doc: clarify that --quantize is not needed for pre-quantized models (…

…#2536)

docs/source/reference/launcher.md

@@ -55,7 +55,9 @@ Options:
 ## QUANTIZE
 ```shell
       --quantize <QUANTIZE>
-          Whether you want the model to be quantized
+          Quantization method to use for the model. It is not necessary to specify this option for pre-quantized models, since the quantization method is read from the model configuration.
+
+          Marlin kernels will be used automatically for GPTQ/AWQ models.
 
           [env: QUANTIZE=]
 

flake.nix

@@ -157,6 +157,7 @@
                 pyright
                 pytest
                 pytest-asyncio
+                redocly
                 ruff
                 syrupy
               ]);

launcher/src/main.rs

@@ -367,7 +367,11 @@ struct Args {
     #[clap(long, env)]
     num_shard: Option<usize>,
 
-    /// Whether you want the model to be quantized.
+    /// Quantization method to use for the model. It is not necessary to specify this option
+    /// for pre-quantized models, since the quantization method is read from the model
+    /// configuration.
+    ///
+    /// Marlin kernels will be used automatically for GPTQ/AWQ models.
     #[clap(long, env, value_enum)]
     quantize: Option<Quantization>,