fix: torch frontend Tensor.__bool__ now returning bools, Tensor.type_…

…as returning new tensor instead of operating inplace and fixed signature for unique_consecutive (ivy-llc#27887)

ivy/functional/frontends/torch/reduction_ops.py

@@ -361,7 +361,7 @@ def unique(input, sorted=True, return_inverse=False, return_counts=False, dim=No
     "torch",
 )
 @to_ivy_arrays_and_back
-def unique_consecutive(input, return_inverse, return_counts, dim):
+def unique_consecutive(input, return_inverse=False, return_counts=False, dim=None):
     output, inverse_indices, counts = ivy.unique_consecutive(input, axis=dim)
     ret = (output,)
     if return_inverse:

ivy/functional/frontends/torch/tensor.py

@@ -949,7 +949,7 @@ def type(self, dtype=None, non_blocking=False, **kwargs):
     @with_unsupported_dtypes({"2.1.2 and below": ("bfloat16",)}, "torch")
     def type_as(self, other):
         if self.dtype != other.dtype:
-            self.ivy_array = ivy.astype(self.ivy_array, other.dtype)
+            return torch_frontend.tensor(ivy.astype(self.ivy_array, other.dtype))
         return self
 
     def byte(self, memory_format=None):
@@ -1185,7 +1185,7 @@ def bernoulli(self, *, generator=None, out=None):
 
     def __bool__(self):
         if len(self.shape) == sum(self.shape):
-            return torch_frontend.tensor(self.ivy_array.to_scalar().__bool__())
+            return self.ivy_array.to_scalar().__bool__()
         raise ValueError(
             "The truth value of an array with more than one element is ambiguous. "
             "Use a.any() or a.all()"