Ururimi AI

Itsinda-Ikibazo

Grouped-Query Attention (GQA) shares key/value heads across groups of query heads.

Umva iki gitaboAI-byavuzwe · ~1 min

Kuriyi page1 min soma
  1. Incamake
  2. Kwibira cyane
  3. Gushyira mu bikorwa Isi
  4. Ejo hazaza h'itsinda-Ikibazo
  5. Kubishyira mubikorwa
  6. Komeza Ubushakashatsi
  7. Ibibazo bikunze kubazwa

Incamake

The original GQA paper studies this as a tradeoff between multi-head and multi-query attention for decoder inference.

Kwibira cyane

The GQA paper proposes grouping query heads so multiple query heads share key/value heads. MHA has one key/value head per query head; MQA shares a single key/value head across all queries; GQA lies between them. The paper evaluates GQA on language modeling and machine translation tasks and reports quality closer to MHA with inference speed closer to MQA in its experiments. It also reports uptraining existing MHA models using about 5% of original pretraining compute. These claims are limited to the paper’s models, tasks and setup.

Gushyira mu bikorwa Isi

Comparing MHA, GQA and MQA head sharing.

Estimating KV-cache size from the number of query heads and key/value groups.

Reading the paper’s reported quality/speed tradeoff on evaluated language-model tasks.

Reviewing the paper’s uptraining recipe for converting an MHA checkpoint to GQA.

Ejo hazaza h'itsinda-Ikibazo

Benchmark head-sharing choices on the target model and hardware. The original paper’s quality/speed results do not establish universal serving speedups or negligible quality loss across all systems.

Kubishyira mubikorwa

Impamvu ari ngombwa

  • Umuvuduko n'igipimo. Ururimi rwakazi rushobora kugenda byihuse nta gutamba guhuzagurika.
  • Kugera no kugera. Yagura uburyo bwindimi nuburyo bwo gutumanaho.
  • Ibyemezo bisobanutse. Amakipe arashobora kumara umwanya munini murubanza mugihe automatike ikora gusubiramo.

Witondere

  • Ibintu bifatika bishobora kwinjiza bucece raporo, gushyigikira imigendekere, cyangwa ibisubizo byubushakashatsi.
  • Kwihuta byihuse birashobora gukora ibisubizo bidahuye mubisabwa bisa.
  • Ibyanditswe byumvikana birashobora kugaragara niba kugenzura kugenzura ari ntege.

Intambwe ikurikira

  1. Sobanura imiterere isohoka, amajwi, hamwe nubuziranenge mbere yo gutangira.
  2. Ibisubizo byibanze hamwe nisoko yizewe igihe cyose ukuri kwingirakamaro.
  3. Komeza kugenzura abantu kugenzura ibisubizo byinshi.
  4. Kurikirana uburyo bwo kunanirwa no kongera imyitozo cyangwa akazi gahoraho.

Komeza Ubushakashatsi

Ibaruwa y'amakuru y'ubuntu

Ibuka amakuru ya buri munsi ya AI

Inkuru eshatu zizewe za AI buri gitondo cy'iminsi y'akazi, zanditse mu rurimi rworoshye. Ubuntu iteka ryose, nta kwamamaza.

Imeri imwe buri munsi y'akazi. Kwiyimura mu buryo bumwe. Ntitwigera tugurisha cyangwa ngo dusangire aderesi yawe.

Isuzume

Take the Grouped-Query Attention quiz

Ibisubizo ako kanya kuri buri kibazo, n'icyemezo gishobora gusangirwa gifite ID yagenzurwa umaze gutsinda isomo.

Tangira ikibazo

Fasha uburezi bwa AI bw'ubuntu. AI Understanding ni umuryango utegamiye kuri leta wa 501(c)(3) — nta kwamamaza, nta kwishyura, iteka ryose. Tanga impano

Ibibazo bikunze kubazwa

What is Grouped-Query Attention?

Grouped-Query Attention (GQA) shares key/value heads across groups of query heads. The original GQA paper studies this as a tradeoff between multi-head and multi-query attention for decoder inference.

Mumatsinda-Ikibazo Cyitonderwa, niki gisangiwe mumatsinda yibibazo?

GQA ibika imitwe yibibazo bitandukanye ariko ikareka buri tsinda ryabo bagabana urufunguzo rumwe rwimitwe nagaciro, kugabanya Kache cache.

GQA isobanurwa neza nkimpamvu yo hagati hagati yizo ntambwe ebyiri?

Multi-head itanga buri mutwe wacyo KV; ibibazo byinshi bisangiye KV imwe kumitwe yose; GQA yicaye hagati hamwe nitsinda rito rya KV.

Ni ikihe gice cyo gufata umwanzuro GQA ikora cyane cyane?

Kuzigama bigaragarira mugihe cyibisekuru, aho gusoma KV cache aribiganza kwibuka-umurongo mugari.

Niba icyitegererezo gifite imitwe 32 yibibazo hamwe nitsinda 8 KV, ubwihisho bwa KV bugabanuka nikihe kintu?

32 ibibazo byibibazo bigabanijwe nitsinda 8 risangiwe KV ritanga hafi kugabanuka inshuro enye murufunguzo rwihishwa nagaciro.

What quality/speed tradeoff did the GQA paper report in its experiments?

The GQA paper reports quality closer to multi-head attention (MHA) with inference speed closer to multi-query attention (MQA) in its evaluated tasks. This is an empirical result, not a universal guarantee.