Skip to content

Commit

Permalink
[Kernel][Model] Varlen prefill + Prefill chunking support for mamba k…
Browse files Browse the repository at this point in the history 
…ernels and Jamba model (vllm-project#8533)
  • Loading branch information
@mzusman @liuyanyi
mzusman authored and liuyanyi committed Oct 6, 2024
1 parent 8b228f5 commit 56e8cea
Show file tree
Hide file tree
Showing 13 changed files with 1,176 additions and 894 deletions.
527 changes: 211 additions & 316 deletions csrc/mamba/causal_conv1d/causal_conv1d.cu
View file

Large diffs are not rendered by default.

10 changes: 10 additions & 0 deletions csrc/mamba/causal_conv1d/causal_conv1d.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,7 @@ struct ConvParamsBase {
index_t out_c_stride;
index_t out_l_stride;

int conv_state_len;
index_t conv_state_batch_stride;
index_t conv_state_c_stride;
index_t conv_state_l_stride;
Expand All @@ -35,6 +36,10 @@ struct ConvParamsBase {
void *__restrict__ out_ptr;

void *__restrict__ conv_state_ptr;
void *__restrict__ query_start_loc_ptr;
void *__restrict__ has_initial_state_ptr;
void *__restrict__ cache_indices_ptr;
int32_t *__restrict__ cache_seqlens;

// For the continuous batching case. Makes it so that the mamba state for
// the current batch doesn't need to be a contiguous tensor.
Expand All @@ -52,6 +57,11 @@ struct ConvParamsBase {
index_t final_states_batch_stride;
index_t final_states_l_stride;
index_t final_states_c_stride;

void * conv_states_ptr;
index_t conv_states_batch_stride;
index_t conv_states_l_stride;
index_t conv_states_c_stride;
};


Expand Down
29 changes: 9 additions & 20 deletions csrc/mamba/mamba_ssm/selective_scan.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -54,10 +54,14 @@ struct SSMParamsBase {
void *__restrict__ delta_ptr;
void *__restrict__ delta_bias_ptr;
void *__restrict__ out_ptr;
void *__restrict__ x_ptr;
void *__restrict__ ssm_states_ptr;
void *__restrict__ z_ptr;
void *__restrict__ out_z_ptr;
void *__restrict__ index_ptr;

void *__restrict__ query_start_loc_ptr;
void *__restrict__ cache_indices_ptr;
void *__restrict__ has_initial_state_ptr;

};


Expand Down Expand Up @@ -201,7 +205,7 @@ inline __device__ void load_input(typename Ktraits::input_t *u,
typename Ktraits::input_t (&u_vals)[Ktraits::kNItems],
typename Ktraits::BlockLoadT::TempStorage &smem_load,
int seqlen) {
if constexpr (Ktraits::kIsEvenLen) {
if constexpr (Ktraits::kIsEvenLen && !Ktraits::kVarlen) {
auto& smem_load_vec = reinterpret_cast<typename Ktraits::BlockLoadVecT::TempStorage&>(smem_load);
using vec_t = typename Ktraits::vec_t;
typename Ktraits::BlockLoadVecT(smem_load_vec).Load(
Expand All @@ -217,21 +221,6 @@ inline __device__ void load_input(typename Ktraits::input_t *u,
}
}

template<typename Ktraits>
inline __device__ void load_index(int *u,
int (&u_vals)[Ktraits::kNItems],
typename Ktraits::BlockLoadIndexT::TempStorage &smem_load_index,
int seqlen) {
if constexpr (Ktraits::kIsEvenLen) {
auto& smem_load_index_vec = reinterpret_cast<typename Ktraits::BlockLoadIndexVecT::TempStorage&>(smem_load_index);
Ktraits::BlockLoadIndexVecT(smem_load_index_vec).Load(
reinterpret_cast<uint4*>(u),
reinterpret_cast<uint4(&)[Ktraits::kNLoadsIndex]>(u_vals)
);
} else {
Ktraits::BlockLoadIndexT(smem_load_index).Load(u, u_vals, seqlen, 0);
}
}

template<typename Ktraits>
inline __device__ void load_weight(typename Ktraits::input_t *Bvar,
Expand All @@ -240,7 +229,7 @@ inline __device__ void load_weight(typename Ktraits::input_t *Bvar,
int seqlen) {
constexpr int kNItems = Ktraits::kNItems;
typename Ktraits::input_t B_vals_load[kNItems];
if constexpr (Ktraits::kIsEvenLen) {
if constexpr (Ktraits::kIsEvenLen && !Ktraits::kVarlen) {
auto& smem_load_weight_vec = reinterpret_cast<typename Ktraits::BlockLoadWeightVecT::TempStorage&>(smem_load_weight);
using vec_t = typename Ktraits::vec_t;
typename Ktraits::BlockLoadWeightVecT(smem_load_weight_vec).Load(
Expand All @@ -263,7 +252,7 @@ inline __device__ void store_output(typename Ktraits::input_t *out,
typename Ktraits::input_t write_vals[Ktraits::kNItems];
#pragma unroll
for (int i = 0; i < Ktraits::kNItems; ++i) { write_vals[i] = out_vals[i]; }
if constexpr (Ktraits::kIsEvenLen) {
if constexpr (Ktraits::kIsEvenLen && !Ktraits::kVarlen) {
auto& smem_store_vec = reinterpret_cast<typename Ktraits::BlockStoreVecT::TempStorage&>(smem_store);
using vec_t = typename Ktraits::vec_t;
typename Ktraits::BlockStoreVecT(smem_store_vec).Store(
Expand Down
297 changes: 179 additions & 118 deletions csrc/mamba/mamba_ssm/selective_scan_fwd.cu
View file

