濟(ji)南2024年7月12日 /美通(tong)社/ -- 近日，2024世界(jie)人工(gong)(gong)智(zhi)(zhi)能(neng)大(da)(da)會暨人工(gong)(gong)智(zhi)(zhi)能(neng)全球(qiu)治(zhi)理高級別會議在上海舉辦，集中呈現大(da)(da)模(mo)型、算力、機器人、自動駕駛等(deng)重點(dian)領(ling)域的創新成(cheng)果。作為國內較(jiao)早探索工(gong)(gong)業(ye)大(da)(da)模(mo)型的主(zhu)力廠商之一，浪潮(chao)云洲"知業(ye)大(da)(da)模(mo)型"獲(huo)評(ping)2024年度工(gong)(gong)業(ye)智(zhi)(zhi)能(neng)創新應用典(dian)型產品。浪潮(chao)云洲工(gong)(gong)業(ye)互聯網首席技術官、副總經理商廣勇受邀出席活動。

當前，隨著以大(da)模型為(wei)代(dai)表(biao)的通(tong)用人工智能技術發(fa)展，工業(ye)大(da)模型已成(cheng)為(wei)新型工業(ye)化新引擎和(he)市場競爭制高點，也是工業(ye)企業(ye)推進數智化轉型重要路徑。知業(ye)大(da)模型是由浪潮云洲開發(fa)的工業(ye)領域(yu)(yu)行業(ye)大(da)模型，融(rong)合云洲工業(ye)互聯網平臺海量工業(ye)數據，在通(tong)用大(da)模型的基(ji)礎上(shang)進行工業(ye)領域(yu)(yu)知識(shi)對齊、增強，具備工業(ye)數據分析、工業(ye)知識(shi)推理、工業(ye)代(dai)碼生成(cheng)等(deng)優勢能力。

在場景應(ying)用上(shang)，基于(yu)國(guo)家級(ji)中小企(qi)(qi)業(ye)(ye)(ye)數字化(hua)(hua)轉(zhuan)型(xing)促進(jin)中心(xin)及中小企(qi)(qi)業(ye)(ye)(ye)數字化(hua)(hua)轉(zhuan)型(xing)試點(dian)城市等沉淀的6000余家企(qi)(qi)業(ye)(ye)(ye)數字化(hua)(hua)診(zhen)斷(duan)數據，浪潮云(yun)洲訓練形成了數字化(hua)(hua)診(zhen)斷(duan)大模型(xing)，目前已經在國(guo)家級(ji)中小企(qi)(qi)業(ye)(ye)(ye)數字化(hua)(hua)轉(zhuan)型(xing)促進(jin)中心(xin)的中小企(qi)(qi)業(ye)(ye)(ye)數字化(hua)(hua)轉(zhuan)型(xing)公共服務平(ping)臺(tai)上(shang)線，面向企(qi)(qi)業(ye)(ye)(ye)提供數字化(hua)(hua)診(zhen)斷(duan)的咨詢和(he)問答服務；國(guo)內(nei)某(mou)鋁行(xing)業(ye)(ye)(ye)龍(long)頭企(qi)(qi)業(ye)(ye)(ye)基于(yu)云(yun)洲知業(ye)(ye)(ye)大模型(xing)平(ping)臺(tai)核心(xin)能力，融合企(qi)(qi)業(ye)(ye)(ye)數據資產打造了知業(ye)(ye)(ye)?鋁業(ye)(ye)(ye)大模型(xing)，并以知業(ye)(ye)(ye)?鋁業(ye)(ye)(ye)大模型(xing)為底(di)座開(kai)發了"鋁行(xing)業(ye)(ye)(ye)知識平(ping)臺(tai)"等智慧應(ying)用平(ping)臺(tai)，實現企(qi)(qi)業(ye)(ye)(ye)生產管理效率提升38%，產線鋁產量(liang)平(ping)均提升12%。

會議期(qi)間，浪潮云洲聯合中信農業(ye)旗下華智生(sheng)(sheng)物(wu)(wu)聯合中國(guo)農業(ye)科學院生(sheng)(sheng)物(wu)(wu)技術研究所(suo)等(deng)(deng)單位共同發起生(sheng)(sheng)物(wu)(wu)育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)大(da)(da)模(mo)型(xing)(xing)(xing)聯合共創啟動儀式，探討智能(neng)(neng)(neng)時代(dai)新趨勢和(he)變革升級(ji)新路徑。生(sheng)(sheng)物(wu)(wu)育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)大(da)(da)模(mo)型(xing)(xing)(xing)將基于種(zhong)(zhong)(zhong)(zhong)(zhong)質(zhi)資源信息，基因(yin)組及多(duo)(duo)組學、表型(xing)(xing)(xing)、環境型(xing)(xing)(xing)等(deng)(deng)多(duo)(duo)模(mo)態大(da)(da)樣本(ben)數據，構建(jian)育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)預訓練(lian)大(da)(da)模(mo)型(xing)(xing)(xing)，為用(yong)戶提供(gong)種(zhong)(zhong)(zhong)(zhong)(zhong)質(zhi)資源精(jing)準評價，育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)親本(ben)智能(neng)(neng)(neng)推薦(jian)(jian)，產量、品(pin)質(zhi)和(he)抗性精(jing)準預測，育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)材料智能(neng)(neng)(neng)鑒定，品(pin)種(zhong)(zhong)(zhong)(zhong)(zhong)最適生(sheng)(sheng)態型(xing)(xing)(xing)推薦(jian)(jian)，育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)知識(shi)問答系(xi)統等(deng)(deng)多(duo)(duo)功能(neng)(neng)(neng)服(fu)務(wu)，實(shi)現生(sheng)(sheng)物(wu)(wu)育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)的智能(neng)(neng)(neng)化，助力我國(guo)生(sheng)(sheng)物(wu)(wu)育(yu)種(zhong)(zhong)(zhong)(zhong)(zhong)高質(zhi)量發展。

下一步(bu)，浪潮云洲(zhou)知業(ye)大(da)模(mo)型(xing)(xing)(xing)將繼續立足行(xing)業(ye)，深(shen)入企業(ye)，打造(zao)"1個基(ji)礎模(mo)型(xing)(xing)(xing)+N個行(xing)業(ye)模(mo)型(xing)(xing)(xing)+X種場景應(ying)用"的模(mo)式，通過"模(mo)型(xing)(xing)(xing)工(gong)具+知識服務(wu)"，助力形(xing)成企業(ye)的"生產大(da)腦"和(he)"經營大(da)腦"，賦能企業(ye)數字化(hua)轉型(xing)(xing)(xing)。