Upload LoRA adapter (README written by author)

README.md

@@ -1,7 +1,7 @@
 ---
 base_model: Qwen/Qwen3-4B-Instruct-2507
 datasets:
-- u-10bei/structured_data_with_cot_dataset_512_v5
+- u-10bei/structured_data_with_cot_dataset_512_v4
 language:
 - en
 license: apache-2.0
@@ -13,7 +13,7 @@ tags:
 - structured-output
 ---
 
-qwen3-4b-structured-output-lora_v5x2+αlearn
+qwen3-4b-structured-output-lora_v4x2+βlearn
 
 This repository provides a **LoRA adapter** fine-tuned from
 **Qwen/Qwen3-4B-Instruct-2507** using **QLoRA (4-bit, Unsloth)**.
@@ -34,7 +34,7 @@ while intermediate reasoning (Chain-of-Thought) is masked.
 - Base model: Qwen/Qwen3-4B-Instruct-2507
 - Method: QLoRA (4-bit)
 - Max sequence length: 512
-- Epochs: 2
+- Epochs: 3
 - Learning rate: 1e-06
 - LoRA: r=64, alpha=128
 
@@ -59,7 +59,7 @@ model = PeftModel.from_pretrained(model, adapter)
 
 ## Sources & Terms (IMPORTANT)
 
-Training data: u-10bei/structured_data_with_cot_dataset_512_v5
+Training data: u-10bei/structured_data_with_cot_dataset_512_v4
 
 Dataset License: MIT License. This dataset is used and distributed under the terms of the MIT License.
 Compliance: Users must comply with the MIT license (including copyright notice) and the base model's original terms of use.

adapter_config.json

@@ -33,13 +33,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "gate_proj",
+    "v_proj",
     "o_proj",
-    "up_proj",
-    "k_proj",
-    "down_proj",
     "q_proj",
-    "v_proj"
+    "k_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f6340b9e50d5642059855a7885221c0b70ae43f3fefaa9b5721a44e980d7c1bd
+oid sha256:213063cb4c910b63bba96ffa7d0a122327401ce824a5983fb767c499f582a8a6
 size 528550256