From 03dab2ba70e2750ad5c2ef30ca5ace9e054f5e27 Mon Sep 17 00:00:00 2001
From: Haris Mahmood <hmahmood.bee16@gmail.com>
Date: Thu, 5 Sep 2024 02:01:44 +0500
Subject: [PATCH] fix: Fixed the frontend torch.Tensor class such that all
 inplace methods now follow a consistent policy for inplace updating the
 underlying ivy array attr

---
 ivy/functional/frontends/torch/tensor.py | 333 ++++++++++++++---------
 1 file changed, 207 insertions(+), 126 deletions(-)

diff --git a/ivy/functional/frontends/torch/tensor.py b/ivy/functional/frontends/torch/tensor.py
index 1af5260ec5d4..942a427eece0 100644
--- a/ivy/functional/frontends/torch/tensor.py
+++ b/ivy/functional/frontends/torch/tensor.py
@@ -185,7 +185,8 @@ def all(self, dim=None, keepdim=False):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def add_(self, other, *, alpha=1):
-        self.ivy_array = self.add(other, alpha=alpha).ivy_array
+        ret = self.add(other, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -194,7 +195,8 @@ def addmm(self, mat1, mat2, *, beta=1, alpha=1):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def addmm_(self, mat1, mat2, *, beta=1, alpha=1):
-        self.ivy_array = self.addmm(mat1, mat2, beta=beta, alpha=alpha).ivy_array
+        ret = self.addmm(mat1, mat2, beta=beta, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -203,9 +205,8 @@ def addmv(self, mat, vec, *, beta=1, alpha=1):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def addmv_(self, mat, vec, *, beta=1, alpha=1):
-        self.ivy_array = torch_frontend.addmv(
-            self, mat, vec, beta=beta, alpha=alpha
-        ).ivy_array
+        ret = torch_frontend.addmv(self, mat, vec, beta=beta, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -214,12 +215,14 @@ def addbmm(self, batch1, batch2, *, beta=1, alpha=1):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def addbmm_(self, batch1, batch2, *, beta=1, alpha=1):
-        self.ivy_array = self.addbmm(batch1, batch2, beta=beta, alpha=alpha).ivy_array
+        ret = self.addbmm(batch1, batch2, beta=beta, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def subtract_(self, other, *, alpha=1):
-        self.ivy_array = self.sub(other, alpha=alpha).ivy_array
+        ret = self.sub(other, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -228,7 +231,8 @@ def asin(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def asin_(self):
-        self.ivy_array = self.asin().ivy_array
+        ret = self.asin()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def float_power(self, exponent):
@@ -245,7 +249,8 @@ def sin(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def sin_(self):
-        self.ivy_array = self.sin().ivy_array
+        ret = self.sin()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -254,7 +259,8 @@ def sinh(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def sinh_(self):
-        self.ivy_array = self.sinh().ivy_array
+        ret = self.sinh()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -263,7 +269,8 @@ def cos(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def cos_(self):
-        self.ivy_array = self.cos().ivy_array
+        ret = self.cos()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -272,7 +279,8 @@ def cosh(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def cosh_(self):
-        self.ivy_array = self.cosh().ivy_array
+        ret = self.cosh()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -281,7 +289,8 @@ def atan(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def atan_(self):
-        self.ivy_array = self.atan().ivy_array
+        ret = self.atan()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
@@ -336,7 +345,8 @@ def asinh(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def asinh_(self):
-        self.ivy_array = self.asinh().ivy_array
+        ret = self.asinh()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -345,7 +355,8 @@ def tan(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def tan_(self):
-        self.ivy_array = self.tan().ivy_array
+        ret = self.tan()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -354,7 +365,8 @@ def tanh(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def tanh_(self):
-        self.ivy_array = self.tanh().ivy_array
+        ret = self.tanh()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -363,7 +375,8 @@ def atanh(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def atanh_(self):
-        self.ivy_array = self.atanh().ivy_array
+        ret = self.atanh()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -372,7 +385,8 @@ def log(self):
 
