yamayos/qwen3-4b-structured-output-lora-comp-2026feb-run09

basemodel: unsloth/qwen3-4b-instruct-2507-unsloth-bnb-4bit libraryname: peft pipeline_tag: text-generation tags:

• —lora
• —transformers
• —unsloth
• —structured-output ---

Model Card for structeval RUN09

This adapter was trained for the StructEval competition using supervised fine-tuning (SFT).

Base model

unsloth/qwen3-4b-instruct-2507-unsloth-bnb-4bit

Training configuration

RUNID: structeval202603011139223a8b212d

Learning rate: 2.4e-6 Epochs: 2 Max seq length: 1536 Batch size: 2 Gradient accumulation: 4

LoRA: r = 64 alpha = 128 dropout = 0.0

Upsampling: {"toml": 3.0}

Validation loss: 0.9636380076408386

Purpose

This LoRA adapter was trained to improve structured output generation (JSON / YAML / TOML / XML / CSV) for the StructEval competition.

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base = "unsloth/qwen3-4b-instruct-2507-unsloth-bnb-4bit"
adapter = "yamayos/qwen3-4b-structured-output-lora-comp-2026feb-run09"

model = AutoModelForCausalLM.from_pretrained(base)
model = PeftModel.from_pretrained(model, adapter)

tokenizer = AutoTokenizer.from_pretrained(base)