No Python file under artifactbench/ or scripts/ contains a maintainer path (/home/unohee, /media/unohee): FSTModel takes

README.md

@@ -229,7 +229,7 @@ python scripts/check_docs.py
 
 Tests marked `slow` need network access, model weights, or inference, and are
 not run in CI; run them with `python -m pytest -m slow`. `check_docs.py` checks
-that every relative link in `README.md`, `NOTICE.md`, `docs/*.md` and
+that every relative link in `README.md`, `NOTICE.md`, `RESULTS_8WAY.md`, `docs/*.md` and
 `paper/README.md` resolves, and that every `--flag` in a documented
 `python -m ...` or `python scripts/....py` command is accepted by that program.
 Use `python scripts/check_docs.py --skip-flags` to check links only.

RESULTS_8WAY.md

@@ -15,8 +15,23 @@ v1.1 purged test partition (n = 2,224; see `v1.1/RESULTS_v1.1.md`).
 | DeepFense EAT+Nes2Net (FakeMusicCaps) | — | [`DeepFense/FakeMusicCaps_EAT_Nes2Net_NoAug_Seed42`](https://huggingface.co/DeepFense/FakeMusicCaps_EAT_Nes2Net_NoAug_Seed42) (Kheir et al., arXiv:2604.08450) |
 
 Adapters: `artifactbench/models/{ast_60s,deepfense,deezer_ismir,fst,spectttra_variants}.py`.
-Run with `--model ast_60s --model deepfense --model deezer_ismir --model fst --model spectttra_beta5s`
-alongside the existing `--model artifactnet --model spectttra --model clam`.
+CLAM and FST import their upstream code, so both need a local clone (with the checkpoints obtained as
+each upstream README describes); FST reads its clone from `--fst-repo` or `ARTIFACTBENCH_FST_REPO`.
+
+```bash
+git clone https://github.com/StarkVision-AI/MoM-CLAM ./MoM-CLAM
+git clone https://github.com/Mippia/FST-AI-music-detection ./FST-AI-music-detection
+
+python -m artifactbench.bench \
+    --model artifactnet --model spectttra --model clam \
+    --model ast_60s --model deepfense --model deezer_ismir --model fst --model spectttra_beta5s \
+    --clam-repo ./MoM-CLAM \
+    --clam-ckpt ./MoM-CLAM/model_wts/best_model_triplet_loss_margin_0.2.pth \
+    --fst-repo ./FST-AI-music-detection \
+    --manifest v1.1/artifactbench_v1.1_test_purged.json \
+    --split test \
+    --output results/eight_way
+```
 
 ## Results (n = 2,104: 1,388 AI / 716 real, τ = 0.5)
 

artifactbench/bench.py

@@ -9,8 +9,8 @@
     # Four-model comparison (CLAM requires --clam-repo + --clam-ckpt)
     python -m artifactbench.bench \
         --model artifactnet --model spectttra --model clam \
-        --clam-repo ~/dev/MoM-CLAM \
-        --clam-ckpt ~/dev/MoM-CLAM/model_wts/best_model_triplet_loss_margin_0.2.pth \
+        --clam-repo ./MoM-CLAM \
+        --clam-ckpt ./MoM-CLAM/model_wts/best_model_triplet_loss_margin_0.2.pth \
         --manifest manifest.json --output out/
 
     # Smoke test with tiny sample
@@ -39,6 +39,7 @@
     BenchModel,
     CLAMModel,
     DeezerISMIRModel,
+    FSTModel,
     SpecTTTraModel,
 )
 from .report.markdown import comparison_report, single_model_report
@@ -64,6 +65,12 @@ def build_model(model_name: str, args) -> "BenchModel":
             onnx_path=args.deezer_onnx,
             hf_revision=args.deezer_revision,
         )
+    if model_name == "fst":
+        return FSTModel(
+            fst_repo=args.fst_repo,
+            stage1_ckpt=args.fst_stage1_ckpt,
+            stage2_ckpt=args.fst_stage2_ckpt,
+        )
     cls = MODEL_REGISTRY[model_name]
     return cls()
 
@@ -169,7 +176,7 @@ def run_single_model(model, entries, by_source, args):
     }
 
