Merge branch 'master' into master

docker/multicuda_framework_directory.py

@@ -28,9 +28,10 @@ def install_pkg(path, pkg, base="fw/"):
             f"pip3 install {pkg} --default-timeout=100   --no-cache-dir", shell=True
         )
 
-if __name__=="__main__":
-    arg_lis=sys.argv
-    directory_generator(arg_lis[1:],"")
+
+if __name__ == "__main__":
+    arg_lis = sys.argv
+    directory_generator(arg_lis[1:], "")
 
 
 # torch_req = ["torch/1.13.1"]

docker/multiversion_framework_directory.py

@@ -25,38 +25,41 @@ def install_pkg(path, pkg, base="fw/"):
         )
     elif pkg.split("==")[0] == "tensorflow":
         subprocess.run(
-            f"pip3 install tensorflow-cpu=={pkg.split('==')[1]} --default-timeout=100  --target={path} --no-cache-dir", shell=True
+            f"pip3 install tensorflow-cpu=={pkg.split('==')[1]} --default-timeout=100  --target={path} --no-cache-dir",
+            shell=True,
         )
     else:
         subprocess.run(
-            f"pip3 install {pkg} --default-timeout=100  --target={path} --no-cache-dir", shell=True
+            f"pip3 install {pkg} --default-timeout=100  --target={path} --no-cache-dir",
+            shell=True,
         )
 
 
-if __name__=="__main__":
-    arg_lis=sys.argv
+if __name__ == "__main__":
+    arg_lis = sys.argv
 
-    if 'backend' in arg_lis:
+    if "backend" in arg_lis:
         for i in arg_lis[2:]:
-            if i.split('/')[0]=='torch':
+            if i.split("/")[0] == "torch":
                 subprocess.run(
                     f"pip3 install torch=={i.split('/')[1]} --default-timeout=100 --extra-index-url https://download.pytorch.org/whl/cpu  --no-cache-dir",
                     shell=True,
                 )
-            elif i.split('/')[0]=='tensorflow':
+            elif i.split("/")[0] == "tensorflow":
                 print(i, "heere")
                 subprocess.run(
                     f"pip3 install tensorflow-cpu=={i.split('/')[1]} --default-timeout=100   --no-cache-dir",
-                    shell=True
+                    shell=True,
                 )
-            elif i.split('/')[0]=='jaxlib':
+            elif i.split("/")[0] == "jaxlib":
                 subprocess.run(
                     f"pip3 install {i} --default-timeout=100 -f https://storage.googleapis.com/jax-releases/jax_releases.html  --no-cache-dir",
                     shell=True,
                 )
             else:
                 subprocess.run(
-                    f"pip3 install {i.split('/')[0]}=={i.split('/')[1]} --default-timeout=100   --no-cache-dir", shell=True
+                    f"pip3 install {i.split('/')[0]}=={i.split('/')[1]} --default-timeout=100   --no-cache-dir",
+                    shell=True,
                 )
         try:
             import tensorflow
@@ -87,6 +90,3 @@ def install_pkg(path, pkg, base="fw/"):
             )
     else:
         directory_generator(arg_lis[1:])
-
-
-

docs/overview/contributing/setting_up.rst

@@ -185,22 +185,26 @@ Windows
 #. Install `WSL 2 <https://docs.microsoft.com/en-us/windows/wsl/install>`_.
    For most, it will only require running the command :code:`wsl --install` in powershell admin mode.
    Visit the link if it doesn't.
-#. Get the latest Docker Image for Ivy by:
-
-   a. Running Docker desktop.
-   b. Opening cmd, and running the command: :code:`docker pull unifyai/ivy:latest`
-
 #. Install `Pycharm Professional Version <https://www.jetbrains.com/pycharm/>`_, make sure to only install the Professional version of PyCharm, not the Community version.
 #. Open pycharm with your cloned Ivy repository.
    Add the remote python interpreter by:
 
