diff --git a/docs/faqs.mdx b/docs/faqs.mdx index d0dcaa8edb0..7836ef9d3c1 100644 --- a/docs/faqs.mdx +++ b/docs/faqs.mdx @@ -146,6 +146,234 @@ Most Continue features will work normally, including autocomplete and chat. Howe For more details about this requirement, see the [LanceDB issue #2195](https://github.com/lancedb/lance/issues/2195). +## Ollama Issues + +For a comprehensive guide on setting up and troubleshooting Ollama, see the [Ollama Guide](/guides/ollama-guide). + +### Unable to connect to local Ollama instance + +If you're getting "Unable to connect to local Ollama instance" errors: + +1. **Verify Ollama is running**: Check http://localhost:11434 in your browser - you should see "Ollama is running" +2. **Start Ollama properly**: Use `ollama serve` (not just `ollama run model-name`) +3. **Check your config**: Ensure your `config.yaml` has the correct setup: + +```yaml +models: + - name: llama3 + provider: ollama + model: llama3:latest +``` + +### Connection failed to remote Ollama (EHOSTUNREACH/ECONNREFUSED) + +When connecting to Ollama on another machine: + +1. **Configure Ollama to listen on all interfaces**: + - Set environment variable: `OLLAMA_HOST=0.0.0.0:11434` + - For systemd: Edit `/etc/systemd/system/ollama.service` and add under `[Service]`: + ``` + Environment="OLLAMA_HOST=0.0.0.0:11434" + Environment="OLLAMA_ORIGINS=*" + ``` + - Restart Ollama: `sudo systemctl restart ollama` + +2. **Update your Continue config**: +```yaml +models: + - name: llama3 + provider: ollama + apiBase: http://192.168.1.136:11434 # Use your server's IP + model: llama3:latest +``` + +3. **Check firewall settings**: Ensure port 11434 is open on the server + +### Ollama not working in WSL + +For WSL users having connection issues: + +#### Windows 11 22H2+ (Recommended) +Create or edit `%UserProfile%\.wslconfig`: +```ini +[wsl2] +networkingMode=mirrored +``` +Then restart WSL: `wsl --shutdown` + +#### Older Windows/WSL versions +In PowerShell (as Administrator): +```powershell +# Add firewall rules +New-NetFireWallRule -DisplayName 'WSL Ollama' -Direction Inbound -LocalPort 11434 -Action Allow -Protocol TCP +New-NetFireWallRule -DisplayName 'WSL Ollama' -Direction Outbound -LocalPort 11434 -Action Allow -Protocol TCP + +# Get WSL IP (run 'ip addr' in WSL to find eth0 IP) +# Then add port proxy (replace with your actual IP) +netsh interface portproxy add v4tov4 listenport=11434 listenaddress=0.0.0.0 connectport=11434 connectaddress= +``` + +### Docker container can't connect to host Ollama + +When running Continue or other tools in Docker that need to connect to Ollama on the host: + +**Windows/Mac**: Use `host.docker.internal`: +```yaml +models: + - name: llama3 + provider: ollama + apiBase: http://host.docker.internal:11434 + model: llama3:latest +``` + +**Linux**: Use the Docker bridge IP (usually `172.17.0.1`): +```yaml +models: + - name: llama3 + provider: ollama + apiBase: http://172.17.0.1:11434 + model: llama3:latest +``` + +**Docker run command**: Add host mapping: +```bash +docker run -d --add-host=host.docker.internal:host-gateway ... +``` + +### Parse errors with remote Ollama + +If you're getting parse errors with remote Ollama: + +1. **Verify the model is installed on the remote**: + ```bash + OLLAMA_HOST=192.168.1.136:11434 ollama list + ``` + +2. **Install missing models**: + ```bash + OLLAMA_HOST=192.168.1.136:11434 ollama pull llama3 + ``` + +3. **Check URL format**: Ensure you're using `http://` not `https://` for local network addresses + +## Local Assistant + +### Managing Local Secrets and Environment Variables + +For running Continue completely offline without internet access, see the [Running Continue Without Internet guide](/guides/running-continue-without-internet). + +Continue supports multiple methods for managing secrets locally, searched in this order: + +1. **Workspace `.env` files**: Place a `.env` file in your workspace root directory +2. **Workspace Continue folder**: Place a `.env` file in `/.continue/.env` +3. **Global `.env` file**: Place a `.env` file in `~/.continue/.env` for user-wide secrets +4. **Process environment variables**: Use standard system environment variables + +#### Creating `.env` files + +Create a `.env` file in one of these locations: +- **Per-workspace**: `/.env` or `/.continue/.env` +- **Global**: `~/.continue/.env` + +Example `.env` file: +``` +OPENAI_API_KEY=sk-... +ANTHROPIC_API_KEY=sk-ant-... +CUSTOM_API_URL=https://api.example.com +``` + +#### Using secrets in config.yaml + +Reference your local secrets using the `secrets` namespace: +```yaml +models: + - provider: openai + apiKey: ${{ secrets.OPENAI_API_KEY }} +``` + +#### Hub-managed secrets + +For centralized team secret management, use `${{ inputs.SECRET_NAME }}` syntax in your config.yaml and manage them at [https://hub.continue.dev/settings/secrets](https://hub.continue.dev/settings/secrets): +```yaml +models: + - provider: openai + apiKey: ${{ inputs.OPENAI_API_KEY }} +``` + +#### Important notes + +- **Never commit `.env` files** to version control - add them to `.gitignore` +- The `.env` file uses standard dotenv format (KEY=value, no quotes needed) +- Secrets are loaded when Continue starts, so restart your IDE after changes +- Local `.env` files take precedence over Hub secrets when both exist + +#### Troubleshooting secrets + +If your API keys aren't being recognized: +1. Check the `.env` file is in the correct location +2. Ensure there are no quotes around values in the `.env` file +3. Restart your IDE after adding/changing secrets +4. Verify the variable name matches exactly (case-sensitive) +5. Check that your `.env` file has proper line endings (LF, not CRLF on Windows) + +### Using Model Addons Locally + +You can leverage model addons from the Continue Hub in your local assistant configurations using the `uses:` syntax. This allows you to reference pre-configured model blocks without duplicating configuration. + +#### Requirements + +- You must be logged in to Continue +- Internet connection is required (model addons are fetched from the hub) + +#### Usage + +In your local `config.yaml`, reference model addons using the format `provider/model-name`: + +```yaml +name: My Local Assistant +version: 0.0.1 +schema: v1 +models: + - uses: ollama/llama3.1-8b + - uses: anthropic/claude-3.5-sonnet + - uses: openai/gpt-4 +``` + +#### With local configuration + +You can combine hub model addons with local models: + +```yaml +name: My Local Assistant +version: 0.0.1 +schema: v1 +models: + # Hub model addon + - uses: anthropic/claude-3.5-sonnet + + # Local model configuration + - name: Local Ollama + provider: ollama + model: codellama:latest + apiBase: http://localhost:11434 +``` + +#### Override addon settings + +You can override specific settings from the model addon: + +```yaml +models: + - uses: ollama/llama3.1-8b + override: + apiBase: http://192.168.1.100:11434 # Use remote Ollama server + roles: + - chat + - autocomplete +``` + +This feature allows you to maintain consistent model configurations across teams while still allowing local customization when needed. + ## How do I reset the state of the extension? Continue stores its data in the `~/.continue` directory (`%USERPROFILE%\.continue` on Windows). diff --git a/docs/features/plan/how-it-works.mdx b/docs/features/plan/how-it-works.mdx index cb608f18254..8555e68ffb2 100644 --- a/docs/features/plan/how-it-works.mdx +++ b/docs/features/plan/how-it-works.mdx @@ -85,8 +85,7 @@ Plan mode uses a dedicated system message that: - Encourages thorough analysis before suggesting changes - The system message can be customized per model. See [How to - Customize](/features/plan/how-to-customize) for details or see it [directly in + The system message can be customized per model. See it [directly in the code](https://github.com/continuedev/continue/blob/main/core/llm/defaultSystemMessages.ts). diff --git a/docs/features/plan/quick-start.mdx b/docs/features/plan/quick-start.mdx index 40bbc80d017..05961273cf4 100644 --- a/docs/features/plan/quick-start.mdx +++ b/docs/features/plan/quick-start.mdx @@ -11,7 +11,7 @@ Plan mode provides a safe environment for understanding and constructing plans w You can switch to `Plan` in the mode selector below the chat input box. -![How to select plan mode](/images/mode-select-plan.png) +![How to select plan mode](/images/plan-mode-selector.png) If Plan is disabled with a `Not Supported` message, the selected model or diff --git a/docs/hub/assistants/use-an-assistant.mdx b/docs/hub/assistants/use-an-assistant.mdx index 85257d25063..f6d19b8360e 100644 --- a/docs/hub/assistants/use-an-assistant.mdx +++ b/docs/hub/assistants/use-an-assistant.mdx @@ -16,4 +16,4 @@ After saving, open your IDE extension and: - Select the assistant from the **assistant dropdown** in the Continue extension. - Begin using it for chat, code generation, or other configured capabilities. -[Extension Assistant Selector](/images/assistant-extension-select-319492a06d6249e5389747687341dfdb.png) +![Extension Assistant Selector](/images/hub/assistants/images/assistant-extension-select.png)