implement markdown-based hint system and add MarkdownHint component
This commit is contained in:
86
app/src/main/assets/hints/example_hint.md
Normal file
86
app/src/main/assets/hints/example_hint.md
Normal file
@@ -0,0 +1,86 @@
|
||||
# How to Scan for AI Models
|
||||
|
||||
This guide explains how to use the **Scan** feature to discover and add AI models to your app.
|
||||
|
||||
## How Scanning Works
|
||||
|
||||
The scan feature searches for available AI models on your device or network.
|
||||
|
||||
> **Note:** Results depend on your API key permissions.
|
||||
|
||||
### Key Points
|
||||
|
||||
- Only public models are shown by default
|
||||
- Private models require additional setup
|
||||
- Try again if no models are found
|
||||
|
||||
## Why Some Models Are Missing
|
||||
|
||||
Some models may not appear in the scan results due to:
|
||||
|
||||
| Reason | Description | Icon |
|
||||
|--------|-------------|------|
|
||||
| Restricted access | Model requires special permissions | 🔒 |
|
||||
| Not suitable | Model type not supported | ⚠️ |
|
||||
| Text only | Only text-based models are supported | ✓ |
|
||||
|
||||
### Model Tiers
|
||||
|
||||
We recommend these tiers for optimal performance:
|
||||
|
||||
- **Nano** - Fastest, for simple tasks
|
||||
- **Mini** - Balanced speed and capability
|
||||
- **Small** - Good for most tasks
|
||||
- **Medium** - More capable, slower
|
||||
- **Large** - Most capable, paid only
|
||||
|
||||
## Tips for Success
|
||||
|
||||
1. **Verify your API key** is active and has correct permissions
|
||||
2. **Select the correct organization** from your account
|
||||
3. **Type model names manually** if scanning doesn't find them
|
||||
4. **Prefer instruct or chat models** for text generation
|
||||
|
||||
```kotlin
|
||||
// Example: Manual model addition
|
||||
val model = Model(
|
||||
name = "llama3.2",
|
||||
type = ModelType.TEXT,
|
||||
provider = "ollama"
|
||||
)
|
||||
```
|
||||
|
||||
## Visual Guide
|
||||
|
||||
### Step 1: Initiate Scan
|
||||
|
||||
Click the scan button to search for available models.
|
||||
|
||||
### Step 2: Select Model Type
|
||||
|
||||
Choose between different model categories:
|
||||
|
||||
- **Text Chat** - For conversational AI
|
||||
- **Instruct** - For direct instructions
|
||||
- **Complete** - For text completion
|
||||
|
||||
### Step 3: Add & Validate
|
||||
|
||||
Add the selected model and validate it works correctly.
|
||||
|
||||
---
|
||||
|
||||
## Can't Find Your Model?
|
||||
|
||||
If your model doesn't appear in the scan results:
|
||||
|
||||
1. Check if the model is running locally or accessible via API
|
||||
2. Verify network connectivity
|
||||
3. Try adding it manually by entering the model details
|
||||
|
||||
> **Pro Tip:** You can always add models manually by clicking the "+" button in the models screen.
|
||||
|
||||
---
|
||||
|
||||
*Last updated: 2024-01-15*
|
||||
*For more help, visit our documentation website.*
|
||||
Reference in New Issue
Block a user