[debug] fix bug for benchmark

tools/benchmark.py

@@ -7,7 +7,8 @@ from mmcv.runner import load_checkpoint, wrap_fp16_model
 
 from mmdet3d.datasets import build_dataloader, build_dataset
 from mmdet3d.models import build_fusion_model
-
+from torchpack.utils.config import configs
+from mmdet3d.utils import recursive_eval
 
 def parse_args():
     parser = argparse.ArgumentParser(description="MMDet benchmark a model")
@@ -19,11 +20,11 @@ def parse_args():
     args = parser.parse_args()
     return args
 
-
 def main():
     args = parse_args()
 
-    cfg = Config.fromfile(args.config)
+    configs.load(args.config, recursive=True)
+    cfg = Config(recursive_eval(configs), filename=args.config)
     # set cudnn_benchmark
     if cfg.get("cudnn_benchmark", False):
         torch.backends.cudnn.benchmark = True