Intangiriro

Ubwiyongere bugenda bwiyongera bwa sisitemu igezweho ya software yatumye kwiyongera inshuro nuburemere bwibyabaye, bituma iperereza ryibyabaye no gukemura ikibazo gikomeye cya DevOps. Ubwenge bwa artificiel (AI) hamwe no kwiga imashini (ML) byagaragaye nkibikoresho bikomeye byo kuzamura imiyoboro ya DevOps, bigafasha gukoresha moteri ya AI hamwe na orchestre kumenya no gukemura ibyabaye mbere yuko bigira ingaruka mubucuruzi. Muri iyi ngingo, tuzasuzuma uburyo bwo gushyira mu bikorwa imiyoboro ikoreshwa na AI ikoresha DevOps ikoresheje DSPy na LangChain, hibandwa ku iperereza ryabaye no gukemura ibibazo hakoreshejwe moderi ya AIOps na RAG, no guhuza na AWS DevOps Agent mu iperereza ryakozwe mu buryo bwikora.

Incamake ya DSPy na LangChain

DSPy ni isomero rya Python ritanga uburyo bworoshye kandi bunoze bwo kubaka no gukoresha imiyoboro ya AI ikoreshwa na AI. Iyemerera abaterankunga guhuza imiterere itandukanye ya AI na ML, harimo gutunganya ururimi karemano (NLP), iyerekwa rya mudasobwa, hamwe nisesengura risesuye, mumiyoboro yabo ya DevOps. Ku rundi ruhande, LangChain, ni isomero rya Python rifasha abitezimbere kubaka no gukoresha imashini nini ya AI ikoresheje uburyo bwububiko kandi bunini.

Gushyira mubikorwa imiyoboro ya AI ikoreshwa na DSPy na LangChain

Kugira ngo dushyire mu bikorwa imiyoboro ikoreshwa na AI hamwe na DSPy na LangChain, dukeneye gukurikiza izi ntambwe:

1. ** Ikusanyamakuru **: Kusanya amakuru ajyanye n'amasoko atandukanye, nk'ibiti, ibipimo, na dosiye y'iboneza.
2. ** Gutunganya amakuru **: Gutunganya amakuru yakusanyijwe kugirango uyategure gusesengura no kwerekana imiterere.
3. ** Amahugurwa yicyitegererezo **: Hugura moderi ya AI na ML ukoresheje amakuru yatunganijwe mbere yo kumenya no guhanura ibyabaye.
4. ** Kohereza icyitegererezo **: Shyira moderi yatojwe mumuyoboro wa DevOps ukoresheje DSPy na LangChain.
5. ** Iperereza ryibyabaye **: Koresha moderi zoherejwe kugirango ukore iperereza kubyabaye no kumenya ibitera.
6. ** Gukemura ibyabaye **: Koresha ubushishozi bwakuwe mubushakashatsi bwakozwe kugirango ukemure ibyabaye.

Kode Urugero: Gushyira mubikorwa Moderi yoroshye ya RAG hamwe na DSPy na LangChain

Dore urugero rworoshye rwa kode yerekana uburyo bwo gushyira mubikorwa urugero rwa RAG (Imizi Impamvu Isesengura) ukoresheje DSPy na LangChain: `` python kwinjiza dspy Kuva kumurongo winjiza LLMChain

Fungura amakuru yatunganijwe

amakuru = pd.isoma_csv ("data.csv")

Sobanura icyitegererezo cya RAG

rag_model = dspy.RAGModel ( ibyinjijwe_ibikorwa = ["log_level", "ikosa_ubutumwa"], ibisohoka_ibikorwa = ["umuzi_impamvu"] )

Hugura icyitegererezo cya RAG

rag_model.imyitozo (amakuru)

Kohereza icyitegererezo cya RAG

llm_umurongo = LLMChain ( icyitegererezo = "davinci: 002", max_uburebure = 2048 ) rag_model.deploy (llm_chain)

Koresha moderi ya RAG yoherejwe kugirango ukore iperereza kubyabaye

ibyabaye_data = pd.DataFrame ({ "log_level": ["IKOSA"], "error_message": ["Ikintu kitagenze neza"] }) umuzi_impamvu = rag_model.teganya (ibyabaye_data) icapa (imizi_impamvu) `` ` Uru rugero rwimyandikire yerekana uburyo bwo gushyira mubikorwa byoroshye RAG ukoresheje DSPy na LangChain. Moderi ya RAG ifata urwego rwubutumwa nubutumwa bwibibazo nkibintu byinjira kandi bikavuga intandaro yibyabaye.

