Use 4-bit quantized models

Author: DePasqualeOrg

Libraries/MLXVLM/VLMModelFactory.swift

@@ -169,14 +169,14 @@ public class VLMRegistry: AbstractModelRegistry, @unchecked Sendable {
         extraEOSTokens: ["<end_of_turn>"]
     )
 
-    static public let gemma3n_E2B_instruct = ModelConfiguration(
-        id: "mlx-community/gemma-3n-E2B-it-bf16",
+    static public let gemma3n_E2B = ModelConfiguration(
+        id: "mlx-community/gemma-3n-E2B-it-4bit",
         defaultPrompt: "Describe this image.",
         extraEOSTokens: ["<end_of_turn>"]
     )
 
-    static public let gemma3n_E4B_instruct = ModelConfiguration(
-        id: "mlx-community/gemma-3n-E4B-it-bf16",
+    static public let gemma3n_E4B = ModelConfiguration(
+        id: "mlx-community/gemma-3n-E4B-it-4bit",
         defaultPrompt: "Describe this image.",
         extraEOSTokens: ["<end_of_turn>"]
     )
@@ -196,8 +196,8 @@ public class VLMRegistry: AbstractModelRegistry, @unchecked Sendable {
             gemma3_4B_qat_4bit,
             gemma3_12B_qat_4bit,
             gemma3_27B_qat_4bit,
-            gemma3n_E2B_instruct,
-            gemma3n_E4B_instruct,
+            gemma3n_E2B,
+            gemma3n_E4B,
             smolvlm,
         ]
     }

Tools/llm-tool/LLMTool.swift

@@ -302,9 +302,9 @@ struct EvaluateCommand: AsyncParsableCommand {
         let modelFactory: ModelFactory
         let defaultModel: ModelConfiguration
 
-        // Always use VLM factory and gemma3n_E2B_instruct for testing
+        // Always use VLM factory and gemma3n_E2B for testing
         modelFactory = VLMModelFactory.shared
-        defaultModel = MLXVLM.VLMRegistry.gemma3n_E2B_instruct
+        defaultModel = MLXVLM.VLMRegistry.gemma3n_E2B
 
         // Load the model
         let modelContainer = try await memory.start { [args] in