qwen3-vl-2b-iconclass-swift-test

LoRA adapter for Qwen/Qwen3-VL-2B-Instruct fine-tuned with ms-swift on davanstrien/iconclass-vlm-brillfull (200 training rows, config sft).

epochs	1.0
learning rate	0.0001
LoRA rank / alpha	8 / 32
effective batch	16
max length / pixels	2048 / 1003520
vision tower	frozen
final train loss	1.847952651977539
eval loss	1.8682091236114502
training time	1.6 min

Use

swift infer --model Qwen/Qwen3-VL-2B-Instruct --adapters davanstrien/qwen3-vl-2b-iconclass-swift-test --use_hf true --stream true

Reproduction

Produced on Hugging Face Jobs (gpu) with the swift-vlm-sft.py recipe from uv-scripts. Run it yourself:

hf jobs uv run --flavor a10g-large --timeout 1h --secrets HF_TOKEN https://huggingface.co/datasets/uv-scripts/finetune/raw/main/swift-vlm-sft.py davanstrien/iconclass-vlm-brillfull davanstrien/qwen3-vl-2b-iconclass-swift-test --config sft --max-samples 200 --epochs 1