KwaiYi
v1.0
KwaiYii large model, KwaiYii's Chinese name is Kuaiyi large model, is a Kuaishou big prediction model that has recently started internal testing. Users who want to experience it early can apply for internal testing qualifications. By then, users can enter the KwaiYii large model to experience more content. .
The Kuaishou AI team will continue to iterate the "Ruyi" large model. On the one hand, it will continue to optimize model performance and develop multi-modal capabilities. On the other hand, it will also promote the implementation of more C-side and B-side business scenarios. Kuaishou's self-developed large model "KwaiYii" was unveiled, ranking first in CMMLU Chinese.
KwaiYii is a series of large-scale language models (LLM) independently developed by the Kuaishou AI team from scratch. It currently includes models of multiple parameter sizes and covers pre-trained models (KwaiYii -Base), dialogue model (KwaiYii-Chat). Here we introduce the 13B scale series model KwaiYii-13B. Its main features include
The KwaiYii-13B-Base pre-trained model has excellent general technical base capabilities and has achieved State-Of-The-Art effect under the same model size on most authoritative Chinese/English Benchmarks. For example, the KwaiYii-13B-Base pre-trained model is currently at the leading level of the same model size on MMLU, CMMLU, C-Eval, HumanEval and other benchmarks.
The KwaiYii-13B-Chat dialogue model has excellent language understanding and generation capabilities, supporting a wide range of tasks such as content creation, information consultation, mathematical logic, code writing, multi-round conversations, etc. Human evaluation results show that KwaiYii-13B-Chat exceeds mainstream open source models , and is close to the same level as ChatGPT (3.5) in content creation, information consultation and mathematical problem solving.
