Merge pull request #154 from uw-ssec/147-fix-anshul-changes-for-workshop-37

fix: Reflecting feedback for 3_7 Workshop

Author: atambay37

AI_Postdoc_Workshop/module1/setup.md

@@ -15,7 +15,8 @@ Codespaces.
 ## 🐱‍💻 GitHub Codespaces
 
 ```{note}
-Indefinite Access, but Limited Resources
+Indefinite Access, but Limited Resources.
+If your Codespace gives you a low disk space error, please refresh the tab.
 ```
 
 This tutorial is available to run within

AI_Postdoc_Workshop/module2/1-retrieval-augmented-text-generation.ipynb

@@ -18,26 +18,6 @@
     "Throughout this notebook we will be exploring RAG, what it is, how it works, and why it's so exciting."
    ]
   },
-  {
-   "cell_type": "markdown",
-   "id": "eea1041c",
-   "metadata": {},
-   "source": [
-    "## Why RAG?\n",
-    "\n",
-    "Although trained on large datasets, stale data can severely limit LLMs. It faces several challenges:\n",
-    "\n",
-    "1. The models are trained on internet content, so they might not generate relevant output when prompted for information that is not publicly available on the internet.\n",
-    "\n",
-    "2. The models are trained up to a certain date, they might not generate relevant output when prompted for content and information that has happened after the training completion date of the model.\n",
-    "\n",
-    "3. The models are trained to be more generalized. This means that they can only produce generic outputs and might not perform as expected when prompted for specific deep-dive concepts related to a particular topic.\n",
-    "\n",
-    "One way to dynamically integrate relevant external information is retrieval-augmented generation (RAG), which can help improve the reliability of LLM outputs.\n",
-    "\n",
-    "Going back to our original question of how this can be utilized in our own work or organization on [module 1, section 3](../module1/3-domain-specific-question-answering.ipynb) of this workshop. RAG Framework can really be useful in the scenario where there may be a set of documents, GitHub repositories, research papers, and domain-specific knowledge bases that you might want to search through quickly."
-   ]
-  },
   {
    "cell_type": "markdown",
    "id": "7",
@@ -490,7 +470,7 @@
     "        documents.extend(loader.load())\n",
     "\n",
     "for each in documents:\n",
-    "    #     # print(each.page_content) # Uncomment this line to see the individual page_content\n",
+    "    # print(each.page_content) # Uncomment this line to see the individual page_content\n",
     "    print(each.metadata)"
    ]
   },

AI_Postdoc_Workshop/module2/2-olmo-chat-rag.ipynb

@@ -22,30 +22,10 @@
    "id": "53cb355b",
    "metadata": {},
    "source": [
-    "For this demo we will use Panel, a powerful open source Python library to create interactive dashboards in Jupyter notebooks or standalone apps.\n",
+    "For this demo we will use [Panel](https://panel.holoviz.org/), a powerful open source Python library to create interactive dashboards in Jupyter notebooks or standalone apps.\n",
     " "
    ]
   },
-  {
-   "cell_type": "markdown",
-   "id": "9cf1c043",
-   "metadata": {},
-   "source": [
-    "## What is Panel?\n",
-    "\n",
-    "[Panel](https://panel.holoviz.org/) is a part of the [HoloViz](https://holoviz.org/) ecosystem, which is a set of tools that make it easier to work with large datasets and create interactive visualizations. This ecosystem consists of the following tools beyond Panel:\n",
-    "\n",
-    "- [hvPlot](https://hvplot.holoviz.org/) to quickly generate interactive plots from your data\n",
-    "- [HoloViews](https://holoviews.org/) to help you make all of your data instantly visualizable\n",
-    "- [GeoViews](https://geoviews.org/) to extend HoloViews for geographic data\n",
-    "- [Datashader](https://datashader.org/) for rendering even the largest datasets\n",
-    "- [Lumen](https://lumen.holoviz.org/) to build data-driven dashboards from a simple YAML specification\n",
-    "- [Param](https://param.holoviz.org/) to create declarative user-configurable objects\n",
-    "- [Colorcet](https://colorcet.holoviz.org/) for perceptually uniform colormaps\n",
-    "\n",
-    "The nice thing about this ecosystem is that all of these tools are designed to work together as part of the [Pandata scalable open-source analysis stack](https://github.com/panstacks/pandata?tab=readme-ov-file), so you can easily integrate them as needed."
-   ]
-  },
   {
    "cell_type": "markdown",
    "id": "9f31fff8",