     @with_supported_dtypes({"2.2 and below": ("float32", "float64")}, "torch")
     def log2_(self):
-        self.ivy_array = self.log2().ivy_array
+        ret = self.log2()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
@@ -381,12 +395,14 @@ def logit(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "uint16")}, "torch")
     def copy_(self, other, non_blocking=False):
-        self._ivy_array = torch_frontend.tensor(other).ivy_array
+        ret = torch_frontend.tensor(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def log_(self):
-        self.ivy_array = self.log().ivy_array
+        ret = self.log()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -416,7 +432,8 @@ def abs(self):
         return torch_frontend.abs(self)
 
     def abs_(self):
-        self.ivy_array = self.abs().ivy_array
+        ret = self.abs()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
@@ -427,7 +444,10 @@ def logical_not(self, *, out=None):
         return torch_frontend.logical_not(self, out=out)
 
     def logical_not_(self):
-        self.ivy_array = ivy.astype(self.logical_not().ivy_array, self.dtype)
+        ret = self.logical_not()
+        self.ivy_array = ivy.inplace_update(
+            self.ivy_array, ivy.astype(ret.ivy_array, self.dtype)
+        )
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
@@ -453,7 +473,8 @@ def bitwise_left_shift(self, other):
 
     @with_supported_dtypes({"2.2 and below": ("integer",)}, "torch")
     def bitwise_or_(self, other):
-        self.ivy_array = self.bitwise_or(other).ivy_array
+        ret = self.bitwise_or(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def contiguous(self, memory_format=None):
@@ -519,7 +540,8 @@ def not_equal(self, other, *, out=None):
         "torch",
     )
     def not_equal_(self, other, *, out=None):
-        self.ivy_array = self.not_equal(other).ivy_array
+        ret = self.not_equal(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def eq(self, other):
@@ -536,7 +558,8 @@ def erf(self, *, out=None):
         {"2.2 and below": ("float32", "float64", "bfloat16")}, "torch"
     )
     def erf_(self, *, out=None):
-        self.ivy_array = self.erf(out=out).ivy_array
+        ret = self.erf(out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_supported_device_and_dtypes(
@@ -551,7 +574,8 @@ def erfc(self, *, out=None):
         "torch",
     )
     def erfc_(self, *, out=None):
-        self.ivy_array = self.erfc(out=out).ivy_array
+        ret = self.erfc(out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def new_zeros(
@@ -645,7 +669,8 @@ def acos(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def acos_(self):
-        self.ivy_array = self.acos().ivy_array
+        ret = self.acos()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def new_tensor(
@@ -692,7 +717,8 @@ def detach(self):
         )
 
     def detach_(self):
-        self.ivy_array = self.detach().ivy_array
+        ret = self.detach()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("uint16",)}, "torch")
@@ -702,7 +728,8 @@ def unsqueeze(self, dim):
 
     @numpy_to_torch_style_args
     def unsqueeze_(self, dim):
-        self.ivy_array = self.unsqueeze(dim).ivy_array
+        ret = self.unsqueeze(dim)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def ravel(self):
@@ -836,7 +863,8 @@ def unflatten(self, dim, sizes):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def pow_(self, exponent):
-        self.ivy_array = self.pow(exponent).ivy_array
+        ret = self.pow(exponent)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def size(self, dim=None):
@@ -923,14 +951,16 @@ def transpose(self, dim0, dim1):
         return torch_frontend.transpose(self, dim0=dim0, dim1=dim1)
 
     def transpose_(self, dim0, dim1):
-        self.ivy_array = self.transpose(dim0, dim1).ivy_array
+        ret = self.transpose(dim0, dim1)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def t(self):
         return torch_frontend.t(self)
 
