From b71eab255559f3d02f9f1a15d8904fe8de292ec1 Mon Sep 17 00:00:00 2001
From: Katarzyna Mitrus <katarzyna.mitrus@intel.com>
Date: Tue, 21 May 2024 16:20:35 +0200
Subject: [PATCH] [Spec][Internal][Op] Specification for RMS internal op
 (#24564)

### Details:
- Dev documentation for the existing
[internal::RMS](https://github.com/openvinotoolkit/openvino/blob/70142121c1c632b6d710026150b0e6deb20b0e68/src/common/transformations/include/ov_ops/rms.hpp)
op
(specified as is without any additional features, to be developed and
updated if needed)
- RMS was created initially as a [custom gpu
RMS](https://github.com/openvinotoolkit/openvino/blob/9fadb5ac64b76297c46b25983cd3b8fe1d038ff9/src/plugins/intel_gpu/include/intel_gpu/op/rms.hpp)
operation, and it's going to be moved to op::internal::RMS to be
available for common transformations
- Similar approach as it was done for AUGRUCell/AUGRUSeqence, so the
place for such documents has been already agreed and it's not a part of
the official web docs tree

### Tickets:
 - 134914, dicsussion 129027
---
 .../operation-specs/internal/rms.rts          | 84 +++++++++++++++++++
 1 file changed, 84 insertions(+)
 create mode 100644 docs/articles_en/documentation/openvino-ir-format/operation-sets/operation-specs/internal/rms.rts

diff --git a/docs/articles_en/documentation/openvino-ir-format/operation-sets/operation-specs/internal/rms.rts b/docs/articles_en/documentation/openvino-ir-format/operation-sets/operation-specs/internal/rms.rts
new file mode 100644
index 00000000000..9b445b1f64b
--- /dev/null
+++ b/docs/articles_en/documentation/openvino-ir-format/operation-sets/operation-specs/internal/rms.rts
@@ -0,0 +1,84 @@
+.. {#openvino_docs_ops_internal_RMS}
+
+RMS
+===
+
+
+.. meta::
+  :description: Learn about RMS a normalization operation.
+
+**Versioned name**: *RMS*
+
+**Category**: *Normalization*
+
+**Short description**: Calculates Root Mean Square (RMS) normalization of the input tensor.
+
+**Detailed description**
+
+*RMS* operation performs Root Mean Square (RMS) normalization on a given input ``data`` along the last dimension of the input.
+`Reference <https://arxiv.org/abs/1910.07467>`__.
+
+
+.. code-block:: py
+
+    (x / Sqrt(ReduceMean(x^2, -1) + eps)) * scale
+
+
+**Attributes**
+
+* *epsilon*
+
+  * **Description**: A very small value added to the variance for numerical stability. Ensures that division by zero does not occur for any normalized element.
+  * **Range of values**: a positive floating-point number
+  * **Type**: ``float``
+  * **Required**: *yes*
+
+* *output_type*
+
+  * **Description**: The precision for output type conversion, after scaling. It's used for output type compression to f16.
+  * **Range of values**: Supported floating point type: "f16", "undefined"
+  * **Type**: ``string``
+  * **Default value**: "undefined" (means that output type is set to the same as the input type)
+  * **Required**: *no*
+
+
+**Inputs**
+
+* **1**: ``data`` - Input data to be normalized. A tensor of type *T* and arbitrary shape. **Required.**
+
+* **2**: ``scale`` - A tensor of type *T* containing the scale values for . The shape should be broadcastable to the shape of ``data`` tensor. **Required.**
+
+
+**Outputs**
+
+* **1**: Output tensor of the same shape as the ``data`` input tensor and type specified by *output_type* attribute.
+
+**Types**
+
+* *T*: any floating point type.
+
+**Example**
+
+.. code-block:: xml
+   :force:
+
+   <layer ... type="RMS"> <!-- normalization always over the last dimension [-1] -->
+       <data eps="1e-6"/>
+       <input>
+           <port id="0">
+               <dim>12</dim>
+               <dim>25</dim>
+               <dim>512</dim>
+           </port>
+           <port id="1">
+               <dim>512</dim>
+           </port>
+       </input>
+       <output>
+           <port id="2">
+               <dim>12</dim>
+               <dim>25</dim>
+               <dim>512</dim>
+           </port>
+       </output>
+   </layer>