逐一修改，给映射列表按照torch api排序

hd9568 · hd9568 · commit 8a07b306f551 · 2025-10-28T07:46:47.000Z
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.Tensor.to_sparse.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.Tensor.to_sparse.md
@@ -3,7 +3,7 @@
 ### [torch.Tensor.to_sparse](https://pytorch.org/docs/stable/tensors.html#torch.Tensor.to_sparse)
 
 ```python
-torch.Tensor.to_sparse(sparseDims)
+torch.Tensor.to_sparse(sparse_dim)
 ```
 
 ### [paddle.Tensor.to_sparse_coo](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/Tensor/to_sparse_coo_cn.html#paddle/Tensor/to_sparse_coo_cn#cn-api-paddle-Tensor-to_sparse_coo)
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.Function.forward.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.Function.forward.md
@@ -31,7 +31,7 @@ class cus_tanh(torch.autograd.Function):
         return grad
 
 # Paddle 写法
-class cus_tanh(paddle.autograd.Function):
+class cus_tanh(paddle.autograd.PyLayer):
     @staticmethod
     def forward(ctx, x, func=paddle.square):
         ctx.func = func
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.mark_non_differentiable.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.mark_non_differentiable.md
@@ -32,7 +32,7 @@ class cus_func(torch.autograd.Function):
         return grad_x
 
 # Paddle 写法
-class cus_func(paddle.autograd.Function):
+class cus_func(paddle.autograd.PyLayer):
     @staticmethod
     def forward(ctx, x):
         a = x + x
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.md
@@ -31,7 +31,7 @@ class MyFunction(torch.autograd.Function):
         return grad_input
 
 # Paddle 写法
-class MyFunction(paddle.autograd.Function):
+class MyFunction(paddle.autograd.PyLayer):
     @staticmethod
     def forward(ctx, x):
         ctx.x = x
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.save_for_backward.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.save_for_backward.md
@@ -31,7 +31,7 @@ class cus_tanh(torch.autograd.Function):
         return grad
 
 # Paddle 写法
-class cus_tanh(paddle.autograd.Function):
+class cus_tanh(paddle.autograd.PyLayer:
     @staticmethod
     def forward(ctx, x):
         y = paddle.nn.functional.tanh(x=x)
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.saved_tensors.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.saved_tensors.md
@@ -32,7 +32,7 @@ class cus_tanh(torch.autograd.Function):
         return grad
 
 # Paddle 写法
-class cus_tanh(paddle.autograd.Function):
+class cus_tanh(paddle.autograd.PyLayer):
     @staticmethod
     def forward(ctx, x):
         y = paddle.nn.functional.tanh(x=x)
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.set_materialize_grads.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.autograd.function.FunctionCtx.set_materialize_grads.md
@@ -30,7 +30,7 @@ class cus_tanh(torch.autograd.Function):
         return grad
 
 # Paddle 写法
-class cus_tanh(paddle.autograd.Function):
+class cus_tanh(paddle.autograd.PyLayer):
     @staticmethod
     def forward(ctx, x):
         ctx.set_materialize_grads(False)
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.backends.cudnn.version.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.backends.cudnn.version.md
@@ -18,14 +18,9 @@ paddle.device.get_cudnn_version()
 
 ```python
 # PyTorch 写法
-import torch
-
 result = torch.backends.cudnn.version() == torch.backends.cudnn.version()
 
-
 # Paddle 写法
-import paddle
-
 result = paddle.device.get_cudnn_version() == paddle.device.get_cudnn_version()
 
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.cuda.is_bf16_supported.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.cuda.is_bf16_supported.md
@@ -3,13 +3,13 @@
 ### [torch.cuda.is_bf16_supported](https://pytorch.org/docs/stable/generated/torch.cuda.is_bf16_supported.html#torch.cuda.is_bf16_supported)
 
