docs cleanup

This commit is contained in:
AI Christianson 2025-02-14 17:25:18 -05:00
parent feecfea3b8
commit 193d4c6955
2 changed files with 102 additions and 102 deletions

View File

@ -1,84 +1,52 @@
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
# Recommended Config
# Open Models Configuration
This configuration combines the strengths of multiple AI models to provide the best experience:
RA.Aid supports various open source model providers and configurations. This guide shows you how to configure and use different open models with RA.Aid.
- Anthropic Sonnet excels at driving the agent's core reasoning and planning
- OpenAI's models provide robust debugging and logical analysis capabilities
- Tavily web search integration allows the agent to find relevant information online
## Supported Providers
:::info
RA.Aid must be installed before using these configurations. If you haven't installed it yet, please see the [Installation Guide](installation).
:::
<Tabs groupId="model-provider">
<TabItem value="deepseek" label="DeepSeek" default>
## Getting API Keys
### DeepSeek Models
To use RA.Aid with the recommended configuration, you'll need to obtain API keys from the following services:
To use DeepSeek models, you'll need a DeepSeek API key. Set it in your environment:
1. **OpenAI API Key**: Create an account at [OpenAI's platform](https://platform.openai.com) and generate an API key from your dashboard.
2. **Anthropic API Key**: Sign up at [Anthropic's Console](https://console.anthropic.com), then generate an API key from the API Keys section.
3. **Tavily API Key** (optional): Create an account at [Tavily](https://app.tavily.com/sign-in) and get your API key from the dashboard.
Please keep your API keys secure and never share them publicly. Each service has its own pricing and usage terms.
## Configuration
Configure your API keys:
```bash
export DEEPSEEK_API_KEY=your_api_key_here
# For OpenAI (required)
export OPENAI_API_KEY=your_api_key_here
# For Anthropic (required)
export ANTHROPIC_API_KEY=your_api_key_here
# For web search capability (optional)
export TAVILY_API_KEY=your_api_key_here
```
Then run RA.Aid with the deepseek provider and model:
## Basic Usage
Start RA.Aid in interactive chat mode:
```bash
ra-aid -m "Your task" --provider deepseek --model deepseek-reasoner
ra-aid --chat
```
You can also access DeepSeek models through OpenRouter:
Or run with a single command:
```bash
ra-aid -m "Your task" --provider openrouter --model deepseek/deepseek-r1
ra-aid -m "Help me understand this code"
```
</TabItem>
<TabItem value="openrouter" label="OpenRouter">
### OpenRouter Integration
OpenRouter provides access to various open source models. First, set your API key:
```bash
export OPENROUTER_API_KEY=your_api_key_here
```
Example using Mistral:
```bash
ra-aid -m "Your task" --provider openrouter --model mistralai/mistral-large-2411
```
</TabItem>
<TabItem value="expert" label="Expert Configuration">
### Expert Tool Configuration
The expert tool can be configured to use open models for complex logic and debugging tasks:
```bash
# Use DeepSeek for expert tool
export EXPERT_DEEPSEEK_API_KEY=your_deepseek_api_key
ra-aid -m "Your task" --expert-provider deepseek --expert-model deepseek-reasoner
# Use OpenRouter for expert
export EXPERT_OPENROUTER_API_KEY=your_openrouter_api_key
ra-aid -m "Your task" --expert-provider openrouter --expert-model mistralai/mistral-large-2411
```
</TabItem>
</Tabs>
## Environment Variables
Here are all the environment variables supported for open model configuration:
- `OPENROUTER_API_KEY`: Required for OpenRouter provider
- `DEEPSEEK_API_KEY`: Required for DeepSeek provider
- `EXPERT_OPENROUTER_API_KEY`: API key for expert tool using OpenRouter provider
- `EXPERT_DEEPSEEK_API_KEY`: API key for expert tool using DeepSeek provider
## Notes and Best Practices
- Set environment variables in your shell's configuration file (e.g., `~/.bashrc` or `~/.zshrc`) for persistence
- Consider using different models for different types of tasks (e.g., DeepSeek for reasoning, Mistral for general tasks)
- Review model performance and adjust based on your specific needs
- Keep your API keys secure and never commit them to version control

