Gabatarwa

Haɗin kai na Artificial Intelligence (AI) cikin aikace-aikacen software ya kawo sauyi yadda muke fuskantar warware matsala, yanke shawara, da sarrafa kansa. Koyaya, yayin da aikace-aikacen AI-kore ke ƙara haɓakawa, buƙatar amintaccen gine-ginen gine-ginen ya zama abin damuwa. A cikin wannan labarin, za mu bincika haɗin kai na AI, injiniyan software, da DevOps, tare da ƙayyadaddun mahimmanci akan yin amfani da kayan aikin orchestration kamar DSPy da LangChain don haɗa ma'aikatan AI a cikin bututun DevOps.

Kalubalen Gine-ginen Software na AI-Driven

Aikace-aikacen software masu amfani da AI suna haifar da ƙalubale na musamman ga injiniyoyin software da ƙungiyoyin DevOps. Wasu daga cikin manyan ƙalubalen sun haɗa da:

• ** Scalability ***: Samfuran AI suna buƙatar mahimman albarkatun ƙididdigewa, wanda zai haifar da lamuran haɓakawa a cikin yanayin samarwa. ** Tsaro ***: Samfuran AI na iya zama masu rauni ga hare-hare, kamar gubar bayanai da harin juyar da samfur. ** Haɗin kai ***: samfuran AI suna buƙatar haɗa su tare da tsarin software na yanzu, wanda zai iya zama ɗawainiya mai rikitarwa. ** Kulawa da Kulawa ***: Samfuran AI suna buƙatar ci gaba da sa ido da kulawa don tabbatar da sun kasance daidai kuma abin dogaro.

Kayan aikin Orchestration: DSPy da LangChain

Kayan aikin Orchestration irin su DSPy da LangChain suna ba da mafita ga waɗannan ƙalubalen ta hanyar ba da damar haɗin kai na wakilan AI cikin bututun DevOps. Waɗannan kayan aikin suna ba da tsari don ginawa, ƙaddamarwa, da sarrafa samfuran AI, tabbatar da amintaccen gine-gine masu ƙima don aikace-aikacen AI-kore.

DSPy

DSPy kayan aikin kade-kade ne na bude-bude da aka gina a saman Python. Yana ba da API mai sauƙi da fahimta don gini, turawa, da sarrafa samfuran AI. DSPy yana goyan bayan faffadan tsarin AI, gami da TensorFlow, PyTorch, da scikit-learn.

LangChain

LangChain wani kayan aiki ne mai buɗe ido wanda ke ba da tsari don gini, turawa, da sarrafa samfuran AI. An gina LangChain a saman Rust kuma yana samar da API mai girma don haɗawa da wakilan AI a cikin bututun DevOps.

Aiwatar da Amintaccen Tsarin Gine-ginen Software na AI-Driven tare da DSPy

A cikin wannan sashe, za mu bincika misali mai sauƙi na aiwatar da amintaccen kuma mai daidaita tsarin gine-ginen software na AI tare da DSPy.

Misali: Gina Samfurin AI mai Sauƙi tare da DSPy

Bari mu yi la'akari da misali mai sauƙi na gina ƙirar koyon injin ta amfani da DSPy. Za mu yi amfani da ɗakin karatu na scikit-koyi don gina ƙirar koma baya ta layi.

`` Python shigo da dspy a matsayin dsp daga sklearn.linear_model shigo da LinearRegression daga sklearn.datasets shigo da load_diabetes

Load ɗin bayanan ciwon sukari

dataset = load_diabetes()

Raba bayanan zuwa tsarin horo da gwaji

X_train, X_test, y_train, y_test = dsp.split_dataset(dataset.data, dataset.target, test_size=0.2)

Gina samfurin koma baya na layi

model = LinearRegression()

Horar da samfurin akan bayanan horo

model.fit (X_train, jirgin kasa)

Kimanta samfurin akan bayanan gwaji

daidaito = model.score (X_test, y_test) bugawa (f" daidaiton samfur: {daidai:.2f}")

Wannan toshe lambar yana nuna yadda ake gina ƙirar koyan injin mai sauƙi ta amfani da DSPy. Muna loda saitin bayanan ciwon sukari, mu raba shi zuwa tsarin horo da gwaji, gina tsarin jujjuyawar layi, horar da shi akan bayanan horo, da kimanta daidaitonsa akan bayanan gwaji.

## Aiwatar da Amintaccen Tsarin Gine-ginen Software na AI-Driven tare da LangChain

A cikin wannan sashe, za mu bincika misali mai sauƙi na aiwatar da amintaccen tsarin gine-ginen software na AI-kore tare da LangChain.

### Misali: Gina Samfuran AI mai Sauƙi tare da LangChain

Bari mu yi la'akari da misali mai sauƙi na gina ƙirar koyon injin ta amfani da LangChain. Za mu yi amfani da ɗakin karatu na TensorFlow don gina ƙirar hanyar sadarwa ta jijiyoyi.

