refactor(compressors): reorganize SAM utilities into proposal module

mini-nav/compressors/model_loader.py

@@ -0,0 +1,59 @@
+"""Model loading utilities for DINO, SAM2 and HashCompressor."""
+
+from typing import TYPE_CHECKING, Any
+
+
+import torch
+from transformers import AutoImageProcessor, AutoModel, pipeline, MaskGenerationPipeline
+
+from utils import get_device
+
+if TYPE_CHECKING:
+    from compressors.hash_compressor import HashCompressor
+
+
+def load_sam_model(
+    model_name: str = "facebook/sam2.1-hiera-large",
+) -> MaskGenerationPipeline:
+    device = get_device()
+
+    return pipeline(
+        task="mask-generation",
+        model=model_name,
+        device=device,
+    )
+
+
+def load_dino_model(
+    model_name: str = "facebook/dinov2-large",
+) -> tuple[Any, Any]:
+    device = get_device()
+
+    processor = AutoImageProcessor.from_pretrained(model_name)
+    dino = AutoModel.from_pretrained(model_name).to(device)
+    dino.eval()
+
+    return processor, dino
+
+
+def get_dino_dim(model_name: str) -> int:
+    if "large" in model_name.lower():
+        return 1024
+    return 768
+
+
+def load_hash_compressor(
+    input_dim: int = 1024,
+    hash_bits: int = 512,
+    compressor_path: str | None = None,
+) -> "HashCompressor":
+    from compressors.hash_compressor import HashCompressor
+
+    device = get_device()
+    compressor = HashCompressor(input_dim=input_dim, hash_bits=hash_bits).to(device)
+
+    if compressor_path is not None:
+        compressor.load_state_dict(torch.load(compressor_path, map_location=device))
+        print(f"[OK] Loaded HashCompressor from {compressor_path}")
+
+    return compressor