fix: Load grounding model without dtype parameter to avoid errors

Changed from using dtype parameter in from_pretrained to manually
converting dtype after loading. This avoids potential JSON serialization
and compatibility issues with the dtype parameter.

Pattern: load() → .to(dtype=bfloat16) → .eval() → .to(device)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

medrax/tools/grounding.py

@@ -67,14 +67,13 @@ class XRayPhraseGroundingTool(BaseTool):
         super().__init__()
         self.device = torch.device(device) if device else "cuda"
 
-        # Load model following transformers 4.56.0 API
-        # Use 'dtype' instead of deprecated 'torch_dtype'
+        # Load model - convert to bfloat16 after loading to avoid dtype parameter issues
+        # Load with default dtype, then manually convert to bfloat16
         self.model = AutoModelForCausalLM.from_pretrained(
             model_path,
             cache_dir=cache_dir,
             trust_remote_code=True,
-            dtype=torch.bfloat16,
-        ).eval().to(self.device)
+        ).to(dtype=torch.bfloat16).eval().to(self.device)
 
         self.processor = AutoProcessor.from_pretrained(
             model_path,