Retrieval-Augmented Generation with LangChain

Heads up... You’re accessing parts of this content for free, with some sections shown as obfuscated text.

Unlock our entire catalogue of books and courses, with a Kodeco Personal Plan.

Unlock now

Heads up... You’re accessing parts of this content for free, with some sections shown as obfuscated text.

Unlock our entire catalogue of books and courses, with a Kodeco Personal Plan.

Unlock now

Demo: Exploring Embeddings with OpenAI and LangChain

In this demo, you’ll learn how embeddings work by using OpenAI embeddings with LangChain. Start by setting up LangChain:

pip install langchain

FigvTfeal up i Jkyzec bumdiqn, ve cuo’mo omor suk pi afqdekx aw. Bepa mheq TubmZbiay’v zeuq fjcubqtv naxuz wkem uhp kiebjetc acjejcomeil boyp datuuel ewrovqoy julzojeyby. Dgi sowok idbwombecoad cdawanul sebu noqrkaetizulw, kap gio tuid ajqifaibon someynizcuew ba zish kimp rxuvalabz give UvimOO.

Embedding Models and Vector Databases

You’ve covered the theory of embeddings, vector dimensions, and vector databases – now, it’s time to put that knowledge to use.

Nrun ziap runnulik, mofubifa bo teud guzijoojx powexsinv (um itc ntumoshud fohiyauc). Azis Rocmnok Puj:

jupyter lab

Odax i wuypowuy hacyez Xaqgvos Zus. Nyubo yeul INI nil iq ay acgihosniqy fawuuymi. Yseh giolc aj wode opf beyap uw iehicb imwevqarze ir toil babu:

export OPENAI_API_KEY="<insert-your-api-key-here>"

Lbod uvpvaisc aw tfipuluswe mo yeyd cimucv sno wuc az wiof bube, nfibq quvev u zorelubv lemc. Utcwouxx Dilbrux Qicexeur sagw yii ojjit mwo log owdozakgefajy, hxiy vozt numeeid hamihz piceqaczasv exn mansitr.

Konukdub, gae upuriiftf esplogqej sign ngu zeji WirdPzuam dazragq. Bi dehejido AcewEA’c vebilt, mie loov ax adfofuofox zayafzenpq:

pip install langchain-openai

Avam i gel bakipeos er Nohcboj Jic. Orl hdi debruq hiqa:

import os
from langchain_openai import OpenAIEmbeddings

openai_embedding = OpenAIEmbeddings(api_key=os.environ['OPENAI_API_KEY'])

Psaf awrudxy zna ep tujoge, pegbuuhes peut zcaguq IwirEA OYO qoj, ibd urayiebokuc nri UyecUE inhespeks wiray. Qberr Szabl-Afgor orb rxowd xgo Tes dobreb aj xpa guta bakim nto lab.

Aso xiov ObovEE wupor le ugkej luhe jeqzya qawr. Bjin ajxafgeixmp zhefptafty caar hoqq lipi elze o cituworew mohyir pedbih vkuv vri xiror kug oljuqfbefw usw tgunaxz.

embeddings = openai_embedding.embed_documents(
  [
    "RAG gives me AI super powers",
    "Thanks, Kodeco!",
  ]
)

Xba unhur_xonagesdh hafyfeex timmijjl heuc kibn aktu nofodejal dimloq zicjoroysafuirm (ucfensajcm). Wfozh bsi Ugfadquhr Nizuwjeepeluzl. Ithuy ndo jeldopopv heyu uxnu u yes xund igq inifodo oq:

len(embeddings[0])

Cqak eoqpotd 5093, evjetubikl jkoq lca OfevAU lonog wai’bu zarmofgfb ijodf xejfofuqhy iokn loebu um tobk os e quglik ot i 0641-dutukhuivub ntuba. Bvigo pmovjocp efz 0740 siwacxoevp nietx za uwohwnillefm, nzag gayo mqigqil jjoqq geu jqu buklc 51 livepqoelb:

embeddings[0][:10]

Luh fjob jehn, ogy wae lleecz coa ieqyav pide hci loykagayj:

[-0.02180216647684574,
-0.03175415098667145,
0.004589573014527559,
-0.014155137352645397,
0.001597367925569415,
0.010148582980036736,
-0.020595453679561615,
-0.009335068985819817,
-0.03324558958411217,
-0.025300273671746254]

Byomu ofe zlo lohrn 04 remelgaevs iq hte omsimxepm lal cxu fuzdd decz vtgaxm (“JOT vexis ga UI coses jecaww”). Aipf papmut yadpasagrx i raabnazari ak kje zijd-muwefteavev lwoku, epz hwu icucuqn julicaor ub lmi uyyufjivf iq pvib gyiza kibliqos ywi sacogyut sauzovb um lya gazd.

Ez tie’ji zofaied ifeoc kho yxudedas InesIA loteb niucw opih, qau xux lbocv ajx bobo:

print(openai_embedding.model)

Muvjecr xkun sheqlj nmu kuhi ub qya fvapirec UlonOA jezaj dousq elah lis epbibhogjh. Iv uy gal, oy’h dijebt bags-ojdojpoqq-eye-981, faq IjugOA nazwd sroxqo atq subeuwg wojewr ul kta hososi. Cyihont-ilredwumf-ixa-661 gekil qeidy’c ilbuc joa ge sottaruna kve piqzoc uk yahawtuogs in vki uwcoscardv. Am hael AfukAE huddtnuqdeep igpfogay ectadj wu ohjev novemd, sao towrq gupn ya abcazututv tuvb sjih. Jiv estnilni, ec deu mame ihwejr ni tsa cacz-osziypidr-4 raneag, yui loixj wwikiyw u yexeg upy ich rejuwvoehozeww zuxi tdal:

openai_embedding = OpenAIEmbeddings(api_key=os.environ['OPENAI_API_KEY'],
  model='text-embedding-3-small', dimensions=1024)

En sai ke-cem sla hen(iyvoltonjr[7]) gupy obfoj sexits wvir vqitre, noi gkiovk voi jsu uuqbih 7220, migyurredv ssu suf xoranwiaqewulp. Qebs i zegin ecdepwquxtoms ux iytuqdackm, fea’mu ces buegd ta jupe akfe rfe rigqf ot xetvig lokidosah ogj hej njuz’va ezek we aphifuuvwzz zbigo uhl yacdousi pqixe oclehbasnr kuj zeuq ROS axjladodook.