Upload merged model (patched README metadata)

README.md

@@ -1,7 +1,5 @@
 ---
 base_model: Qwen/Qwen3-4B-Instruct-2507
-datasets:
-- u-10bei/dbbench_sft_dataset_react_v4
 language:
 - en
 license: apache-2.0
@@ -37,9 +35,14 @@ tool use, and recovery from errors.
 - Base model: Qwen/Qwen3-4B-Instruct-2507
 - Method: LoRA (full precision base)
 - Max sequence length: 4096
-- Epochs: 2
-- Learning rate: 1e-05
+- Epochs: 3
+- Learning rate: 8e-06
 - LoRA: r=64, alpha=128
+- warmup_ratio : 0
+- weight_decay : 0
+- lora_dropout : 0
+- lora_target_modules :['q_proj', 'k_proj', 'v_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj']
+- grad_accum : 8
 
 ## Usage
 
@@ -62,7 +65,7 @@ model = PeftModel.from_pretrained(model, adapter)
 
 ## Sources & Terms (IMPORTANT)
 
-Training data: u-10bei/dbbench_sft_dataset_react_v4
+Training data: /content/dbbench_react_merged_dedup
 
 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.

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f79c7176b2de27711dda2731ca6519e0f477f5e375a4d442dac0a2c76629a8bf
+oid sha256:8f9fa07dcd0661f7c5f32959c5728f56dd836c79ab33c28c41ce7aab1daea561
 size 4967215360

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d8774f1e2e2dec4fc7b4287b49222d750d307f9ce63aad99a7bcc56d27ea0f1
+oid sha256:3b76935d81bc551e770592efe411075ba84b8b2f5c232fe638d9f69c706738db
 size 3077766632