Science Question Answering on ScienceQA IMG 38
81.3AccuracyMipha-3B
Evaluation Results
| Method | Links | |
|---|---|---|
| Mipha-3BLM=Phi-2 (2.7B), Res.=384, PT=558K, IT=665K2024.07 | 81.3 | |
| MoE-LLaVA-3.6BLM=Phi-2 (2.7B), Res.=384, PT=558k, IT=1.59M2024.07 | 79.9 | |
| TinyLLaVALM=Phi-2 (2.7B), Res.=384, PT=1.2M, IT=665k2024.07 | 79.9 | |
| Bunny-3BLM=Phi-2 (2.7B), Res.=384, PT=2M, IT=695K2024.07 | 79.8 | |
| Imp-v1LM=Phi-2 (2.7B), Res.=384, PT=558K, IT=665K2024.07 | 79.5 | |
| Mipha-3B+ (Ours)LM=Phi-2 (2.7B), Res.=384, PT=558K, IT=665K2024.07 | 71.8 | |
| LLaVA-PhiLM=Phi-2 (2.7B), Res.=336, PT=558k, IT=665K2024.07 | 71.4 | |
| LLaVA-1.5+ (Ours)LM=Vicuna(7B), Res.=336, PT=558K, IT=665K2024.07 | 69.5 | |
| mPLUG-Owl2LM=LLAMA (7B), Res.=448, PT=400M, IT=1.23M2024.07 | 68.7 | |
| Qwen-VL-ChatLM=Qwen (7B), Res.=448, PT=1.4B, IT=50M2024.07 | 68.2 | |
| Qwen-VLLM=Qwen (7B), Res.=448, PT=1.4B, IT=50M2024.07 | 67.1 | |
| LLaVA-1.5LM=Vicuna (7B), Res.=336, PT=558K, IT=665K2024.07 | 66.8 | |
| MC-LLaVALM=Phi-2 (2.7B), Res.=384, PT=558k, IT=665K2024.07 | 64.2 | |
| InstructBLIPLM=Vicuna (13B), Res.=224, PT=129M, IT=1.2M2024.07 | 63.1 | |
| Mobile VLM-v2-3BLM=M-LLaMA (2.7B), Res.=336, PT=1.2M, IT=2.4M2024.07 | 61.1 | |
| BLIP-2LM=Vicuna (13B), Res.=224, PT=129M, IT=-2024.07 | 61 | |
| InstructBLIPLM=Vicuna (7B), Res.=224, PT=129M, IT=1.2M2024.07 | 60.5 | |
| LAF-7BLM=Vicuna (7B), Res.=336, PT=558K, IT=665K2024.07 | 60.1 | |
| Mobile VLM-v2-1.7BLM=M-LLaMA (1.4B), Res.=336, PT=1.2M, IT=2.4M2024.07 | 59.3 | |
| Mobile VLM-3BLM=M-LLaMA (2.7B), Res.=336, PT=558K, IT=665K2024.07 | 59 | |
| Mobile VLM-1.7BLM=M-LLaMA (1.4B), Res.=336, PT=558K, IT=665K2024.07 | 56.1 |