Fix DLRM OOM issue (#299)

Signed-off-by: changwa1 <chang1.wang@intel.com>

examples/pytorch/recommendation/dlrm/quantization/ptq/eager/dlrm_s_pytorch_tune.py

@@ -842,7 +842,7 @@ def loss_fn_wrap(Z, T, use_gpu, device):
             args.print_freq = ld_nbatches
             args.test_freq = 0
 
-        del ld_model
+        del(ld_model)
 
         print(
             "Saved at: epoch = {:d}/{:d}, batch = {:d}/{:d}, ntbatch = {:d}".format(

examples/pytorch/recommendation/dlrm/quantization/ptq/fx/dlrm_s_pytorch_tune.py

@@ -847,7 +847,7 @@ def loss_fn_wrap(Z, T, use_gpu, device):
             args.print_freq = ld_nbatches
             args.test_freq = 0
 
-        del ld_model
+        del(ld_model)
 
         print(
             "Saved at: epoch = {:d}/{:d}, batch = {:d}/{:d}, ntbatch = {:d}".format(

examples/pytorch/recommendation/dlrm/quantization/ptq/ipex/dlrm_s_pytorch.py

@@ -821,7 +821,7 @@ def run():
             )
         )
         print("Testing state: accuracy = {:3.3f} %".format(ld_acc_test * 100))
-        del ld_model
+        del(ld_model)
 
     ext_dist.barrier()
     print("time/loss/accuracy (if enabled):")

neural_compressor/adaptor/pytorch.py

@@ -2584,7 +2584,7 @@ def _get_quantizable_ops_recursively(self, model, prefix, quantizable_ops):
                 self.q_dataloader.batch(batch_size)
                 logger.info('Recovery `calibration.dataloader.batchsize` {} according \
                             to config.yaml'                                           .format(batch_size))
-            del init_model
+            del(init_model)
         with open(self.ipex_config_path, 'r') as f:
             self.cfgs = json.load(f)
             if self.version.release < Version("1.12.0").release:
@@ -2776,6 +2776,7 @@ def quantize(self, tune_cfg, model, dataloader, q_func=None):
         from torch.quantization.quantize_fx import prepare_fx, convert_fx, prepare_qat_fx
         try:
             q_model = copy.deepcopy(model)
+            q_model.fp32_model = model.fp32_model
         except Exception as e:  # pragma: no cover
             logger.warning("Fail to deep copy the model due to {}, inplace is used now.".format(
                 repr(e)))
@@ -2983,6 +2984,13 @@ def _pre_hook_for_qat(self, dataloader=None):
         # so set it to None.
         example_inputs = None
 
+        # For export API, deepcopy fp32_model
+        try:
+            self.model.fp32_model = copy.deepcopy(self.model.fp32_model)
+        except Exception as e:  # pragma: no cover
+            logger.warning("Fail to deep copy the model due to {}, inplace is used now.".format(
+                repr(e)))
+
         if self.sub_module_list is None:
             if self.version.release >= Version("1.13.0").release:  # pragma: no cover
                 # pylint: disable=E1123

neural_compressor/model/torch_model.py

@@ -45,12 +45,7 @@ def __init__(self, model, **kwargs):
         self.q_config = None
         self._workspace_path = ''
         self.is_quantized = False
-        try:
-            self.fp32_model = copy.deepcopy(model)
-        except Exception as e:  # pragma: no cover
-            logger.warning("Fail to deep copy the model due to {}, inplace is used now.".format(
-                repr(e)))
-            self.fp32_model = model
+        self.fp32_model = model
         self.kwargs = kwargs if kwargs else None
 
     def __repr__(self):
@@ -93,6 +88,16 @@ def model(self, model):
         """ Setter to model """
         self._model = model
 
+    @property
+    def fp32_model(self):
+        """ Getter to model """
+        return self._fp32_model
+
+    @fp32_model.setter
+    def fp32_model(self, fp32_model):
+        """ Setter to model """
+        self._fp32_model = fp32_model
+
     def register_forward_pre_hook(self):
         self.handles.append(
                 self._model.register_forward_pre_hook(self.generate_forward_pre_hook()))