Hejuru 5-10 Amabwiriza Yakoreshejwe Byinshi / Uduce

Dore top 5-10 zikoreshwa cyane / ibisobanuro kuri DSPy na LangChain:

1. `dspy.
2. dspy.RAGModel.train (data): Hugura moderi ya RAG ukoresheje amakuru yatanzwe.
3. dspy.RAGModel.deploy (llm_chain): Kohereza moderi ya RAG yatojwe ukoresheje LangChain LLMChain.
4. Llm_chain = LLMChain (icyitegererezo, max_uburebure): Kurema LangChain LLMChain ifite icyitegererezo cyihariye n'uburebure bwa max.
5. rag_model.predict (data): Koresha moderi ya RAG yoherejwe kugirango umenye intandaro yibyabaye.
6. dspy.DataPreprocessor (input_data): Sobanura umuyoboro utunganya amakuru ukoresheje DSPy.
7. dspy.ModelTrainer (icyitegererezo, amakuru): Hugura moderi ya AI cyangwa ML ukoresheje DSPy.
9. langchain.LLMChain.predict (input_data): Koresha LangChain LLMChain kugirango uhanure umusaruro wicyitegererezo.
10. dspy.RAGModel. Suzuma (data): Suzuma imikorere ya moderi ya RAG ukoresheje DSPy.

Kugaragara

Dore urugero rwo kubona amashusho ukoresheje Imbonerahamwe.js: json { "ubwoko": "akabari", "amakuru": { "ibirango": ["Ubwoko bw'ibyabaye A", "Ubwoko bw'ibyabaye B", "Ubwoko bw'ibyabaye C"], "imibare": [ { "label": "Impamvu Imizi A", "amakuru": [10, 15, 7], "backgroundColor": "rgba (255, 99, 132, 0.2)", "umupaka Ibara": "rgba (255, 99, 132, 1)", "Ubugari bwumupaka": 1 }, { "label": "Impamvu Imizi B", "amakuru": [8, 12, 9], "backgroundColor": "rgba (54, 162, 235, 0.2)", "umupaka Ibara": "rgba (54, 162, 235, 1)", "Ubugari bwumupaka": 1 }, { "label": "Impamvu Imizi C", "amakuru": [14, 18, 11], "backgroundColor": "rgba (255, 206, 86, 0.2)", "umupaka Ibara": "rgba (255, 206, 86, 1)", "Ubugari bwumupaka": 1 } ] }, "amahitamo": { "umutwe": { "kwerekana": ukuri, "inyandiko": "Isesengura ry'Imizi" }, "umunzani": { "yAxes": [ { "amatiku": { "gutangiraAtZero": ukuri } } ] } } } ` Iyi visualisation yerekana ikwirakwizwa ryintandaro yubwoko butandukanye bwibyabaye.

Umwanzuro

Muri iki kiganiro, twasuzumye uburyo twashyira mu bikorwa imiyoboro ikoreshwa na AI ikoresha DevOps dukoresheje DSPy na LangChain, twibanze ku iperereza ry’ibyabaye no gukemura hakoreshejwe icyitegererezo cya AIOps na RAG, no guhuza na AWS DevOps Agent mu iperereza ryakozwe mu buryo bwikora. Twerekanye kode yoroshye urugero ishyira mubikorwa RAG ikoresheje DSPy na LangChain, tunatanga urutonde rwambere 5-10 rwakoreshejwe cyane / ibisobanuro kuri DSPy na LangChain. Twashyizemo kandi urugero rwo kureba dukoresheje Imbonerahamwe.js. Mugukurikiza intambwe zavuzwe muriyi ngingo, abashinzwe iterambere barashobora kubaka no gukoresha imiyoboro ikoreshwa na AI ikoreshwa na DevOps itahura kandi igakemura ibyabaye ku buryo bwitondewe, bikazamura imikorere rusange no kwizerwa bya sisitemu zabo.