Merge pull request #8 from maytaldahan/main

Added section headers and fixed minor navigation consistencies

Author: ajamthetacc

tutorials/Fine_Tuning_Vision/01-intro-fine-tuning.md

@@ -1,4 +1,4 @@
-## Fine-Tuning Vision Models Using Ultralytics and Tapis
+## Section 8: Fine-Tuning Vision Models Using Ultralytics and Tapis
 
 This Tapis application allows users to fine-tune Vision models such as YOLO using Ultralytics running in a Singularity container on Vista. It is designed to run on High-Performance Computing (HPC) systems via Tapis, leveraging GPU acceleration for training tasks.
 
@@ -93,12 +93,12 @@ These parameters define the hardware footprint requested from the Slurm schedule
 
 Follow these steps to submit the job using the Tapis UI:
 
-### Step 1: Initiate Submission
+### Step 8.1: Initiate Submission
 Navigate to the **Apps** list and select the `ultralytics-fine-tune` app. Click the button to initiate a JSON-based submission.
 
 ![Step 1 - Selecting the Ultralytics App in UI](/tutorials/images/Step1-Ultralyticsapp.png)
 
-### Step 2: Edit the JSON Payload
+### Step 8.2: Edit the JSON Payload
 Paste the job JSON provided below into the editor and click `Submit`.
 
 <div style="max-height:400px; overflow:auto; border:1px solid #ddd; padding:10px;">
@@ -148,21 +148,21 @@ These are specific flags passed to the Slurm scheduler on the Vista system:
 
 ![Step 2 - Pasting the Job JSON into the Editor and Submit](/tutorials/images/Fine-tune-job.png)
 
-### Step 3: Monitor Job Progress
+### Step 8.3: Monitor Job Progress
 After clicking **Submit**, navigate to the **Jobs** tab. You can monitor the status as it moves from `PENDING` to `RUNNING` and finally `FINISHED`.
 
 ![Step 3 - Monitoring Job Status in the Dashboard](/tutorials/images/fine-tune-job-running.png) 
 Since we are running 100 Epochs to fine-tune, this will take around 10-12 minutes to finish.
 
 
-### Step 4: Job Output and Results location
+### Step 8.4: Job Output and Results location
  Once the job finishes, you should see output similar to the image below.
  ![Step 4 - After Job completes](/tutorials/images/fine-tune-job-completion.png) 
 
  You should see a train directory with weights. Inside the `weights` directory, you can access the `best.pt model`. This file will be accessible in Jupyter Notebook.
 
 
-### Step 5: Finding best.pt file from Jupyter 
+### Step 8.5: Finding best.pt file from Jupyter 
 
 Inside your work directory, you should see a `vista` folder. In there you can find the directory named with your `jobUUID`. In the above image the job uuid is highlighted. In the job directory you will find the outputs of the job archived with the same train directory containing the best.pt file.
 `$WORK -> vista -> jobUUID -> train -> weights -> best.pt`

tutorials/Fine_Tuning_Vision/02-evaluation.md

@@ -1,14 +1,22 @@
-## Evaluating the Fine-tuned Model 
+## Section 9: Evaluating the Fine-tuned Model 
 
 To try this part of the tutorial, we will use the fine-tuned model `best.pt` created by the ultralytics-fine-tune job.
 
-Your fine-tuned model will be accessible through Jupyter in this location
+### Step 9.1: Locate your Fine-tuned model
+
+Once your job is complete, your fine-tuned model will be accessible through Jupyter in this location
 `$WORK -> vista -> jobUUID -> train -> weights -> best.pt`
 
+Please locate your file location. 
+
+### Step 9.2: Point your Jupyter notebook code at your Fine-tuned model location
+
 You should already have the generated code that we tested accuracy on the baseline model.
 
 Now, just change the path of your model to the fine-tuned model path and re-run the code to test the accuracy.
 