    a. Going to the settings -> Build, Execution, Deployment -> Docker
       Click the "+" on top left and it should add a docker connection.
    b. Going to settings -> project -> Python Interpreter
-   c. Clicking add interpreter (currently by clicking the ⚙ icon by the right side) which should open a new window.
-   d. Choosing "Docker" from the left panel.
-      Type python3 (with the number) in python interpreter path and press ok.
-   e. Opening "Edit Run/Debug configurations" dialog -> "Edit Configurations..." and making sure that "Working directory" is empty in case of getting the "Can't run process: the working directory '\ivy' is invalid, it needs to be an absolute path" error.
+   c. Clicking add interpreter (currently by clicking the ⚙ icon by the right side) which should open a new small drop down menu. Select "On Docker...".A         window will open which will have three steps.
+#. It will ask to create a new Docker target, at this step you have to select the following:
+
+   a. Docker image -> Docker
+   b. Image -> Pull
+   c. Image tag -> unifyai/ivy:latest
+   d. Select "Next"
+#. The image will start pulling. It will take a respectible amount of time to complete. Once you see "Introspection Completed" message, select "Next". 
+#. Another window will appear, at this step select the following:
+
+   a. In the left panel select "System Interpreter".
+   b. For Interpreter, select the default option which will be "/usr/bin/python3" the select "Create".   
+#. Opening "Edit Run/Debug configurations" dialog -> "Edit Configurations..." and making sure that "Working directory" is empty in case of getting the "Can't run process: the working directory '\ivy' is invalid, it needs to be an absolute path" error.
 
 Once these steps are finished, your interpreter should be set up correctly!
 If Docker's latest version causes error, try using an earlier version by visiting `Docker release note <https://docs.docker.com/desktop/release-notes/>`_.
@@ -316,8 +320,20 @@ Ubuntu
    d. Choosing "Docker" from the left panel.
       Type python3 (with the number) in python interpreter path and press ok.
 
-For questions, please reach out on `discord`_ in the `docker channel`_!
+**Docker Connection not Successfull**
 
+This is a common error which you might face. If you are not successfully able to connect docker with Pycharm(point 4a) and your docker is also running, the issue is that you are not able to use your docker socket. So, executing the below two commands should solve this.
+
+    .. code-block:: none
+        
+        sudo chmod a+rwx /var/run/docker.sock
+        
+    .. code-block:: none
+    
+        sudo chmod a+rwx /var/run/docker.pid  
+
+
+For questions, please reach out on `discord`_ in the `docker channel`_!
 
 **Video**
 

docs/partial_conf.py

@@ -47,10 +47,6 @@
 # Only generate docs for index.rst
 # That resolved a bug of autosummary generating docs for code-block examples
 # of autosummary
-autosummary_generate = ['index.rst']
+autosummary_generate = ["index.rst"]
 
-skippable_method_attributes = [
-    {
-        "__qualname__": "_wrap_function.<locals>.new_function"
-    }
-]
+skippable_method_attributes = [{"__qualname__": "_wrap_function.<locals>.new_function"}]

ivy/data_classes/container/base.py

@@ -3149,7 +3149,9 @@ def cont_map(
         """
         return_dict = self if inplace else dict()
         for key, value in self.items():
-            this_key_chain = key if key_chain == "" else (key_chain + "/" + key)
+            this_key_chain = (
+                key if key_chain == "" else (str(key_chain) + "/" + str(key))
+            )
             if isinstance(value, ivy.Container):
                 ret = value.cont_map(
                     func,

ivy/functional/backends/jax/experimental/layers.py

@@ -508,8 +508,7 @@ def interpolate(
         "area",
         "nearest_exact",
         "tf_area",
-        "bicubic_tensorflow"
-        "bicubic",
+        "bicubic_tensorflow" "bicubic",
         "mitchellcubic",
         "lanczos3",
         "lanczos5",
@@ -526,8 +525,7 @@ def interpolate(
     mode = (
         "nearest"
         if mode == "nearest-exact"
-        else "bicubic"
-        "bicubic"
+        else "bicubic" "bicubic"
         if mode == "bicubic_tensorflow"
         else mode
     )