``tsatsa
amfani da langchain kamar lc;
yi amfani da tensorflow::gabatarwa::*;

// Ƙayyade samfurin hanyar sadarwa mai sauƙi
fn model (shigarwar: & Tensor) -> Tensor {
    bari fitarwa = shigarwar.matmul (& Tensor :: sabo (1, 1));
    fitarwa
}

// Ƙirƙiri wakilin LangChain
bari wakili = lc :: Wakili :: sabon (samfurin);

// Ƙayyade saitin bayanai
bari dataset = lc :: Dataset :: sabo(vec![
    lc:: Misali:: sabo(vec![1.0, 2.0], vec![3.0]),
    lc:: Misali:: sabo(vec![4.0, 5.0], vec![6.0]),
]);

// Horon samfurin akan ma'aunin bayanai
agent.train (&dataset);

// Ƙimar samfurin akan bayanan bayanai
bari daidaito = wakili. kimanta (&dataset);
println! ("Tsarin Samfura: {}", daidaito);

Wannan toshe lambar yana nuna yadda ake gina ƙirar koyan injin mai sauƙi ta amfani da LangChain. Muna ayyana ƙirar hanyar sadarwa mai sauƙi, ƙirƙira wakili na LangChain, ayyana saitin bayanai, horar da ƙirar akan saitin bayanai, da kimanta daidaiton sa akan saitin bayanai.

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

Anan akwai manyan 5-10 da aka fi amfani da umarni/snippets don DSPy da LangChain:

DSPy

• dsp.split_dataset(bayanai, manufa, test_size=0.2): Yana raba bayanan zuwa tsarin horo da gwaji.
• dsp.build_model(model_name, model_params): Yana Gina samfurin koyan na'ura ta amfani da ƙayyadadden suna da sigogi.
• `dsp.train_model(samfurin, bayanai): Yana horar da ƙirar na'ura akan takamaiman bayanai.
• `dsp.evaluate_model(samfuri, bayanai): Yana kimanta aikin ƙirar koyan na'ura akan ƙayyadaddun bayanai.
• dsp.deploy_model(samfurin, uwar garken_url): Yana tura samfurin koyo na inji zuwa ƙayyadadden URL na sabar.

LangChain

• `lc :: Wakili :: sabon (samfurin): Yana ƙirƙira wakilin LangChain ta amfani da ƙayyadaddun ƙirar.
• lc :: Dataset :: sababbi(misali): Ƙirƙirar saitin bayanai na LangChain ta amfani da ƙayyadaddun misalan.
• `lc :: jirgin kasa(wakili, dataset): Yana horar da wakilin LangChain akan ƙayyadadden saitin bayanai.
• `lc :: auna (wakili, dataset): Yana kimanta aikin wakilin LangChain akan ƙayyadadden saitin bayanai.
• lc :: tura (wakili, uwar garken_url): Yana tura wakilin LangChain zuwa URL ɗin uwar garken ƙayyadaddun.

Kallon gani

Anan akwai sauƙi mai sauƙi na gine-ginen software da AI ke motsawa ta amfani da Chart.js:

''javascript const ctx = document.getElementById('myChart').getContext('2d'); const ginshiƙi = sabon Chart (ctx, { irin: 'bar', data: { Takaddun: ['Lokacin Horo','Lokacin Ƙira',' Daidaiton Model'], bayanai: [{ Tag: 'DSPy', bayanai: [10, 5, 90], bayaLayi: 'rgba(255, 99, 132, 0.2)', iyakaLauni: 'rgba(255, 99, 132, 1)', iyakar iyaka: 1 }, { Tag: 'LangChain', bayanai: [8, 6, 92], bayaLayi: 'rgba(54, 162, 235, 0.2)', iyakaLauni: 'rgba(54, 162, 235, 1)', iyakar iyaka: 1 }] }, zabin: { sikeli: { y: { startAtZero: gaskiya } } } });

Wannan toshe lambar yana ƙirƙirar ginshiƙi mai sauƙi ta amfani da Chart.js don ganin ayyukan DSPy da LangChain.

## Kammalawa

A cikin wannan labarin, mun bincika hanyar haɗin yanar gizo na AI, injiniyan software, da DevOps, tare da musamman mahimmanci akan yin amfani da kayan aikin ƙungiyar kamar DSPy da LangChain don haɗawa da wakilan AI a cikin bututun DevOps. Mun nuna yadda za a aiwatar da amintaccen tsarin gine-ginen software na AI da ke amfani da shi ta amfani da DSPy da LangChain, kuma mun ba da misali mai sauƙi na gina ƙirar koyon injin ta amfani da kowane kayan aiki. Mun kuma bayar da manyan umarni 5-10 da aka fi amfani da su / snippets ga kowane kayan aiki da sauƙi na gani na gine-ginen software na AI-kore ta amfani da Chart.js.