Okwu mmalite
Mmụba ngwa ngwa nke ọgụgụ isi Artificial (AI) na Machine Learning (ML) emeela ka mmepe nke AI AI, ụdị AI nke nwere ike ime ihe ma mee mkpebi n'onwe ya. Nkà na ụzụ a nwere ike ịgbanwe ụzọ anyị si abịaru nso n'ịrụ ụlọ ọrụ ngwanrọ yana automation DevOps. N'isiokwu a, anyị ga-enyocha njikọ nke AI AI, Ụdị Ụdị Asụsụ buru ibu (LLMs), na ngwá ọrụ nhazi dị ka LangChain na DSPy iji rụọ ọrụ nhazi ngwa ngwa, akụrụngwa, na usoro DevOps.
Gịnị bụ AI Agent?
AI Agentic na-ezo aka n'ụdị AI nke nwere ike ime ihe ma mee mkpebi n'onwe ya. Nke a dị iche na sistemụ AI ọdịnala, nke a na-emekarị ka ọ rụọ ọrụ dị iche iche yana soro usoro iwu akọwapụtara. N'aka nke ọzọ, a na-emepụta usoro AI nke AI ka ọ bụrụ nke kwadoro onwe ya na ime mgbanwe, na-enye ha ohere ịmụta ihe site na gburugburu ebe obibi ha ma mee mkpebi dabere na ebumnuche na ebumnobi ha.
Ọrụ LLM na AI Agent
Ụdị Asụsụ buru ibu (LLM) bụ ụdị AI a zụrụ azụ na nnukwu data ederede ma nwee ike ịmepụta asụsụ dị ka mmadụ. LLM bụ akụkụ bụ isi nke sistemụ AI na-arụ ọrụ, ebe ha na-enye ike ịghọta na imepụta asụsụ mmadụ. Nke a na-enye ohere ka sistemụ AI na-emekọrịta ihe na ụmụ mmadụ na sistemụ ndị ọzọ n'ụzọ okike na nghọta.
Ngwa ## Orchestration: LangChain na DSPy
Ngwá ọrụ orchestration dị ka LangChain na DSPy ka emebere iji jikwaa na ịhazi mmekọrịta dị n'etiti usoro na ihe dị iche iche. Ngwá ọrụ ndị a na-enye ụzọ iji kọwaa na mebie usoro ọrụ, nke nwere ike ịgụnye ọrụ dịka nhazi data, ọzụzụ nlereanya, na ntinye.
LangChain
LangChain bụ ngwá ọrụ na-emepe emepe na-emepe emepe nke na-enye usoro na-agbanwe agbanwe na nke nwere ike ịmepụta maka iwulite na ijikwa usoro ọrụ. Emebere LangChain ka ọ rụọ ọrụ na sistemụ na akụrụngwa dị iche iche, gụnyere LLM, ọdụ data, na API.
DSPy
DSPy bụ ngwa orchestration sitere na Python nke na-enye ụzọ dị mfe na nghọta iji kọwaa na mebie usoro ọrụ. Emebere DSPy ka ọ rụọ ọrụ na sistemụ na akụrụngwa dị iche iche, gụnyere LLM, ọdụ data, na API.
Nhazi ngwanrọ akpaaka na usoro DevOps
Agentic AI, LLMs, na ngwa orchestration dị ka LangChain na DSPy nwere ike iji megharịa ọtụtụ ụdị nhazi ngwanrọ na usoro DevOps. Ụfọdụ ọmụmaatụ gụnyere:
** Ntụle koodu akpaaka ***: Enwere ike iji sistemu AI Agent iji nyochaa koodu wee chọpụta nsogbu ndị nwere ike, dị ka chinchi na adịghị ike nchekwa.
- **Nnwale akpaaka ***: Enwere ike iji sistemụ AI na-emezigharị nnwale, gụnyere nnwale otu, nnwale nnabata, na nnwale ngwụcha ruo ọgwụgwụ. ** Nbunye akpaaka ***: Enwere ike iji sistemu AI Agent rụọ ọrụ na-akpaghị aka, gụnyere ntinye koodu, nhazi na akụrụngwa ndị ọzọ. ** Nleba anya akpaghị aka ***: Enwere ike iji sistemu AI Agent mee ka nleba anya na-akpaghị aka, gụnyere nleba anya arụmọrụ sistemu, nchekwa na metrik ndị ọzọ.
Mmejuputa bara uru
Na ngalaba a, anyị ga-enye ihe atụ dị mfe nke otu esi eji LangChain na DSPy arụ ọrụ arụ ọrụ ngwanrọ na usoro DevOps.
Ọmụmaatụ: Nyochaa koodu akpaaka
N'ihe atụ a, anyị ga-eji LangChain na DSPy megharịa nyocha koodu. Anyị ga-akọwapụta usoro ọrụ nke gụnyere usoro ndị a:
- ** Nyochaa koodu ***: Usoro ọrụ ga-eji LLM nyochaa koodu ahụ ma chọpụta ihe ndị nwere ike ime.
