fix for #2 -- CodeLlama crashes

- add replacement tokenizer class for unknown tokenizers
- fix quantization for models that don't have lm_head quantized

Requires https://github.com/ml-explore/mlx-swift/pull/28

Libraries/LLM/Gemma.swift

@@ -202,9 +202,11 @@ public class GemmaModelInner: Module {
 
 public class GemmaModel: Module, LLMModel {
 
+    public let vocabularySize: Int
     let model: GemmaModelInner
 
     public init(_ args: GemmaConfiguration) {
+        self.vocabularySize = args.vocabularySize
         self.model = GemmaModelInner(args)
     }