【inplace api】batch add inplace api paddle.log_, paddle.i0_, paddle.nn…

….functional.leaky_relu_... (#55576) * batch add inplace api * add inplace test * add activation inplace * fix test * remove atan2 ge, gt, le, lt, nq * remove atan2 ge, gt, le, lt, nq * fix windows ci error * rerun ci * fix typro * fix bugs --------- Co-authored-by: zhangrui34 <v_zhangrui34@baidu.com>
PaddlePaddle · Jul 27, 2023 · 58a03d4 · 58a03d4
1 parent da25896
commit 58a03d4
Show file tree

Hide file tree

Showing 10 changed files with 434 additions and 14 deletions.
diff --git a/paddle/phi/api/yaml/legacy_ops.yaml b/paddle/phi/api/yaml/legacy_ops.yaml
@@ -907,6 +907,7 @@
     func : TrilInferMeta
   kernel :
     func : tril
+  inplace: (x -> out)
   backward : tril_grad
 
 - op : tril_indices
@@ -928,6 +929,7 @@
     func : TriuInferMeta
   kernel :
     func : triu
+  inplace: (x -> out)
   backward : triu_grad
 
 - op : triu_indices

diff --git a/paddle/phi/api/yaml/ops.yaml b/paddle/phi/api/yaml/ops.yaml
@@ -665,11 +665,12 @@
 
 - op : digamma
   args : (Tensor x)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
   kernel :
     func : digamma
+  inplace: (x -> out)
   backward : digamma_grad
 
 - op : dirichlet
@@ -1107,12 +1108,13 @@
 
 - op : hardtanh
   args : (Tensor x, float t_min=0, float t_max=24)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
     param : [x]
   kernel :
     func : hardtanh
+  inplace: (x -> out)
   backward : hardtanh_grad
 
 - op : heaviside
@@ -1149,6 +1151,7 @@
     func : UnchangedInferMeta
   kernel :
     func : i0
+  inplace: (x -> out)
   backward : i0_grad
 
 - op : i0e
@@ -1361,12 +1364,13 @@
 
 - op : leaky_relu
   args : (Tensor x, float negative_slope = 0.02f)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
     param : [x]
   kernel :
     func : leaky_relu
+  inplace: (x -> out)
   backward : leaky_relu_grad
 
 - op : lerp
@@ -1386,6 +1390,7 @@
     func : UnchangedInferMeta
   kernel :
     func : lgamma
+  inplace: (x -> out)
   backward : lgamma_grad
 
 - op : linear_interp
@@ -1413,38 +1418,42 @@
 
 - op : log
   args : (Tensor x)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
   kernel :
     func : log
+  inplace: (x -> out)
   backward: log_grad
 
 - op : log10
   args : (Tensor x)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
   kernel :
     func : log10
+  inplace: (x -> out)
   backward: log10_grad
 
 - op : log1p
   args : (Tensor x)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
   kernel :
     func : log1p
+  inplace: (x -> out)
   backward: log1p_grad
 
 - op : log2
   args : (Tensor x)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
   kernel :
     func : log2
+  inplace: (x -> out)
   backward: log2_grad
 
 - op : log_loss
@@ -1517,12 +1526,13 @@
 
 - op : logit
   args : (Tensor x, float eps = 1e-6f)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
     param : [x]
   kernel :
     func : logit
+  inplace: (x -> out)
   backward : logit_grad
 
 - op : logsigmoid
@@ -1895,6 +1905,7 @@
     param: [x]
   kernel :
     func : polygamma
+  inplace: (x -> out)
   backward : polygamma_grad
 
 - op : pow
@@ -2494,12 +2505,13 @@
 
 - op : thresholded_relu
   args : (Tensor x, float threshold = 1.0)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
     param : [x]
   kernel :
     func : thresholded_relu
+  inplace: (x -> out)
   backward : thresholded_relu_grad
 
 - op : topk
@@ -2546,11 +2558,12 @@
 
 - op : trunc
   args : (Tensor input)
-  output : Tensor
+  output : Tensor(out)
   infer_meta :
     func : UnchangedInferMeta
   kernel :
     func : trunc
+  inplace: (input -> out)
   backward : trunc_grad
 