View File

@ -1,52 +1,84 @@
# Recommended Config
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
This configuration combines the strengths of multiple AI models to provide the best experience:
# Open Models Configuration
- Anthropic Sonnet excels at driving the agent's core reasoning and planning
- OpenAI's models provide robust debugging and logical analysis capabilities
- Tavily web search integration allows the agent to find relevant information online
RA.Aid supports various open source model providers and configurations. This guide shows you how to configure and use different open models with RA.Aid.
:::info
RA.Aid must be installed before using these configurations. If you haven't installed it yet, please see the [Installation Guide](installation).
:::
## Supported Providers
## Getting API Keys
<Tabs groupId="model-provider">
<TabItem value="deepseek" label="DeepSeek" default>
To use RA.Aid with the recommended configuration, you'll need to obtain API keys from the following services:
### DeepSeek Models
1. **OpenAI API Key**: Create an account at [OpenAI's platform](https://platform.openai.com) and generate an API key from your dashboard.
2. **Anthropic API Key**: Sign up at [Anthropic's Console](https://console.anthropic.com), then generate an API key from the API Keys section.
3. **Tavily API Key** (optional): Create an account at [Tavily](https://app.tavily.com/sign-in) and get your API key from the dashboard.
Please keep your API keys secure and never share them publicly. Each service has its own pricing and usage terms.
## Configuration
Configure your API keys:
To use DeepSeek models, you'll need a DeepSeek API key. Set it in your environment:
```bash
# For OpenAI (required)
export OPENAI_API_KEY=your_api_key_here
# For Anthropic (required)
export ANTHROPIC_API_KEY=your_api_key_here
# For web search capability (optional)
export TAVILY_API_KEY=your_api_key_here
export DEEPSEEK_API_KEY=your_api_key_here
```
## Basic Usage
Start RA.Aid in interactive chat mode:
Then run RA.Aid with the deepseek provider and model:
```bash
ra-aid --chat
ra-aid -m "Your task" --provider deepseek --model deepseek-reasoner
```
Or run with a single command:
You can also access DeepSeek models through OpenRouter:
```bash
ra-aid -m "Help me understand this code"
ra-aid -m "Your task" --provider openrouter --model deepseek/deepseek-r1
```
</TabItem>
<TabItem value="openrouter" label="OpenRouter">
### OpenRouter Integration
OpenRouter provides access to various open source models. First, set your API key:
```bash
export OPENROUTER_API_KEY=your_api_key_here
```
Example using Mistral:
```bash
ra-aid -m "Your task" --provider openrouter --model mistralai/mistral-large-2411
```
</TabItem>
<TabItem value="expert" label="Expert Configuration">
### Expert Tool Configuration
The expert tool can be configured to use open models for complex logic and debugging tasks:
```bash
# Use DeepSeek for expert tool
export EXPERT_DEEPSEEK_API_KEY=your_deepseek_api_key
ra-aid -m "Your task" --expert-provider deepseek --expert-model deepseek-reasoner
# Use OpenRouter for expert
export EXPERT_OPENROUTER_API_KEY=your_openrouter_api_key
ra-aid -m "Your task" --expert-provider openrouter --expert-model mistralai/mistral-large-2411
```
</TabItem>
</Tabs>
## Environment Variables
Here are all the environment variables supported for open model configuration:
- `OPENROUTER_API_KEY`: Required for OpenRouter provider
- `DEEPSEEK_API_KEY`: Required for DeepSeek provider
- `EXPERT_OPENROUTER_API_KEY`: API key for expert tool using OpenRouter provider
- `EXPERT_DEEPSEEK_API_KEY`: API key for expert tool using DeepSeek provider
## Notes and Best Practices
- Set environment variables in your shell's configuration file (e.g., `~/.bashrc` or `~/.zshrc`) for persistence
- Consider using different models for different types of tasks (e.g., DeepSeek for reasoning, Mistral for general tasks)
- Review model performance and adjust based on your specific needs
- Keep your API keys secure and never commit them to version control