Initial commit (clean history)HEADmain

1 files changed, 76 insertions, 0 deletions

diff --git a/scripts/pull_models.py b/scripts/pull_models.py
new file mode 100644
index 0000000..1e4abc3
--- /dev/null
+++ b/scripts/pull_models.py
@@ -0,0 +1,76 @@
+from __future__ import annotations
+
+import argparse
+from pathlib import Path
+import sys
+
+ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT / "src"))
+
+from huggingface_hub import snapshot_download  # type: ignore
+
+from personalization.config.settings import load_local_models_config
+
+
+def pull_one(repo_id: str, dest: Path, force: bool = False) -> None:
+    dest.mkdir(parents=True, exist_ok=True)
+    if any(dest.iterdir()) and not force:
+        print(f"[skip] {dest} already populated. Use --force to overwrite.")
+        return
+    print(f"[pull] {repo_id} -> {dest}")
+    snapshot_download(repo_id=repo_id, local_dir=str(dest), local_dir_use_symlinks=False)
+    print(f"[done] {repo_id}")
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--target",
+        choices=[
+            "llm",
+            "preference_extractor",
+            "embed_qwen3",
+            "embed_nemotron",
+            "embedders",
+            "reranker_qwen3",
+            "rerankers",
+            "all",
+        ],
+        default="all",
+    )
+    parser.add_argument("--force", action="store_true")
+    args = parser.parse_args()
+
+    cfg = load_local_models_config()
+    if args.target in ("llm", "all"):
+        pull_one(cfg.llm.hf_id, ROOT / cfg.llm.local_path, force=args.force)
+    if args.target in ("preference_extractor", "all"):
+        pull_one(
+            cfg.preference_extractor.hf_id,
+            ROOT / cfg.preference_extractor.local_path,
+            force=args.force,
+        )
+    if args.target in ("embed_qwen3", "embedders", "all") and cfg.embedding and cfg.embedding.qwen3:
+        pull_one(
+            cfg.embedding.qwen3.hf_id,
+            ROOT / cfg.embedding.qwen3.local_path,
+            force=args.force,
+        )
+    if args.target in ("embed_nemotron", "embedders", "all") and cfg.embedding and cfg.embedding.nemotron:
+        pull_one(
+            cfg.embedding.nemotron.hf_id,
+            ROOT / cfg.embedding.nemotron.local_path,
+            force=args.force,
+        )
+    if args.target in ("reranker_qwen3", "rerankers", "all") and cfg.reranker and cfg.reranker.qwen3_8b:
+        pull_one(
+            cfg.reranker.qwen3_8b.hf_id,
+            ROOT / cfg.reranker.qwen3_8b.local_path,
+            force=args.force,
+        )
+
+
+if __name__ == "__main__":
+    main()
+
+