 
-def main():
+def build_parser() -> argparse.ArgumentParser:
     parser = argparse.ArgumentParser(description="ArtifactBench v2-compatible runner")
     parser.add_argument("--model", action="append", required=True,
                         choices=list(MODEL_REGISTRY.keys()),
@@ -208,8 +215,18 @@ def main():
                         help="path to cloned MoM-CLAM repository")
     parser.add_argument("--clam-ckpt", default=None,
                         help="path to CLAM checkpoint (.pth)")
-
-    args = parser.parse_args()
+    parser.add_argument("--fst-repo", default=None,
+                        help="path to cloned FST-AI-music-detection repository "
+                             "(default: $ARTIFACTBENCH_FST_REPO)")
+    parser.add_argument("--fst-stage1-ckpt", default=None,
+                        help="FST stage-1 checkpoint (default: <fst-repo>/checkpoints/backbone_stage1.ckpt)")
+    parser.add_argument("--fst-stage2-ckpt", default=None,
+                        help="FST stage-2 checkpoint (default: <fst-repo>/checkpoints/classifier_stage2.ckpt)")
+    return parser
+
+
+def main(argv=None):
+    args = build_parser().parse_args(argv)
 
     np.random.seed(args.seed)
     torch.manual_seed(args.seed)

artifactbench/models/fst.py

@@ -15,6 +15,7 @@
 Paper:     arXiv:2601.13647
 """
 import gc
+import os
 import sys
 from contextlib import contextmanager
 from pathlib import Path
@@ -63,7 +64,9 @@ def pinned_mert_loaders(module, snapshot):
             setattr(module, n, originals[n])
 
 
-FST_REPO_DEFAULT = "/home/unohee/dev/FST-AI-music-detection"
+# Location of a clone of https://github.com/Mippia/FST-AI-music-detection. There is no default: the
+# adapter imports that repository's code, so it must be told where the clone lives.
+FST_REPO_ENV = "ARTIFACTBENCH_FST_REPO"
 STAGE1_CKPT_DEFAULT = "checkpoints/backbone_stage1.ckpt"
 STAGE2_CKPT_DEFAULT = "checkpoints/classifier_stage2.ckpt"
 
@@ -77,6 +80,17 @@ def pinned_mert_loaders(module, snapshot):
 MAX_DURATION_SEC = 90
 
 
+def resolve_fst_repo(fst_repo: str | Path | None = None) -> Path:
+    """The FST clone: the argument, else ``$ARTIFACTBENCH_FST_REPO``; an error names both when neither is set."""
+    location = fst_repo if fst_repo else os.environ.get(FST_REPO_ENV)
+    if not location:
+        raise ValueError(
+            "FST needs a clone of https://github.com/Mippia/FST-AI-music-detection: pass --fst-repo "
+            f"(FSTModel(fst_repo=...)) or set {FST_REPO_ENV}"
+        )
+    return Path(location).expanduser()
+
+
 class FSTModel(BenchModel):
     """FST (Mippia 2025) — two-stage MERT + Fusion Segment Transformer."""
 
@@ -87,12 +101,12 @@ class FSTModel(BenchModel):
     paper_ref = "arXiv:2601.13647"
 
     def __init__(self,
-                 fst_repo: str = FST_REPO_DEFAULT,
+                 fst_repo: str | None = None,
                  stage1_ckpt: str | None = None,
                  stage2_ckpt: str | None = None,
                  mert_path: str | None = None,
                  beat_checkpoint: str | None = None):
-        self.fst_repo = Path(fst_repo)
+        self.fst_repo = resolve_fst_repo(fst_repo)
         self.stage1_ckpt = stage1_ckpt or str(self.fst_repo / STAGE1_CKPT_DEFAULT)
         self.stage2_ckpt = stage2_ckpt or str(self.fst_repo / STAGE2_CKPT_DEFAULT)
         # Pinned local MERT snapshot (v12.model_assets); None keeps the upstream Hub lookup.

artifactbench/rerun_current.py

@@ -3,6 +3,7 @@
 from collections import Counter, defaultdict
 import hashlib
 import json
+import os
 from pathlib import Path
 import random
 import subprocess
@@ -29,18 +30,39 @@ def write_json(path, value):
     path.write_text(json.dumps(value, ensure_ascii=False, indent=2) + "\n")
 