     def t_(self):
-        self.ivy_array = self.t().ivy_array
+        ret = self.t()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def flatten(self, start_dim=0, end_dim=-1):
@@ -944,7 +974,8 @@ def cumsum(self, dim, *, dtype=None):
     @numpy_to_torch_style_args
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def cumsum_(self, dim, *, dtype=None):
-        self.ivy_array = self.cumsum(dim, dtype=dtype).ivy_array
+        ret = self.cumsum(dim, dtype=dtype)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
@@ -957,7 +988,8 @@ def neg(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("bool",)}, "torch")
     def neg_(self):
-        self.ivy_array = torch_frontend.negative(self).ivy_array
+        ret = torch_frontend.negative(self)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     __neg__ = neg
@@ -968,7 +1000,8 @@ def negative(self):
 
     @with_unsupported_dtypes({"2.0.1 and below": ("bool", "bfloat16")}, "torch")
     def negative_(self):
-        self.ivy_array = torch_frontend.negative(self).ivy_array
+        ret = torch_frontend.negative(self)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def int(self, memory_format=None):
@@ -1007,7 +1040,8 @@ def squeeze(self, dim=None):
     @numpy_to_torch_style_args
     @with_unsupported_dtypes({"2.2 and below": ("uint16",)}, "torch")
     def squeeze_(self, dim=None):
-        self.ivy_array = self.squeeze(dim).ivy_array
+        ret = self.squeeze(dim)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def flip(self, dims):
@@ -1023,7 +1057,8 @@ def tril(self, diagonal=0):
         return torch_frontend.tril(self, diagonal=diagonal)
 
     def tril_(self, diagonal=0):
-        self.ivy_array = self.tril(diagonal=diagonal).ivy_array
+        ret = self.tril(diagonal=diagonal)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def index_select(self, dim, index):
@@ -1035,7 +1070,8 @@ def clamp(self, min=None, max=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "complex")}, "torch")
     def clamp_(self, min=None, max=None):
-        self.ivy_array = self.clamp(min=min, max=max).ivy_array
+        ret = self.clamp(min=min, max=max)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes(
@@ -1045,7 +1081,8 @@ def clamp_min(self, min=None):
         return torch_frontend.clamp(self, min=min)
 
     def clamp_min_(self, min=None):
-        self.ivy_array = self.clamp_min(min).ivy_array
+        ret = self.clamp_min(min)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
@@ -1058,12 +1095,14 @@ def rsqrt(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
     def rsqrt_(self):
-        self.ivy_array = self.rsqrt().ivy_array
+        ret = self.rsqrt()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
     def sqrt_(self):
-        self.ivy_array = self.sqrt().ivy_array
+        ret = self.sqrt()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def where(self, condition, other):
@@ -1083,7 +1122,8 @@ def masked_fill(self, mask, value):
         )
 
