feat(kernel): add vocab embedding CUDA kernels#645

Open

huangzhenhua111 wants to merge 2 commits intoUbiquitousLearning:mainfrom

huangzhenhua111:feat/vocab-embedding

Commits on Mar 2, 2026

feat(kernel): add vocab embedding CUDA kernels
huangzhenhua111
committed

Commits on Mar 3, 2026

feat(kernel): add vocab embedding CUDA kernels with benchmarks
huangzhenhua111
committed