ivy/functional/backends/jax/general.py

@@ -424,4 +424,4 @@ def vmap(
     func = ivy.output_to_native_arrays(func)
     return ivy.inputs_to_native_arrays(
         jax.vmap(func, in_axes=in_axes, out_axes=out_axes)
-    )
+    )

ivy/functional/backends/jax/searching.py

@@ -96,9 +96,10 @@ def where(
 # ----- #
 
 
-def argwhere(x: JaxArray,
-             /,
-             *,
-             out: Optional[JaxArray] = None,
+def argwhere(
+    x: JaxArray,
+    /,
+    *,
+    out: Optional[JaxArray] = None,
 ) -> JaxArray:
     return jnp.argwhere(x)

ivy/functional/backends/numpy/searching.py

@@ -101,9 +101,10 @@ def where(
 # ----- #
 
 
-def argwhere(x: np.ndarray,
-             /,
-             *,
-             out: Optional[np.ndarray] = None,
+def argwhere(
+    x: np.ndarray,
+    /,
+    *,
+    out: Optional[np.ndarray] = None,
 ) -> np.ndarray:
     return np.argwhere(x)

ivy/functional/backends/paddle/creation.py

@@ -146,7 +146,7 @@ def asarray(
         if dtype is None:
             dtype = ivy.default_dtype(item=obj)
         with ivy.ArrayMode(False):
-            return ivy.squeeze(paddle.to_tensor(obj,dtype=dtype),0)
+            return ivy.squeeze(paddle.to_tensor(obj, dtype=dtype), 0)
 
     else:
         dtype = ivy.as_native_dtype((ivy.default_dtype(dtype=dtype, item=obj)))

ivy/functional/backends/paddle/data_type.py

@@ -174,9 +174,7 @@ def broadcast_to(
         paddle.uint8,
         paddle.float16,
     ]:
-        return paddle.broadcast_to(x.cast("float32"), shape).cast(
-            x.dtype
-        )
+        return paddle.broadcast_to(x.cast("float32"), shape).cast(x.dtype)
     elif x.dtype in [paddle.complex64, paddle.complex128]:
         x_real = paddle.broadcast_to(x.real(), shape)
         x_imag = paddle.broadcast_to(x.imag(), shape)

ivy/functional/backends/paddle/experimental/elementwise.py

@@ -85,15 +85,15 @@ def float_power(
 
 @with_unsupported_device_and_dtypes(
     {"2.4.2 and below": {"cpu": ("uint16", "bfloat16")}}, backend_version
-) 
+)
 def exp2(
     x: Union[paddle.Tensor, float, list, tuple],
     /,
     *,
     out: Optional[paddle.Tensor] = None,
 ) -> paddle.Tensor:
     with ivy.ArrayMode(False):
-        return ivy.pow(2,x)
+        return ivy.pow(2, x)
 
 
 @with_unsupported_device_and_dtypes(
@@ -107,9 +107,9 @@ def copysign(
     out: Optional[paddle.Tensor] = None,
 ) -> paddle.Tensor:
     with ivy.ArrayMode(False):
-        x2 = ivy.where(ivy.equal(x2, 0),ivy.divide(1, x2),x2)
+        x2 = ivy.where(ivy.equal(x2, 0), ivy.divide(1, x2), x2)
         signs = ivy.sign(x2)
-        return ivy.multiply(ivy.abs(x1),signs)
+        return ivy.multiply(ivy.abs(x1), signs)
 