     def masked_fill_(self, mask, value):
-        self.ivy_array = self.masked_fill(mask, value).ivy_array
+        ret = self.masked_fill(mask, value)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def masked_select(self, mask):
@@ -1103,25 +1143,24 @@ def masked_scatter_(self, mask, source):
         flat_source = torch_frontend.flatten(source)
         indices = torch_frontend.squeeze(torch_frontend.nonzero(flat_mask), -1)
         flat_self.scatter_(0, indices, flat_source[: indices.shape[0]])
-        self.ivy_array = flat_self.reshape(self.shape).ivy_array
+        ret = flat_self.reshape(self.shape)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
     def index_add_(self, dim, index, source, *, alpha=1):
-        self.ivy_array = torch_frontend.index_add(
-            self, dim, index, source, alpha=alpha
-        ).ivy_array
+        ret = torch_frontend.index_add(self, dim, index, source, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
     def index_add(self, dim, index, source, *, alpha=1):
-        return torch_frontend.index_add(
-            self._ivy_array, dim, index, source, alpha=alpha
-        )
+        return torch_frontend.index_add(self.ivy_array, dim, index, source, alpha=alpha)
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def acosh_(self):
-        self.ivy_array = self.acosh().ivy_array
+        ret = self.acosh()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
@@ -1134,7 +1173,8 @@ def sigmoid(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def sigmoid_(self):
-        self.ivy_array = self.sigmoid().ivy_array
+        ret = self.sigmoid()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -1172,24 +1212,29 @@ def remainder(self, other, *, out=None):
         {"2.2 and below": ("float16", "float32", "float64", "bfloat16")}, "torch"
     )
     def reciprocal_(self):
-        self.ivy_array = torch_frontend.reciprocal(self).ivy_array
+        ret = torch_frontend.reciprocal(self)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def remainder_(self, other, *, out=None):
-        self.ivy_array = torch_frontend.remainder(self, other, out=out).ivy_array
+        ret = torch_frontend.remainder(self, other, out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def bitwise_not_(self):
-        self.ivy_array = self.bitwise_not().ivy_array
+        ret = self.bitwise_not()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def bitwise_and_(self, other):
-        self.ivy_array = self.bitwise_and(other).ivy_array
+        ret = self.bitwise_and(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
     def atan2_(self, other):
-        self.ivy_array = self.atan2(other).ivy_array
+        ret = self.atan2(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
@@ -1215,7 +1260,8 @@ def trunc(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "complex")}, "torch")
     def trunc_(self):
-        self.ivy_array = self.trunc().ivy_array
+        ret = self.trunc()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "complex")}, "torch")
@@ -1224,28 +1270,30 @@ def fix(self):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "complex")}, "torch")
     def fix_(self):
-        self.ivy_array = self.fix().ivy_array
+        ret = self.fix()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def isinf(self):
-        return torch_frontend.isinf(self._ivy_array)
+        return torch_frontend.isinf(self.ivy_array)
 
     def is_complex(self):
-        return torch_frontend.is_complex(self._ivy_array)
+        return torch_frontend.is_complex(self.ivy_array)
 
     @with_unsupported_dtypes({"2.2 and below": ("uint16", "bfloat16")}, "torch")
     def is_floating_point(self):
-        return torch_frontend.is_floating_point(self._ivy_array)
+        return torch_frontend.is_floating_point(self.ivy_array)
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def isreal(self):
-        return torch_frontend.isreal(self._ivy_array)
+        return torch_frontend.isreal(self.ivy_array)
 
     def addr(self, vec1, vec2, *, beta=1, alpha=1, out=None):
         return torch_frontend.addr(self, vec1, vec2, beta=beta, alpha=alpha, out=out)
 
     def addr_(self, vec1, vec2, *, beta=1, alpha=1):
-        self.ivy_array = self.addr(vec1, vec2, beta=beta, alpha=alpha).ivy_array
+        ret = self.addr(vec1, vec2, beta=beta, alpha=alpha)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
@@ -1254,15 +1302,16 @@ def dot(self, tensor):
 
     @with_supported_dtypes({"2.2 and below": ("float32", "float64")}, "torch")
     def bernoulli(self, *, generator=None, out=None):
-        return torch_frontend.bernoulli(self._ivy_array, generator=generator, out=out)
+        return torch_frontend.bernoulli(self.ivy_array, generator=generator, out=out)
 
     @with_supported_dtypes({"2.2 and below": ("float32", "float64")}, "torch")
     def bernoulli_(self, p, *, generator=None, out=None):
-        self.ivy_array = torch_frontend.bernoulli(
+        ret = torch_frontend.bernoulli(
             torch_frontend.full(self.shape, p, dtype=torch_frontend.float64),
             generator=generator,
             out=out,
-        ).ivy_array
+        )
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def numel(self):
@@ -1466,7 +1515,8 @@ def bitwise_xor(self, other):
         return torch_frontend.bitwise_xor(self, other)
 
