prep: ChromBPNet CDF rebuild against chrombpnet_nobias (script + handoff)#67
Merged
lucapinello merged 2 commits intomainfrom Apr 30, 2026
Merged
prep: ChromBPNet CDF rebuild against chrombpnet_nobias (script + handoff)#67lucapinello merged 2 commits intomainfrom
lucapinello merged 2 commits intomainfrom
Conversation
Prep work for the deferred CDF rebuild against `chrombpnet_nobias` that the 0.3.0 audit at `audits/2026-04-28_chrombpnet_slim_mirror/` flagged as a follow-up. Background: chrombpnet_pertrack.npz on huggingface.co/datasets/lucapinello/chorus-backgrounds was built in 0.2.x against the bias-aware `chrombpnet` variant. After 0.3.0 flipped the default to `chrombpnet_nobias` (bias-corrected), user-facing percentile lookups go: `chrombpnet_nobias` predictions → `chrombpnet` empirical CDFs. The bias systematically shifts the mapping. Two changes in this prep commit, no compute yet: 1. scripts/build_backgrounds_chrombpnet.py: add `--model-type` argparse flag (default `chrombpnet_nobias`, matches 0.3+ chorus default). The build script previously called `oracle.load_pretrained_model( fold=args.fold, **spec)` without specifying model_type, so it inherited the oracle's default — which was `chrombpnet` in 0.2.x and is `chrombpnet_nobias` post-0.3. The new flag makes this explicit: re-running today produces the right CDF, and a future maintainer can pin `--model-type chrombpnet` for ablation against the legacy variant. Help text references the rebuild audit dir. 2. audits/2026-04-29_chrombpnet_cdf_rebuild/HANDOFF.md: agent handoff for running the rebuild on a CUDA box. Estimated wall-clock: ~3-5 h ATAC/DNase + ~10-20 h CHIP/BPNet on A100 (vs ~78 h on M3 Ultra Metal — too slow for the macOS dev machine). Includes the exact `--shard / --shard-of` commands for 2-GPU parallelism, the spot-check before upload, and the upload command (HfApi to lucapinello/chorus-backgrounds, dataset repo). The actual rebuild (compute + upload) gets done on the user's lab CUDA box in a separate run. This commit is just the prep so the script is unambiguous and the handoff is documented. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
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.
Prep PR for the ChromBPNet CDF rebuild flagged as a deferred follow-up at the 0.3.0 audit (
audits/2026-04-28_chrombpnet_slim_mirror/).Why
CDFs on
huggingface.co/datasets/lucapinello/chorus-backgroundswere built against the bias-awarechrombpnetvariant. 0.3.0 flipped the default tochrombpnet_nobias. User-facing percentile lookups now dochrombpnet_nobiaspredictions →chrombpnetempirical CDFs — the bias systematically shifts the mapping.What's in this PR
No compute yet. Two prep changes:
scripts/build_backgrounds_chrombpnet.pygets a--model-typeargparse flag (defaultchrombpnet_nobias). Previously the script inherited the oracle's default model_type, which waschrombpnetin 0.2.x. The flag makes 0.3+ runs unambiguous and lets a future maintainer pin--model-type chrombpnetfor ablation against the legacy CDF.audits/2026-04-29_chrombpnet_cdf_rebuild/HANDOFF.md— full agent handoff doc for running the rebuild on a CUDA box. Estimated wall-clock: ~3–5 h ATAC/DNase + ~10–20 h CHIP/BPNet on A100. Includes--shard / --shard-ofcommands for 2-GPU parallelism, spot-check before upload, and the HfApi upload command forlucapinello/chorus-backgrounds.Why split prep + compute
CDF rebuild needs ~13–25 h on CUDA, not feasible on the macOS dev machine (~78 h on Metal). Separating prep (mergeable now, lets maintainers verify the script + plan) from compute (runs on the lab box, output is a single NPZ + audit report).
Test plan
mamba run -n chorus python scripts/build_backgrounds_chrombpnet.py --helpshows the new--model-typeflag with the expected choices and default.effect_counts > 0, no NaN/Inf).🤖 Generated with Claude Code