Gabatarwa

Ci gaba cikin sauri a cikin Intelligence Artificial (AI) da injiniyan software sun haifar da haɓaka tsarin tsarin software masu rikitarwa waɗanda ke haɗa wakilan AI da tsarin gine-ginen aiki na zamani. Waɗannan tsarin suna da yuwuwar yin juyin juya hali yadda muke kusanci DevOps aiki da kai da kade-kade. Koyaya, ƙirƙira ma'auni da tsarin software na AI-kore yana buƙatar zurfin fahimtar sabbin ci gaba a cikin AI, injiniyan software, da gine-gine.

A cikin wannan labarin, za mu bincika ƙa'idodin ƙira da mafi kyawun ayyuka don ƙirƙirar tsarin software na zamani, mai daidaitawa, da ɗabi'a waɗanda ke haɗa wakilan AI da tsarin gine-ginen aiki na zamani. Za mu mai da hankali kan haɗa sabbin ci gaba a cikin AI (Large Language Models (LLMs), Retrieval-Augmented Generation (RAG)), injiniyan software (Python, Tsatsa, Go), da gine-gine don ƙirƙirar tsarukan ƙarfi da kiyayewa.

Modular Workflow Architecture

Tsarin gine-ginen aiki na yau da kullun tsari ne na ƙira wanda ke rushe hadaddun tsarin zuwa ƙanƙanta, na'urori masu zaman kansu waɗanda za'a iya kiyaye su cikin sauƙi, sabunta su, da ƙima. Kowane tsari yana da alhakin takamaiman aiki ko aiki, kuma suna sadarwa tare da juna ta hanyar ingantattun hanyoyin sadarwa.

Anan ga misali mai sauƙi na tsarin gine-ginen aiki na zamani a Python: `` Python

aikin aiki.py

Tsarin Aiki na aji: def init(kai): self.modules = []

def add_module (kai, module):
    kai.modules.append (module)

def run(self):
    don module a self.modules:
        module.run()

module1.py

Module 1: def run(self): buga ("Module 1 yana gudana")

module2.py

Module 2: def run(self): buga ("Module 2 yana gudana")

babban.py

idan name == "babban": aiki = Gudun aiki () aikin aiki.add_module(Module1()) aikin aiki.add_module (Module2()) aikin aiki.run ()

A cikin wannan misali, muna da ajin 'Tsarin Aiki' wanda ke sarrafa jerin kayayyaki. Kowane samfurin yana da alhakin takamaiman aiki, kuma ana ƙara su zuwa aikin aiki ta hanyar 'add_module'. Ana amfani da hanyar 'run' don aiwatar da kowane tsari a cikin aikin aiki.

## AI Agents

Ma'aikatan AI su ne kayan aikin software waɗanda ke amfani da algorithms koyon injin don yanke shawara da ɗaukar ayyuka a cikin tsarin. Ana iya amfani da su don sarrafa ayyuka, haɓaka ƙwarewar mai amfani, da haɓaka aikin tsarin.

Akwai nau'ikan wakilan AI da yawa, gami da:

**Wakilan tushen doka**: Waɗannan wakilai suna amfani da ƙayyadaddun ƙa'idodi don yanke shawara.
**Masu aikin koyo na inji**: Waɗannan wakilai suna amfani da algorithms na koyon injin don koyo daga bayanai da yanke shawara.
* **Masu haɗe-haɗe**: Waɗannan wakilai suna haɗa hanyoyin koyon ƙa'ida da na'ura don yanke shawara.

Anan akwai misalin wakilin AI mai sauƙi a cikin Python ta amfani da algorithm koyan inji:
`` Python
# wakili.py
daga sklearn.ensemble shigo da RandomForestClassifier
daga sklearn.model_selection shigo da jirgin kasa_test_split
daga sklearn.datasets shigo da load_iris

Wakilin aji:
    def __init__(kai):
        self.model = RandomForestClassifier()

    def jirgin kasa (kai, X, y):
        X_train, X_test, y_train, y_test = jirgin_test_split (X, y, test_size=0.2, random_state=42)
        self.model.fit (X_train, jirgin kasa)

    def tsinkaya (kai, X):
        mayar da kai.model.predict(X)

# babban.py
idan __name__ == "__babban__":
    iris = load_iris()
    X = iris.data
    y = iris. manufa
    wakili = wakili ()
    agent.train (X, y)
    tsinkaya = wakili.predict(X)
    bugu ( tsinkaya)

A cikin wannan misalin, muna da wakilin AI mai sauƙi wanda ke amfani da mai rarraba gandun daji don yin tsinkaya. An horar da wakili a kan bayanan iris kuma yana yin tsinkaya akan saitin bayanai iri ɗaya.

Haɗin Agents AI da Tsarin Gine-gine na Aiki na Modular

Haɗa wakilan AI da tsarin gine-ginen aiki na yau da kullun yana ba mu damar ƙirƙirar rikitattun tsare-tsare waɗanda ke haɗa shawarar yanke shawara ta AI tare da ƙirar ƙira mai ƙima.

