diff --git a/03.image_classification/README.cn.md b/03.image_classification/README.cn.md
index d28269f8d93670f4cf8a4e1dafd6a32f713b9051..9b82a2635593e052fff1bbbf38dc4147f9b7be6b 100644
--- a/03.image_classification/README.cn.md
+++ b/03.image_classification/README.cn.md
@@ -538,11 +538,7 @@ with fluid.scope_guard(inference_scope):
     [inference_program, feed_target_names,
      fetch_targets] = fluid.io.load_inference_model(params_dirname, exe)
 
-        # The input's dimension of conv should be 4-D or 5-D.
-        # Use inference_transpiler to speedup
-    inference_transpiler_program = inference_program.clone()
-    t = fluid.transpiler.InferenceTranspiler()
-    t.transpile(inference_transpiler_program, place)
+
 
         # Construct feed as a dictionary of {feed_target_name: feed_target_data}
         # and results will contain a list of data corresponding to fetch_targets.
@@ -550,14 +546,6 @@ with fluid.scope_guard(inference_scope):
                       feed={feed_target_names[0]: img},
                       fetch_list=fetch_targets)
 
-    transpiler_results = exe.run(inference_transpiler_program,
-                                 feed={feed_target_names[0]: img},
-                                 fetch_list=fetch_targets)
-
-    assert len(results[0]) == len(transpiler_results[0])
-    for i in range(len(results[0])):
-        numpy.testing.assert_almost_equal(
-            results[0][i], transpiler_results[0][i], decimal=5)
 
     # infer label
     label_list = [
diff --git a/03.image_classification/README.md b/03.image_classification/README.md
index 661ff53fb22abeacc725281315a23806d94c3d50..7526cb132e4b8b6b5a338e7715feaafadb8d02a5 100644
--- a/03.image_classification/README.md
+++ b/03.image_classification/README.md
@@ -541,11 +541,7 @@ with fluid.scope_guard(inference_scope):
     [inference_program, feed_target_names,
      fetch_targets] = fluid.io.load_inference_model(params_dirname, exe)
 
-        # The input's dimension of conv should be 4-D or 5-D.
-        # Use inference_transpiler to speedup
-    inference_transpiler_program = inference_program.clone()
-    t = fluid.transpiler.InferenceTranspiler()
-    t.transpile(inference_transpiler_program, place)
+
 
         # Construct feed as a dictionary of {feed_target_name: feed_target_data}
         # and results will contain a list of data corresponding to fetch_targets.
@@ -553,14 +549,7 @@ with fluid.scope_guard(inference_scope):
                       feed={feed_target_names[0]: img},
                       fetch_list=fetch_targets)
 
-    transpiler_results = exe.run(inference_transpiler_program,
-                                 feed={feed_target_names[0]: img},
-                                 fetch_list=fetch_targets)
 
-    assert len(results[0]) == len(transpiler_results[0])
-    for i in range(len(results[0])):
-        numpy.testing.assert_almost_equal(
-            results[0][i], transpiler_results[0][i], decimal=5)
 
     # infer label
     label_list = [
diff --git a/03.image_classification/index.cn.html b/03.image_classification/index.cn.html
index 376054e06c099d546af78eb641f61c75c7a3fa0f..3a66d5c98968b14c256e77e699a386f07cc0b806 100644
--- a/03.image_classification/index.cn.html
+++ b/03.image_classification/index.cn.html
@@ -580,11 +580,7 @@ with fluid.scope_guard(inference_scope):
     [inference_program, feed_target_names,
      fetch_targets] = fluid.io.load_inference_model(params_dirname, exe)
 
-        # The input's dimension of conv should be 4-D or 5-D.
-        # Use inference_transpiler to speedup
-    inference_transpiler_program = inference_program.clone()
-    t = fluid.transpiler.InferenceTranspiler()
-    t.transpile(inference_transpiler_program, place)
+
 
         # Construct feed as a dictionary of {feed_target_name: feed_target_data}
         # and results will contain a list of data corresponding to fetch_targets.
@@ -592,14 +588,6 @@ with fluid.scope_guard(inference_scope):
                       feed={feed_target_names[0]: img},
                       fetch_list=fetch_targets)
 
-    transpiler_results = exe.run(inference_transpiler_program,
-                                 feed={feed_target_names[0]: img},
-                                 fetch_list=fetch_targets)
-
-    assert len(results[0]) == len(transpiler_results[0])
-    for i in range(len(results[0])):
-        numpy.testing.assert_almost_equal(
-            results[0][i], transpiler_results[0][i], decimal=5)
 
     # infer label
     label_list = [
diff --git a/03.image_classification/index.html b/03.image_classification/index.html
index fcf8c744a402236b9a623096dea05094ec0f248f..7eaca4f528a41895139ce3c9f80b48dc59223e98 100644
--- a/03.image_classification/index.html
+++ b/03.image_classification/index.html
@@ -583,11 +583,7 @@ with fluid.scope_guard(inference_scope):
     [inference_program, feed_target_names,
      fetch_targets] = fluid.io.load_inference_model(params_dirname, exe)
 
-        # The input's dimension of conv should be 4-D or 5-D.
-        # Use inference_transpiler to speedup
-    inference_transpiler_program = inference_program.clone()
-    t = fluid.transpiler.InferenceTranspiler()
-    t.transpile(inference_transpiler_program, place)
+
 
         # Construct feed as a dictionary of {feed_target_name: feed_target_data}
         # and results will contain a list of data corresponding to fetch_targets.
@@ -595,14 +591,7 @@ with fluid.scope_guard(inference_scope):
                       feed={feed_target_names[0]: img},
                       fetch_list=fetch_targets)
 
-    transpiler_results = exe.run(inference_transpiler_program,
-                                 feed={feed_target_names[0]: img},
-                                 fetch_list=fetch_targets)
 
-    assert len(results[0]) == len(transpiler_results[0])
-    for i in range(len(results[0])):
-        numpy.testing.assert_almost_equal(
-            results[0][i], transpiler_results[0][i], decimal=5)
 
     # infer label
     label_list = [
diff --git a/03.image_classification/train.py b/03.image_classification/train.py
index 78bf1ce45d9e176690f73cc70c8c172df2dbf141..47df1809b055e0b90c126816c49521e4166cc3f3 100644
--- a/03.image_classification/train.py
+++ b/03.image_classification/train.py
@@ -191,12 +191,6 @@ def infer(use_cuda, params_dirname=None):
         [inference_program, feed_target_names,
          fetch_targets] = fluid.io.load_inference_model(params_dirname, exe)
 
-        # The input's dimension of conv should be 4-D or 5-D.
-        # Use inference_transpiler to speedup
-        inference_transpiler_program = inference_program.clone()
-        t = fluid.transpiler.InferenceTranspiler()
-        t.transpile(inference_transpiler_program, place)
-
         # Construct feed as a dictionary of {feed_target_name: feed_target_data}
         # and results will contain a list of data corresponding to fetch_targets.
         results = exe.run(
@@ -204,16 +198,6 @@ def infer(use_cuda, params_dirname=None):
             feed={feed_target_names[0]: img},
             fetch_list=fetch_targets)
 
-        transpiler_results = exe.run(
-            inference_transpiler_program,
-            feed={feed_target_names[0]: img},
-            fetch_list=fetch_targets)
-
-        assert len(results[0]) == len(transpiler_results[0])
-        for i in range(len(results[0])):
-            numpy.testing.assert_almost_equal(
-                results[0][i], transpiler_results[0][i], decimal=5)
-
         # infer label
         label_list = [
             "airplane", "automobile", "bird", "cat", "deer", "dog", "frog",