--- license: apache-2.0 language: - en base_model: - Qwen/Qwen2.5-0.5B - facebook/dinov2-small pipeline_tag: visual-question-answering tags: - multimodal --- pretrain stage only, 1180 epochs [MMMU](https://tinyllava-factory.readthedocs.io/en/latest/Evaluation.html#mmmu) | Category | # Samples | Accuracy | |---------------------------------|-----------|----------| | Overall | 900 | 0.303 | | Overall-Art and Design | 120 | 0.308 | | Art | 30 | 0.200 | | Art Theory | 30 | 0.300 | | Design | 30 | 0.400 | | Music | 30 | 0.333 | | Overall-Business | 150 | 0.247 | | Accounting | 30 | 0.200 | | Economics | 30 | 0.333 | | Finance | 30 | 0.267 | | Management | 30 | 0.267 | | Marketing | 30 | 0.167 | | Overall-Science | 150 | 0.253 | | Biology | 30 | 0.233 | | Chemistry | 30 | 0.200 | | Geography | 30 | 0.200 | | Math | 30 | 0.267 | | Physics | 30 | 0.367 | | Overall-Health and Medicine | 150 | 0.327 | | Basic Medical Science | 30 | 0.433 | | Clinical Medicine | 30 | 0.233 | | Diagnostics and Laboratory Med. | 30 | 0.200 | | Pharmacy | 30 | 0.400 | | Public Health | 30 | 0.367 | | Overall-Humanities and Soc. Sci.| 120 | 0.367 | | History | 30 | 0.367 | | Literature | 30 | 0.567 | | Sociology | 30 | 0.333 | | Psychology | 30 | 0.200 | | Overall-Tech and Engineering | 210 | 0.324 | | Agriculture | 30 | 0.300 | | Architecture and Engineering | 30 | 0.300 | | Computer Science | 30 | 0.267 | | Electronics | 30 | 0.267 | | Energy and Power | 30 | 0.467 | | Materials | 30 | 0.500 | | Mechanical Engineering | 30 | 0.167 |