+### Step 9.3: Re-run the notebook
+
 When the code runs on the test data, you should see improved accuracy
 
 ```

tutorials/Intro_Tapis/01-intro-to-tapis.md

@@ -1,4 +1,4 @@
-## AI, The National CI Ecosystem, and the Tapis API Platform
+## Section 1: AI, The National CI Ecosystem, and the Tapis API Platform
 
 [Lecture Slides](https://docs.google.com/presentation/d/1BVLnUbyiWjsaS33zMshW3TXqtfvv6zGklaCNBeX7Go0/edit?slide=id.g3cd6a51b6a2_0_21#slide=id.g3cd6a51b6a2_0_21)
 ​

tutorials/Intro_Tapis/02-initial-tapis-ui.md

@@ -6,9 +6,9 @@ img {
 </style>
 
 
-# Initial Steps with TapisUI
+# Section 2: Initial Steps with TapisUI
 
-## Logging In
+## Step 2.1: Logging In
 
 Login to [https://public.tapis.io](https://public.tapis.io). Visit the site and press the "Proceed to login" area.
 
@@ -25,7 +25,7 @@ You will then be prompted to log-in with username, password, and MFA token.
 <img src="/tutorials/assets/demo/login-sso.png" style="max-width:50%;">
 
 
-### Inspect JWT
+###  Step 2.2: Inspect JWT
 
 Once logged in, you can inspect your JSON Web Token (JWT). The JWT is the authentication token that Tapis uses to verify your identity across all API calls. It contains encoded claims about your user session — including your username, tenant, and token expiration. You can view your token in TapisUI to understand how Tapis manages identity and access.
 
@@ -37,7 +37,7 @@ A Tapis **system** represents a storage or execution resource — such as a VM o
 
 In this tutorial, instead of creating systems from scratch, we will use a **public system** that has been pre-registered for you on TACC's Vista cluster.
 
-### Navigate the Systems List
+### Step 2.3: Navigate the Systems List
 
 When you log in to TapisUI and click on **Systems** from the left-hand menu, you should see one public system available to you. This system has been pre-registered for the tutorial.
 
@@ -46,7 +46,7 @@ When you log in to TapisUI and click on **Systems** from the left-hand menu, you
 {: .note}
 > ⚠️ Note: The system is visible but you are **not yet authenticated** to access files or run jobs on it. You must first add TMS credentials.
 
-### Add TMS Credentials for the NAIRR Vista Public System
+### Step 2.4: Add TMS Credentials for the NAIRR Vista Public System
 
 To access the public system running on Vista, you need to add **TMS (Trust Management System)** credentials. TMS credentials are temporary credentials generated by the TMS service and stored in the Tapis Security Kernel (SK). They allow Tapis services and applications to securely access external resources on your behalf. Instead of storing permanent usernames or passwords, Tapis retrieves the required credentials from TMS at runtime — improving security by keeping sensitive information encrypted and centrally managed.
 
@@ -56,7 +56,7 @@ Click on **Authenticate with TMS Keys** on the Vista system.
 
 Once your TMS credentials are added, your system is ready to use.
 
-### Use "Go to Home" to Get to Files
+### Step 2.5: Use "Go to Home" to Get to Files
 
 After authenticating, you can verify your system access by clicking the **View Files** button, which will take you to the file listing on the Vista system.
 
@@ -68,13 +68,13 @@ After authenticating, you can verify your system access by clicking the **View F
 
 The Tapis Files service provides a unified interface for managing files across any registered Tapis system. Once you have authenticated to a system, you can browse, upload, download, and manage files directly from TapisUI.
 
-### Navigate the Files in Vista System
+### Step 2.6: Navigate the Files in Vista System
 
 From the left-hand menu, click on the **Files** tab and select the Vista system. You should see a listing of files on the system. You can navigate directories, view file details, and open files directly.
 
 ![System View Files](/tutorials/images/System_View_Files.png)
 
-### Test Uploading a File
+### Step 2.7: Test Uploading a File
 
 Try uploading a file to the Vista system using the upload functionality in the Files tab. This confirms that your credentials are working and that you have write access to the system.
 
@@ -86,11 +86,11 @@ A Tapis **application** represents all the information required to run a Tapis j
 
 In this tutorial, instead of creating an application from scratch, we will use the **FlexServ** public application that has already been registered for you.
 
-### Navigate to the Tapis Apps List
+### Step 2.8: Navigate to the Tapis Apps List
 
 Click on **Apps** from the left-hand menu in TapisUI. You will see a list of available applications. Toggle between "My Apps" and "Public Apps" to see the full list.
 
