rhai-1448 copy edits and image updates

Author: MelissaFlinn

assets/usecase/knowledge-tuning/Data Preprocessing.png

@@ -2,7 +2,7 @@
 # The environment variables required are listed below on a step-by-step basis.
 # Please set them in your own .env file before running the notebooks.
 
-# Data Preprocessing Configuration
+# Data Processing Configuration
 
 
 # Knowledge Tuning Configuration

assets/usecase/knowledge-tuning/Data Processing.png

@@ -20,23 +20,23 @@ To use the Knowledge Tuning example, follow these steps to set up your working e
 3. [Create a workbench](#create-a-workbench)
 4. [Clone the example Git repository](#clone-the-example-git-repository)
 
-For this example, you create an OpenShift project and a workbench. From the workench, you launch an JupyterLab integrated development environment. In JupyterLab, you run the provided notebooks that guide you through the example data processing, knowledge generation and mixing, model training, and evaluation workflow.
+For this example, you create an OpenShift project and a workbench. From the workench, you launch a JupyterLab integrated development environment. In JupyterLab, you run the provided notebooks that guide you through the Knowledge Tuning example end-to-end workflow.
 
 ## Configure resources on the OpenShift cluster
 
 The notebooks provided in the Knowledge Tuning example require the following resources:
 
-- **GPUs:** GPUs are optional for the preprocessing and mixing steps. For the model training step, fine-tuning large models requires at least one NVIDIA A100/40GB or similar. Training smaller student models requires 8–16 GB GPU.
+- **GPUs:** GPUs are optional for the data processing and mixing steps. For the model training step, fine-tuning large models requires at least one NVIDIA A100/40GB or similar. Training smaller student models requires 8–16 GB GPU.
 
 - **Persistent Volumes:** Attach a persistent volume with at least 200 GB.
 
-To run all of the notebooks in the Knowledge Training example, you must ensure that the following resources are configured on the OpenShift cluster:
+To run all of the notebooks in the Knowledge Tuning example, you must ensure that the following resources are configured on the OpenShift cluster:
 
 - Your Red Hat OpenShift cluster administrator must configure the following resources:
 
    - A persistent volume with a storage capacity of 200 GB or greater.
 
-   - Installed and enabled a NVIDIA GPU as described in [Enabling accelerators]https://docs.redhat.com/en/documentation/red_hat_openshift_ai_self-managed/3.0/html/working_with_accelerators/enabling-accelerators_accelerators.
+   - An NVIDIA GPU that is installed and enabled, as described in [Enabling accelerators](https://docs.redhat.com/en/documentation/red_hat_openshift_ai_self-managed/3.0/html/working_with_accelerators/enabling-accelerators_accelerators).
 
 - An OpenShift AI administrator must create a hardware profile that allocates the following resources, as described in [Creating a hardware profile](https://docs.redhat.com/en/documentation/red_hat_openshift_ai_self-managed/3.0/html/working_with_accelerators/working-with-hardware-profiles_accelerators#creating-a-hardware-profile_accelerators).
 
@@ -74,7 +74,7 @@ You can see your project's initial state.
 
 ## Create a workbench
 
-A workbench is an instance of your development and experimentation environment. When you create a workbench, you select a workbench image that has the tools, libraries, and an integrated development environment, such as JupyetrLab, that you need for developing AI models.
+A workbench is an instance of your development and experimentation environment. When you create a workbench, you select a workbench image that has the tools, libraries, and an integrated development environment, such as JupyterLab, that you need for developing AI models.
 
 The JupyterLab environment is a web-based environment, but everything you do inside it happens on Red Hat OpenShift AI and is powered by the OpenShift cluster. This means that, without having to install and maintain anything on your own computer, and without using valuable local resources such as CPU, GPU and RAM, you can conduct your work in this powerful and stable managed environment.
 
@@ -90,9 +90,9 @@ The JupyterLab environment is a web-based environment, but everything you do ins
 
 3. Fill out the name and description.
 
-   Red Hat OpenShift AI provides several supported workbench images. In the **Workbench image** section, you can select one of the default images or a custom image that an administrator has set up for you. The **Jupyter | Minimal | CUDA | Python 3.12** has the libraries needed for this example.
+   Red Hat OpenShift AI provides several supported workbench images.
 
-4. Select the latest **Jupyter | Minimal | CUDA | Python 3.12** image.
+4. For this example, select the **Jupyter | Minimal | CUDA | Python 3.12** image.
 
 5. Select the latest version: **2025.2**.
 
@@ -138,7 +138,7 @@ The files for each step in the Knowledge Tuning example end-to-end workflow are
 
 In the file browser, view the notebooks that you cloned from Git.
 
-Congratulations! Your workbench is configured and ready for the knowledge training example. The notebooks and supporting README files provide details about each step in the knowledge training workflow.
+Congratulations! Your workbench is configured and ready for the Knowledge Tuning example. The notebooks and supporting README files provide details about each step in the Knowledge Tuning workflow.
 
 ## Next step
 

assets/usecase/knowledge-tuning/Detailed Flow.png

@@ -14,7 +14,7 @@
 
 ## Evaluate the base model
 
-Before you train a base model, evaluate it's preliminary performance. Later, after you train the model with your data, you can objectively compare how effective the knowledge tuning tasks are for customizing the model.
+In the Teacher-Student model paradigm, the student model is also known as the base model. Before you train a base model, evaluate it's preliminary performance. Later, after you train the model with your data, you can objectively compare how effective the knowledge tuning tasks are for customizing the model.
 
 ### Prerequisites
 

…nowledge-tuning/Knowledge Genertaion.png …nowledge-tuning/Knowledge Generation.pngassets/usecase/knowledge-tuning/Knowledge Genertaion.png renamed to assets/usecase/knowledge-tuning/Knowledge Generation.png assets/usecase/knowledge-tuning/Knowledge Genertaion.png renamed to assets/usecase/knowledge-tuning/Knowledge Generation.png

@@ -9,7 +9,7 @@
     "\n",
     "\n",
     "## Overview\n",
-    "Evaluation is a crucial step in the knowldedge tuning workflow. It allows you to measure the performance and generalization ability of your model. In this notebook, you systematically evaluate the base model by using appropriate metrics and validation datasets before you fine tune the model on the example Bank of Montreal (BMO) data.\n",
+    "Evaluation is a crucial step in the Knowldedge Tuning workflow. It allows you to measure the performance and generalization ability of your model. In this notebook, you systematically evaluate the base model by using appropriate metrics and validation datasets before you fine tune the model on the example Bank of Montreal (BMO) data.\n",
     "\n",
     "Throughout this notebook, you use visualizations and quantitative metrics to analyze performance of the base model."
    ]