Ƙirƙiri wakilin AI na tattaunawa ta amfani da LangChain | B4N1

Ƙirƙiri wakilin AI na tattaunawa ta amfani da LangChain

3 months ago · b4n1 Read time: 5 min

** Haɗa Wakilan AI tare da DevOps Automation: Buɗe Ƙarfin LangChain da AWS DevOps Agent**

** Gabatarwa ***

Haɗin kai na bayanan wucin gadi (AI), DevOps aiki da kai, da kuma gine-gine na asali na girgije yana canza yadda ƙungiyoyin haɓaka software ke tunkarar tushen bincike da haɓaka. Ta hanyar yin amfani da wakilai na AI, ƙungiyoyi za su iya sarrafa ayyuka masu rikitarwa, inganta lokutan amsa lamarin, da haɓaka amincin tsarin gaba ɗaya. A cikin wannan labarin, za mu bincika haɗin kai na LangChain, ɗakin karatu na Python don gina wakilan AI na tattaunawa, tare da AWS DevOps Agent, kayan aiki don sarrafa kayan aikin software da ƙaddamarwa. Za mu zurfafa cikin fa'idodin wannan haɗin kai, mu ba da misali mai amfani na yadda ake aiwatar da shi, da kuma haskaka manyan umarni da snippets na wannan fasaha.

Fa'idodin AI-Driven DevOps Automation

Kayan aikin sarrafa kayan aikin DevOps na gargajiya suna mai da hankali kan sarrafa ayyuka masu maimaitawa, kamar turawa da gwaji. Duk da haka, waɗannan kayan aikin sau da yawa ba su da ikon yin nazarin halayen tsarin rikitarwa da gano tushen abubuwan da ke haifar da al'amura. Ma'aikatan AI-kore, a gefe guda, na iya yin nazarin ɗimbin bayanai, gano alamu, da yin tsinkaya game da halayen tsarin. Ta hanyar haɗa wakilan AI tare da kayan aikin DevOps na atomatik, ƙungiyoyi na iya:

  • Inganta lokutan amsa al'amura ta hanyar gano tushen al'amura da sauri
  • Haɓaka amincin tsarin ta hanyar tsinkaya da hana gazawa
  • Kaddamar da hadaddun ayyuka, kamar gyara matsala da gyara kurakurai
  • Inganta haɗin gwiwa tsakanin ƙungiyoyin haɓakawa da ƙungiyoyin aiki ta hanyar samar da fahimtar juna game da halayen tsarin

** LangChain: Laburaren Python don Gina Wakilan AI na Tattaunawa ***

LangChain ɗakin karatu ne na Python wanda ke ba masu haɓaka damar gina wakilan AI na tattaunawa ta amfani da dabarun sarrafa harshe na halitta (NLP) da dabarun koyon injin (ML). Tare da LangChain, masu haɓakawa na iya ƙirƙirar wakilai waɗanda zasu iya fahimta da amsa tambayoyin mai amfani, bincika bayanan rubutu, da yin tsinkaya game da halayen tsarin. LangChain yana ba da kewayon fasali, gami da:

  • Binciken rubutu da sarrafawa
  • NLP da haɗin ML
  • Tsarin AI na tattaunawa
  • Haɗin kai tare da shahararrun tsarin AI, kamar TensorFlow da PyTorch

** Wakilin AWS DevOps: Kayan aiki don Aiwatar da Isar da Software da Aiwatarwa ***

AWS DevOps Agent kayan aiki ne wanda ke ba masu haɓaka damar sarrafa isar da software da turawa akan AWS. Wakilin yana ba da kewayon fasali, gami da:

  • Ƙaddamarwa ta atomatik da gwaji
  • Haɗin kai da ci gaba da bayarwa (CI / CD)
  • Haɗin kai tare da shahararrun kayan aikin DevOps, kamar Jenkins da GitLab
  • Taimako don harsunan shirye-shirye da yawa da tsarin aiki

** Haɗa LangChain tare da AWS DevOps Agent**

Don haɗa LangChain tare da AWS DevOps Agent, za mu ƙirƙiri rubutun Python mai sauƙi wanda ke amfani da LangChain don nazarin rajistan ayyukan tsarin da kuma gano abubuwan da za su iya yiwuwa. Za mu yi amfani da AWS DevOps Agent don sarrafa aikin tura rubutun kuma mu haɗa shi da bututun mu na CI/CD.

