Nhanganyaya
Iko kukurumidza kufambira mberi muArtificial Intelligence (AI) uye software engineering zvakakonzera kugadzirwa kweakaomesesa software masisitimu anobatanidza maAI vamiririri uye modular workflow architecture. Aya masisitimu ane mukana wekushandura nzira yatinosvika nayo DevOps otomatiki uye orchestration. Nekudaro, kugadzira scalable uye ehutsika AI-inotyairwa software masisitimu inoda kunzwisisa kwakadzama kweazvino kufambira mberi muAI, software engineering, uye zvivakwa.
Muchinyorwa chino, isu tichaongorora misimboti yekugadzira uye akanakisa maitiro ekugadzira modular, scalable, uye ehutsika software masisitimu anobatanidza AI vamiririri uye modular workflow architecture. Tichatarisa kusanganisa kufambira mberi kwazvino muAI (Makuru Mutauro Models (LLMs), Retrieval-Augmented Generation (RAG)), software engineering (Python, Rust, Go), uye zvivakwa zvekugadzira akasimba uye anochengeteka masisitimu.
Modular Workflow Architecture
Modular workflow architecture idhizaini yedhizaini inopwanya masisitimu akaomarara kuita madiki, akazvimirira mamodule anogona kuchengetedzwa zviri nyore, kuvandudzwa, uye kuyerwa. Imwe neimwe module ine basa rerimwe basa kana basa, uye ivo vanotaurirana kuburikidza neakanyatso kutsanangurwa nzvimbo.
Heino muenzaniso wakapusa we modular workflow architecture muPython:
# workflow.py
kirasi Workflow:
def __init__(iwe):
self.modules = []
def add_module(pachako, module):
self.modules.append(module)
def run (iwe pachako):
yemodule mune self.modules:
module.run()
# module1.py
kirasi Module1:
def run (iwe pachako):
print ("Module 1 iri kushanda")
# module2.py
kirasi Module2:
def run (iwe pachako):
print ("Module 2 iri kushanda")
# chikuru.py
kana __zita__ == "__main__":
workflow = Workflow ()
workflow.add_module(Module1())
workflow.add_module(Module2())
workflow.run()
Mumuenzaniso uyu, isu tine Workflow kirasi inogadzirisa rondedzero yemamodule. Imwe neimwe module ine basa rerimwe basa, uye ivo vanowedzerwa kune mafambiro ebasa vachishandisa iyo add_module nzira. Iyo run nzira inoshandiswa kuita imwe neimwe module mukufambiswa kwebasa.
AI Agents
AI vamiririri zvikamu zvesoftware zvinoshandisa muchina kudzidza algorithms kuita sarudzo uye kutora zviito muhurongwa. Anogona kushandiswa kuita otomatiki mabasa, kuvandudza ruzivo rwemushandisi, uye kukwidziridza mashandiro ehurongwa.
Kune akati wandei marudzi eAI vamiririri, kusanganisira:
- Rule-based agents: Vamiririri ava vanoshandisa mitemo yakafanotsanangurwa kuita sarudzo.
- ** Muchina wekufunda mumiririri **: Vamiririri ava vanoshandisa muchina kudzidza algorithms kudzidza kubva kudata uye kuita sarudzo.
- Hybrid agents: Vamiririri ava vanobatanidza mitemo-yakavakirwa uye nzira dzekudzidza muchina kuita sarudzo.
Heino muenzaniso weakareruka AI mumiriri muPython achishandisa muchina kudzidza algorithm:
# mumiririri.py
kubva sklearn.ensemble import RandomForestClassifier
kubva sklearn.model_selection import train_test_split
kubva sklearn.datasets import load_iris
kirasi Agent:
def __init__(iwe):
self.model = RandomForestClassifier()
def chitima(pachako, 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 kufanotaura (pachako, X):
return self.model.predict(X)
# chikuru.py
kana __zita__ == "__main__":
iris = load_iris()
X = iris.data
y = iris.target
mumiririri = Mumiririri ()
agent.train(X, y)
kufanotaura = agent.predict(X)
dhinda (predictions)
Mumuenzaniso uyu, isu tine yakapfava AI mumiririri anoshandisa isina kujairika sango classifier kuita fungidziro. Mumiririri anodzidziswa pane iris dataset uye anoita fungidziro pane imwechete dhata.
Kubatanidza AI Agents uye Modular Workflow Architecture
Kubatanidza maAI agents uye modular workflow architecture inotibvumira kugadzira masisitimu akaomarara anobatanidza AI-inotungamirwa kuita sarudzo ine modular, scalable architecture.