 - op : unbind

diff --git a/paddle/phi/kernels/funcs/activation_functor.h b/paddle/phi/kernels/funcs/activation_functor.h
@@ -2032,7 +2032,7 @@ struct LogFunctor : public BaseActivationFunctor<T> {
 
   template <typename Device, typename X, typename Out>
   void operator()(Device d, X x, Out out) const {
-    out.device(d) = x.template cast<U>().unaryExpr(Log<U>());
+    out.device(d) = x.template cast<U>().unaryExpr(Log<U>()).eval();
   }
 };
 
@@ -2076,7 +2076,7 @@ struct Log2Functor : public BaseActivationFunctor<T> {
 
   template <typename Device, typename X, typename Out>
   void operator()(Device d, X x, Out out) const {
-    out.device(d) = x.template cast<U>().unaryExpr(Log2<U>());
+    out.device(d) = x.template cast<U>().unaryExpr(Log2<U>()).eval();
   }
 };
 
@@ -2121,7 +2121,7 @@ struct Log10Functor : public BaseActivationFunctor<T> {
 
   template <typename Device, typename X, typename Out>
   void operator()(Device d, X x, Out out) const {
-    out.device(d) = x.template cast<U>().unaryExpr(Log10<U>());
+    out.device(d) = x.template cast<U>().unaryExpr(Log10<U>()).eval();
   }
 };
 
@@ -2166,7 +2166,7 @@ struct Log1pFunctor : public BaseActivationFunctor<T> {
 
   template <typename Device, typename X, typename Out>
   void operator()(Device d, X x, Out out) const {
-    out.device(d) = x.template cast<U>().unaryExpr(Log1p<U>());
+    out.device(d) = x.template cast<U>().unaryExpr(Log1p<U>()).eval();
   }
 };
 

diff --git a/python/paddle/__init__.py b/python/paddle/__init__.py
@@ -110,7 +110,9 @@
 from .tensor.creation import full  # noqa: F401
 from .tensor.creation import full_like  # noqa: F401
 from .tensor.creation import triu  # noqa: F401
+from .tensor.creation import triu_  # noqa: F401
 from .tensor.creation import tril  # noqa: F401
+from .tensor.creation import tril_  # noqa: F401
 from .tensor.creation import meshgrid  # noqa: F401
 from .tensor.creation import empty  # noqa: F401
 from .tensor.creation import empty_like  # noqa: F401
@@ -224,14 +226,18 @@
 from .tensor.math import cumprod  # noqa: F401
 from .tensor.math import logcumsumexp  # noqa: F401
 from .tensor.math import logit  # noqa: F401
+from .tensor.math import logit_  # noqa: F401
 from .tensor.math import exp  # noqa: F401
 from .tensor.math import expm1  # noqa: F401
 from .tensor.math import expm1_  # noqa: F401
 from .tensor.math import floor  # noqa: F401
 from .tensor.math import increment  # noqa: F401
 from .tensor.math import log  # noqa: F401
+from .tensor.math import log_  # noqa: F401
+from .tensor.math import log2_  # noqa: F401
 from .tensor.math import log2  # noqa: F401
 from .tensor.math import log10  # noqa: F401
+from .tensor.math import log10_  # noqa: F401
 from .tensor.math import multiplex  # noqa: F401
 from .tensor.math import pow  # noqa: F401
 from .tensor.math import pow_  # noqa: F401
@@ -279,6 +285,7 @@
 from .tensor.math import logaddexp  # noqa: F401
 from .tensor.math import inverse  # noqa: F401
 from .tensor.math import log1p  # noqa: F401
+from .tensor.math import log1p_  # noqa: F401
 from .tensor.math import erf  # noqa: F401
 from .tensor.math import erf_  # noqa: F401
 from .tensor.math import addmm  # noqa: F401
@@ -294,9 +301,13 @@
 from .tensor.math import broadcast_shape  # noqa: F401
 from .tensor.math import conj  # noqa: F401
 from .tensor.math import trunc  # noqa: F401
+from .tensor.math import trunc_  # noqa: F401
 from .tensor.math import digamma  # noqa: F401
+from .tensor.math import digamma_  # noqa: F401
 from .tensor.math import neg  # noqa: F401
+from .tensor.math import neg_  # noqa: F401
 from .tensor.math import lgamma  # noqa: F401
+from .tensor.math import lgamma_  # noqa: F401
 from .tensor.math import acosh  # noqa: F401
 from .tensor.math import acosh_  # noqa: F401
 from .tensor.math import asinh  # noqa: F401
@@ -317,6 +328,7 @@
 from .tensor.math import outer  # noqa: F401
 from .tensor.math import heaviside  # noqa: F401
 from .tensor.math import frac  # noqa: F401
+from .tensor.math import frac_  # noqa: F401
 from .tensor.math import sgn  # noqa: F401
 from .tensor.math import take  # noqa: F401
 from .tensor.math import frexp  # noqa: F401
@@ -326,10 +338,12 @@
 from .tensor.math import vander  # noqa: F401
 from .tensor.math import nextafter  # noqa: F401
 from .tensor.math import i0  # noqa: F401
+from .tensor.math import i0_  # noqa: F401
 from .tensor.math import i0e  # noqa: F401
 from .tensor.math import i1  # noqa: F401
 from .tensor.math import i1e  # noqa: F401
 from .tensor.math import polygamma  # noqa: F401
+from .tensor.math import polygamma_  # noqa: F401
 
