feat(sleep): cross-model scaling results (+11.9 on nano) + dual-model experiment runner#89
Merged
Merged
Conversation
…ram ablation Update RESULTS.md with: - §2: GPT-5.4-nano target yields +11.9 pt (0.560→0.679) on SearchQA — 2× the GPT-5.5 gain, demonstrating bigger benefit where headroom exists - §4: Hyperparameter sweep confirms shipped defaults are optimal Co-Authored-By: Claude Opus 4 <noreply@anthropic.com>
Yif-Yang
force-pushed
the
sleep/nano-results
branch
from
407c485 to
ce354f6
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
dream_factor,rollouts,per_night,nights— every direction away from the shipped defaults hurts. Users get optimal performance out of the box.run_nightly.py: New--target-model/--optimizer-modelflags enabling split-model experiments viaDualBackend.Changes
docs/sleep/RESULTS.mdskillopt_sleep/experiments/run_nightly.pydocs/sleep/blog_runs/sweep_nano/*.jsonTest plan
run_nightly.pyimports cleanly:PYTHONPATH=. python -c "from skillopt_sleep.experiments.run_nightly import main"--target-model gpt-5.4-nano --optimizer-model gpt-5.5 --benchmarks searchqa --gate on --replay-mode cumulative --nights 1 --per-night 6 --test-limit 40🤖 Generated with Claude Code