 
-def main():
+ARTIFACTNET_REPO_ENV = "ARTIFACTBENCH_ARTIFACTNET_REPO"
+FMA_FULL_ROOT_ENV = "ARTIFACTBENCH_FMA_FULL_ROOT"
+
+
+def parse_args(argv=None):
+    """Parse the CLI; the private ArtifactNet tree and the local fma_full root have no built-in location."""
     ap = argparse.ArgumentParser(description=__doc__)
     ap.add_argument("--output", required=True, type=Path)
     ap.add_argument("--device", default="cuda")
     ap.add_argument("--limit", type=int, default=0, help="0 = full corrected test set")
     ap.add_argument("--codec-pairs", type=int, default=50)
     ap.add_argument("--recover-fma-from", type=Path,
                     help="reuse verified track results and retry HTML-disguised FMA files using local fma_full")
-    args = ap.parse_args()
+    ap.add_argument("--artifactnet-repo", type=Path, default=os.environ.get(ARTIFACTNET_REPO_ENV),
+                    help=f"private ArtifactNet source tree (default: ${ARTIFACTNET_REPO_ENV})")
+    ap.add_argument("--fma-full-root", type=Path, default=os.environ.get(FMA_FULL_ROOT_ENV),
+                    help=f"local fma_full root used by --recover-fma-from (default: ${FMA_FULL_ROOT_ENV})")
+    args = ap.parse_args(argv)
+    if args.artifactnet_repo is None:
+        ap.error(f"the ArtifactNet source tree is required: pass --artifactnet-repo or set {ARTIFACTNET_REPO_ENV}")
+    if args.recover_fma_from is not None and args.fma_full_root is None:
+        ap.error(f"--recover-fma-from needs the local fma_full root: pass --fma-full-root or set {FMA_FULL_ROOT_ENV}")
+    args.artifactnet_repo = Path(args.artifactnet_repo).expanduser()
+    if args.fma_full_root is not None:
+        args.fma_full_root = Path(args.fma_full_root).expanduser()
+    return args
+
+
+def main(argv=None):
+    args = parse_args(argv)
     out = args.output
     out.mkdir(parents=True, exist_ok=True)
-    rd = Path('/home/unohee/dev/ArtifactNet')
+    rd = args.artifactnet_repo
     sys.path.insert(0, str(rd))
     import src.pipeline.infer as infer
     from src.pipeline import model_config as config
@@ -103,7 +125,7 @@ def main():
                 assert p.stem.isdigit() and len(p.stem) == 6
                 with p.open('rb') as fh:
                     assert fh.read(30).lstrip().lower().startswith(b'<!doctype html')
-                original = Path('/media/unohee/Archive/dataset/fma_full') / p.stem[:3] / p.name
+                original = args.fma_full_root / p.stem[:3] / p.name
                 assert original.is_file()
                 recovered[r['path']] = str(original)
         parent_files = [parent / 'provenance.json', parent / 'rows.jsonl', snapshot]

artifactbench/v12/freeze.py

@@ -8,6 +8,7 @@
 from urllib.parse import urlsplit, urlunsplit
 
 from .common import digest, manifest_rows, rank, write_once
+from .model_assets import asset_location
 from .duplicates import Groups, entry_key, compare_waveforms
 from .inventory import EXPOSURE_INPUTS
 from .review_duplicates import full_audio
@@ -54,8 +55,10 @@ def recording_groups(entries, audit, review):
     return labels, representatives
 
 
-def local_exposure_audit(entries, inventory):
-    root = Path('/home/unohee/dev/ArtifactNet')
+def local_exposure_audit(entries, inventory, artifactnet_repo=None):
+    # The exposure inputs are training manifests inside the private ArtifactNet tree; its location comes
+    # from --artifactnet-repo or ARTIFACTBENCH_ARTIFACTNET_REPO (hash checks below still pin the content).
+    root = asset_location('ARTIFACTBENCH_ARTIFACTNET_REPO', artifactnet_repo)
     sys.path.insert(0, str(root))
     from src.data.dataset_guard import stem_key, same_seed_different_generator
     findings, input_hashes = [], {}
@@ -113,6 +116,9 @@ def main():
     ap = argparse.ArgumentParser(description=__doc__)
     for argument in ('selection', 'validation', 'duplicates', 'review', 'inventory', 'variants', 'variant-validation', 'output'):
         ap.add_argument('--'+argument, required=True, type=Path)
+    ap.add_argument('--artifactnet-repo', type=Path, default=None,
+                    help='private ArtifactNet tree holding the exposure manifests '
+                         '(default: $ARTIFACTBENCH_ARTIFACTNET_REPO)')
     args = ap.parse_args()
     entries = json.loads(args.selection.read_text())['bench']
     variations = json.loads(args.variants.read_text())['bench']
@@ -136,7 +142,7 @@ def main():
     if review['pairs_reviewed'] != audit['perceptual_candidates']:
         raise ValueError('All fingerprint candidates need explicit review')
     inventory = json.loads(args.inventory.read_text())
-    exposure = local_exposure_audit(entries, inventory)
+    exposure = local_exposure_audit(entries, inventory, args.artifactnet_repo)
     if exposure['candidate_collisions']:
         raise ValueError('Unresolved exposure ID candidates require investigation')
     groups, representatives = recording_groups(entries, audit, review)

artifactbench/v12/model_assets.py

@@ -7,9 +7,16 @@
 from .common import digest
 
 
-def asset_location(variable, default):
-    """경로만 기계별로 바꾸며, 실제 구현과 가중치의 해시 검사는 유지한다."""
-    return Path(os.environ.get(variable, default)).expanduser().resolve()
+def asset_location(variable, default=None):
+    """Machine-specific location of an external asset; hash checks on its code and weights still apply.
+
+    The environment variable wins; ``default`` is used only when it is given. Without either the call
+    raises, naming the variable, rather than looking in a directory that exists on one machine only.
+    """
+    location = os.environ.get(variable) or default
+    if not location:
+        raise ValueError(f'Set {variable} to the location of this asset (no default is assumed)')
+    return Path(location).expanduser().resolve()
 
