Upload fine-tuned intent detection model - 2025-06-24 14:04:19

0000050_adapters.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:944413997cab576d5f7619f1cfb5c368e9e8eb7644ea5b01ee244f24fe2be3c4
+size 2497239

0000100_adapters.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a4c814a70c1e82e10d9cf893bd7a3fa3b6c90cb1ee42cbd1e962de8e1a4df65
+size 2497239

README.md

@@ -65,8 +65,8 @@ let model = try await LLMModelFactory.shared.loadContainer(
 ## Training Details
 
 - **Fine-tuning Method**: LoRA with model fusion
-- **Export Date**: 2025-06-24T13:30:04.165855
-- **Fusion Completed**: True
+- **Export Date**: 2025-06-24T14:04:19.734671
+- **Fusion Completed**: False
 
 ## Expected Outputs
 

adapter_config.json

@@ -0,0 +1,35 @@
+{
+    "adapter_path": "/Users/hebertg/iOSDev/KnowledgeBase/MLX-Fine-Tuning-Pipeline/outputs/knowledgebase-adapter",
+    "batch_size": 2,
+    "config": null,
+    "data": "/Users/hebertg/iOSDev/KnowledgeBase/MLX-Fine-Tuning-Pipeline/outputs/mlx_data",
+    "fine_tune_type": "lora",
+    "grad_checkpoint": false,
+    "iters": 100,
+    "learning_rate": 0.0001,
+    "lora_parameters": {
+        "rank": 8,
+        "dropout": 0.0,
+        "scale": 20.0
+    },
+    "lr_schedule": null,
+    "mask_prompt": false,
+    "max_seq_length": 512,
+    "model": "mlx-community/Qwen2.5-1.5B-Instruct-4bit",
+    "num_layers": 16,
+    "optimizer": "adam",
+    "optimizer_config": {
+        "adam": {},
+        "adamw": {}
+    },
+    "resume_adapter_file": null,
+    "save_every": 50,
+    "seed": 0,
+    "steps_per_eval": 25,
+    "steps_per_report": 10,
+    "test": false,
+    "test_batches": 500,
+    "train": true,
+    "val_batches": 25,
+    "wandb": null
+}

adapters.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a4c814a70c1e82e10d9cf893bd7a3fa3b6c90cb1ee42cbd1e962de8e1a4df65
+size 2497239

model_config.json

@@ -1,11 +1,13 @@
 {
   "model_info": {
-    "model_id": "knowledgebaseLLM",
+    "model_id": "mlx-community/Qwen2.5-1.5B-Instruct-4bit",
     "model_type": "intent_detection",
     "version": "1.0",
-    "base_model": "mlx-community/Qwen2.5-1.5B-Instruct-4bit",
-    "model_format": "complete_merged",
-    "adapter_merged": true
+    "base_model": "mlx-community/Qwen2.5-1.5B-Instruct-4bit"
+  },
+  "adapter_files": {
+    "adapters": "adapters.safetensors",
+    "config": "adapter_config.json"
   },
   "inference_settings": {
     "max_tokens": 256,
@@ -45,13 +47,11 @@
     "examples": 5000,
     "iterations": 100,
     "training_completed": true,
-    "export_timestamp": "2025-06-24T13:30:04.165855",
-    "fusion_completed": true
+    "export_timestamp": "2025-06-24T14:04:19.734671"
   },
   "ios_integration": {
-    "model_type": "complete_bundled",
-    "load_method": "local_bundle_only",
-    "requires_download": false,
+    "adapter_path": "Model/adapters.safetensors",
+    "config_path": "Model/adapter_config.json",
     "system_prompt": "You are a helpful AI assistant for a knowledge-management app on an iPhone."
   }
 }