 
 def count_nonzero(
@@ -214,24 +214,32 @@ def nan_to_num(
     neginf: Optional[Union[float, int]] = None,
     out: Optional[paddle.Tensor] = None,
 ) -> paddle.Tensor:
-     with ivy.ArrayMode(False):
+    with ivy.ArrayMode(False):
         if ivy.is_int_dtype(x):
-            if posinf==None:
+            if posinf == None:
                 posinf = ivy.iinfo(x).max
-            if neginf==None:
-                neginf= ivy.iinfo(x).min
+            if neginf == None:
+                neginf = ivy.iinfo(x).min
         elif ivy.is_float_dtype(x) or ivy.is_complex_dtype(x):
-            if posinf==None:
+            if posinf == None:
                 posinf = ivy.finfo(x).max
-            if neginf==None:
-                neginf=ivy.finfo(x).min     
+            if neginf == None:
+                neginf = ivy.finfo(x).min
         ret = ivy.where(ivy.isnan(x), paddle.to_tensor(nan, dtype=x.dtype), x)
-        ret = ivy.where(ivy.logical_and(ivy.isinf(ret), ret > 0), paddle.to_tensor(posinf, dtype=x.dtype), ret)
-        ret = ivy.where(ivy.logical_and(ivy.isinf(ret), ret < 0), paddle.to_tensor(neginf, dtype=x.dtype), ret)
+        ret = ivy.where(
+            ivy.logical_and(ivy.isinf(ret), ret > 0),
+            paddle.to_tensor(posinf, dtype=x.dtype),
+            ret,
+        )
+        ret = ivy.where(
+            ivy.logical_and(ivy.isinf(ret), ret < 0),
+            paddle.to_tensor(neginf, dtype=x.dtype),
+            ret,
+        )
         if copy:
             return ret.clone()
         else:
-            x= ret
+            x = ret
             return x
 
 

ivy/functional/backends/paddle/manipulation.py

@@ -200,9 +200,7 @@ def squeeze(
             "tried to squeeze a zero-dimensional input by axis {}".format(axis)
         )
     if x.dtype in [paddle.int16, paddle.float16]:
-        return paddle.squeeze(x.cast("float32"), axis=axis).cast(
-            x.dtype
-        )
+        return paddle.squeeze(x.cast("float32"), axis=axis).cast(x.dtype)
     return paddle.squeeze(x, axis=axis)
 
 

ivy/functional/backends/paddle/random.py

@@ -154,7 +154,5 @@ def shuffle(
             shuffled_real = paddle.index_select(x.real(), indices)
             shuffled_imag = paddle.index_select(x.imag(), indices)
             return shuffled_real + 1j * shuffled_imag
-        return paddle.index_select(x.cast("float32"), indices).cast(
-            x.dtype
-        )
+        return paddle.index_select(x.cast("float32"), indices).cast(x.dtype)
     return paddle.index_select(x, indices)

ivy/functional/backends/torch/searching.py

@@ -111,9 +111,10 @@ def where(
 # ----- #
 
 
-def argwhere(x: torch.Tensor,
-             /,
-             *,
-             out: Optional[torch.Tensor] = None,
+def argwhere(
+    x: torch.Tensor,
+    /,
+    *,
+    out: Optional[torch.Tensor] = None,
 ) -> torch.Tensor:
     return torch.argwhere(x)

ivy/functional/frontends/jax/_src/__init__.py

@@ -1,4 +1,4 @@
 # flake8: noqa
 from . import numpy
 from . import array
-from . import api
+from . import api

ivy/functional/frontends/jax/_src/api.py

@@ -1,11 +1,12 @@
 import ivy
-from ivy.functional.frontends.jax.func_wrapper import to_ivy_arrays_and_back, outputs_to_native_arrays
-def vmap(fun,
-         in_axes=0,
-         out_axes=0,
-         axis_name=None,
-         axis_size=None,
-         spmd_axis_name=None
-         ):
+from ivy.functional.frontends.jax.func_wrapper import (
+    to_ivy_arrays_and_back,
+    outputs_to_native_arrays,
+)
+
+
+def vmap(
+    fun, in_axes=0, out_axes=0, axis_name=None, axis_size=None, spmd_axis_name=None
+):
     fun = outputs_to_native_arrays(fun)
-    return to_ivy_arrays_and_back(ivy.vmap(fun, in_axes=in_axes, out_axes=out_axes))
+    return to_ivy_arrays_and_back(ivy.vmap(fun, in_axes=in_axes, out_axes=out_axes))