mirror of
https://github.com/ml-explore/mlx.git
synced 2025-12-16 01:49:05 +08:00
Gather qmm batched kernel and refactoring of quantized (#2078)
This commit is contained in:
committed by
GitHub
parent
99eefd2ec0
commit
5de6d94a90
@@ -269,4 +269,21 @@ MTL::ComputePipelineState* get_quantized_kernel(
|
||||
return d.get_kernel(kernel_name);
|
||||
}
|
||||
|
||||
MTL::ComputePipelineState* get_gather_qmm_kernel(
|
||||
metal::Device& d,
|
||||
const std::string& kernel_name,
|
||||
const std::string& hash_name,
|
||||
const metal::MTLFCList& func_consts,
|
||||
const array&,
|
||||
int,
|
||||
int,
|
||||
int,
|
||||
int,
|
||||
int,
|
||||
int,
|
||||
int,
|
||||
bool) {
|
||||
return d.get_kernel(kernel_name, "mlx", hash_name, func_consts);
|
||||
}
|
||||
|
||||
} // namespace mlx::core
|
||||
|
||||
Reference in New Issue
Block a user