Call torch.cuda.empty_cache to release device memory (#114663)

Summary:
X-link: pytorch/pytorch#114663
Approved by: https://github.com/eellison

Reviewed By: osalpekar

Differential Revision: D52062377

Pulled By: desertfire

fbshipit-source-id: a4d6fdc530046d8d7c2ab259398f6deb1daa3ded

userbenchmark/dynamo/dynamobench/common.py

@@ -1942,8 +1942,6 @@ def validate_model(self, model, example_inputs):
             raise RuntimeError("Eager run failed") from e
 
     def maybe_cast(self, model, example_inputs):
-        model = self.deepcopy_model(model)
-        example_inputs = clone_inputs(example_inputs)
         model, example_inputs = self.cast_based_on_args(model, example_inputs)
         return model, example_inputs
 
@@ -2145,7 +2143,8 @@ def record_status(accuracy_status, dynamo_start_stats):
                 self.args.cosine = True
                 fp64_outputs = None
             finally:
-                del model_fp64
+                del model_fp64, inputs_fp64
+                torch.cuda.empty_cache()
 
             tolerance, cos_similarity = self.get_tolerance_and_cosine_flag(
                 self.args.training, current_device, name
@@ -2174,6 +2173,7 @@ def record_status(accuracy_status, dynamo_start_stats):
                 return record_status(accuracy_status, dynamo_start_stats=start_stats)
             finally:
                 del model_copy
+                torch.cuda.empty_cache()
 
             # Rerun native pytorch
             reset_rng_state()
@@ -2194,6 +2194,7 @@ def record_status(accuracy_status, dynamo_start_stats):
                 return record_status(accuracy_status, dynamo_start_stats=start_stats)
             finally:
                 del model_copy
+                torch.cuda.empty_cache()
 
             # Two eager runs should have exactly same result
             is_same = True