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/LLMModel.swift

@@ -6,6 +6,9 @@ import MLXNN
 
 // Interface for all LLM Models
 public protocol LLMModel: Module {
+
+    var vocabularySize: Int { get }
+
     func callAsFunction(_ inputs: MLXArray, cache: [(MLXArray, MLXArray)]?) -> (
         MLXArray, [(MLXArray, MLXArray)]
     )