whisper.cpp / ggml-cuda

Commit History

sync : llama.cpp (skip)
88282d1
unverified

ggerganov HF Staff commited on

ggml : mul_mat_id use the same tensor for all the experts (llama/6387)
26fdc9f
unverified

slaren ggerganov HF Staff commited on

sync : ggml (#2001)
cbbfa9e
unverified

ggerganov HF Staff commited on