Fix missing link (#386)

First of all, kudos for this project.
It's the only project I found that properly supports modern models like llama-3.1 out of the box.
Also the speed, and other factors, seem better.

Fixing in this PR a small bug - the previous link led to a missing page.

README.md

@@ -179,7 +179,7 @@ model = GPTQModel.from_quantized(quant_output_dir)
 print(tokenizer.decode(model.generate(**tokenizer("gptqmodel is", return_tensors="pt").to(model.device))[0]))
 ```
 
-For more advanced features of model quantization, please reference to [this script](examples/quantization/quant_with_alpaca.py)
+For more advanced features of model quantization, please reference to [this script](https://github.com/ModelCloud/GPTQModel/blob/main/examples/quantization/basic_usage_wikitext2.py)
 
 ### How to Add Support for a New Model