Configurable LLM

.gitignore

@@ -167,3 +167,5 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+
+.streamlit/

.vscode/launch.json

@@ -0,0 +1,19 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Debug Streamlit",
+            "type": "debugpy",
+            "request": "launch",
+            "module": "streamlit",
+            "args": [
+                "run",
+                "${file}"
+            ],
+            "justMyCode": false,
+        }
+    ]
+}

README.md

@@ -6,24 +6,6 @@
 
 #### Generate publication highlights using AI
 
-### Setting Up OpenAI API Key
-
-To use the OpenAI API, you need to set up an API key and set it as an environment variable.
-
-1. Obtain your OpenAI API key from the [OpenAI website](https://platform.openai.com/api-keys).
-
-2. Set the API key as an environment variable:
-    - On Windows:
-        ```bash
-        set OPENAI_API_KEY=your_api_key_here
-        ```
-    - On macOS and Linux:
-        ```bash
-        export OPENAI_API_KEY=your_api_key_here
-        ```
-
-Replace `your_api_key_here` with your actual OpenAI API key.
-
 ### Installation
 
 #### Clone this repository
@@ -59,7 +41,7 @@ To install this Python package in a virtual environment, you can use either pip
 
 #### Using Anaconda
 
-1. Create a virtual environment:
+1. Create a virtual environment (**Note:** Support for models provided by Dartmouth requires Python 3.10 or higher.):
     ```bash
     conda create --name highlight_env python=3.9
     ```
@@ -69,10 +51,44 @@ To install this Python package in a virtual environment, you can use either pip
     conda activate highlight_env
     ```
 
-3. Install the package from the cloned `highlight` directory:
+3. Install the package from the cloned `highlight` directory.
     ```bash
     pip install .
     ```
+    By default, only OpenAI's models are supported. You can add support for other providers by installing one ore more of the package's optional dependencies:
+    ```bash
+    pip install ".[anthropic, dartmouth, google, mistral, ollama]"
+    ```
+
+### Configuring the LLMs
+You can configure a variety of LLMs, both remote and local, to choose from in the app. To do this, edit the file `config.toml` in the root of this repo.
+
+#### Setting Up API Keys
+If any of the configured models require an API key, you need to set them up as Streamlit secrets.
+
+First, you have to acquire the API key from the provider of your choice (e.g., [OpenAI](https://platform.openai.com/api-keys), [Google](https://ai.google.dev/gemini-api/docs/api-key), [Dartmouth](https://developer.dartmouth.edu/keys), [Mistral](https://console.mistral.ai/api-keys/), or [Anthropic](https://console.anthropic.com/settings/keys)).
+
+Then, create a folder named `.streamlit` in the root of this repo. Inside that folder, create a file named `secrets.toml`.
+
+Finally, store your API key(s) in the following format:
+
+```toml
+KEY_NAME = "your_api_key_here"
+```
+
+Replace `KEY_NAME` with the variable name corresponding to your provider:
+
+| Provider    | Variable name       |
+| --------    | -------             |
+| Anthropic   | `ANTHROPIC_API_KEY` |
+| Dartmouth   | `DARTMOUTH_API_KEY` |
+| Google      | `GOOGLE_API_KEY`    |
+| Mistral     | `MISTRAL_API_KEY`    |
+| OpenAI      | `OPENAI_API_KEY`    |
+
+
+Replace `your_api_key_here` with your actual API key.
+
 
 ### Running the App