Ikibazo gihuza ibibazo · Biragoye Urwego
Cache-Yongerewe Igisekuru (CAG) Ikibazo
Gerageza gusobanukirwa kubyerekeranye na cache-yongerewe ibisekuruza: uburyo KV cache yabanjirijwe kandi ikoreshwa, cache igarura hagati yibibazo, hamwe nubunini bwa corpus.
Ikibazo 1 Bya 8
CAG ikoresha iki aho kubaza buri kibazo?
Reba buri kibazo8 ibibazo
Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.
- CAG ikoresha iki aho kubaza buri kibazo?
- A keyword search engine
- Fine-tuning the model on the documents
- A new vector index per question
- Preloading the whole knowledge base into the KV cache once and reusing it
- Niki cache cache muri transformateur?
- A database of user accounts
- The stored attention keys and values computed for tokens the model has already read
- A list of document URLs
- The model's trained weights
- Kuki sisitemu ya CAG igabanya cache nyuma ya buri gisubizo?
- To save the answer for training
- To delete the documents permanently
- To restore the documents-only state so the next query does not see the previous question and answer
- To switch to a different model
- Ni ubuhe butumwa bukomeye bwa CAG?
- The entire corpus must fit in the model's context window
- It requires a vector database
- It cannot answer questions quickly
- Ihindura uburemere bwikitegererezo
- Kuki CAG isubiza vuba kuruta gushira inyandiko mubisabwa byose?
- It skips generating output tokens
- The model only needs to compute the new question tokens on top of the cached prefix
- Ikoresha icyitegererezo gito
- It retrieves fewer documents
- Which knowledge base is the best fit for CAG?
- A news archive updated every hour
- Millions of customer records with per-user permissions
- An entire web crawl
- A small, stable product manual that rarely changes
- Which statement about CAG is a misconception?
- CAG removes retrieval errors because all documents are present
- Changing documents requires rebuilding the cache
- CAG fine-tunes the model's weights on the documents
- Hosted prompt caching is related to the idea
- Which techniques reduce the memory needed for a large KV cache?
- Grouped-query attention and cache quantization
- Ongeraho imitwe myinshi
- Using longer prompts
- Kuzamura ubushyuhe bw'icyitegererezo