 ```python
-torch.cuda.is_bf16_supported(including_emulation)
+torch.cuda.is_bf16_supported(including_emulation=True)
 ```
 
 ### [paddle.amp.is_bfloat16_supported](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/amp/is_bfloat16_supported_cn.html#paddle/amp/is_bfloat16_supported_cn#cn-api-paddle-amp-is_bfloat16_supported)
 
 ```python
-paddle.amp.is_bfloat16_supported(device)
+paddle.amp.is_bfloat16_supported(device=None)
 ```
 
 两者功能一致，但调用方式不一致，具体如下：
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.cuda.stream.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.cuda.stream.md
@@ -18,8 +18,11 @@ paddle.device.stream_guard(stream)
 
 ```python
 # PyTorch 写法
-context = torch.cuda.stream(stream=s)
+with torch.cuda.stream(stream=s):
+    result = data1 + data2
 
 # Paddle 写法
-context = paddle.device.stream_guard(stream=s)
+with paddle.device.stream_guard(stream=s):
+    result = data1 + data2
+
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.distributed.ReduceOp.PRODUCT.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.distributed.ReduceOp.PRODUCT.md
@@ -3,13 +3,13 @@
 ### [torch.distributed.ReduceOp.PRODUCT](https://pytorch.org/docs/stable/distributed.html#torch.distributed.ReduceOp.PRODUCT)
 
 ```python
-torch.distributed.ReduceOp.PRODUCT()
+torch.distributed.ReduceOp.PRODUCT
 ```
 
 ### [paddle.distributed.ReduceOp.PROD](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/distributed/ReduceOp/PROD_cn.html#paddle/distributed/ReduceOp/PROD_cn#cn-api-paddle-distributed-ReduceOp-PROD)
 
 ```python
-paddle.distributed.ReduceOp.PROD()
+paddle.distributed.ReduceOp.PROD
 ```
 
 两者功能一致，但调用方式不一致，具体如下：
@@ -18,8 +18,12 @@ paddle.distributed.ReduceOp.PROD()
 
 ```python
 # PyTorch 写法
-dist.reduce_scatter(
-    data1, [data1, data2], op=dist.ReduceOp.PRODUCT, group=None, async_op=False
+torch.distributed.reduce_scatter(
+    data1,
+    [data1, data2],
+    op=torch.distributed.ReduceOp.PRODUCT,
+    group=None,
+    async_op=False
 )
 
 # Paddle 写法
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.distributed.ReduceOp.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.distributed.ReduceOp.md
@@ -3,13 +3,13 @@
 ### [torch.distributed.ReduceOp](https://pytorch.org/docs/stable/generated/torch.distributed.ReduceOp.html#torch.distributed.ReduceOp)
 
 ```python
-torch.distributed.ReduceOp(pybind11_builtins.pybind11_object)
+torch.distributed.ReduceOp
 ```
 
 ### [paddle.distributed.ReduceOp](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/distributed/ReduceOp_cn.html#paddle/distributed/ReduceOp_cn#cn-api-paddle-distributed-ReduceOp)
 
 ```python
-paddle.distributed.ReduceOp(*args, **kwargs)
+paddle.distributed.ReduceOp
 ```
 
 两者功能一致，但调用方式不一致，具体如下：
@@ -18,8 +18,8 @@ paddle.distributed.ReduceOp(*args, **kwargs)
 
 ```python
 # PyTorch 写法
-dist.init_process_group(backend="nccl")
-print(dist.ReduceOp)
+torch.distributed.init_process_group(backend="nccl")
+print(torch.distributed.ReduceOp)
 
 # Paddle 写法
 paddle.distributed.init_parallel_env()
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.is_inference.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.is_inference.md
@@ -3,7 +3,7 @@
 ### [torch.is_inference](https://pytorch.org/docs/stable/generated/torch.is_inference.html)
 
 ```python
-torch.is_inference()
+torch.is_inference(input)
 ```
 
