Upload README.md with huggingface_hub

README.md

@@ -1,65 +1,63 @@
 ---
 base_model: Qwen/Qwen3-4B-Instruct-2507
 datasets:
-- u-10bei/structured_data_with_cot_dataset_512_v2
+- u-10bei/dpo-dataset-qwen-cot
 language:
 - en
 license: apache-2.0
-library_name: peft
+library_name: transformers
 pipeline_tag: text-generation
 tags:
-- qlora
-- lora
-- structured-output
+- dpo
+- unsloth
+- qwen
+- alignment
 ---
 
-qwen3-4b-structured-output-lora
+# qwen3-4b-structured-output-lora-merged-dpo
 
-This repository provides a **LoRA adapter** fine-tuned from
-**Qwen/Qwen3-4B-Instruct-2507** using **QLoRA (4-bit, Unsloth)**.
+This model is a fine-tuned version of **Qwen/Qwen3-4B-Instruct-2507** using **Direct Preference Optimization (DPO)** via the **Unsloth** library.
 
-This repository contains **LoRA adapter weights only**.
-The base model must be loaded separately.
+This repository contains the **full-merged 16-bit weights**. No adapter loading is required.
 
 ## Training Objective
-
-This adapter is trained to improve **structured output accuracy**
-(JSON / YAML / XML / TOML / CSV).
-
-Loss is applied only to the final assistant output,
-while intermediate reasoning (Chain-of-Thought) is masked.
+This model has been optimized using DPO to align its responses with preferred outputs, focusing on improving reasoning (Chain-of-Thought) and structured response quality based on the provided preference dataset.
 
 ## Training Configuration
-
-- Base model: Qwen/Qwen3-4B-Instruct-2507
-- Method: QLoRA (4-bit)
-- Max sequence length: 512
-- Epochs: 1
-- Learning rate: 1e-06
-- LoRA: r=64, alpha=128
+- **Base model**: Qwen/Qwen3-4B-Instruct-2507
+- **Method**: DPO (Direct Preference Optimization)
+- **Epochs**: 1
+- **Learning rate**: 1e-07
+- **Beta**: 0.1
+- **Max sequence length**: 1024
+- **LoRA Config**: started from SFT adapter (r=64, alpha=128), updated with DPO, then merged into base (16-bit)
 
 ## Usage
+Since this is a merged model, you can use it directly with `transformers`.
 
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
-from peft import PeftModel
 import torch
 
-base = "Qwen/Qwen3-4B-Instruct-2507"
-adapter = "your_id/your-repo"
+model_id = "your_id/your-repo-name"
 
-tokenizer = AutoTokenizer.from_pretrained(base)
+tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
-    base,
+    model_id,
     torch_dtype=torch.float16,
-    device_map="auto",
+    device_map="auto"
 )
-model = PeftModel.from_pretrained(model, adapter)
-```
 
-## Sources & Terms (IMPORTANT)
+# Test inference
+prompt = "Your question here"
+inputs = tokenizer.apply_chat_template([{"role": "user", "content": prompt}], tokenize=True, add_generation_prompt=True, return_tensors="pt").to("cuda")
+outputs = model.generate(**inputs, max_new_tokens=512)
+print(tokenizer.decode(outputs[0]))
+
+```
 
-Training data: u-10bei/structured_data_with_cot_dataset_512_v2
+## Sources & License (IMPORTANT)
 
-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.
+* **Training Data**: [u-10bei/dpo-dataset-qwen-cot]
+* **License**: MIT License. (As per dataset terms).
+* **Compliance**: Users must follow the original base model's license terms.