Heino muenzaniso weiyo system inosanganisa AI vamiririri uye modular workflow architecture:
# workflow.py
kirasi Workflow:
def __init__(iwe):
self.modules = []
self.agent = Hapana
def add_module(pachako, module):
self.modules.append(module)
def add_agent(iwe pachako, mumiririri):
self.agent = mumiririri
def run (iwe pachako):
yemodule mune self.modules:
module.run()
kana self.agent:
self.agent.predict()
# module1.py
kirasi Module1:
def run (iwe pachako):
print ("Module 1 iri kushanda")
# module2.py
kirasi Module2:
def run (iwe pachako):
print ("Module 2 iri kushanda")
# mumiririri.py
kubva sklearn.ensemble import RandomForestClassifier
kubva sklearn.model_selection import train_test_split
kubva sklearn.datasets import load_iris
kirasi Agent:
def __init__(iwe):
self.model = RandomForestClassifier()
def chitima(pachako, 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 kufanotaura (pachako, X):
return self.model.predict(X)
# chikuru.py
kana __zita__ == "__main__":
workflow = Workflow ()
workflow.add_module(Module1())
workflow.add_module(Module2())
mumiririri = Mumiririri ()
agent.train(X, y)
workflow.add_agent(mumiririri)
workflow.run()
Mumuenzaniso uyu, isu tine sisitimu inosanganisa maAI agents uye modular workflow architecture. Iyo sisitimu ine mamodule maviri uye mumiriri weAI uyo anoshandisa isina kurongeka sango classifier kuita fungidziro. Mumiririri anowedzerwa kune mafambiro ebasa uchishandisa add_agent nzira, uye inofembera mushure mekunge ma module apera.
Yepamusoro 5-10 Yakanyanya Kushandiswa Mirairo / Zvimedu
Heano mamwe einonyanya kushandiswa mirairo uye snippets kugadzira scalable uye tsika AI-inotyairwa software masisitimu ane modular workflow architecture uye AI vamiririri:
- Python:
import pandas as pd(importing the pandas library) - Python:
pd.read_csv('data.csv')(kuverenga faira reCSV mupandas dataframe) - Python:
agent.train(X, y)(kudzidzisa mumiririri weAI pane dataset) - Python:
agent.predict(X)(kuita fungidziro uchishandisa mumiririri weAI) - Ngura:
regai mut mumiririri = Mumiririri::mutsva()(kugadzira mutsva weAI muRust) - Ngura:
agent.train(X, y)(kudzidzisa mumiririri weAI pane dataset muRust) - Enda:
mumiririri := NewAgent()(kugadzira mutsva weAI muGo) - Enda:
agent.Train(X, y)(kudzidzisa mumiririri weAI pane dataset muGo) - Chati.js:
const chart = new Chart(ctx, { type: 'bar', data: { labels: ['Label 1', 'Label 2'], datasets: [{ label: 'Dataset 1', data: [10, 20] }] }});(kugadzira bhawa chati) uchishandisa Chati - Chati.js:
chart.data.datasets[0].data.push(30)(kuwedzera poindi itsva yedata pachati uchishandisa Chart.js)
Kuona
Heino muenzaniso webhawa chati uchishandisa Chart.js:
{
"type": "bha",
"data": {
"labels": ["Label 1", "Label 2"],
"datasets": [
{
"label": "Dataset 1",
"data": [10, 20]
}
]
}
}
Iyi chati ine mavara maviri uye imwe dataset ine mapoinzi maviri edata. Iyo chati inogona kugadzirwa uchishandisa akasiyana sarudzo uye marongero.
Mhedziso
Kugadzira scalable uye tsika AI-inotyairwa software masisitimu ane modular workflow architecture uye AI vamiririri inoda kunzwisisa kwakadzama kwekufambira mberi kwazvino muAI, software engineering, uye zvivakwa. Nekusanganisa maAI agents uye modular workflow architecture, tinogona kugadzira masisitimu akaomarara anobatanidza AI-inotungamirwa kuita sarudzo ne modular, scalable architecture.
Muchinyorwa chino, takaongorora misimboti yekugadzira uye akanakisa maitiro ekugadzira modular, scalable, uye ehutsika software masisitimu anobatanidza maAI vamiririri uye modular workflow architecture. Isu takapawo mienzaniso yemanyorerwo eAI akareruka uye modular workflow architecture muPython, Rust, uye Go, pamwe nemuenzaniso wekuona uchishandisa Chart.js.
Nekutevera nhungamiro uye maitiro akanakisa anotsanangurwa muchinyorwa chino, vagadziri vanogona kugadzira yakasimba uye inochengeteka AI-inotyairwa software masisitimu inosangana nezvinodiwa zvemazuva ano maapplication nemaindasitiri.