Large diffs are not rendered by default.

31 changes: 18 additions & 13 deletions csrc/ops.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -215,25 +215,30 @@ void moe_align_block_size(torch::Tensor topk_ids, int64_t num_experts,
torch::Tensor experts_ids,
torch::Tensor num_tokens_post_pad);

std::vector<torch::Tensor> selective_scan_fwd(
const torch::Tensor& u, const torch::Tensor& delta, const torch::Tensor& A,
const torch::Tensor& B, const torch::Tensor& C,
const c10::optional<torch::Tensor>& D_,
const c10::optional<torch::Tensor>& z_,
const c10::optional<torch::Tensor>& delta_bias_, bool delta_softplus,
const c10::optional<torch::Tensor>& index_,
const c10::optional<torch::Tensor>& x);
void selective_scan_fwd(const torch::Tensor& u, const torch::Tensor& delta,
const torch::Tensor& A, const torch::Tensor& B,
const torch::Tensor& C,
const c10::optional<torch::Tensor>& D_,
const c10::optional<torch::Tensor>& z_,
const c10::optional<torch::Tensor>& delta_bias_,
bool delta_softplus,
const c10::optional<torch::Tensor>& query_start_loc,
const c10::optional<torch::Tensor>& cache_indices,
const c10::optional<torch::Tensor>& has_initial_state,
const torch::Tensor& ssm_states);

at::Tensor causal_conv1d_update(
const at::Tensor& x, const at::Tensor& conv_state, const at::Tensor& weight,
const c10::optional<at::Tensor>& bias, bool silu_activation,
const c10::optional<at::Tensor>& conv_state_indices);
const c10::optional<at::Tensor>& bias_, bool silu_activation,
const c10::optional<at::Tensor>& cache_seqlens_,
const c10::optional<at::Tensor>& conv_state_indices_);

at::Tensor causal_conv1d_fwd(const at::Tensor& x, const at::Tensor& weight,
const c10::optional<at::Tensor>& bias_,
const c10::optional<at::Tensor>& seq_idx_,
const c10::optional<at::Tensor>& initial_states_,
const c10::optional<at::Tensor>& final_states_out_,
const c10::optional<at::Tensor>& conv_states,
const c10::optional<at::Tensor>& query_start_loc,
const c10::optional<at::Tensor>& cache_indices,
const c10::optional<at::Tensor>& has_initial_state,
bool silu_activation);

#ifndef USE_ROCM
Expand Down
17 changes: 11 additions & 6 deletions csrc/torch_bindings.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -273,26 +273,31 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
ops.def(
"selective_scan_fwd(Tensor! u, Tensor! delta,"
"Tensor! A, Tensor! B, Tensor! C,"
"Tensor? D_, Tensor? z_, Tensor? delta_bias_,"
"Tensor? D_, Tensor!? z_, Tensor? delta_bias_,"
"bool delta_softplus,"
"Tensor? index_, Tensor!? x) -> Tensor[]");
"Tensor? query_start_loc,"
"Tensor? cache_indices,"
"Tensor? has_initial_state,"
"Tensor! ssm_states) -> ()");
ops.impl("selective_scan_fwd", torch::kCUDA, &selective_scan_fwd);

ops.def(
"causal_conv1d_update(Tensor! x,"
"Tensor! conv_state,"
"Tensor! weight,"
"Tensor? bias,"
"Tensor? bias_,"
"bool silu_activation,"
"Tensor? cache_seqlens_,"
"Tensor? conv_state_indices) -> Tensor");
ops.impl("causal_conv1d_update", torch::kCUDA, &causal_conv1d_update);

ops.def(
"causal_conv1d_fwd(Tensor! x, Tensor! weight,"
"Tensor? bias_,"
"Tensor? seq_idx_,"
"Tensor? initial_states_,"
"Tensor!? final_states_out_,"
"Tensor!? conv_states,"
"Tensor? query_start_loc,"
"Tensor? cache_indices,"
"Tensor? has_initial_state,"
"bool silu_activation) -> Tensor");
ops.impl("causal_conv1d_fwd", torch::kCUDA, &causal_conv1d_fwd);
#endif
Expand Down
Loading

0 comments on commit 56e8cea

Please sign in to comment.