---
base_model: Qwen/Qwen3-4B-Instruct-2507
datasets:
- u-10bei/sft_alfworld_trajectory_dataset_v5
- u-10bei/sft_alfworld_trajectory_dataset_v4
- u-10bei/dbbench_sft_dataset_react_v4
- u-10bei/dbbench_sft_dataset_react_v3
language:
- en
license: apache-2.0
library_name: transformers
pipeline_tag: text-generation
tags:
- agent
- tool-use
- alfworld
- dbbench
---

# qwen3-4b-agent-full-v3

This repository provides a **fully fine-tuned model** based on
**Qwen/Qwen3-4B-Instruct-2507**.

Because this model underwent full parameter fine-tuning, this repository contains the **full model weights**.
You can load it directly without needing to merge it with the base model.

## Training Objective

This model is trained to improve **multi-turn agent task performance**
on ALFWorld (household tasks) and DBBench (database operations).

Loss is applied to **all assistant turns** in the multi-turn trajectory,
enabling the model to learn environment observation, action selection,
tool use, and recovery from errors.

## Training Configuration

- Base model: Qwen/Qwen3-4B-Instruct-2507
- Method: full
- Max sequence length: 4096
- Epochs: 2
- Learning rate: 2e-06

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

base = "Qwen/Qwen3-4B-Instruct-2507"

tokenizer = AutoTokenizer.from_pretrained(base)
model = AutoModelForCausalLM.from_pretrained(
    base,
    torch_dtype=torch.float16,
    device_map="auto",
)
```

## Sources & Terms (IMPORTANT)

Training data: u-10bei/sft_alfworld_trajectory_dataset_v4, u-10bei/sft_alfworld_trajectory_dataset_v5, u-10bei/dbbench_sft_dataset_react_v4, u-10bei/dbbench_sft_dataset_react_v3

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.