##Chiyambi

Kupita patsogolo kwachangu mu Artificial Intelligence (AI) ndi uinjiniya wamapulogalamu kwapangitsa kuti pakhale makina ovuta apulogalamu omwe amaphatikiza othandizira a AI ndi kamangidwe kamayendedwe kantchito. Makinawa ali ndi kuthekera kosintha momwe timayendera ma automation ndi orchestration ya DevOps. Komabe, kupanga madongosolo owopsa komanso abwino oyendetsedwa ndi AI kumafuna kumvetsetsa mwakuya zakupita patsogolo kwaposachedwa mu AI, uinjiniya wamapulogalamu, ndi zomangamanga.

M'nkhaniyi, tiwona mfundo zamapangidwe ndi njira zabwino zopangira mapulogalamu amtundu wa modular, scalable, komanso amakhalidwe abwino omwe amaphatikiza othandizira a AI ndi kamangidwe kamayendedwe kantchito. Tidzayang'ana pakuphatikiza kupita patsogolo kwaposachedwa mu AI (Large Language Models (LLMs), Retrieval-Augmented Generation (RAG)), uinjiniya wamapulogalamu (Python, Rust, Go), ndi zomangamanga kuti apange machitidwe olimba komanso osungika.

Zomangamanga za Modular Workflow

Zomangamanga zamayendedwe amodular ndi njira yopangira yomwe imaphwanya machitidwe ovuta kukhala ma module ang'onoang'ono, odziyimira pawokha omwe amatha kusamalidwa, kusinthidwa, komanso kukulitsidwa. Mutu uliwonse uli ndi udindo pa ntchito inayake kapena ntchito inayake, ndipo amalumikizana wina ndi mnzake kudzera m'malo odziwika bwino.

Nachi chitsanzo chosavuta cha kamangidwe kamayendedwe kantchito ku Python:

# workflow.py
Gawo la ntchito:
    def __init__(yekha):
        self.modules = []

    def add_module(yekha, gawo):
        self.modules.append(module)

    def run (self):
        za module mu self.modules:
            module.run()

# module1.py
kalasi Module1:
    def run (self):
        print ("Module 1 ikugwira ntchito")

# module2.py
kalasi Module2:
    def run (self):
        print ("Module 2 ikugwira ntchito")

# chachikulu.py
ngati __name__ == "__main__":
    ntchito = mayendedwe ()
    workflow.add_module(Module1())
    workflow.add_module(Module2())
    workflow.run()

Mu chitsanzo ichi, tili ndi kalasi ya Workflow yomwe imayang'anira mndandanda wa ma module. Gawo lirilonse limagwira ntchito inayake, ndipo amawonjezedwa kumayendedwe a ntchito pogwiritsa ntchito njira ya add_module. Njira ya run imagwiritsidwa ntchito pochita gawo lililonse pakayendetsedwe ka ntchito.

AI Agents

Othandizira AI ndi zida zamapulogalamu zomwe zimagwiritsa ntchito makina ophunzirira makina kupanga zisankho ndikuchitapo kanthu pamakina. Atha kugwiritsidwa ntchito kusinthiratu ntchito, kupititsa patsogolo luso la ogwiritsa ntchito, ndikuwongolera magwiridwe antchito adongosolo.

Pali mitundu ingapo ya othandizira a AI, kuphatikiza:

• Nthawi zotengera malamulo: Othandizirawa amagwiritsa ntchito malamulo omwe adalembedweratu popanga zisankho.
• Othandizira kuphunzira pamakina: Othandizirawa amagwiritsa ntchito njira zophunzirira pamakina kuphunzira kuchokera pa data ndikupanga zisankho.
• Hybrid agents: Othandizirawa amaphatikiza njira zozikidwa pamalamulo ndi makina ophunzirira kupanga zisankho.

Nachi chitsanzo cha wothandizira wa AI wosavuta ku Python pogwiritsa ntchito makina ophunzirira makina:

# wothandizira.py
kuchokera ku sklearn.ensemble import RandomForestClassifier
kuchokera ku sklearn.model_selection import train_test_split
kuchokera ku sklearn.datasets import load_iris

Class Agent:
    def __init__(yekha):
        self.model = RandomForestClassifier()

    def sitima (yekha, X, y):
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
        self.model.fit(X_train, y_train)

    def kulosera (yekha, X):
        bwererani nokha.model.predict(X)

# chachikulu.py
ngati __name__ == "__main__":
    iris = load_iris ()
    X = iris.data
    y = iris.target
    wothandizira = Wothandizira ()
    agent.train(X, y)
    kulosera = agent.predict(X)
    sindikiza (zoneneratu)

Muchitsanzo ichi, tili ndi wothandizira wa AI wosavuta yemwe amagwiritsa ntchito gulu la nkhalango mwachisawawa kuti alosere. Wothandizira amaphunzitsidwa pa dataset ya iris ndipo amalosera pa dataset yomweyo.

Kuphatikiza Ma AI Agents ndi Modular Workflow Architecture

Kuphatikiza ma AI othandizira ndi ma modular workflow architecture amatilola kupanga makina ovuta omwe amaphatikiza kupanga zisankho koyendetsedwa ndi AI ndi zomanga modular, scalable.