 ### [paddle.Tensor.stop_gradient](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/Tensor/stop_gradient_cn.html#paddle/Tensor/stop_gradient_cn#cn-api-paddle-Tensor-stop_gradient)
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.apply.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.apply.md
@@ -18,12 +18,16 @@ paddle.nn.Layer.apply(fn)
 
 ```python
 # PyTorch 写法
+def init_weights(m):
+    pass
 net = torch.nn.Sequential(
     torch.nn.Linear(2, 2, bias=False), torch.nn.Linear(2, 2, bias=False)
 )
 net.apply(init_weights)
 
 # Paddle 写法
+def init_weights(m):
+    pass
 net = paddle.nn.Sequential(
     paddle.nn.Linear(in_features=2, out_features=2, bias_attr=False),
     paddle.nn.Linear(in_features=2, out_features=2, bias_attr=False),
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.children.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.children.md
@@ -18,14 +18,12 @@ paddle.nn.Layer.children()
 
 ```python
 # PyTorch 写法
-net = torch.nn.Sequential(l, l)
-result = torch.Tensor([0, 0])
-for i, j in enumerate(net.children()):
-    result = j(result)
+model = torch.nn.Sequential(net1, net2)
+for layer in model.children():
+    print(layer)
 
 # Paddle 写法
-net = paddle.nn.Sequential(l, l)
-result = paddle.Tensor([0, 0])
-for i, j in enumerate(net.children()):
-    result = j(result)
+model = paddle.nn.Sequential(net1, net2)
+for layer in model.children():
+    print(layer)
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.eval.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.eval.md
@@ -18,18 +18,10 @@ paddle.nn.Layer.eval()
 
 ```python
 # PyTorch 写法
-class TheModelClass(torch.nn.Module):
-    def forward(self, x):
-        return x
-
-model = TheModelClass()
+model = torch.nn.Linear(10, 10)
 model.eval()
 
 # Paddle 写法
-class TheModelClass(paddle.nn.Layer):
-    def forward(self, x):
-        return x
-
-model = TheModelClass()
+model = paddle.nn.Linear(10, 10)
 model.eval()
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.md
@@ -22,44 +22,24 @@ class MLP(torch.nn.Module):
     def __init__(self, input_size, hidden_size, output_size):
         super(MLP, self).__init__()
         self.fc1 = torch.nn.Linear(input_size, hidden_size)
-        self.relu = torch.nn.ReLU()
-        self.fc2 = torch.nn.Linear(hidden_size, output_size)
         with torch.no_grad():
             self.fc1.weight.fill_(1.0)
             self.fc1.bias.fill_(0.1)
-            self.fc2.weight.fill_(1.0)
-            self.fc2.bias.fill_(0.1)
 
     def forward(self, x):
         x = self.fc1(x)
-        x = self.relu(x)
-        x = self.fc2(x)
         return x
 
-model = MLP(784, 256, 10)
-inputs = torch.ones([64, 784])
-result = model(inputs)
-
 # Paddle 写法
 class MLP(paddle.nn.Layer):
     def __init__(self, input_size, hidden_size, output_size):
         super(MLP, self).__init__()
         self.fc1 = paddle.nn.Linear(in_features=input_size, out_features=hidden_size)
-        self.relu = paddle.nn.ReLU()
-        self.fc2 = paddle.nn.Linear(in_features=hidden_size, out_features=output_size)
         with paddle.no_grad():
             self.fc1.weight.fill_(1.0)
             self.fc1.bias.fill_(0.1)
-            self.fc2.weight.fill_(1.0)
-            self.fc2.bias.fill_(0.1)
 
     def forward(self, x):
         x = self.fc1(x)
-        x = self.relu(x)
-        x = self.fc2(x)
         return x
-
-model = MLP(784, 256, 10)
-inputs = paddle.ones([64, 784])
-result = model(inputs)
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.modules.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.modules.md
@@ -18,14 +18,12 @@ paddle.nn.Layer.sublayers(include_self=False)
 
