|
#include <sycl/sycl.hpp> |
|
#include "wkv6.hpp" |
|
|
|
constexpr int WKV_BLOCK_SIZE = 64; |
|
|
|
|
|
static void rwkv_wkv_f32_kernel( |
|
const int B, const int T, const int C, const int H, |
|
const float* k, const float* v, const float* r, |
|
const float* tf, const float* td, const float* s, |
|
float* dst, const sycl::nd_item<3>& item_ct1, float* shared_mem) { |
|
|
|
const int tid = item_ct1.get_local_id(2); |
|
const int bid = item_ct1.get_group(2); |
|
|
|
const int head_size = WKV_BLOCK_SIZE; |
|
const int batch_i = bid / H; |
|
const int head_i = bid % H; |
|
const int state_size = C * head_size; |
|
const int n_seq_tokens = T / B; |
|
|
|
|
|
float* _k = shared_mem; |
|
float* _r = _k + head_size; |
|
float* _tf = _r + head_size; |
|
float* _td = _tf + head_size; |
|
|
|
|
|
float state[WKV_BLOCK_SIZE]; |
|
|
|
|
|
#pragma unroll |
|
for (int i = 0; i < head_size; i++) { |
|
state[i] = s[batch_i * state_size + head_i * head_size * head_size + i * head_size + tid]; |
|
} |
|
|
|
|
|
item_ct1.barrier(sycl::access::fence_space::local_space); |
|
|
|
|
|
_tf[tid] = tf[head_i * head_size + tid]; |
|
item_ct1.barrier(sycl::access::fence_space::local_space); |
|
|
|
|
|
for (int t = batch_i * n_seq_tokens * C + head_i * head_size + tid; |
|
t < (batch_i + 1) * n_seq_tokens * C + head_i * head_size + tid; |
|
t += C) { |
|
|
|
item_ct1.barrier(sycl::access::fence_space::local_space); |
|
|
|
|
|
_k[tid] = k[t]; |
|
_r[tid] = r[t]; |
|
_td[tid] = td[t]; |
|
|
|
item_ct1.barrier(sycl::access::fence_space::local_space); |
|
|
|
const float _v = v[t]; |
|
float y = 0; |
|
|
|
|
|
sycl::float4 k4, r4, tf4, td4, s4; |
|
#pragma unroll |
|
for (int j = 0; j < head_size; j += 4) { |
|
|
|
k4 = sycl::float4(_k[j], _k[j+1], _k[j+2], _k[j+3]); |
|
r4 = sycl::float4(_r[j], _r[j+1], _r[j+2], _r[j+3]); |
|
tf4 = sycl::float4(_tf[j], _tf[j+1], _tf[j+2], _tf[j+3]); |
|
td4 = sycl::float4(_td[j], _td[j+1], _td[j+2], _td[j+3]); |
|
s4 = sycl::float4(state[j], state[j+1], state[j+2], state[j+3]); |
|
|
|
|
|
sycl::float4 kv4 = k4 * _v; |
|
|
|
|
|
y += sycl::dot(r4, tf4 * kv4 + s4); |
|
|
|
|
|
s4 = s4 * td4 + kv4; |
|
|
|
|
|
state[j] = s4.x(); |
|
state[j+1] = s4.y(); |
|
state[j+2] = s4.z(); |
|
state[j+3] = s4.w(); |
|
} |
|
|
|
dst[t] = y; |
|
} |
|
|
|
|
|
#pragma unroll |
|
for (int i = 0; i < head_size; i++) { |
|
dst[T * C + batch_i * state_size + head_i * head_size * head_size + i * head_size + tid] = state[i]; |
|
} |
|
} |
|
|
|
void ggml_sycl_op_rwkv_wkv6(ggml_backend_sycl_context& ctx, ggml_tensor* dst) { |
|
|
|
const ggml_tensor *src0 = dst->src[0]; |
|
const ggml_tensor *src1 = dst->src[1]; |
|
|
|
const float* k_d = (const float*)dst->src[0]->data; |
|
const float* v_d = (const float*)dst->src[1]->data; |
|
const float* r_d = (const float*)dst->src[2]->data; |
|
const float* tf_d = (const float*)dst->src[3]->data; |
|
const float* td_d = (const float*)dst->src[4]->data; |
|
const float* s_d = (const float*)dst->src[5]->data; |
|
float* dst_d = (float*)dst->data; |
|
|
|
const int64_t B = dst->src[5]->ne[1]; |
|
const int64_t T = dst->src[0]->ne[2]; |
|
const int64_t C = dst->ne[0]; |
|
const int64_t H = dst->src[0]->ne[1]; |
|
|
|
GGML_ASSERT(dst->src[5]->type == GGML_TYPE_F32); |
|
GGML_ASSERT(C % H == 0); |
|
GGML_ASSERT(C / H == WKV_BLOCK_SIZE); |
|
|
|
dpct::queue_ptr stream = ctx.stream(); |
|
|
|
|
|
const size_t shared_mem_size = WKV_BLOCK_SIZE * 4 * sizeof(float); |
|
sycl::range<3> block_dims(1, 1, C / H); |
|
sycl::range<3> grid_dims(1, 1, B * H); |
|
|
|
|
|
stream->submit([&](sycl::handler& cgh) { |
|
sycl::local_accessor<float, 1> shared_mem_acc(shared_mem_size, cgh); |
|
|
|
cgh.parallel_for( |
|
sycl::nd_range<3>(grid_dims * block_dims, block_dims), |
|
[=](sycl::nd_item<3> item_ct1) { |
|
rwkv_wkv_f32_kernel( |
|
B, T, C, H, k_d, v_d, r_d, tf_d, td_d, s_d, dst_d, |
|
item_ct1, (float*)shared_mem_acc.get_multi_ptr<sycl::access::decorated::no>().get() |
|
); |
|
}); |
|
}); |
|
|
|
GGML_UNUSED(src0); |
|
GGML_UNUSED(src1); |
|
} |
|
|