Overview
Maximum Marginal Relevance (MMR) inzira yekuisa zvinzvimbo zvakare inoyera kuti mhedzisiro inoenderana nekusiyana kwayakaita kubva kune zvakatosarudzwa. Izvo zvine basa nekuti kwakachena kuenderana chinzvimbo kunowanzo dzorera padhuze-duplicate ndima iyo inoparadza nzvimbo muRAG mamiriro hwindo.
Maximum Marginal Relevance chikamu chemutauro-AI stack inoshandiswa kuverenga, kugadzira, kuronga, uye kushandura zvinyorwa uye kutaura pamwero.
Deep Dive
Kana hurongwa hwekutsvaga huchiwana magwaro chete nekukoshera kumubvunzo, mibairo yepamusoro inowanzo shaya basa - ndima shanu dzese dzichitaura chinhu chimwe. MMR, yakaunzwa naCarbonell naGoldstein muna 1998, inogadzirisa izvi nekusarudza mhinduro imwe panguva. Panhanho imwe neimwe inotora mumiriri anokwirisa musanganiswa wakayerwa: lambda inopeta kukosha kwayo kumubvunzo, kubvisa (1 minus lambda) nguva dzayakafanana zvakanyanya kune chero chatosarudzwa kare. Lambda iri pedyo ne1 inofarira kukosha kwakachena; pedyo ne0 inofarira kusiyana-siyana. Muchizvarwa chekutsvaga-chakawedzerwa, MMR inofarirwa nekutora zvakasiyana-siyana zvezvimedu kuitira kuti modhi yemutauro ione humbowo hunopindirana pane kuti chokwadi chimwechete chidzokororwe, kuvandudza kuvharika pasina kuwedzera mamiriro.
Technical Insight
MMR inokara, iterative algorithm. Zvese zvine chekuita uye kufanana kwepakati-gwaro kunowanzo kuverengerwa se cosine kufanana pakati pekumisikidza vheji. Formula yezvibodzwa ndeiyi: MMR = argmax pamusoro pezvinyorwa zvasara zve [ lambda * sim(doc, query) - (1 - lambda) * max sim(doc, yakasarudzwa)]. Nekuti inoongorora zvakare ichipikisa iri kukura yakasarudzwa seti yega yega, inotsamira-kurongeka uye inomhanya muinenge O(k*n) kuenzanisa kwekuenzanisa kwe k picks kubva kun vavhoti.
Mastering Maximum Marginal Relevance
Maximum Marginal Relevance (MMR) inzira yekuisa zvinzvimbo zvakare inoyera kuti mhedzisiro inoenderana nekusiyana kwayakaita kubva kune zvakatosarudzwa. Izvo zvine basa nekuti kwakachena kuenderana chinzvimbo kunowanzo dzorera padhuze-duplicate ndima iyo inoparadza nzvimbo muRAG mamiriro hwindo. Maximum Marginal Relevance chikamu chemutauro-AI stack inoshandiswa kuverenga, kugadzira, kuronga, uye kushandura zvinyorwa uye kutaura pamwero. Kuti uvake kunzwisisa kwakadzama, bata Maximum Marginal Relevance semuenzaniso wekushandisa, kwete chinhu chimwe chete: tsanangura zvinodiwa, kujekesa fungidziro, uye patsanura izvo zvinogona kuitwa nehurongwa hwakavimbika kubva kune izvo zvichiri kuda kutonga kwenyanzvi.
Mukuita, zvikwata zvakasimba zvinoshandisa Maximum Marginal Relevance dhizaini zvinokurudzira, kudzoreredza, uye kuongorora zvishwe seimwe yakabatanidzwa yekutaurirana system. Ivo vanonyora zvakajeka maitiro ebudiriro, bvunzo vachipokana ne data rechokwadi uye mafambiro ebasa, uye iterate zvichibva pane zvakacherechedzwa maitiro ekutadza kwete kuhwina-nguva imwe chete yebhenji. Apa ndipo apo kunzwisisa kwe theoretical kunoshanduka kuve kugona kwakasimba pane chigadzirwa, mutemo, uye mashandiro.