-### Find the FlexServ Public App
+### Step 2.9: Find the FlexServ Public App
 
 Look for the application **FlexServ-vista-nairr version 1.4.0** in the public apps list. FlexServ is a TACC-owned inference server for running AI models on HPC systems. It has been pre-registered as a public app for all users to submit jobs.
 

tutorials/Jupyter-Notebook/intro-to-jupyter.md

@@ -1,21 +1,13 @@
-Introduction to the Jupyter Notebook Environment
-===================================================
-
+## Section 6: Launching the Jupyter Notebook Environment
 
 Jupyter is an open source project that provides a webapp interface for writing code and documents. Throughout this tutorial, we will be using a Jupyter Notebook environment for making Tapis User Requests. 
 
-### Download Notebook
-
-Dowload the Jupyter Notebook by clicking the link below:
-[Download Notebook](sentiment_analysis.ipynb)
-
-### Starting up your Jupyter Notebook Environment
+### Step 6.1: Starting up your Jupyter Notebook Environment
 
 For this tutorial, we will use [TACC's Public JupyterHub](https://public.jupyter.tacc.cloud) 
 You may login with your TACC accounts.
 
-
-### Navigating to the $WORK File System
+### Step 6.2: Navigating to the $WORK File System
 
 On successful login, ensure that you have access to a folder, `work`, within the Jupyter file system. 
 

tutorials/Tapis_FlexServ/01-lecture.md

@@ -1,4 +1,4 @@
-## Classes of AI Models, Commercial Registries and Flexserv
+## Section 3: Classes of AI Models, Commercial Registries and Flexserv
 
 [Lecture Slides](https://docs.google.com/presentation/d/1BVLnUbyiWjsaS33zMshW3TXqtfvv6zGklaCNBeX7Go0/edit?slide=id.g3cd6a51b6a2_0_26#slide=id.g3cd6a51b6a2_0_26)
 

tutorials/Tapis_FlexServ/01b-running-flexserv.md

@@ -1,4 +1,6 @@
-## Adding TMS Credentials on the Vista system.
+## Section 4:  Executing Large Models on Vista with Tapis and Flexserv
+
+### Step 4.1: Adding TMS Credentials on the Vista system.
 
 To access the public system running on Vista, first you will need to add TMS credentials on the system. TMS (Trust Management System) credentials on Tapis systems are temporary credentials generated by the TMS System and stored in the Tapis Security Kernel (SK) that allow services or applications to securely access external resources on behalf of a user. Instead of storing permanent usernames or passwords, Tapis retrieves the required credentials from the TMS service at runtime. This approach improves security by keeping sensitive information encrypted and centrally managed while enabling automated job execution on Tapis systems.
 
@@ -14,7 +16,7 @@ You can now view files on Vista, by clicking on the `View Files ` button.
 ![View Files](/tutorials/images/ViewFiles.png)
 
 
-## Running FlexServ Application on Vista
+### Step 4.2:  Running FlexServ Application on Vista
 
 The following app runs the FlexServ on TACC's Vista System. For the purposes of this tutorial, the application has already been registered with Tapis and is available as a public app for all users to submit jobs.
 
@@ -242,7 +244,7 @@ The following app runs the FlexServ on TACC's Vista System. For the purposes of
 You should see the Flex Server application already registered in your Tapis UI: **FlexServ-vista-nairr version 1.4.0**
 ![FlexServ Application](/tutorials/images/Flexserv_app.png)
 
-## Submit FlexServ Job using TAPIS UI
+### Step 4.3: Submit FlexServ Job using TAPIS UI
 
 **1. Initiate Submission**
 
@@ -370,13 +372,12 @@ Step 5b) Once the `tapisjob.out` opens, look at the ACCESS INFORMATION Section t
 ![Step 5b: Vista_url_token](/tutorials/images/GetFlexServerPortToken.png)
 
 
-### What's next?
+### Step 4.4: Using the FlexServ UI 
 
 If you made it this far, you are successfully running FlexServ, you can explore the FlexServ UI next and try to send your first chat. 
 
-### Using the FlexServ UI 
-Go to the URL hhtps://vista.tacc.utexas.edu:`Port number from above` and enter the TAP token from the tapisjob.out as shown in figure below.
+Go to the URL `htps://vista.tacc.utexas.edu`:`Port number from above` and enter the TAP token from the tapisjob.out as shown in figure below.
 
 ![FlexServ UI](/tutorials/images/FlexservUI.png)
 
-Next, we will use the flexserv to generate a code for small animal detection using vision models and evaluate the model performance.
+Next, we will use the flexserv to generate a code for small animal detection using vision models and evaluate the model performance.

tutorials/Tapis_FlexServ/01c-code-gen-flexserv.md

@@ -1,20 +1,23 @@
-## Prompt Engineering and Generating Image Detection Code