 from .tensor.random import bernoulli  # noqa: F401
 from .tensor.random import poisson  # noqa: F401
@@ -473,6 +487,7 @@
     'logaddexp',
     'logcumsumexp',
     'logit',
+    'logit_',
     'LazyGuard',
     'sign',
     'is_empty',
@@ -561,6 +576,7 @@
     'rand',
     'less_equal',
     'triu',
+    'triu_',
     'sin',
     'sin_',
     'dist',
@@ -582,6 +598,7 @@
     'abs',
     'abs_',
     'tril',
+    'tril_',
     'pow',
     'pow_',
     'zeros_like',
@@ -608,7 +625,9 @@
     'broadcast_shape',
     'conj',
     'neg',
+    'neg_',
     'lgamma',
+    'lgamma_',
     'lerp',
     'erfinv',
     'inner',
@@ -693,13 +712,19 @@
     'floor',
     'cosh',
     'log',
+    'log_',
     'log2',
+    'log2_',
     'log10',
+    'log10_',
     'concat',
     'check_shape',
     'trunc',
+    'trunc_',
     'frac',
+    'frac_',
     'digamma',
+    'digamma_',
     'standard_normal',
     'diagonal',
     'broadcast_tensors',
@@ -741,8 +766,10 @@
     'unflatten',
     'nextafter',
     'i0',
+    'i0_',
     'i0e',
     'i1',
     'i1e',
     'polygamma',
+    'polygamma_',
 ]
diff --git a/python/paddle/nn/functional/__init__.py b/python/paddle/nn/functional/__init__.py
@@ -21,9 +21,11 @@
 from .activation import gelu  # noqa: F401
 from .activation import hardshrink  # noqa: F401
 from .activation import hardtanh  # noqa: F401
+from .activation import hardtanh_  # noqa: F401
 from .activation import hardsigmoid  # noqa: F401
 from .activation import hardswish  # noqa: F401
 from .activation import leaky_relu  # noqa: F401
+from .activation import leaky_relu_  # noqa: F401
 from .activation import log_sigmoid  # noqa: F401
 from .activation import maxout  # noqa: F401
 from .activation import prelu  # noqa: F401
@@ -44,6 +46,7 @@
 from .activation import tanh_  # noqa: F401
 from .activation import tanhshrink  # noqa: F401
 from .activation import thresholded_relu  # noqa: F401
+from .activation import thresholded_relu_  # noqa: F401
 from .activation import log_softmax  # noqa: F401
 from .activation import glu  # noqa: F401
 from .activation import gumbel_softmax  # noqa: F401
@@ -153,9 +156,11 @@
     'gelu',
     'hardshrink',
     'hardtanh',
+    'hardtanh_',
     'hardsigmoid',
     'hardswish',
     'leaky_relu',
+    'leaky_relu_',
     'log_sigmoid',
     'maxout',
     'prelu',
@@ -176,6 +181,7 @@
     'tanh_',
     'tanhshrink',
     'thresholded_relu',
+    'thresholded_relu_',
     'log_softmax',
     'glu',
     'gumbel_softmax',