     def bitwise_xor_(self, other):
-        self.ivy_array = self.bitwise_xor(other).ivy_array
+        ret = self.bitwise_xor(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def item(self):
@@ -1511,13 +1561,15 @@ def expm1(self):
         {"2.2 and below": ("bfloat16", "float16", "complex")}, "torch"
     )
     def expm1_(self):
-        self.ivy_array = torch_frontend.expm1(self).ivy_array
+        ret = torch_frontend.expm1(self)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     # fmt: off
     @with_unsupported_dtypes({"2.2 and below": ("int8", "int16", "int32", "int64", "uint8", "bool", "float16",)},"torch",)  # noqa
     def exp_(self):
-        self.ivy_array = self.exp().ivy_array
+        ret = self.exp()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
     # fmt: on
 
@@ -1526,14 +1578,14 @@ def mul(self, other):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def ceil_(self):
-        self.ivy_array = torch_frontend.ceil(self).ivy_array
+        ret = torch_frontend.ceil(self)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def mul_(self, other):
-        self.ivy_array = self.mul(other).ivy_array
-        # the return dtype is the same as the input dtype
-        self.ivy_array = self.to(self.dtype).ivy_array
+        ret = self.mul(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "float16")}, "torch")
@@ -1542,7 +1594,8 @@ def round(self, *, decimals=0):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "float16")}, "torch")
     def round_(self, *, decimals=0):
-        self.ivy_array = self.round(decimals=decimals).ivy_array
+        ret = self.round(decimals=decimals)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @numpy_to_torch_style_args
@@ -1561,7 +1614,7 @@ def fill_(self, value):
         ret = torch_frontend.full_like(
             self, value, dtype=self.dtype, device=self.device
         )
-        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def nonzero(self, as_tuple=False):
@@ -1572,7 +1625,7 @@ def mm(self, mat2):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "float16")}, "torch")
     def square(self):
-        return torch_frontend.square(self._ivy_array)
+        return torch_frontend.square(self.ivy_array)
 
     @with_supported_dtypes(
         {
@@ -1592,22 +1645,24 @@ def square(self):
         "torch",
     )
     def square_(self):
-        self.ivy_array = torch_frontend.square(self._ivy_array).ivy_array
+        ret = torch_frontend.square(self.ivy_array)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def log10(self):
-        return torch_frontend.log10(self._ivy_array)
+        return torch_frontend.log10(self.ivy_array)
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def log10_(self):
-        self.ivy_array = self.log10().ivy_array
+        ret = self.log10()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("uint16",)}, "torch")
     def zero_(self):
         ret = torch_frontend.zeros_like(self)
-        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def short(self, memory_format=None):
@@ -1623,24 +1678,27 @@ def div(self, other, *, rounding_mode=None):
         return torch_frontend.div(self, other, rounding_mode=rounding_mode)
 
     def div_(self, other, *, rounding_mode=None):
-        self.ivy_array = self.div(other, rounding_mode=rounding_mode).ivy_array
+        ret = self.div(other, rounding_mode=rounding_mode)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_supported_dtypes(
         {"2.2 and below": ("float16", "float32", "float64", "bfloat16")}, "torch"
     )
     def true_divide_(self, other):
-        self.ivy_array = self.div(other, rounding_mode=None).ivy_array
+        ret = self.div(other, rounding_mode=None)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def normal_(self, mean=0, std=1, *, generator=None):
-        self.ivy_array = ivy.random_normal(
+        ret = ivy.random_normal(
             mean=mean,
             std=std,
             shape=self.ivy_array.shape,
             dtype=self.dtype,
             device=self.device,
         )
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
@@ -1653,18 +1711,20 @@ def addcmul(self, tensor1, tensor2, *, value=1):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def addcmul_(self, tensor1, tensor2, *, value=1):
-        self.ivy_array = self.addcmul(tensor1, tensor2, value=value).ivy_array
+        ret = self.addcmul(tensor1, tensor2, value=value)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     sign_decorator_dtypes = ("float16", "complex", "bool")
 
