The InstructBLIP model was proposed in InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning by Wenliang Dai, Junnan Li, Dongxu Li, Anthony Meng Huat Tiong, Junqi Zhao, Weisheng Wang, Boyang Li, Pascale Fung, Steven Hoi. InstructBLIP leverages the BLIP-2 architecture for visual instruction tuning.
The abstract from the paper is the following:
General-purpose language models that can solve various language-domain tasks have emerged driven by the pre-training and instruction-tuning pipeline. However, building general-purpose vision-language models is challenging due to the increased task discrepancy introduced by the additional visual input. Although vision-language pre-training has been widely studied, vision-language instruction tuning remains relatively less explored. In this paper, we conduct a systematic and comprehensive study on vision-language instruction tuning based on the pre-trained BLIP-2 models. We gather a wide variety of 26 publicly available datasets, transform them into instruction tuning format and categorize them into two clusters for held-in instruction tuning and held-out zero-shot evaluation. Additionally, we introduce instruction-aware visual feature extraction, a crucial method that enables the model to extract informative features tailored to the given instruction. The resulting InstructBLIP models achieve state-of-the-art zero-shot performance across all 13 held-out datasets, substantially outperforming BLIP-2 and the larger Flamingo. Our models also lead to state-of-the-art performance when finetuned on individual downstream tasks (e.g., 90.7% accuracy on ScienceQA IMG). Furthermore, we qualitatively demonstrate the advantages of InstructBLIP over concurrent multimodal models.
InstructBLIP architecture. Taken from the original paper.
This model was contributed by nielsr. The original code can be found here.
InstructBLIP uses the same architecture as BLIP-2 with a tiny but important difference: it also feeds the text prompt (instruction) to the Q-Former.
[[autodoc]] InstructBlipConfig - from_vision_qformer_text_configs
[[autodoc]] InstructBlipVisionConfig
[[autodoc]] InstructBlipQFormerConfig
[[autodoc]] InstructBlipProcessor
[[autodoc]] InstructBlipVisionModel - forward
[[autodoc]] InstructBlipQFormerModel - forward
[[autodoc]] InstructBlipForConditionalGeneration - forward - generate