diff --git a/python/jittor/__init__.py b/python/jittor/__init__.py
index 4322f99e..9510cf14 100644
--- a/python/jittor/__init__.py
+++ b/python/jittor/__init__.py
@@ -9,7 +9,7 @@
 # file 'LICENSE.txt', which is part of this source code package.
 # ***************************************************************
 
-__version__ = '1.2.2.70'
+__version__ = '1.2.2.71'
 from . import lock
 with lock.lock_scope():
     ori_int = int
diff --git a/python/jittor/nn.py b/python/jittor/nn.py
index 9d58d62a..e38c5b82 100644
--- a/python/jittor/nn.py
+++ b/python/jittor/nn.py
@@ -348,6 +348,8 @@ class Identity(Module):
     def execute(self, input):
         return input
 
+def identity(input): return input
+
 class Dropout(Module):
     def __init__(self, p=0.5, is_train=False):
         assert p >= 0 and p <= 1, "dropout probability has to be between 0 and 1, but got {}".format(p)
@@ -1322,6 +1324,14 @@ class Upsample(Module):
                 int(x.shape[3]*self.scale_factor[1])),
             mode=self.mode)
 
+class UpsamplingBilinear2d(Upsample):
+    def __init__(self, scale_factor=None):
+        Upsample.__init__(self, scale_factor, 'bilinear')
+
+class UpsamplingNearest2d(Upsample):
+    def __init__(self, scale_factor=None):
+        Upsample.__init__(self, scale_factor, 'nearest')
+
 class Sequential(Module):
     def __init__(self, *args):
         self.layers = collections.OrderedDict()
diff --git a/python/jittor/pool.py b/python/jittor/pool.py
index 1714a835..42e3922c 100644
--- a/python/jittor/pool.py
+++ b/python/jittor/pool.py
@@ -215,6 +215,36 @@ class AdaptiveAvgPool2d(Module):
         ])
         return xx.reduce("mean", [4,5])
 
+class AdaptiveMaxPool2d(Module):
+    def __init__(self, output_size):
+        self.output_size = output_size
+
+    def execute(self, x):
+        if isinstance(self.output_size, int):
+            oh = self.output_size
+            ow = self.output_size
+        elif isinstance(self.output_size, tuple) or isinstance(self.output_size, list):
+            oh = x.shape[2] if self.output_size[0] is None else self.output_size[0]
+            ow = x.shape[3] if self.output_size[1] is None else self.output_size[1]
+        else:
+            raise TypeError(f"AdaptiveMaxPool2d only support int, tuple or list input. Not support {type(self.output_size)} yet.")
+        if oh == 1 and ow == 1:
+            return x.reduce("maximum", [2,3], keepdims=True)
+        N,C,H,W = x.shape
+        self.sh = math.floor(H / oh)
+        self.sw = math.floor(W / ow)
+        self.ksh = H - (oh - 1) * self.sh
+        self.ksw = W - (ow - 1) * self.sw
+        h = (H-self.ksh)//self.sh+1
+        w = (W-self.ksw)//self.sw+1
+        xx = x.reindex([N,C,h,w,self.ksh,self.ksw], [
+            "i0", # Nid
+            "i1", # Cid
+            f"i2*{self.sh}+i4", # Hid
+            f"i3*{self.sw}+i5", # Wid
+        ])
+        return xx.reduce("maximum", [4,5])
+
 def pool(x, kernel_size, op, padding=0, stride=None):
     return Pool(kernel_size, stride, padding, op=op)(x)
 
diff --git a/python/jittor/utils/pytorch_converter.py b/python/jittor/utils/pytorch_converter.py
index f6257453..b91dfc30 100644
--- a/python/jittor/utils/pytorch_converter.py
+++ b/python/jittor/utils/pytorch_converter.py
@@ -217,18 +217,6 @@ pjmap = {
         'links': {},
         'extras': {},
     },
-    'kaiming_normal_': {
-        'pytorch': {
-            'args': "tensor, a=0, mode='fan_in', nonlinearity='leaky_relu'", 
-        },
-        'jittor': {
-            'module': 'init',
-            'name': 'relu_invariant_gauss_',
-            'args': 'var, mode="fan_in"'
-        },
-        'links': {'tensor': 'var'},
-        'extras': {},
-    },
     'constant_': {
         'pytorch': {
             'args': "tensor, val", 
@@ -382,10 +370,10 @@ unsupport_ops = [
     'TransformerDecoder', 'TransformerEncoderLayer', 'TransformerDecoderLayer', 'Identity', 'Bilinear', 
     'Dropout3d', 'AlphaDropout', 'EmbeddingBag', 'CosineSimilarity', 'PairwiseDistance', 'CTCLoss', 'NLLLoss', 'PoissonNLLLoss', 'KLDivLoss', 'BCEWithLogitsLoss', 
     'MarginRankingLoss', 'HingeEmbeddingLoss', 'MultiLabelMarginLoss', 'SmoothL1Loss', 'SoftMarginLoss', 
-    'MultiLabelSoftMarginLoss', 'CosineEmbeddingLoss', 'MultiMarginLoss', 'TripletMarginLoss', 'UpsamplingNearest2d', 
-    'UpsamplingBilinear2d', 'DataParallel', 'DistributedDataParallel', 'clip_grad_norm_', 'clip_grad_value_', 
-    'parameters_to_vector', 'vector_to_parameters', 'BasePruningMethod', 'PruningContainer', 'Identity', 
-    'RandomUnstructured', 'L1Unstructured', 'RandomStructured', 'LnStructured', 'CustomFromMask', 'identity', 
+    'MultiLabelSoftMarginLoss', 'CosineEmbeddingLoss', 'MultiMarginLoss', 'TripletMarginLoss', # 'DataParallel', 'DistributedDataParallel', 
+    'clip_grad_norm_', 'clip_grad_value_', 
+    'parameters_to_vector', 'vector_to_parameters', 'BasePruningMethod', 'PruningContainer', 
+    'RandomUnstructured', 'L1Unstructured', 'RandomStructured', 'LnStructured', 'CustomFromMask', 
     'random_unstructured', 'l1_unstructured', 'random_structured', 'ln_structured', 'global_unstructured', 
     'custom_from_mask', 'remove', 'is_pruned', 'weight_norm', 'remove_weight_norm', 'spectral_norm', 
     'remove_spectral_norm', 'PackedSequence', 'pack_padded_sequence', 'pad_packed_sequence', 'pad_sequence', 'pack_sequence'