Update readme for LM-Studio Usage
Browse files
README.md
CHANGED
|
@@ -136,9 +136,21 @@ All evaluations were conducted using **[simple-eval](https://github.com/Nikityyy
|
|
| 136 |
|
| 137 |
## 🚀 How to Use
|
| 138 |
|
| 139 |
-
|
| 140 |
|
| 141 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 142 |
|
| 143 |
```bash
|
| 144 |
pip install simpleai-sdk
|
|
@@ -169,7 +181,7 @@ response = model.generate(prompt, max_new_tokens=50, temperature=0.9)
|
|
| 169 |
print(f"\n--- Completion Example ---\n{prompt}{response}")
|
| 170 |
```
|
| 171 |
|
| 172 |
-
###
|
| 173 |
|
| 174 |
You can also use the model directly with the `transformers` library for more advanced use cases.
|
| 175 |
|
|
|
|
| 136 |
|
| 137 |
## 🚀 How to Use
|
| 138 |
|
| 139 |
+
There are several ways to use the Lille models, from easy-to-use graphical interfaces to advanced programmatic control.
|
| 140 |
|
| 141 |
+
### 1. LM Studio (Easiest for Chat)
|
| 142 |
+
|
| 143 |
+
LM Studio provides a simple graphical interface to run LLMs on your local machine. It's the easiest way to start chatting with Lille.
|
| 144 |
+
|
| 145 |
+
1. **Download & Install:** Get [LM Studio](https://lmstudio.ai/) for your operating system (Windows, Mac, or Linux).
|
| 146 |
+
2. **Search for the Model:** Open LM Studio and click the **magnifying glass** icon on the left.
|
| 147 |
+
3. **Find Lille:** In the search bar, type `Lille` or `Nikity`. You will find the models I have uploaded.
|
| 148 |
+
4. **Download a GGUF:** On the right-hand side, you'll see a list of GGUF files. Download a recommended version like `lille-130m-instruct-f16.gguf`.
|
| 149 |
+
5. **Chat:** Click the **speech bubble** icon on the left. At the top, select the model you just downloaded. Now you can start a conversation!
|
| 150 |
+
|
| 151 |
+
### 2. SimpleAI SDK (Recommended for Programmatic Use)
|
| 152 |
+
|
| 153 |
+
The easiest way to use Lille programmatically is with the `simpleai-sdk`, which handles all the boilerplate for you and provides a simple, high-level API for both Hugging Face and ONNX backends.
|
| 154 |
|
| 155 |
```bash
|
| 156 |
pip install simpleai-sdk
|
|
|
|
| 181 |
print(f"\n--- Completion Example ---\n{prompt}{response}")
|
| 182 |
```
|
| 183 |
|
| 184 |
+
### 3. Standard Hugging Face Transformers (this also needs `simpleai-sdk` currently)
|
| 185 |
|
| 186 |
You can also use the model directly with the `transformers` library for more advanced use cases.
|
| 187 |
|