Mutauro workflows inogona kufamba nekukurumidza pasina kupira kuenderana. Panguva imwecheteyo, chokwadi cheHallucified chinogona kupinda chinyararire mishumo, kuyerera kwetsigiro, kana kutsvagisa zvinobuda. Nzira yakatsiga ndeyekubatanidza kukurumidza kuyedza nekutonga: mhanyisa vatyairi vendege, tora humbowo, buritsa matanda esarudzo, uye urambe uchivandudza chengetedzo semaitiro emuenzaniso, zvinotarisirwa nemushandisi, uye zvinodikanwa zvekutonga.
Strategic Impact
Mutauro workflows inogona kufamba nekukurumidza pasina kupira kuenderana.
Mutauro workflows inogona kufamba nekukurumidza pasina kupira kuenderana. Mukutumirwa kwemhando yepamusoro, izvi zvinoshandurirwa kuita mitemo inoyerwa yekushanda, miganhu yevaridzi, uye tsika dzekudzokorora dzinodzokororwa kuitira kuti zvikwata zvikwire kuvimba pane kukwidza kusajeka.
Inopamhidzira kupinda mumitauro yese nemataera ekutaurirana.
Inopamhidzira kupinda mumitauro yese nemataera ekutaurirana. Mukutumirwa kwemhando yepamusoro, izvi zvinoshandurirwa kuita mitemo inoyerwa yekushanda, miganhu yevaridzi, uye tsika dzekudzokorora dzinodzokororwa kuitira kuti zvikwata zvikwire kuvimba pane kukwidza kusajeka.
Zvikwata zvinogona kupedza nguva yakawanda pakutonga uku otomatiki ichibata kudzokorora.
Zvikwata zvinogona kupedza nguva yakawanda pakutonga uku otomatiki ichibata kudzokorora. Mukutumirwa kwemhando yepamusoro, izvi zvinoshandurirwa kuita mitemo inoyerwa yekushanda, miganhu yevaridzi, uye tsika dzekudzokorora dzinodzokororwa kuitira kuti zvikwata zvikwire kuvimba pane kukwidza kusajeka.
Real-World Implementation
RAG chatbot inoshandisa MMR kudzoreredza saka machunks ayo mashanu epamusoro anovhara akasiyana maitirwo emutemo panzvimbo pezvirevo zvishanu zvendima imwe chete.
Chishandiso chekupfupisa chetsvagiro chinoshandisa MMR kusarudza ndima dzinoderedza kupindirana, dzichigadzira pfupiso, isingadzokororwe.
Muunganidzi wenhau anoisa zvinyorwa neMMR kuratidza zvakasiyana siyana zvechiitiko kwete zvitoro gumi zvinodzokorora nyaya imwe chete.
LangChain's vector store retriever inofumura search_type='mmr' nefetch_k uye lambda_mult kusiyanisa zvinyorwa zvakadzorerwa.
Maitiro Ekuita
Maximum Marginal Relevance mukuita
RAG chatbot inoshandisa MMR kudzoreredza saka machunks ayo mashanu epamusoro anovhara akasiyana maitirwo emutemo panzvimbo pezvirevo zvishanu zvendima imwe chete.
A RAG chatbot inoshandisa MMR kudzoreredza kuitira kuti machunks ayo mashanu epamusoro avhare zvakasiyana siyana zvegwaro pachinzvimbo chezvirevo zvishanu zvendima imwechete Matimu anowanzo kuwana mhedzisiro iri nani kana achinge atsanangura emhando yepamusoro kumberi, chengetedza nzira yekukwira kwevanhu yemakesi ekumucheto, uye kuteedzera zvese zvakawanikwa zvechigadzirwa nemitengo yekukanganisa nekufamba kwenguva.