     @with_unsupported_dtypes({"2.2 and below": sign_decorator_dtypes}, "torch")
     def sign(self):
-        return torch_frontend.sign(self._ivy_array)
+        return torch_frontend.sign(self.ivy_array)
 
     @with_unsupported_dtypes({"2.2 and below": sign_decorator_dtypes}, "torch")
     def sign_(self):
-        self.ivy_array = self.sign().ivy_array
+        ret = self.sign()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @numpy_to_torch_style_args
@@ -1679,14 +1739,15 @@ def fmod(self, other, *, out=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
     def fmod_(self, other):
-        self.ivy_array = self.fmod(other).ivy_array
+        ret = self.fmod(other)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def norm(self, p="fro", dim=None, keepdim=False, dtype=None):
         return torch_frontend.norm(self, p=p, dim=dim, keepdim=keepdim, dtype=dtype)
 
     def tolist(self):
-        return self._ivy_array.to_list()
+        return self.ivy_array.to_list()
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def multiply(self, other, *, out=None):
@@ -1694,7 +1755,8 @@ def multiply(self, other, *, out=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def multiply_(self, other, *, out=None):
-        self.ivy_array = torch_frontend.multiply(self, other, out=out).ivy_array
+        ret = torch_frontend.multiply(self, other, out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @numpy_to_torch_style_args
@@ -1706,13 +1768,14 @@ def topk(self, k, dim=None, largest=True, sorted=True):
 
     @with_unsupported_dtypes({"2.2 and below": rshift_dtypes}, "torch")
     def bitwise_right_shift(self, other, *, out=None):
-        return torch_frontend.bitwise_right_shift(self._ivy_array, other)
+        return torch_frontend.bitwise_right_shift(self.ivy_array, other)
 
     @with_supported_dtypes(
         {"2.2 and below": ("uint8", "int8", "int32", "int64")}, "torch"
     )
     def bitwise_right_shift_(self, other, *, out=None):
-        self.ivy_array = self.bitwise_right_shift(other, out=out).ivy_array
+        ret = self.bitwise_right_shift(other, out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("float16", "bfloat16")}, "torch")
@@ -1730,7 +1793,8 @@ def copysign(self, other, *, out=None):
         {"2.2 and below": ("float16", "float32", "float64")}, "torch"
     )
     def copysign_(self, other, *, out=None):
-        self.ivy_array = self.copysign(other, out=out).ivy_array
+        ret = self.copysign(other, out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes(
@@ -1741,7 +1805,8 @@ def greater(self, other, *, out=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "bool")}, "torch")
     def greater_(self, other):
-        self.ivy_array = ivy.astype(self.greater(other).ivy_array, self.dtype)
+        ret = ivy.astype(self.greater(other).ivy_array, self.dtype)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     @with_unsupported_dtypes(
@@ -1752,7 +1817,10 @@ def greater_equal(self, other, *, out=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "bool")}, "torch")
     def greater_equal_(self, other):
-        self.ivy_array = ivy.astype(self.greater_equal(other).ivy_array, self.dtype)
+        ret = self.greater_equal(other)
+        self.ivy_array = ivy.inplace_update(
+            self.ivy_array, ivy.astype(ret.ivy_array, self.dtype)
+        )
         return self
 
     @with_unsupported_dtypes(
@@ -1763,7 +1831,10 @@ def less(self, other, *, out=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "bool")}, "torch")
     def less_(self, other):
-        self.ivy_array = ivy.astype(self.less(other).ivy_array, self.dtype)
+        ret = self.less(other)
+        self.ivy_array = ivy.inplace_update(
+            self.ivy_array, ivy.astype(ret.ivy_array, self.dtype)
+        )
         return self
 