 ```python
 # PyTorch 写法
-net = torch.nn.Sequential(l, l)
-result = torch.Tensor([0, 0])
-for i, j in enumerate(net.modules()):
-    result = j(result)
+model = torch.nn.Sequential(net1, net2)
+for layer in model.modules():
+    print(layer)
 
 # Paddle 写法
-net = paddle.nn.Sequential(l, l)
-result = paddle.Tensor([0, 0])
-for i, j in enumerate(net.sublayers()):
-    result = j(result)
+model = paddle.nn.Sequential(net1, net2)
+for layer in model.sublayers():
+    print(layer)
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.named_children.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.named_children.md
@@ -19,13 +19,11 @@ paddle.nn.Layer.named_children()
 ```python
 # PyTorch 写法
 model = torch.nn.Sequential(OrderedDict([("wfs", l), ("wfs1", l1)]))
-result = torch.Tensor([0, 0])
 for name, module in model.named_children():
-    result = module(result)
+    print(name, module)
 
 # Paddle 写法
 model = paddle.nn.Sequential(OrderedDict([("wfs", l), ("wfs1", l1)]))
-result = paddle.Tensor([0, 0])
 for name, module in model.named_children():
-    result = module(result)
+    print(name, module)
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.train.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.nn.Module.train.md
@@ -18,18 +18,10 @@ paddle.nn.Layer.train()
 
 ```python
 # PyTorch 写法
-class TheModelClass(torch.nn.Module):
-    def forward(self, x):
-        return x
-
-model = TheModelClass()
+model = torch.nn.Linear(10, 20)
 model.train()
 
 # Paddle 写法
-class TheModelClass(paddle.nn.Layer):
-    def forward(self, x):
-        return x
-
-model = TheModelClass()
+model = paddle.nn.Linear(10, 20)
 model.train()
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.numel.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.numel.md
@@ -3,7 +3,7 @@
 ### [torch.numel](https://pytorch.org/docs/stable/generated/torch.numel.html)
 
 ```python
-torch.numel()
+torch.numel(input)
 ```
 
 ### [paddle.Tensor.size](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/Tensor/size_cn.html#paddle/Tensor/size_cn#cn-api-paddle-Tensor-size)
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.optim.Optimizer.step.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torch.optim.Optimizer.step.md
@@ -18,10 +18,18 @@ paddle.optimizer.Optimizer.step()
 
 ```python
 # PyTorch 写法
-optim = torch.optim.Optimizer([theta], defaults={"learning_rate": 1.0})
-result = type(optim.step)
+sgd = torch.optim.SGD(
+    params=linear.parameters(),
+    lr=0.1,
+    weight_decay=0.01
+)
+sgd.step()
 
 # Paddle 写法
-optim = paddle.optimizer.Optimizer(parameters=[theta], **{"learning_rate": 1.0})
-result = type(optim.step)
+sgd = paddle.optimizer.SGD(
+     learning_rate=0.1,
+     parameters=linear.parameters(),
+     weight_decay=0.01
+)
+sgd.step()
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torchvision.transforms.RandomAffine.md b/docs/guides/model_convert/convert_from_pytorch/api_difference/invok_only_diff/torchvision.transforms.RandomAffine.md
@@ -34,9 +34,9 @@ transform = paddle.vision.transforms.RandomAffine(
     (0.1, 0.1),
     (0.5, 1.5),
     20,
-    torchvision.transforms.InterpolationMode.BILINEAR,
+    "bilinear",
     255,
-    (2, 2),
+    (2, 2)
 )
 
 ```
diff --git a/docs/guides/model_convert/convert_from_pytorch/tools/generate_pytorch_api_mapping.py b/docs/guides/model_convert/convert_from_pytorch/tools/generate_pytorch_api_mapping.py
diff --git a/docs/guides/model_convert/convert_from_pytorch/tools/utils.py b/docs/guides/model_convert/convert_from_pytorch/tools/utils.py
