Refresh Lemonade Playbook

playbooks/dependencies/lemonade.md

@@ -22,9 +22,10 @@ msiexec /i lemonade-server-minimal.msi /qn
 <!-- @os:end -->
 
 <!-- @os:linux -->
-**Ubuntu (snap):**
+**Ubuntu:**
 ```bash
-sudo snap install lemonade-server
+sudo add-apt-repository ppa:lemonade-team/stable
+sudo apt install lemonade-server
 ```
 
 **Arch Linux (AUR):**
@@ -50,54 +51,19 @@ lemonade version x.y.z
 
 If you see a version number, Lemonade is installed correctly and ready to go.
 
-> **Tip**: Use `lemonade --help` to display help information.
-
-
-#### Checking Lemonade Server Status
-
-Open a terminal and run:
-```bash
-lemonade status
-```
-
-You should see the output showing that the server is running, typically on port `13305`. 
-
-The default OpenAI-compatible API base URL is: `http://localhost:13305/api/v1`.
-
-
-#### Loading and Chatting with a Model
-
-To load a model and open the Lemonade web app in your browser to chat with the model:
-```bash
-lemonade run Gemma-3-4b-it-GGUF
-```
-
-> **Tip**: Use `lemonade list` to see available models, or `lemonade pull <MODEL_NAME>` to download new ones.
-
-
-#### Loading a Model with a Specific Backend
-
-To load a llama.cpp model with AMD ROCm™ software backend:
-```bash
-lemonade run <MODEL_NAME> --llamacpp rocm
-```
-
-To load a llama.cpp model with Vulkan:
-```bash
-lemonade run <MODEL_NAME> --llamacpp vulkan
-```
-
-
-#### Setting a Default llama.cpp Backend
-
-If you want Lemonade to use a default backend (for example, ROCm) for llama.cpp models, run:
-```bash
-lemonade config set llamacpp.backend=rocm
-```
-
-You can inspect the current config anytime with:
-```bash
-lemonade config
-```
-
-For the latest Lemonade server options or troubleshooting, please refer to the [official Lemonade documentation](https://lemonade-server.ai/docs/lemonade-cli/).
+For quick reference, here are common Lemonade CLI commands:
+
+| Command | What it does |
+| --- | --- |
+| `lemonade --help` | Shows all available commands and flags. |
+| `lemonade --version` | Prints the installed Lemonade version. |
+| `lemonade status` | Confirms whether the Lemonade server is running and reachable. The default OpenAI-compatible API base URL is `http://localhost:13305/api/v1`. |
+| `lemonade list` | Lists models available to your Lemonade setup. |
+| `lemonade pull <MODEL_NAME>` | Downloads a model without launching it. |
+| `lemonade run <MODEL_NAME>` | Downloads the model if needed, then starts it for inference/chat. |
+| `lemonade run <MODEL_NAME> --llamacpp rocm` | Starts a llama.cpp model with the ROCm backend. |
+| `lemonade run <MODEL_NAME> --llamacpp vulkan` | Starts a llama.cpp model with the Vulkan backend. |
+| `lemonade config` | Displays the current Lemonade configuration values. |
+| `lemonade config set llamacpp.backend=rocm` | Sets the default llama.cpp backend to ROCm. |
+
+For the latest Lemonade server options or troubleshooting, please refer to the [official Lemonade documentation](https://lemonade-server.ai/docs/lemonade-cli/).