Update README.md
Browse files
README.md
CHANGED
@@ -42,4 +42,48 @@ This project implements the Mixture-of-Agents architecture proposed in the follo
|
|
42 |
}
|
43 |
```
|
44 |
|
45 |
-
For more information about the Mixture-of-Agents concept, please refer to the [original research paper](https://arxiv.org/abs/2406.04692) and the [Together AI blog post](https://www.together.ai/blog/together-moa).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
}
|
43 |
```
|
44 |
|
45 |
+
For more information about the Mixture-of-Agents concept, please refer to the [original research paper](https://arxiv.org/abs/2406.04692) and the [Together AI blog post](https://www.together.ai/blog/together-moa).
|
46 |
+
|
47 |
+
----
|
48 |
+
|
49 |
+
## App Structure
|
50 |
+
|
51 |
+
The app is structured into several main sections:
|
52 |
+
|
53 |
+
1. **Configuration Sidebar**:
|
54 |
+
- Allows users to select the main model, number of cycles, and main model temperature.
|
55 |
+
- Provides a JSON editor for configuring layer agents.
|
56 |
+
- Includes a button to apply recommended configurations.
|
57 |
+
|
58 |
+
2. **Main Chat Interface**:
|
59 |
+
- Displays the chat history and allows users to input questions.
|
60 |
+
- Shows streamed responses from the MOA system, including intermediate outputs from layer agents.
|
61 |
+
|
62 |
+
3. **Current Configuration Display**:
|
63 |
+
- Shows the current MOA configuration in an expandable section.
|
64 |
+
|
65 |
+
4. **Information Sections**:
|
66 |
+
- Displays credits and related information.
|
67 |
+
- Shows a visualization of the MOA workflow.
|
68 |
+
|
69 |
+
## Configuration Options
|
70 |
+
|
71 |
+
- **Main Model**: Select from a list of available Groq LLMs.
|
72 |
+
- **Number of Layers**: Choose the number of intermediate processing layers (1-10).
|
73 |
+
- **Main Model Temperature**: Set the temperature for the main model's output (0.0-1.0).
|
74 |
+
- **Layer Agent Configuration**: Customize the behavior of each layer agent using a JSON editor.
|
75 |
+
|
76 |
+
## How It Works
|
77 |
+
|
78 |
+
1. The app initializes with default or previously set configurations.
|
79 |
+
2. Users can update the configuration using the sidebar.
|
80 |
+
3. Questions are input through the chat interface.
|
81 |
+
4. The MOA system processes the input:
|
82 |
+
- Layer agents process the input in parallel for each cycle.
|
83 |
+
- The main model generates the final response.
|
84 |
+
5. Responses are streamed back to the user, showing both intermediate and final outputs.
|
85 |
+
|
86 |
+
## Notes
|
87 |
+
|
88 |
+
- The app uses session state to maintain conversation history and configuration across interactions.
|
89 |
+
- The MOA system is reset when the configuration is updated.
|