Upload LoRA adapter (README written by author)

README.md

@@ -1,8 +1,7 @@
 ---
 base_model: Qwen/Qwen3-4B-Instruct-2507
 datasets:
-- u-10bei/structured_data_with_cot_dataset_512_v5
-- daichira/structured-hard-sft-4k
+- daichira/structured-5k-mix-sft
 language:
 - en
 license: apache-2.0
@@ -14,7 +13,6 @@ tags:
 - structured-output
 ---
 
-
 ＜Standard SFT Implementation for Improving Structured Output in Qwen3-4B＞
 
 This repository provides a **LoRA adapter** fine-tuned from
@@ -35,9 +33,9 @@ while intermediate reasoning (Chain-of-Thought) is masked.
 
 - Base model: Qwen/Qwen3-4B-Instruct-2507
 - Method: QLoRA (4-bit)
-- Max sequence length: 512
-- Epochs: 1
-- Learning rate: 2e-04
+- Max sequence length: 1024
+- Epochs: 2
+- Learning rate: 1e-05
 - LoRA: r=64, alpha=128
 
 ## Usage
@@ -61,7 +59,7 @@ model = PeftModel.from_pretrained(model, adapter)
 
 ## Sources & Terms (IMPORTANT)
 
-Training data: u-10bei/structured_data_with_cot_dataset_512_v5,daichira/structured-hard-sft-4k
+Training data: daichira/structured-5k-mix-sft
 
 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

@@ -34,12 +34,12 @@
   "revision": null,
   "target_modules": [
     "down_proj",
-    "gate_proj",
     "q_proj",
-    "up_proj",
-    "k_proj",
     "o_proj",
-    "v_proj"
+    "v_proj",
+    "k_proj",
+    "gate_proj",
+    "up_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94b53e45bd43b00ac9e7cfab065e34aea747dcf1ee5747333408f8cf386c9db9
+oid sha256:91d57fba50d34504c4ba1594b907d44ad550cf0695d37924608521f329f45f5e
 size 528550256