[PYTHON API] Tensor.data property for low precisions + packing (#11131)

* rebase old branch with master

* Fix doc style

* fix test

* update tests

* Add missed param

* Rewrite docstring for tensor and refactor set_input_tensors test

* update python exclusives

* keep compatibility

* remove notes about slices

* fix code style

* Fix code style

docs/OV_Runtime_UG/Python_API_exclusives.md

@@ -28,12 +28,6 @@ Python API allows passing data as tensors. `Tensor` object holds a copy of the d
 
 @snippet docs/snippets/ov_python_exclusives.py tensor_shared_mode
 
-### Slices of array's memory
-
-One of the `Tensor` class constructors allows to share the slice of array's memory. When `shape` is specified in the constructor that has the numpy array as first argument, it triggers the special shared memory mode.
-
-@snippet docs/snippets/ov_python_exclusives.py tensor_slice_mode
-
 ## Running inference
 
 Python API supports extra calling methods to synchronous and asynchronous modes for inference.
@@ -76,6 +70,14 @@ The callback of `AsyncInferQueue` is uniform for every job. When executed, GIL i
 
 @snippet docs/snippets/ov_python_exclusives.py asyncinferqueue_set_callback
 
+### Working with u1, u4 and i4 element types
+
+Since openvino supports low precision element types there are few ways how to handle them in python.
+To create an input tensor with such element types you may need to pack your data in new numpy array which byte size matches original input size:
+@snippet docs/snippets/ov_python_exclusives.py packing_data
+
+To extract low precision values from tensor into numpy array you can use next helper:
+@snippet docs/snippets/ov_python_exclusives.py unpacking
 
 ### Releasing the GIL
 

docs/snippets/ov_python_exclusives.py

@@ -50,17 +50,6 @@ shared_tensor.data[0][2] = 0.6
 assert data_to_share[0][2] == 0.6
 #! [tensor_shared_mode]
 
-#! [tensor_slice_mode]
-data_to_share = np.ones(shape=(2,8))
-
-# Specify slice of memory and the shape
-shared_tensor = ov.Tensor(data_to_share[1][:] , shape=ov.Shape([8]))
-
-# Editing of the numpy array affects Tensor's data
-data_to_share[1][:] = 2
-assert np.array_equal(shared_tensor.data, data_to_share[1][:])
-#! [tensor_slice_mode]
-
 infer_request = compiled.create_infer_request()
 data = np.random.randint(-5, 3 + 1, size=(8))
 
@@ -132,6 +121,23 @@ infer_queue.wait_all()
 assert all(data_done)
 #! [asyncinferqueue_set_callback]
 
+unt8_data = np.ones([100])
+
+#! [packing_data]
+from openvino.helpers import pack_data
+
+packed_buffer = pack_data(unt8_data, ov.Type.u4)
+# Create tensor with shape in element types
+t = ov.Tensor(packed_buffer, [1, 128], ov.Type.u4)
+#! [packing_data]
+
+#! [unpacking]
+from openvino.helpers import unpack_data
+
+unpacked_data = unpack_data(t.data, t.element_type, t.shape)
+assert np.array_equal(unpacked_data , unt8_data)
+#! [unpacking]
+
 #! [releasing_gil]
 import openvino.runtime as ov
 import cv2 as cv