Nachi chitsanzo cha dongosolo lomwe limaphatikiza othandizira a AI ndi ma modular workflow mamangidwe:

# workflow.py
Gawo la ntchito:
    def __init__(yekha):
        self.modules = []
        self.agent = Palibe

    def add_module(yekha, gawo):
        self.modules.append(module)

    def add_agent(yekha, wothandizira):
        self.agent = wothandizira

    def run (self):
        za module mu self.modules:
            module.run()
        ngati self.agent:
            self.agent.predict()

# module1.py
kalasi Module1:
    def run (self):
        print ("Module 1 ikugwira ntchito")

# module2.py
kalasi Module2:
    def run (self):
        print ("Module 2 ikugwira ntchito")

# wothandizira.py
kuchokera ku sklearn.ensemble import RandomForestClassifier
kuchokera ku sklearn.model_selection import train_test_split
kuchokera ku sklearn.datasets import load_iris

Class Agent:
    def __init__(yekha):
        self.model = RandomForestClassifier()

def sitima (yekha, X, y):
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
        self.model.fit(X_train, y_train)

    def kulosera (yekha, X):
        bwererani nokha.model.predict(X)

# chachikulu.py
ngati __name__ == "__main__":
    ntchito = mayendedwe ()
    workflow.add_module(Module1())
    workflow.add_module(Module2())
    wothandizira = Wothandizira ()
    agent.train(X, y)
    workflow.add_agent(wothandizira)
    workflow.run()

Muchitsanzo ichi, tili ndi dongosolo lomwe limaphatikiza othandizira a AI ndi ma modular workflow mamangidwe. Dongosololi lili ndi ma module awiri ndi wothandizira wa AI yemwe amagwiritsa ntchito gulu la nkhalango mwachisawawa kuti alosere. Wothandizira amawonjezedwa kumayendedwe a ntchito pogwiritsa ntchito njira ya add_agent, ndipo amalosera ma module akatha.

Apamwamba 5-10 Omwe Amagwiritsidwa Ntchito Kwambiri / Zidutswa

Nawa ena mwa malamulo omwe amagwiritsidwa ntchito kwambiri popanga mapulogalamu owopsa komanso oyendetsedwa ndi AI okhala ndi ma modular workflow architecture ndi othandizira a AI:

1. Python: import pandas as pd (kulowetsa pandas library)
2. Python: pd.read_csv('data.csv') (kuwerenga fayilo ya CSV mu pandas dataframe)
3. Python: agent.train(X, y) (kuphunzitsa wothandizira AI pa dataset)
4. Python: agent.predict(X) (kulosera pogwiritsa ntchito AI)
5. ** Dzimbiri**: let mut agent = Agent::new() (kupanga mtundu watsopano wa AI mu Dzimbiri)
6. ** Dzimbiri**: agent.train(X, y) (kuphunzitsa wothandizira AI pa dataset ku Rust)
7. Pitani: wothandizira := NewAgent() (kupanga wothandizira AI watsopano mu Go)
8. Pitani: agent.Train(X, y) (kuphunzitsa wothandizira AI pa data mu Go)
9. Chart.js: const chart = new Chart(ctx, {mtundu: 'bar', data: {labels: ['Label 1', 'Label 2'], datasets: [{ label: 'Dataset 1', data: [10, 20] }] } }); (kupanga tchati cha bar) pogwiritsa ntchito tchati.j
10. Chart.js: chart.data.datasets[0].data.push(30) (kuwonjezera mfundo yatsopano ku tchati pogwiritsa ntchito Chart.js)

# Kuwona

Nachi chitsanzo cha tchati chogwiritsa ntchito Chart.js:

{
  "type": "bar",
  "data": {
    "labels": ["Label 1", "Label 2"],
    "ma dataset": [
      {
        "label": "Dataset 1",
        "data": [10, 20]
      }
    ]
  }
}

Tchatichi chili ndi zilembo ziwiri ndi seti imodzi yokhala ndi mfundo ziwiri. Tchaticho chikhoza kusinthidwa pogwiritsa ntchito zosankha ndi zoikamo zosiyanasiyana.

##Mapeto

Kupanga machitidwe owopsa komanso abwino omwe amayendetsedwa ndi AI okhala ndi ma modular workflow architecture ndi othandizira a AI amafunikira kumvetsetsa kwakuya zakupita patsogolo kwaposachedwa mu AI, uinjiniya wamapulogalamu, ndi zomangamanga. Mwa kuphatikiza othandizira a AI ndi ma modular workflow architecture, titha kupanga machitidwe ovuta omwe amaphatikiza kupanga zisankho koyendetsedwa ndi AI ndi zomanga modular, scalable.

M'nkhaniyi, tawonanso mfundo zamapangidwe ndi njira zabwino zopangira mapulogalamu a modular, scalable, komanso amakhalidwe abwino omwe amaphatikiza othandizira a AI ndi kamangidwe kamayendedwe kantchito. Taperekanso zitsanzo za othandizira a AI osavuta komanso kamangidwe kamayendedwe kantchito ku Python, Rust, ndi Go, komanso chitsanzo chowonera pogwiritsa ntchito Chart.js.

Potsatira malangizo ndi machitidwe abwino omwe afotokozedwa m'nkhaniyi, opanga mapulogalamu amatha kupanga mapulogalamu olimba komanso osasunthika omwe amayendetsedwa ndi AI omwe amakwaniritsa zofunikira za mapulogalamu ndi mafakitale amakono.