diff --git a/ppcls/arch/__init__.py b/ppcls/arch/__init__.py
index d43bc488059681881ea0f28333668d430f0a2f6c..f2c2e412c757c269d1475734e7f378312f4e7775 100644
--- a/ppcls/arch/__init__.py
+++ b/ppcls/arch/__init__.py
@@ -26,7 +26,7 @@ from .utils import *
 from ppcls.arch.backbone.base.theseus_layer import TheseusLayer
 from ppcls.utils import logger
 from ppcls.utils.save_load import load_dygraph_pretrain
-from ppcls.engine.slim import prune_model, quantize_model
+from ppcls.arch.slim import prune_model, quantize_model
 
 
 __all__ = ["build_model", "RecModel", "DistillationModel"]
diff --git a/ppcls/engine/slim/__init__.py b/ppcls/arch/slim/__init__.py
similarity index 86%
rename from ppcls/engine/slim/__init__.py
rename to ppcls/arch/slim/__init__.py
index de3d857b213dcb6734916ba79c2d08bd64d76fa7..e2842472244a719619ffaf45d3759166fac63893 100644
--- a/ppcls/engine/slim/__init__.py
+++ b/ppcls/arch/slim/__init__.py
@@ -12,5 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ppcls.engine.slim.prune import prune_model
-from ppcls.engine.slim.quant import quantize_model
+from ppcls.arch.slim.prune import prune_model
diff --git a/ppcls/engine/slim/prune.py b/ppcls/arch/slim/prune.py
similarity index 100%
rename from ppcls/engine/slim/prune.py
rename to ppcls/arch/slim/prune.py
diff --git a/ppcls/engine/slim/quant.py b/ppcls/arch/slim/quant.py
similarity index 100%
rename from ppcls/engine/slim/quant.py
rename to ppcls/arch/slim/quant.py
diff --git a/ppcls/engine/engine.py b/ppcls/engine/engine.py
index 22b3e058972b75d271ca3ccd8e2f07155a0754f7..c53594889ad2468dca4f823c99599f9f2cd34317 100644
--- a/ppcls/engine/engine.py
+++ b/ppcls/engine/engine.py
@@ -186,8 +186,6 @@ class Engine(object):
 
         # build model
         self.model = build_model(self.config)
-        self.quanted = self.config.get("Slim", {}).get("quant", False)
-        self.pruned = self.config.get("Slim", {}).get("prune", False)
         # set @to_static for benchmark, skip this by default.
         apply_to_static(self.config, self.model)
 
@@ -368,7 +366,7 @@ class Engine(object):
         model.eval()
         save_path = os.path.join(self.config["Global"]["save_inference_dir"],
                                  "inference")
-        if self.quanted:
+        if model.quanter:
             model.quanter.save_quantized_model(
                 model.base_model,
                 save_path,