** Misalin Aiki: Yin Nazartar Rubutun Tsari tare da LangChain ***

`` Python shigo da langchain daga langchain.chains.shigowar taɗi ConversationalChain daga langchain.llms shigo da OpenAI daga langchain.embeddings shigo da HuggingFaceEmbeddings

wakili = ConversationalChain ( llm=BuɗeAI (zazzabi = 0.7), embeddings=HuggingFaceEmbeddings(samfurin_name = "masu juye-juye/duk-MiniLM-L6-v2") )

Ƙayyade aiki don nazarin rajistan ayyukan

def analyze_logs(tambayoyi): # Yi amfani da LangChain don bincika rajistan ayyukan da gano abubuwan da za su iya faruwa amsa = agent.call(tambaya = "Mene ne matsalolin da za a iya fuskanta tare da waɗannan rajistan ayyukan?") mayar da martani

Ƙayyade aiki don tura rubutun ta amfani da AWS DevOps Agent

def deploy_script(): # Yi amfani da AWS DevOps Agent don tura rubutun da haɗa shi tare da bututun CI / CD ɗin mu #... wuce

Gudanar da rubutun

logs = ["error1", "error2", "error3"] amsa = analyze_logs buga(amsa)

deploy_script()


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

Anan akwai manyan umarni 5-10 da aka fi amfani da su da snippets don LangChain da AWS DevOps Agent:

1. ** LangChain: ** `agent.call (tambaya = "Mene ne matsalolin da za a iya samu tare da waɗannan logs?")'
2. ** LangChain: ** 'ConversationalChain (llm = Buɗe AI (zazzabi = 0.7), sakawa = HuggingFaceEmbeddings (model_name = "sentence-transformers/all-MiniLM-L6-v2"))'
3. **AWS DevOps Agent:** `aws devops-wakili tura --script script.py`
4. ** AWS DevOps Agent:** `aws devops-wakilin haɗe-bututun-sunan'
5. ** LangChain: ** 'HuggingFaceEmbeddings (samfurin_name = "masu jumla-jumla/duk-MiniLM-L6-v2")''
6. ** LangChain: ** 'OpenAI (zazzabi = 0.7)'
7. ** AWS DevOps Agent: ** 'aws devops-matsayin wakili'
8. ** AWS DevOps Agent: ** 'aws devops-agent logs'
9. **LangChain:** `agent.embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
10. **LangChain:** `agent.llm = Buɗe AI(zazzabi = 0.7)`

** Hoton hoto: Chart.js ***

Anan akwai sauƙi mai sauƙi ta amfani da Chart.js don ganin aikin wakilin AI na tushen LangChain:
Json
{
  "type": "bar",
  "data": {
    "labels": ["Saidai", "Madaidaici", "Tunawa"],
    "datasets": [
      {
        "label": "LangChain",
        "bayanai": [0.9, 0.8, 0.7],
        "backgroundColor": "rgba(255, 99, 132, 0.2)",
        "BorderColor": "rgba(255, 99, 132, 1)",
        "BorderWidth": 1
      }
    ]
  },
  "zaɓi": {
    "take": {
      "nuna": gaskiya,
      "rubutu": "Ayyukan Wakilin AI na tushen LangChain"
    },
    "ma'auni": {
      "yAxes": [
        {
          "zama": {
            "beginAtZero": gaskiya
          }
        }
      ]
    }
  }
}

Wannan ginshiƙi yana nuna daidaito, daidaito, da kuma tunawa da wakilin AI na tushen LangChain. Taswirar tana ba da wakilci na gani na aikin wakilin kuma ana iya amfani da shi don gano wuraren da za a inganta.

Kammalawa

Haɗa ma'aikatan AI tare da sarrafa kansa na DevOps ta amfani da LangChain da AWS DevOps Agent hanya ce mai ƙarfi don haɓaka tushen bincike da haɓaka. Ta hanyar yin amfani da ƙarfin fasahohin biyu, ƙungiyoyi za su iya sarrafa ayyuka masu rikitarwa, inganta lokutan amsa abin da ya faru, da haɓaka amincin tsarin. A cikin wannan labarin, mun bincika fa'idodin AI-driven DevOps automation, gabatar da LangChain da AWS DevOps Agent, kuma mun ba da misali mai amfani na yadda ake haɗa fasahar biyu. Mun kuma haskaka manyan umarni da snippets na wannan fasaha kuma mun haɗa da gani ta amfani da Chart.js. Ta bin matakan da aka zayyana a cikin wannan labarin, ƙungiyoyi za su iya buɗe cikakkiyar damar aikin DevOps da AI ke tukawa da haɓaka haɓaka haɓaka software gaba ɗaya da ayyukan turawa.

Have Questions?

Let's Talk

Contact me on LinkedIn