare_pair_live(), typesafe_compare_pair_live(), vertex_compare_pair_live()