Require native BF16 capability in structural smoke

training/run_fable_router_structural_smoke.py

@@ -142,14 +142,21 @@ def gpu_facts(minimum_vram_gib: float) -> dict[str, Any]:
     total_gib = properties.total_memory / 2**30
     if total_gib < minimum_vram_gib:
         raise RuntimeError(f"GPU has {total_gib:.2f} GiB, requires at least {minimum_vram_gib:.2f} GiB")
+    capability = torch.cuda.get_device_capability(0)
+    native_bf16 = capability[0] >= 8 and bool(torch.cuda.is_bf16_supported())
     return {
         "name": properties.name,
         "totalVramGiB": total_gib,
-        "capability": list(torch.cuda.get_device_capability(0)),
-        "bf16Supported": bool(torch.cuda.is_bf16_supported()),
+        "capability": list(capability),
+        "torchBf16Reported": bool(torch.cuda.is_bf16_supported()),
+        "nativeBf16Admitted": native_bf16,
     }
 
 
+def compute_dtype(gpu: dict[str, Any]) -> torch.dtype:
+    return torch.bfloat16 if gpu["nativeBf16Admitted"] else torch.float16
+
+
 def upload_evidence(output: Path, repo_id: str, token: str) -> dict[str, Any]:
     api = HfApi(token=token)
     api.create_repo(repo_id, repo_type="dataset", private=True, exist_ok=True)
@@ -245,7 +252,7 @@ def run(args: argparse.Namespace, result: dict[str, Any]) -> None:
     }
 
     device = torch.device("cuda:0")
-    dtype = torch.bfloat16 if torch.cuda.is_bf16_supported() else torch.float16
+    dtype = compute_dtype(result["gpu"])
     torch.cuda.reset_peak_memory_stats(device)
     load_started = time.perf_counter()
     model = AutoModelForCausalLM.from_pretrained(