femtowin commited on
Commit
4f23b02
·
1 Parent(s): 2ad1ebd

Add configuration options documentation to README

Browse files
Files changed (1) hide show
  1. README.md +28 -2
README.md CHANGED
@@ -14,7 +14,9 @@ tags:
14
  - agent-demo-track
15
  ---
16
 
17
- An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).
 
 
18
 
19
  ## Demo Videos
20
 
@@ -24,4 +26,28 @@ Watch this video to see minion-space in action, demonstrating how it can solve m
24
 
25
  🎥 **[Minion-Space List-Files Demo](https://youtu.be/7neYqi_IQ18)**
26
 
27
- This video showcases the list-files functionality of minion-space, demonstrating how it can efficiently browse and manage file systems.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14
  - agent-demo-track
15
  ---
16
 
17
+ An example chatbot based on minion framework(https://github.com/femto/minion), can handle arbitray type
18
+ of queries, task-aware routing(if don't choose route, it will automatically choose proper route)
19
+
20
 
21
  ## Demo Videos
22
 
 
26
 
27
  🎥 **[Minion-Space List-Files Demo](https://youtu.be/7neYqi_IQ18)**
28
 
29
+ This video showcases the list-files functionality of minion-space, demonstrating how it can efficiently browse and manage file systems.
30
+
31
+ ## Configuration Options
32
+
33
+ ### Route Selection
34
+ - **Auto (empty)**: Automatically selects the most appropriate route based on query type
35
+ - **Raw**: Direct LLM output without additional processing
36
+ - **Native**: Standard minion processing
37
+ - **CoT**: Chain of Thought reasoning for complex problems
38
+ - **DCoT**: Dynamic Chain of Thought with adaptive reasoning
39
+ - **Plan**: Planning-based approach for multi-step tasks
40
+ - **Python**: Code execution environment for computational queries
41
+
42
+ ### Preset Models
43
+ - **gpt-4o**: Latest GPT-4 Omni model for general tasks
44
+ - **gpt-4o-mini**: Lightweight version for faster responses
45
+ - **gpt-4.1**: Advanced reasoning model
46
+ - **o4-mini**: Optimized mini model
47
+ - **Custom**: Configure your own API settings
48
+
49
+ ### Additional Settings
50
+ - **Enable Check**: Activates output verification and validation
51
+ - **Query Type** (Python route only): Choose between `calculate`, `code_solution`, or `generate`
52
+ - **Temperature**: Controls response creativity (0.0-2.0)
53
+ - **Max Tokens**: Maximum response length (100-8000)