 
 def source_identity(path):
@@ -42,7 +49,7 @@ def local(path):
         return str(path)
     if name == 'artifactnet':
         import sys
-        root = asset_location('ARTIFACTBENCH_ARTIFACTNET_REPO', '/home/unohee/dev/ArtifactNet')
+        root = asset_location('ARTIFACTBENCH_ARTIFACTNET_REPO')
         sys.path.insert(0, str(root))
         from src.pipeline import model_config as cfg
         import src.pipeline.infer as infer
@@ -67,13 +74,13 @@ def local(path):
     elif name == 'deezer_ismir':
         model = DeezerISMIRModel(onnx_path=str(Path(snapshot('lofcz/ai-music-detector')) / 'ai_music_detector.onnx'))
     elif name == 'clam':
-        root = asset_location('ARTIFACTBENCH_CLAM_REPO', '/home/unohee/dev/MoM-CLAM')
+        root = asset_location('ARTIFACTBENCH_CLAM_REPO')
         evidence['source_repos']['MoM-CLAM'] = source_identity(root)
         model = CLAMModel(clam_repo=str(root), clam_ckpt=local(root / 'model_wts/best_model_triplet_loss_margin_0.2.pth'),
                     mert_repo=snapshot('m-a-p/MERT-v1-95M'),
                     wav2vec_repo=snapshot('m3hrdadfi/wav2vec2-base-100k-gtzan-music-genres'))
     elif name == 'fst':
-        root = asset_location('ARTIFACTBENCH_FST_REPO', '/home/unohee/dev/FST-AI-music-detection')
+        root = asset_location('ARTIFACTBENCH_FST_REPO')
         evidence['source_repos']['FST'] = source_identity(root)
         model = FSTModel(fst_repo=str(root), stage1_ckpt=local(root / 'checkpoints/backbone_stage1.ckpt'),
                     stage2_ckpt=local(root / 'checkpoints/classifier_stage2.ckpt'),

docs/REPRODUCING_v1.2.md

@@ -78,14 +78,16 @@ absolute paths is not portable by itself; acquire the same revisions into the
 new machine's cache and retain its new location manifest.
 
 Three source checkouts and the BeatThis checkpoint are separate assets. Their
-locations can be set with these environment variables before model creation:
+locations are set with these environment variables before model creation. The
+three source checkouts have no built-in location: when a variable is unset, model
+creation stops with an error naming it.
 
 | Variable | Required content |
 |---|---|
 | `ARTIFACTBENCH_ARTIFACTNET_REPO` | Matching ArtifactNet checkout, `src`, and checkpoint files selected by its pinned `model_config` |
 | `ARTIFACTBENCH_CLAM_REPO` | MoM-CLAM checkout and `model_wts/best_model_triplet_loss_margin_0.2.pth` |
 | `ARTIFACTBENCH_FST_REPO` | FST checkout and `checkpoints/{backbone_stage1,classifier_stage2}.ckpt` |
-| `ARTIFACTBENCH_BEAT_CHECKPOINT` | BeatThis `final0.ckpt` |
+| `ARTIFACTBENCH_BEAT_CHECKPOINT` | BeatThis `final0.ckpt` (default `.tools/beat-this/final0.ckpt`) |
 
 Overrides change paths only: the factory still hashes all used files and source.
 The [external-asset acquisition and verification workflow](EXTERNAL_ASSETS_v1.2.md)

scripts/check_docs.py

@@ -6,7 +6,7 @@
 # Test Status: tests/test_check_docs.py
 """Documentation checker.
 
-Link check: relative ``[text](target)`` links in README.md, NOTICE.md, docs/*.md and
+Link check: relative ``[text](target)`` links in README.md, NOTICE.md, RESULTS_8WAY.md, docs/*.md and
 paper/README.md must point at an existing file or directory (anchors and URLs skipped).
 
 Flag check: in fenced code blocks, commands of the form ``python -m pkg.mod ...`` or
@@ -22,7 +22,7 @@
 from pathlib import Path
 
 ROOT = Path(__file__).resolve().parents[1]
-DOC_GLOBS = ("README.md", "NOTICE.md", "docs/*.md", "paper/README.md")
+DOC_GLOBS = ("README.md", "NOTICE.md", "RESULTS_8WAY.md", "docs/*.md", "paper/README.md")
 
 LINK_RE = re.compile(r"(?<!!)\[[^\]\n]*\]\(([^)\s]+)(?:\s+\"[^\"]*\")?\)")
 FENCE_RE = re.compile(r"^(\s*)(```|~~~)")