quiz bu lëkkale ak tegtal · Diggante Niveau
Yi models by 01.AI Quiz
Xoolaat cosaanu famiy model Yi, focus bilingue, decoder-only design, xeetu contexte yu gudd, ak njàngale-oriented post-training.
Laaj 1 de 8
Kan moo sos 01.AI, liggéeyukaay bi ci ginaaw model Yi yi?
Xoolaatal laaj bu nekk8 laaj
Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.
- Kan moo sos 01.AI, liggéeyukaay bi ci ginaaw model Yi yi?
- Kaifu Lee
- Robin Li
- Jack Ma
- André Ng
- Lan mooy ñaari xeetu Yi yiñ njëkka génne?
- Yi-1 ak Yi-2
- yi-6b ak yi-34b
- yi-100 ak yi-200
- yi-mini ak yi-max
- Ban kàttan lañu gëna xamee xeetu Yi yi?
- Defar ay nataal kese
- Dawal te amul benn GPU
- Liggéey bu am doole ci ñaari làkk, Chinois-Anglais
- Tekki audio buñ wax rek
- Ban xeetu architecture la xeetu Yi yi topp?
- Design LSTM buy baaxoo
- ab xeetu diffusion
- Modèle encodeur bu sell bu melni BERT
- Transformatëru dekodeer bu nuru Llama
- Lu tollu ci ban guddaay la palanteer contexte bu contexte Yi bu gudd bi?
- 200k jetons
- 4K jetons
- 32K jetons
- 1 milioŋ ciy jeton
- What did 01.AI’s Yi technical report identify as an important driver of model performance?
- Bàyyi IA lëmm
- Data quality resulting from data-engineering efforts
- Tabax puce hardware kese
- boole ak OpenAI
- What does supervised fine-tuning plus reinforcement learning from human feedback aim to improve in Yi chat models?
- Compressing retinal images
- Changing the model into an encoder-only network
- Following instructions in conversational use
- Removing all text-generation ability
- What data strategy did 01.AI emphasize for the Yi models?
- Training exclusively on synthetic audio
- Avoiding all data filtering
- Using only programming-language source code
- Careful data quality and curation rather than relying only on sheer scale