[DOCS] nncf changes pass 3 recommend (openvinotoolkit#26873)

Author: kblaszczak-intel

docs/articles_en/openvino-workflow/model-optimization-guide/weight-compression.rst

@@ -1,5 +1,5 @@
-Weight Compression
-==================
+LLM Weight Compression
+=========================
 
 .. toctree::
    :maxdepth: 1
@@ -187,7 +187,7 @@ trade-offs after optimization:
       ratio=0.9,
     )
 
-* ``scale_estimation`` - boolean parameter that enables more accurate estimation of 
+* ``scale_estimation`` - boolean parameter that enables more accurate estimation of
   quantization scales. Especially helpful when the weights of all layers are quantized to
   4 bits. Requires dataset.
 

docs/articles_en/openvino-workflow/model-optimization.rst

@@ -22,7 +22,7 @@ It is a `set of compression algorithms <https://github.com/openvinotoolkit/nncf/
 organized as a Python package, that make your models smaller and faster. Note that NNCF
 is **not part of the OpenVINO package**, so it needs to be installed separately. It supports
 models in **PyTorch**, **TensorFlow** , **ONNX**, and **OpenVINO IR** formats, offering
-the following optimizations:
+the following main optimizations:
 
 .. image:: ../assets/images/WHAT_TO_USE.svg
 
@@ -42,20 +42,41 @@ the following optimizations:
        as Quantization-aware Training. This kind of optimization requires the use of the model's
        original framework, for NNCF, it is either PyTorch or TensorFlow.
 
-A common approach is to perform post-training quantization first, as it is the easiest option.
-If the result proves unsatisfactory, quantization-aware training will give you higher accuracy
-with the same level of performance boost. For the most performant product, adding filter pruning
-will further streamline the model.
 
-To learn about the full scope of the framework, its installation, and technical details, visit
-both `the NNCF repository <https://github.com/openvinotoolkit/nncf?tab=readme-ov-file>`__ and
-`NNCF API documentation <https://openvinotoolkit.github.io/nncf/autoapi/nncf/>`__.
+
+Recommended workflows
+##########################
+
+* A common approach for most cases is to:
+
+  1. Perform post-training quantization first, as it is the easiest option.
+  2. For even better results, combine post-training quantization with filter pruning.
+  3. If the accuracy drop is unacceptable, use quantization-aware training instead. It will give
+     you the same level of performance boost, with a smaller impact on accuracy.
+
+* **Weight compression** works **only with LLMs**. Do not try to use it with other models.
+* For **visual-multimodal** use cases, the encoder / decoder split approach may be recommended.
+
+
+
+
+
 
 
 
 .. image:: ../assets/images/DEVELOPMENT_FLOW_V3_crunch.svg
 
 
+
+Installation and usage
+###########################
+
+To learn about the full scope of the framework, its installation, and technical details, visit
+both `the NNCF repository <https://github.com/openvinotoolkit/nncf?tab=readme-ov-file>`__ and
+`NNCF API documentation <https://openvinotoolkit.github.io/nncf/autoapi/nncf/>`__.
+
+
+
 .. tab-set::
 
    .. tab-item:: Installation