[Layer][Kernel] Merge batchSize and seqLen into one param (tokenSize) in TokenEembedding - Githubissues

intel / xFasterTransformer

Apache License 2.0

344 stars 60 forks source link

[Layer][Kernel] Merge batchSize and seqLen into one param (tokenSize) in TokenEembedding #350

Closed pujiang2018 closed 4 months ago