     @with_unsupported_dtypes(
@@ -1774,13 +1845,17 @@ def less_equal(self, other, *, out=None):
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16", "bool")}, "torch")
     def less_equal_(self, other):
-        self.ivy_array = ivy.astype(self.less_equal(other).ivy_array, self.dtype)
+        ret = self.less_equal(other)
+        self.ivy_array = ivy.inplace_update(
+            self.ivy_array, ivy.astype(ret.ivy_array, self.dtype)
+        )
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("bfloat16",)}, "torch")
     def eq_(self, other):
-        self.ivy_array = ivy.astype(
-            torch_frontend.eq(self, other).ivy_array, self.dtype
+        ret = torch_frontend.eq(self, other)
+        self.ivy_array = ivy.inplace_update(
+            self.ivy_array, ivy.astype(ret.ivy_array, self.dtype)
         )
         return self
 
@@ -1816,8 +1891,9 @@ def log1p(self):
     @with_supported_dtypes({"2.2 and below": ("float32", "float64")}, "torch")
     def log1p_(self):
         promoted_type = ivy.promote_types(self.dtype, "float32")
-        res = torch_frontend.log1p(self)
-        self.ivy_array = res.to(promoted_type).ivy_array
+        ret = torch_frontend.log1p(self)
+        ret = ret.to(promoted_type)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def baddbmm(self, batch1, batch2, *, beta=1, alpha=1):
@@ -1826,9 +1902,10 @@ def baddbmm(self, batch1, batch2, *, beta=1, alpha=1):
         )
 
     def baddbmm_(self, batch1, batch2, *, beta=1, alpha=1):
-        self.ivy_array = torch_frontend.baddbmm(
+        ret = torch_frontend.baddbmm(
             self, batch1=batch1, batch2=batch2, beta=beta, alpha=alpha
-        ).ivy_array
+        )
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     def bmm(self, mat2):
@@ -1836,7 +1913,8 @@ def bmm(self, mat2):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def floor_(self):
-        self.ivy_array = self.floor().ivy_array
+        ret = self.floor()
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes(
@@ -1868,7 +1946,8 @@ def gather(self, dim, index):
         {"2.2 and below": ("float32", "float64", "int32", "int64")}, "torch"
     )
     def scatter_add_(self, dim, index, src):
-        self.ivy_array = ivy.put_along_axis(self.ivy_array, index, src, dim, mode="sum")
+        ret = ivy.put_along_axis(self.ivy_array, index, src, dim, mode="sum")
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     @with_supported_dtypes(
@@ -1883,9 +1962,8 @@ def scatter_(self, dim, index, src, *, reduce=None):
                 "multiply": "mul",
             }
             reduce = mode_mappings.get(reduce, reduce)
-        self.ivy_array = ivy.put_along_axis(
-            self.ivy_array, index, src, dim, mode=reduce
-        )
+        ret = ivy.put_along_axis(self.ivy_array, index, src, dim, mode=reduce)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     @with_supported_dtypes(
@@ -1894,9 +1972,8 @@ def scatter_(self, dim, index, src, *, reduce=None):
     def scatter_reduce_(self, dim, index, src, reduce, *, include_self=True):
         if reduce == "prod":
             reduce = "mul"
-        self.ivy_array = ivy.put_along_axis(
-            self.ivy_array, index, src, dim, mode=reduce
-        )
+        ret = ivy.put_along_axis(self.ivy_array, index, src, dim, mode=reduce)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     @with_supported_dtypes(
@@ -1925,9 +2002,8 @@ def movedim(self, source, destination):
 
