CoolFace
Apppublic

aail-hf/ensemble_machine

sourceHugging Facemitupdated 2y agoView on Hugging Face
0likes
constants.py59 linesDownload Raw Back to root
1WORD_LIMIT_MIN = 1002WORD_LIMIT_MAX = 5003GROUP_SEPERATION = 10004LOCAL_PARAMS = False5 6SDG_DETAILS = """71. No Poverty: Eradicate all forms of poverty.82. Zero Hunger: End hunger and promote sustainable agriculture.93. Good Health and Well-being: Ensure health and well-being for everyone.104. Quality Education: Guarantee inclusive, quality education for all.115. Gender Equality: Empower women and achieve gender equality.126. Clean Water and Sanitation: Provide sustainable water and sanitation.137. Affordable and Clean Energy: Ensure access to sustainable energy.148. Decent Work and Economic Growth: Foster inclusive economic growth and decent jobs.159. Industry, Innovation and Infrastructure: Build resilient infrastructure and promote innovation.1610. Reduced Inequality: Decrease inequality within and among countries.1711. Sustainable Cities and Communities: Develop safe and sustainable communities.1812. Responsible Consumption and Production: Ensure sustainable consumption and production.1913. Climate Action: Take action against climate change.2014. Life Below Water: Protect oceans and marine life.2115. Life on Land: Sustainably manage ecosystems and biodiversity.2216. Peace and Justice Strong Institutions: Promote peace, justice, and strong institutions.2317. Partnerships to achieve the Goals: Enhance global partnerships for sustainable development.24"""25 26 27GPT_PROMPT_parallel="""28For AI generated/modified parts, our base model is GPT-4o. 29 30For direct generation, we use instructions like following:31"Given the task as: [task_description], provide an answer: "32 33For the merging, we use instructions like following:34"Given the task as: [task_description], there are two answers provided:35The first answer: [human_text]36The second answer: [ai_text]37Merge the two answers into one in a coherent way: "38"""39 40GPT_PROMPT_sequential="""41For AI generated/modified parts, our base model is GPT-4o. 42 43For the modification, we use instructions like following:44"Given the task as :[task_description, the human answer is: [human_text]45Provide an answer of your own but make sure it is coherent and should be based on the human answer: "46"""47 48GPT_PROMPT_reverse_sequential="""49For AI generated/modified parts, our base model is GPT-4o. 50 51For direct generation, we use instructions like following:52"Given the task as: [task_description], provide an answer: "53 54For the merging, we use instructions like following:55"Given the task as: [task_description], there are two answers provided:56The first answer: [human_text]57The second answer: [ai_text]58Merge the two answers into one in a coherent way: "59"""