Upload merged Qwen3-4B-Instruct-2507 model (auto-generated README)

README.md

@@ -17,7 +17,6 @@ tags:
 - agentbench
 - alfworld
 - dbbench
-- db-oversampling
 ---
 
 # qwen3-4b-agentbench-dbalf-lora
@@ -49,13 +48,13 @@ Loss is applied to **all assistant turns** in the trajectory, enabling the model
 - Mixing ratio (pre-merge target): **DB:ALF = 1:1**
 
 ### DB Oversampling (category-aware)
-Enabled: **True**
+Enabled: **False**
 
 DB category weights used during training-data preparation:
 
-- counting: 6
-- comparison: 4
-- ranking: 2
+- counting: 1
+- comparison: 1
+- ranking: 1
 - select: 1
 - insert: 1
 - update: 1
@@ -65,10 +64,10 @@ DB category weights used during training-data preparation:
 
 - Base model: Qwen/Qwen3-4B-Instruct-2507
 - Method: LoRA (full precision base)
-- Max sequence length: 2048
-- Epochs: 2
-- Learning rate: 2e-06
-- LoRA: r=64, alpha=128, dropout=0.0
+- Max sequence length: 4096
+- Epochs: 1
+- Learning rate: 2e-04
+- LoRA: r=32, alpha=64, dropout=0.05
 - Per-device train batch size: 2
 - Gradient accumulation: 4
 
@@ -80,7 +79,7 @@ from peft import PeftModel
 import torch
 
 base = "Qwen/Qwen3-4B-Instruct-2507"
-adapter = "your_id/your-repo"
+adapter = "your_id/your-model-name"
 
 tokenizer = AutoTokenizer.from_pretrained(base)
 model = AutoModelForCausalLM.from_pretrained(

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ce2ce8c52da46bf7471c0d2eef812a191d5be71b1f82e2539a20dafd00148e4f
+oid sha256:3b4f9428ff718d9cb13a3c2edc84fc40c13618001bd765b1ddb70cd26dc7b465
 size 4967215360

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:353c3f776c17902ac161c9d1ab001e4929ffcffd4c9626ce12a4a4362de0ac7f
+oid sha256:6f556fb2a665d7393d46117557f24b9ce0ff12b3f0577e160b9de892b3f11973
 size 3077766632