Anan akwai misalin tsarin da ya haɗu da wakilan AI da tsarin gine-ginen aiki na zamani: `` Python

aikin aiki.py

Tsarin Aiki na aji: def init(kai): self.modules = [] kai.agent = Babu

def add_module (kai, module):
    kai.modules.append (module)

def add_agent (kai, wakili):
    kai.agent = wakili

def run(self):
    don module a self.modules:
        module.run()
    idan kai.agent:
        kai.agent.predict()

module1.py

Module 1: def run(self): buga ("Module 1 yana gudana")

module2.py

Module 2: def run(self): buga ("Module 2 yana gudana")

wakili.py

daga sklearn.ensemble shigo da RandomForestClassifier daga sklearn.model_selection shigo da jirgin kasa_test_split daga sklearn.datasets shigo da load_iris

Wakilin aji: def init(kai): self.model = RandomForestClassifier()

def jirgin kasa (kai, X, y):
    X_train, X_test, y_train, y_test = jirgin_test_split (X, y, test_size=0.2, random_state=42)
    self.model.fit (X_train, jirgin kasa)

def tsinkaya (kai, X):
    mayar da kai.model.predict(X)

babban.py

idan name == "babban": aiki = Gudun aiki () aikin aiki.add_module(Module1()) aikin aiki.add_module (Module2()) wakili = wakili () agent.train (X, y) workflow.add_agent (wakili) aikin aiki.run ()

A cikin wannan misalin, muna da tsarin da ya haɗu da wakilai AI da tsarin gine-ginen aiki na zamani. The system has two modules and an AI agent that uses a random forest classifier to make predictions. Ana ƙara wakili zuwa aikin aiki ta amfani da hanyar `add_agent`, kuma yana yin tsinkaya bayan kayan aikin sun gudana.

## Manyan Dokoki 5-10 da Akafi Amfani da su/Snippets

Anan akwai wasu umarni da aka fi amfani da su da snippets don ƙirƙira madaidaitan tsarin software na AI-kore tare da tsarin gine-ginen aiki na zamani da wakilan AI:

1. **Python**: `shigo da pandas azaman pd` (shigo da ɗakin karatu na pandas)
2. **Python**: `pd.read_csv('data.csv')` (karanta fayil ɗin CSV cikin tsarin bayanan pandas)
3. **Python**: `agent.train (X, y)` (horar da wani AI a kan wani dataset)
4. ** Python **: `agent.predict (X)` (yin tsinkaya ta amfani da wakilin AI)
5. ** Tsatsa ***: `bar mut wakili = Wakili :: sabon ()` (ƙirƙirar sabon wakili na AI a cikin Rust)
6. ** Tsatsa ***: `agent.train (X, y)` (horar da wani AI a kan wani dataset a Tsatsa)
7. ** Tafi **: `wakili: = NewAgent ()` (ƙirƙirar sabon wakilin AI a cikin Go)
8. **Tafi**: `wakili.Train (X, y)` (horar da wani AI a kan dataset a Go)
9. **Chart.js**: `const ginshiƙi = sabon Chart (ctx, {nau'in: 'bar', data: {label: ['Label 1', 'Label 2'], datasets: [{label: 'Dataset 1', data: [10, 20]}]]}}});" (ƙirƙirar taswirar bar).
10. **Chart.js**: `chart.data.datasets[0].data.push(30)` (ƙara sabon ma'anar bayanai zuwa ginshiƙi ta amfani da Chart.js)

## Kallon gani

Ga misalin ginshiƙi mai amfani da Chart.js:
Json
{
  "type": "bar",
  "data": {
    "labarai": ["Label 1", "Label 2"],
    "datasets": [
      {
        "label": "Dataset 1",
        "bayanai": [10, 20]
      }
    ]
  }
}

Wannan ginshiƙi yana da lakabi biyu da saitin bayanai ɗaya tare da maki biyu. Za a iya keɓance ginshiƙi ta amfani da zaɓuɓɓuka da saituna daban-daban.

Kammalawa

Ƙirƙirar tsarin software mai ƙima da ɗabi'a na AI tare da tsarin gine-ginen aiki na yau da kullun da wakilan AI na buƙatar zurfin fahimtar sabbin ci gaba a cikin AI, injiniyan software, da gine-gine. Ta hanyar haɗa wakilan AI da tsarin gine-ginen aiki na yau da kullun, za mu iya ƙirƙirar rikitattun tsare-tsare waɗanda ke haɗa shawarar yanke shawara ta AI tare da ƙirar ƙira mai ƙima.

A cikin wannan labarin, mun bincika ƙa'idodin ƙira da mafi kyawun ayyuka don ƙirƙirar tsarin software na zamani, mai daidaitawa, da ɗabi'a waɗanda ke haɗa wakilan AI da tsarin gine-ginen aiki na zamani. Mun kuma bayar da misalan wakilai masu sauƙi na AI da tsarin gine-ginen aiki na yau da kullun a cikin Python, Rust, da Go, da kuma misalin gani ta amfani da Chart.js.

Ta bin jagororin da mafi kyawun ayyuka da aka zayyana a cikin wannan labarin, masu haɓakawa za su iya ƙirƙirar ƙaƙƙarfan tsarin software na AI-kore waɗanda ke biyan bukatun aikace-aikacen zamani da masana'antu.