intel / xFasterTransformer

Apache License 2.0
355 stars 61 forks source link

[Kenrel] Add FP16 LLaMA YARN rotary_embedding. #412

Closed changqi1 closed 4 months ago

pujiang2018 commented 4 months ago

@abenmao since it changes LlamaYaRNScaledRotaryEmbedding, could you pls help to review the changes?