[Refactor] Remove dead cuda kernels (#42767)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
@@ -92,16 +92,6 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, m) {
|
||||
"bool use_fp32_reduce, bool is_zp_float,"
|
||||
"int thread_k, int thread_n, int blocks_per_sm) -> Tensor");
|
||||
|
||||
m.def(
|
||||
"marlin_gemm_moe(Tensor! a, Tensor! b_q_weights, Tensor! sorted_ids, "
|
||||
"Tensor! topk_weights, Tensor! topk_ids, Tensor! b_scales, Tensor! "
|
||||
"b_zeros, Tensor! g_idx, Tensor! perm, Tensor! workspace, "
|
||||
"int b_q_type, SymInt size_m, "
|
||||
"SymInt size_n, SymInt size_k, bool is_k_full, int num_experts, int "
|
||||
"topk, "
|
||||
"int moe_block_size, bool replicate_input, bool apply_weights)"
|
||||
" -> Tensor");
|
||||
|
||||
m.def(
|
||||
"moe_permute(Tensor input, Tensor topk_ids,"
|
||||
"Tensor token_expert_indices, Tensor? expert_map, int n_expert,"
|
||||
|
||||
Reference in New Issue
Block a user