- ** Nyochaa koodu ***: Usoro ọrụ ga-eji usoro AI na-arụ ọrụ iji nyochaa koodu ahụ ma chọpụta ihe ndị nwere ike ime.
- ** Ntinye koodu ***: Usoro ọrụ ga-eji ngwá ọrụ ntinye aka iji tinye koodu ahụ na gburugburu ebe mmepụta ihe.
Nke a bụ ọmụmaatụ otu esi etinye usoro ọrụ a site na iji LangChain na DSPy: "Python mbubata langchain si langchain mbubata LLM
Kọwaa LLM
llm = LLM ("ederede-davinci-003")
Kọwaa sistemụ AI na-arụ ọrụ
ai_system = langchain.AgenticAI()
Kọwaa usoro ọrụ
usoro ọrụ = langchain. Usoro ọrụ ( nzọụkwụ=[ langchain.Nzọụkwụ( aha = "Nyocha koodu", func=llm.analyze_code, ntinye = {"code": "def hello_world(): ebipụta ('Ndewo Ụwa')"}, ), langchain.Nzọụkwụ( aha = "Ntụle koodu", func=ai_system.review_code, ntinye = {"code": "def hello_world(): ebipụta ('Ndewo Ụwa')"}, ), langchain.Nzọụkwụ( aha = "Nkwanye koodu", func=langchain.deploy_code, ntinye = {"code": "def hello_world(): ebipụta ('Ndewo Ụwa')"}, ), ] )
Mepụta usoro ọrụ
workflow.execute() `` Usoro a na-akọwa usoro ọrụ nke gụnyere usoro atọ: nyocha koodu, nyocha koodu, na ntinye koodu. Usoro a na-arụ ọrụ na-eji LLM nyochaa koodu ahụ, usoro AI na-arụ ọrụ iji nyochaa koodu ahụ, yana ngwá ọrụ ntinye iji tinye koodu ahụ na gburugburu mmepụta ihe.
Top 5-10 Iwu/Snippets kacha eji
Nke a bụ iwu/snippets 5-10 kacha eji eme ihe maka LangChain na DSPy:
- `langchain.Workflow(nzọụkwụ = [...])': Kọwaa usoro ọrụ nwere ọtụtụ usoro.
- `langchain.Step(name="...", func=..., inputs={...})": Kọwaa otu nzọụkwụ na usoro ọrụ.
- `llm.analyze_code( ntinye = {"koodu": ...})': Jiri LLM nyochaa koodu.
- `ai_system.review_code(ntinye = {"koodu": ...})': Jiri sistemụ AI na-arụ ọrụ iji nyochaa koodu.
- `langchain.deploy_code(ntinye = {"koodu": ...})': Nyefee koodu na gburugburu mmepụta.
- `langchain.AgenticAI()': Mepụta ihe atụ nke sistemụ AI na-arụ ọrụ.
langchain.LLM(): Mepụta ihe atụ nke LLM.- `langchain.deploy_code()': Nyefee koodu na gburugburu mmepụta.
langchain.execute_workflow(): Mepụta usoro ọrụ.langchain.get_workflow_status(): Nweta ọkwa nke usoro ọrụ.
Nlere anya
Nke a bụ ọmụmaatụ otu esi eji Chart.js were anya nke uche ya: Json { "ụdị": "ogwe", "data": { "labels": ["Ntụle koodu", "Ntụle koodu", "Nkwanye koodu"], "datasets": [ { "label": "Ụzọ usoro ọrụ", "data": [10, 20, 30], "backgroundAgba": [ "rgba (255, 99, 132, 0.2)", "rgba (54, 162, 235, 0.2)", "rgba (255, 206, 86, 0.2)" ], "BorderAgba": [ "rgba (255, 99, 132, 1)", "rgba (54, 162, 235, 1)", "rgba (255, 206, 86, 1)" ], "BorderWiidth": 1 } ] }, "Nhọrọ": {} } `` Koodu a na-akọwa eserese mmanya nwere akara atọ: Nyocha koodu, Nyochaa koodu, na ntinye koodu. Chaatị ahụ na-egosi ọnụọgụ nke usoro na ọkwa ọ bụla nke usoro ọrụ ahụ.
Mmechi
AI, LLMs, na ngwa orchestration dị ka LangChain na DSPy nwere ikike ịgbanwe ụzọ anyị si abịaru nso nrụpụta ngwanrọ yana automation DevOps. Site n'ime ka ọrụ akpaaka dị ka nyocha koodu, nnwale na mbugharị, anyị nwere ike melite arụmọrụ na ịdị irè nke usoro mmepe anyị. N'isiokwu a, anyị enyochala njikọ nke teknụzụ ndị a ma nye ihe atụ dị mfe nke otu esi eji ha rụọ ọrụ nhazi ngwa ngwa na usoro DevOps.