CoolFace
Modelpublic

IDMedicine/code-graph-trajeval-v2

sourceHugging Facemitupdated 5mo agoView on Hugging Face
0likes
README.md68 linesDownload Raw Back to root
1---2license: mit3---4 5# SWE-Bench Trajectory Eval Bundle (v2)6 7Airgapped-ready bundle for the trajectory-probe downstream eval.8 9## Changes from v110 11- **django dropped** (230 base_commits × 5 agents = 1139 trajectories removed).12  django's blobs were too slow to prefetch at scale; the other 11 repos13  give 1317 trajectories which is plenty.14- **Sparse-checkout blobs only**. For each (repo, base_commit), only the15  files any trajectory's patch touches are fetched. Dramatically smaller16  than the v1 "full-tree" approach; still enough for the graph encoder17  to see the patched files + their containing package.18 19## Contents (570 MB compressed)20 21- `specs.jsonl`: 1317 SWE-Bench Verified agent trajectories (down from22  2456). 46% resolved. 11 repos, 269 unique base_commits.23- `repos/`: 11 blobless clones with sparse-checkout already configured24  per base_commit. Working trees on-disk reflect the UNION of touched25  paths across all trajectories for each base_commit.26- `graphjepa/`: pipeline code (trajectory_pipeline with `extract_patch_paths`27  and `set_sparse_paths`, trajectory_realize, trajectory_probe, checks).28- `harvest.log`.29 30## Dataset31 32| Repo | Trajs | Resolved |33|---|---|---|34| sympy/sympy | 370 | 48% |35| sphinx-doc/sphinx | 207 | 32% |36| matplotlib | 167 | 41% |37| scikit-learn | 158 | 70% |38| astropy | 110 | 42% |39| pydata/xarray | 108 | 51% |40| pytest-dev/pytest | 95 | 49% |41| pylint-dev/pylint | 49 | 22% |42| psf/requests | 38 | 61% |43| mwaskom/seaborn | 10 | 0% |44| pallets/flask | 5 | 80% |45| **total** | **1317** | **46%** |46 47## Downstream workflow (on airgapped big)48 49```bash50huggingface-cli download IDMedicine/code-graph-trajeval-v2 traj_full_bundle.tar.gz \51    --local-dir outputs/traj_real52 53cd /opt/train/code-transformer/code-transformer54tar -xzf outputs/traj_real/traj_full_bundle.tar.gz55rsync -a traj_full/graphjepa/ graphjepa/56mkdir -p outputs/traj_real57cp traj_full/specs.jsonl outputs/traj_real/58mv traj_full/repos outputs/traj_real/repos59 60export BERT_MODEL_PATH=/opt/train/datasets/bert-base-uncased61GPUS=0,2,3,4,5,6,7 bash graphjepa/scripts/trajectory_full.sh62tail -f outputs/traj_real/logs/prebuild_shard*_gpu*.log63```64 65Because the clone already has the right blobs prefetched on laptop,66the airgapped big machine can checkout every base_commit without any67network access.68