JakeyWakey/

pixmo-qwen3.5-0.8b-distillation

This dataset contains 1,497 single-image examples selected from allenai/pixmo-ask-model-anything. Each row includes the validated image, original question, reference answer, and a response from Qwen/Qwen3.5-0.8B for behavior-preservation experiments during LoRA fine-tuning.

from datasets import load_dataset

dataset = load_dataset("JakeyWakey/pixmo-qwen3.5-0.8b-distillation", split="train") example = dataset[0]… See the full description on the dataset page: https://huggingface.co/datasets/JakeyWakey/pixmo-qwen3.5-0.8b-distillation.