+## Section 7: Prompt Engineering and Generating Image Detection Code
 
 [Lecture Slides](https://docs.google.com/presentation/d/1BVLnUbyiWjsaS33zMshW3TXqtfvv6zGklaCNBeX7Go0/edit?slide=id.g3cdba15a02d_6_191#slide=id.g3cdba15a02d_6_191)
 
-
 ### FlexServ API Prompt: YOLO Evaluation Script Generator
 
 #### Task Summary:
 To test the capabilities of the FlexServ inference server, we can provide a complex prompt to the Responses API. This prompt asks the AI to generate a complete Python evaluation script that performs Animal detection on the images from the LILA BC Small Animal dataset. This is a large camera-trap image dataset used for wildlife monitoring and ecological research. It contains millions of images captured by automated cameras, including small mammals and many blank triggers, along with annotations describing the detected species. For training object detection models such as YOLO, the dataset can be downloaded in YOLO format, where each image has a corresponding .txt label file containing bounding-box coordinates in the form <class_id> <x_center> <y_center> <width> <height>.
 
 ---
 
-#### On FlexServ UI
+### Exploring the FlexServ UI 
+
+### Step 7.1: Refresh Model Pool
 
 - Refresh the Model pool so you can see public and private models available for you to run.
 ![Model Pool](/tutorials/images/Model_pool.png)
 
+### Step 7.2: Update the Responses API and Parameters
+
 - Copy and paste the following prompt into the FlexServ UI in the `Responses API`, `Input(Markdown)` section, shown in the image below.
 
 
@@ -72,21 +75,24 @@ After the code, briefly explain how the program works in plain English.
     -   `Qwen/Qwen2.5-Coder32B-Instruct-61.0 GB - Text Generation`
 -  Make sure the `Streams` is checked. 
 -  Uncheck `Multi-turn conversation`
+
+  ### Step 7.3: Run the Responses API
+  
 -  Click `Run`. Within a few minutes, you should see the code generation start in the blue box in the Responses API. Wait for it to complete. After completion, you should see output similar to the image below.
 
 ![Code](/tutorials/images/Code.png)
 
 
-### On Jupyter:
+### Step 7.4: Running Code Detection On Jupyter
 
-Go to the notebook Code-Detection on your Jupyter path. 
+Go to the Jupyter notebook Code-Detection on your Jupyter path. 
 `ai-tutorial-2026 -> notebooks -> Code-Detection.ipynb`
 
 Copy the generated code from FlexServ UI in a new cell below the cell titled `Put your generated code here`. 
 
-Update the variable `DATASET_ROOT` to path `/home/jovyan/work/vista/ai-tutorial-2026/datasets/AnimalEcology.v4i.yolov11` in your generated code
+Update the variable `DATASET_ROOT` to path `/home/jovyan/ai-tutorial-2026/datasets/AnimalEcology.v4i.yolov11` in your generated code
 
-Also update the model path to `/home/jovyan/work/vista/ai-tutorial-2026/models/yolov9t_ep200_bs32_lr0.005_baa22147.pt`
+Also update the model path to `/home/jovyan/ai-tutorial-2026/models/yolov9t_ep200_bs32_lr0.005_baa22147.pt`
 
 Now run the code. On successful run, you should see output similar to below 
 

tutorials/Vision_Models/01-intro-vision-models.md

@@ -1,7 +1,5 @@
-## Computer Vision Models in Scientific Applications 
+## Section 5: Computer Vision Models in Scientific Applications 
 
 [Lecture Slides](https://docs.google.com/presentation/d/1BVLnUbyiWjsaS33zMshW3TXqtfvv6zGklaCNBeX7Go0/edit?slide=id.g3cd6a51b6a2_0_31#slide=id.g3cd6a51b6a2_0_31)
 
 
-### Summary of Topics 
-* 

tutorials/_config.yml

@@ -5,14 +5,14 @@ permalink: pretty
 baseurl: "/tutorials"
 nav:
 - AI/ML, The National CI Ecosysystem, and Tapis:
-  - AI, CI and Tapis: ./Intro_Tapis/01-intro-to-tapis
-  - Initial Steps with TapisUI: ./Intro_Tapis/02-initial-tapis-ui
+  - (Lecture) AI, CI and Tapis: ./Intro_Tapis/01-intro-to-tapis
+  - (Hands-on) Initial Steps with TapisUI: ./Intro_Tapis/02-initial-tapis-ui
 - AI Models and Fleserv:
-  - Classes of AI Models and the AI Ecosysystem: ./Tapis_FlexServ/01-lecture
-  - Hands-on Flexserv TapisUI Tutorial: ./Tapis_FlexServ/01b-running-flexserv
+  - (Lecture) Classes of AI Models, 3rd Party Model Registries, Flexserv: ./Tapis_FlexServ/01-lecture
+  - (Hands-on) Executing Large Models on Vista Via Tapis and Flexserv: ./Tapis_FlexServ/01b-running-flexserv
 - Vision Models, Fine-tuning and Jupyter Notebooks:
   - Vision Models in Science: ./Vision_Models/01-intro-vision-models
-  - Intro to Jupyter: ./Jupyter-Notebook/intro-to-jupyter 
+  - (Hands-on) Launching Juptyer Notebook: ./Jupyter-Notebook/intro-to-jupyter 
   - Prompt Engineering and Detection Code: ./Tapis_FlexServ/01c-code-gen-flexserv
   - Fine-tuning with Ultralytics: ./Fine_Tuning_Vision/01-intro-fine-tuning
 - Evaluating Fine-tuned Models: