evo2_megatron: load Savanna HF checkpoints with weights_only=False

torch 2.6 changed the default of `weights_only` to True. The Savanna
checkpoint pickle includes numpy globals (`numpy.core.multiarray._reconstruct`),
which the safer loader rejects. The converter then exits 0 with no output
written and the error gets buried in stderr — silent failure.

The Savanna repos under arcinstitute/* are trusted sources, so load with
weights_only=False.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: polinabinder1

bionemo-recipes/recipes/evo2_megatron/src/bionemo/evo2/utils/checkpoint/savanna_to_mbridge.py

@@ -135,7 +135,7 @@ def load_savanna_state_dict(path: Path) -> dict[str, torch.Tensor]:
     Returns:
         Flat state dict with keys like 'sequential.{i}.xxx'.
     """
-    raw = torch.load(str(path), map_location="cpu", weights_only=True, mmap=True)
+    raw = torch.load(str(path), map_location="cpu", weights_only=False, mmap=True)
     if "module" in raw:
         raw = raw["module"]