List your company

Scientific Reasoning on MMLU-Sci

54.96ScoreGalactica-30B
25.777633.353840.9348.5062Jan 15, 2024

Evaluation results

MethodLinks
Galactica-30BParameter Scale=30B~32B2024.0154.96
Llama3-8B-InstructEvaluation Protocol=few-shot, Parameter Scale=6B~7B2024.0152.67
ChatGLM3-32B-BaseParameter Scale=30B~32B2024.0150.3
SciGLMBackbone=ChatGLM3-32B-Base, Parameter Scale=30B~32B2024.0149.38
SciGLMBackbone=ChatGLM3-6B-Base, Parameter Scale=6B~7B2024.0145.34
Mistral-7B: MetaMATHEvaluation Protocol=few-shot, Parameter Scale=6B~7B2024.0144.74
Mistral-7B: MetaMATH + SciInstructFine-tuning=SciInstruct, Parameter Scale=6B~7B2024.0142.16
ChatGLM3-6BParameter Scale=6B~7B2024.0141.78
Llama3-8B-Instruct + SciInstructFine-tuning=SciInstruct, Parameter Scale=6B~7B2024.0140.86
ChatGLM3-6B-BaseParameter Scale=6B~7B2024.0140.16
ChatGLM2-6B-BaseParameter Scale=6B~7B2024.0138.06
ChatGLM2-6BParameter Scale=6B~7B2024.0137.09
LLaMA-2-13BParameter Scale=12B~13B2024.0135.85
Vicuna-13BParameter Scale=12B~13B2024.0132.13
Galactica-6.7BParameter Scale=6B~7B2024.0130.68
LLaMA-2-7BParameter Scale=6B~7B2024.0130.41
Mistral-7B: MetaMATHEvaluation Protocol=zero-shot, Parameter Scale=6B~7B2024.0128.25
Llama3-8B-InstructEvaluation Protocol=zero-shot, Parameter Scale=6B~7B2024.0126.9