answer --seed, and runs/spread.sh: a score comes with its spread
Every score so far was one run of a small visitor. `answer --seed n` answers the same tasks with other shuffles and sampling, and runs/spread.sh prints a run's score per seed, so a change can be told from the visitor's mood. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Fable 5.1
parent
ec2f07da77
commit
cc50249d7f
+1
-1
@@ -34,7 +34,7 @@ pub fn run(argv: &[String]) -> anyhow::Result<()> {
|
||||
let mut summary = serde_json::Map::new();
|
||||
for (mode, tasks, skim) in [("full", &tasks_full, false), ("skim", &tasks_skim, true)] {
|
||||
eprintln!("score: the visitor answers {mode}");
|
||||
let answers = crate::answer_blind(&cmd, &std::fs::read_to_string(tasks)?, Some(&site))?;
|
||||
let answers = crate::answer_blind(&cmd, &std::fs::read_to_string(tasks)?, Some(&site), 0)?;
|
||||
let answers_path = out.join(format!("answers-{mode}.jsonl"));
|
||||
std::fs::write(&answers_path, &answers)?;
|
||||
let (tasks_arg, answers_arg) = (s(tasks), s(&answers_path));
|
||||
|
||||
Reference in New Issue
Block a user