Dala isibonelo se-LangChain LLM

Ukusebenzisa I-Architecture evikelekile ngokuzenzakalelayo yezinhlelo zokusebenza zewebhu ezinikwe amandla yi-AI

Isingeniso

Njengoba izinhlelo zokusebenza zewebhu ezinikwe amandla yi-AI ziya ngokuya zanda, isidingo sezakhiwo eziqinile nezivikelekile asikaze sicindezeleke kakhulu. Ngokukhula kokuhlasela okuqhutshwa yi-AI, kubalulekile ukusebenzisa indlela evikelekile ngokuzenzakalelayo engamelana nezinsongo ezingaba khona. Kulesi sihloko, sizohlola ukuthi singahlanganisa kanjani isizindalwazi somthombo ovulekile ovikelekile we-Percona nendlela ka-Chainguard evikelekile ngokuzenzakalelayo futhi sihlanganiswe ne-LangChain yezinhlelo zokusebenza zewebhu ezinikwe amandla yi-AI.

Izinsongo ze-AI-Powered Web Application

Izinhlelo zokusebenza zewebhu ezinikwe amandla yi-AI zisengozini yezinsongo ezihlukahlukene, okuhlanganisa:

• Ubuthi Bedatha: Amamodeli e-AI angasetshenziswa ngokuwapha idatha enonya, okuholela ekuqaguleni okungalungile noma izinqumo.
• Ukuguqulwa Kwemodeli: Abahlaseli bangadlulisela ulwazi olubucayi mayelana nemodeli ye-AI noma idatha yayo yokuqeqeshwa.
• Ukuhlasela Okuphikisayo: Amamodeli e-AI angakhohliswa okokufaka okuklanywe ngokucophelela okudizayinelwe ukudukisa imodeli.

I-Architecture evikelekile ngokuzenzakalelayo

Ukuze sinciphise lezi zinsongo, sizosebenzisa ukwakheka okuvikelekile ngokuzenzakalela okuhlanganisa isizindalwazi esivikelekile somthombo ovulekile we-Percona nendlela ka-Chainguard evikelekile ngokuzenzakalelayo futhi ehlanganisa ne-LangChain yezinhlelo zokusebenza zewebhu ezinikwe amandla yi-AI.

I-Percona's Secure Open-Source Database

I-Percona inikezela ngohlu lwemininingwane evikelekile yomthombo ovulekile, okuhlanganisa i-Percona Server ye-MySQL ne-Percona Server ye-PostgreSQL. Lezi zingosi zolwazi zihlinzeka ngezici zokuphepha eziqinile, njenge:

• Ukubethela: Idatha ibethelwa lapho uphumule futhi isendleleni, iqinisekisa ukuthi ulwazi olubucayi luhlala luyimfihlo.
• Ukulawula Ukufinyelela: Ukulawulwa kokufinyelela okuhlaziywe kahle kuvumela abalawuli ukuthi bakhawulele ukufinyelela kudatha ebucayi futhi baqinisekise ukuthi izisebenzi ezigunyaziwe kuphela ezingashintsha noma zisuse idatha.
• Ukucwaningwa kwamabhuku: Amandla okucwaninga okuningiliziwe anikeza irekhodi eliphelele lawo wonke umsebenzi wesizindalwazi, okwenza kube lula ukuthola nokuphendula izehlakalo zokuphepha.

Indlela Ka-Chainguard Evikelekile Ngokuzenzakalelayo

I-Chainguard inikeza indlela evikelekile ngokuzenzakalelayo eqinisekisa ukuthi izinhlelo zokusebenza zakhiwe ngokuvikeleka engqondweni kusukela ekuqaleni. Le ndlela ihlanganisa:

• Ilungelo Elincane: Izinhlelo zokusebenza ziklanyelwe ukusebenza ngelungelo elincane elidingekayo ukuze zenze imisebenzi yazo, zinciphise indawo yokuhlasela.
• Ingqalasizinda Engaguquki: Ingqalasizinda yakhelwe ngendlela engashintsheki, okwenza kube lula ukuthola nokusabela ezigamekweni zokuphepha.
• Ukucushwa Okuvikelekile: Izinhlelo zokusebenza zilungiswa kucatshangelwa ukuvikeleka, okuhlanganisa okumisiwe okuvikelekile kanye nokucushwa okuncane.

LangChain Integration

I-LangChain iwumtapo wezincwadi wePython ohlinzeka nge-API elula nenembile yokwakha izinhlelo zokusebenza zewebhu ezine-AI. Sizohlanganisa i-LangChain nesizindalwazi esivikelekile somthombo ovulekile we-Percona kanye nendlela ka-Chainguard evikelekile ngokuzenzakalelayo ukuze sakhe isakhiwo esiqinile nesivikelekile.