Maximum Marginal Relevance mukuita
Chishandiso chekupfupisa chetsvagiro chinoshandisa MMR kusarudza ndima dzinoderedza kupindirana, dzichigadzira pfupiso, isingadzokororwe.
Chishandiso chekupfupisa chetsvagiro chinoshandisa MMR kusarudza ndima dzinoderedza kupindirana, kugadzira pfupiso yakafara, isingadzokororwe Matimu anowanzo kuwana mibairo iri nani kana achinge atsanangura hunhu hwepamberi, chengetedza nzira yekukwira kwevanhu yemakesi emupendero, uye kuteedzera zvese zvakawanikwa zvechigadzirwa uye mutengo wekukanganisa nekufamba kwenguva.
Maximum Marginal Relevance mukuita
Muunganidzi wenhau anoisa zvinyorwa neMMR kuratidza zvakasiyana siyana zvechiitiko kwete zvitoro gumi zvinodzokorora nyaya imwe chete.
Muunganidzi wenhau anoisa zvinyorwa neMMR kuratidza kwakasiyana siyana kwechiitiko kwete zvitoro gumi zvinodzokorora waya imwe nyaya Matimu anowanzo kuwana mhedzisiro iri nani kana achinge atsanangura hunhu hwepamberi, chengetedza nzira yekukwira kwevanhu yemakesi emupendero, uye kuteedzera zvese zvakawanikwa zvechigadzirwa uye kukanganisa mutengo nekufamba kwenguva.
Maximum Marginal Relevance mukuita
LangChain's vector store retriever inofumura search_type='mmr' nefetch_k uye lambda_mult kusiyanisa zvinyorwa zvakadzorerwa.
LangChain's vector store Retriever inofumura search_type='mmr' ine fetch_k uye lambda_mult kusiyanisa magwaro akadzoserwa Matimu anowanzo kuwana mibairo iri nani kana achinge atsanangura hunhu hwepamberi, chengetedza nzira yekukwira kwevanhu yemakesi ekumucheto, uye kuteedzera zvese zvakawanikwa zvechigadzirwa nemitengo yekukanganisa nekufamba kwenguva.
Njodzi & Guardrails
Chokwadi chehuroyi chinogona kupinda chinyararire mishumo, kuyerera kwetsigiro, kana tsvakiridzo.
Kunzwa nekukasira kunogona kugadzira mhedzisiro isingaenderane pane zvikumbiro zvakafanana.
Sensitive text data inogona kuburitswa kana zvidhiraivho zvisina kusimba.
Implementation Roadmap
Tsanangura chimiro chekubuda, toni, uye mhando zviyero usati waburitsa.
Tsanangura chimiro chekubuda, toni, uye mhando zviyero usati waburitsa. Bata nhanho yega yega segedhi rehumbowo: kana maitiro asina kusangana, imbomira kuburitsa, vhara gaka, uye wobva wawedzera kushandiswa.
Mhinduro dzepasi neakavimbika masosi pese pazvine basa.
Mhinduro dzepasi neakavimbika masosi pese pazvine basa. Bata nhanho yega yega segedhi rehumbowo: kana maitiro asina kusangana, imbomira kuburitsa, vhara gaka, uye wobva wawedzera kushandiswa.
Chengetedza ongororo yekuongorora yemunhu kune yakakwira-stake zvinobuda.
Chengetedza ongororo yekuongorora yemunhu kune yakakwira-stake zvinobuda. Bata nhanho yega yega segedhi rehumbowo: kana maitiro asina kusangana, imbomira kuburitsa, vhara gaka, uye wobva wawedzera kushandiswa.
Tevera maitiro ekutadza uye dzidzisazve kukurudzira kana mafambiro ebasa nguva nenguva.
Tevera maitiro ekutadza uye dzidzisazve kukurudzira kana mafambiro ebasa nguva nenguva. Bata nhanho yega yega segedhi rehumbowo: kana maitiro asina kusangana, imbomira kuburitsa, vhara gaka, uye wobva wawedzera kushandiswa.