diff --git a/mlx/backend/no_cpu/primitives.cpp b/mlx/backend/no_cpu/primitives.cpp index dba82c6dc..4d373bd1a 100644 --- a/mlx/backend/no_cpu/primitives.cpp +++ b/mlx/backend/no_cpu/primitives.cpp @@ -130,6 +130,7 @@ NO_CPU(View) namespace fast { NO_CPU_MULTI(Quantize) +NO_CPU_MULTI(ConvertFP8) } // namespace fast namespace distributed { diff --git a/mlx/backend/no_gpu/primitives.cpp b/mlx/backend/no_gpu/primitives.cpp index 22a0c8acc..a57df046c 100644 --- a/mlx/backend/no_gpu/primitives.cpp +++ b/mlx/backend/no_gpu/primitives.cpp @@ -154,6 +154,7 @@ NO_GPU_USE_FALLBACK(RMSNorm) NO_GPU_MULTI(RMSNormVJP) NO_GPU_USE_FALLBACK(RoPE) NO_GPU(ScaledDotProductAttention) +NO_GPU_MULTI(ConvertFP8) NO_GPU_MULTI(Quantize) NO_GPU_MULTI(CustomKernel) } // namespace fast