Isibonelo sekhodi

Nasi isibonelo sekhodi esilula esibonisa indlela yokuhlanganisa i-LangChain nesizindalwazi esivikelekile somthombo ovulekile we-Percona kanye nendlela ka-Chainguard evikelekile ngokuzenzakalelayo:

ngenisa i-langchain
kusuka ku-langchain yokungenisa i-LLM


llm = LLM(
    model_name="davinci:2-large",
    idivayisi="cuda",
    max_tokens=1024,
)

# Dala uxhumano lwedatha yePercona
ngenisa i-mysql.connector
cnx = mysql.connector.connect(
    umsebenzisi='igama lomsebenzisi',
    iphasiwedi='iphasiwedi',
    umphathi='127.0.0.1',
    isizindalwazi='isizindalwazi'
)

# Dala ukucushwa okuvikelekile kwe-Chainguard
import chainguard
cfg = chainguard.Config(
    app_name="my_app",
    app_version="1.0",
    env_vars={
        "DB_HOST": "127.0.0.1",
        "DB_USER": "igama lomsebenzisi",
        "DB_PASSWORD": "iphasiwedi",
    },
)

# Sebenzisa i-LangChain ukubuza imininingwane egciniwe
umbuzo = "KHETHA * KUSUKA_kuthebula lami"
umphumela = llm.query(umbuzo, db=cnx, cfg=cfg)

# Phrinta umphumela
phrinta(umphumela)

Lesi sibonelo sekhodi sibonisa ukuthi ungakha kanjani isibonelo se-LangChain LLM, uxhume kusizindalwazi se-Percona, futhi udale ukucushwa okuvikelekile kwe-Chainguard. Ibese isebenzisa i-LangChain ukubuza kusizindalwazi bese iphrinta umphumela.

Imiyalo/Amazwibela Aphezulu Okuyi-5-10

Nayi imiyalo/ amazwibela aphezulu angu-5-10 kulobu buchwepheshe:

1. langchain.LLM(model_name="davinci:2-large", device="cuda", max_tokens=1024) - Dala isibonelo se-LangChain LLM.
2. mysql.connector.connect(umsebenzisi='igama lomsebenzisi', iphasiwedi='iphasiwedi', umsingathi='127.0.0.1', isizindalwazi='isizindalwazi') - Dala uxhumano lwesizindalwazi se-Percona.
3. chainguard.Config(app_name="my_app", app_version="1.0", env_vars={"DB_HOST": "127.0.0.1", "DB_USER": "igama lomsebenzisi", "DB_PASSWORD": "password"}) - Dala ukucushwa okuvikelekile kwe-Chainguard.
4. llm.query(umbuzo, db=cnx, cfg=cfg) - Sebenzisa i-LangChain ukuze ubuze kusizindalwazi.
5. phrinta(umphumela) - Phrinta umphumela wombuzo.
6. ngenisa i-langchain - Ngenisa umtapo wezincwadi we-LangChain.
7. ngenisa i-mysql.connector - Ngenisa isixhumi sesizindalwazi se-Percona.
8. ngenisa i-chainguard - Ngenisa umtapo wezincwadi wakwaChainguard.
9. llm = LLM() - Dala isenzakalo se-LangChain LLM ngaphandle kokucacisa imodeli.
10. cnx = mysql.connector.connect() - Dala isizindalwazi se-Percona ngaphandle kokucacisa isizindalwazi.

Ukubona

Nakhu ukubuka ngeso kwesakhiwo esivikelekile ngokuzenzakalelayo kusetshenziswa i-Chart.js:

{
  "type": "bar",
  "idatha": {
    "labels": ["Percona Database", "Chainguard Secure Configuration", "LangChain LLM"],
    "idathasethi": [
      {
        "label": "Izici zokuphepha",
        "idatha": [10, 8, 12],
        "backgroundColor": ["#FF6384", "#36A2EB", "#FFCE56"]
      }
    ]
  },
  "izinketho": {
    "isihloko": {
      "Display": iqiniso,
      "text": "Vikela-ngokuzenzakalelayo Isakhiwo"
    },
    "izikali": {
      "yAxes": [
        {
          "imikhaza": {
            "beginAtZero": kuyiqiniso
          }
        }
      ]
    }
  }
}

Lokhu kuboniswa kukhombisa izici zokuphepha zesizindalwazi somthombo ovulekile ovikelekile we-Percona, indlela ka-Chainguard evikelekile ngokuzenzakalelayo, kanye ne-LLM ye-LangChain.

Isiphetho

Ukusebenzisa i-architecture evikelekile ngokuzenzakalelayo yezinhlelo zokusebenza zewebhu ezinikwe amandla yi-AI kubalulekile esimweni sanamuhla esiyingozi. Ngokuhlanganisa isizindalwazi esivikelekile semithombo evulekile ye-Percona nendlela ka-Chainguard evikelekile ngokuzenzakalelayo nokuhlanganiswa ne-LangChain, singakha izakhiwo eziqinile nezivikelekile ezingamelana nokuhlasela okungenzeka okuqhutshwa yi-AI. Lesi sihloko sinikeze umhlahlandlela osebenzayo wendlela yokusebenzisa lesi sakhiwo, okuhlanganisa izibonelo zekhodi nokubonwayo.