     @with_unsupported_dtypes({"2.2 and below": ("float16",)}, "torch")
     def addcdiv_(self, tensor1, tensor2, *, value=1):
-        self.ivy_array = self.addcdiv(
-            tensor1=tensor1, tensor2=tensor2, value=value
-        ).ivy_array
+        ret = self.addcdiv(tensor1=tensor1, tensor2=tensor2, value=value)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_supported_dtypes(
@@ -1957,7 +2033,8 @@ def tile(self, *reps):
     def apply_(self, callable, /):
         if self.device != "cpu":
             raise ValueError("apply_ is only supported on cpu tensors")
-        self.ivy_array = callable(self.ivy_array)
+        ret = callable(self.ivy_array)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     def requires_grad_(self, requires_grad=True):
@@ -2079,7 +2156,8 @@ def lcm(self, other, *, out=None):
         "torch",
     )
     def lcm_(self, other, *, out=None):
-        self.ivy_array = self.lcm(other, out=out).ivy_array
+        ret = self.lcm(other, out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes(
@@ -2097,7 +2175,8 @@ def lcm_(self, other, *, out=None):
         "torch",
     )
     def triu_(self, diagonal=0):
-        self.ivy_array = torch_frontend.triu(self, diagonal).ivy_array
+        ret = torch_frontend.triu(self, diagonal)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes(
@@ -2136,9 +2215,10 @@ def random_(
                 to = ivy.finfo(self.dtype).max
             else:
                 to = ivy.iinfo(self.dtype).max
-        self.ivy_array = ivy.random_uniform(
+        ret = ivy.random_uniform(
             low=from_, high=to, shape=self.size(), dtype=self.dtype
         )
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self.ivy_array
 
     @with_unsupported_dtypes(
@@ -2157,14 +2237,12 @@ def uniform_(self, from_=0, to=1, *, generator=None):
         ret = ivy.random_uniform(
             low=from_, high=to, shape=self.shape, dtype=self.dtype, seed=generator
         )
-        self._ivy_array = ivy.inplace_update(
-            self._ivy_array, ivy.astype(ret, self._ivy_array.dtype)
-        )
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ivy.astype(ret, self.dtype))
         return self
 
     @with_supported_dtypes({"2.2 and below": ("float32", "float64")}, "torch")
     def frac(self, name=None):
-        return torch_frontend.frac(self._ivy_array)
+        return torch_frontend.frac(self.ivy_array)
 
     @with_unsupported_dtypes(
         {
@@ -2189,7 +2267,8 @@ def sinc(self):
         "torch",
     )
     def sinc_(self):
-        self.ivy_array = torch_frontend.sinc(self).ivy_array
+        ret = torch_frontend.sinc(self)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes({"2.2 and below": ("uint8",)}, "torch")
@@ -2263,7 +2342,8 @@ def triu(self, diagonal=0):
         "torch",
     )
     def xlogy_(self, *, other, out=None):
-        self.ivy_array = torch_frontend.xlogy(self, other, out=out).ivy_array
+        ret = torch_frontend.xlogy(self, other, out=out)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret.ivy_array)
         return self
 
     @with_unsupported_dtypes(
@@ -2341,7 +2421,8 @@ def rad2deg(self, *, out=None):
         return torch_frontend.rad2deg(self, out=out)
 
     def fill_diagonal_(self, fill_value, wrap=False):
-        self._ivy_array = ivy.fill_diagonal(self._ivy_array, fill_value, wrap=wrap)
+        ret = ivy.fill_diagonal(self.ivy_array, fill_value, wrap=wrap)
+        self.ivy_array = ivy.inplace_update(self.ivy_array, ret)
         return self
 
     @with_supported_dtypes(
@@ -2380,8 +2461,8 @@ def erfinv(self, *, out=None):
     @with_unsupported_dtypes({"2.2 and below": ("float16", "complex")}, "torch")
     def erfinv_(self, *, out=None):
         ret = self.erfinv(out=out)
-        self._ivy_array = ivy.inplace_update(
-            self._ivy_array, ivy.astype(ret.ivy_array, self._ivy_array.dtype)
+        self.ivy_array = ivy.inplace_update(
+            self.ivy_array, ivy.astype(ret.ivy_array, self.dtype)
         )
         return self