Add support for Llama 3.1 (#2359)

* Add Llama 3.1 rope

* Clippy

* Format

* Clippy

* Add support for multiple eos tokens:

* Untagged either

* Remove either dep and fix settings.json

* Make the max positional embeddings configurable

candle-nn/src/var_builder.rs

@@ -264,6 +264,7 @@ impl SimpleBackend for VarMap {
     }
 }
 
+#[allow(dead_code)]
 pub struct SafeTensorWithRouting<'a> {
     routing: HashMap<String, usize>,
     safetensors: Vec<SafeTensors<'a>>,