Skip to content

Commit

Permalink
CANN: RoPE operator optimization (ggerganov#10563)
Browse files Browse the repository at this point in the history 
* [cann] RoPE operator optimization

* [CANN]Code Formatting

---------

Co-authored-by: noemotiovon <[email protected]>
  • Loading branch information
@noemotiovon
noemotiovon and noemotiovon authored Nov 29, 2024
1 parent f095a64 commit 938f608
Show file tree
Hide file tree
Showing 2 changed files with 222 additions and 32 deletions.
241 changes: 221 additions & 20 deletions ggml/src/ggml-cann/aclnn_ops.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2965,7 +2965,7 @@ static void aclnn_cache_init(ggml_backend_cann_context& ctx, ggml_tensor* dst,
aclTensor* acl_cos_repeat_tensor,
aclTensor* acl_sin_repeat_tensor,
float theta_scale, float freq_scale,
bool is_neox) {
float attn_factor, bool is_neox) {
// int sin/cos cache, cache has different repeat method depond on
// @param.is_neox

Expand Down Expand Up @@ -3017,6 +3017,7 @@ static void aclnn_cache_init(ggml_backend_cann_context& ctx, ggml_tensor* dst,
ggml_type_size(src2->type), arange_ne, arange_nb, GGML_MAX_DIMS);
aclnn_div_tensor(ctx, acl_theta_scale_tensor, acl_freq_factors_tensor,
nullptr, true);
ACL_CHECK(aclDestroyTensor(acl_freq_factors_tensor));
}

// position
Expand Down Expand Up @@ -3047,16 +3048,6 @@ static void aclnn_cache_init(ggml_backend_cann_context& ctx, ggml_tensor* dst,
aclnn_mul(ctx, acl_position_tensor, acl_theta_scale_tensor,
acl_theta_tensor);

// // power[] * position[] * freq_scale / freq_factors[]
// ggml_cann_pool_alloc theta_final_allocator(ctx.pool(),
// theta_length *
// sizeof(float_t));
// aclTensor* acl_theat_final_tensor = aclnn_zero(
// ctx, theta_final_allocator.get(), sizeof(float_t) * theta_length,
// theta_ne, GGML_MAX_DIMS, ACL_FLOAT, sizeof(float_t));
// aclnn_inplace_addcdiv(ctx, acl_theat_final_tensor, acl_theta_tensor,
// acl_freq_factors_tensor, freq_scale);

// permute: [0,1,2,3]->[0,2,1,3]
int64_t permute_ne[] = {arange_length, 1, position_length, 1};
size_t permute_nb[GGML_MAX_DIMS];
Expand Down Expand Up @@ -3092,6 +3083,12 @@ static void aclnn_cache_init(ggml_backend_cann_context& ctx, ggml_tensor* dst,
GGML_MAX_DIMS, ACL_FORMAT_ND);
aclnn_cos(ctx, acl_permute_tensor, acl_cos_tensor);

// attn_factor
if (attn_factor != 1) {
aclnn_muls(ctx, acl_sin_tensor, attn_factor, nullptr, true);
aclnn_muls(ctx, acl_cos_tensor, attn_factor, nullptr, true);
}

// repeat
if (is_neox) {
int64_t repeatsArray[] = {1, 1, 1, 2};
Expand Down Expand Up @@ -3155,15 +3152,11 @@ void ggml_cann_rope(ggml_backend_cann_context& ctx, ggml_tensor* dst) {
memcpy(&beta_fast, (int32_t*)dst->op_params + 9, sizeof(float));
memcpy(&beta_slow, (int32_t*)dst->op_params + 10, sizeof(float));

// TODO: attn_factor != 1
GGML_ASSERT(attn_factor == 1);
// TODO: n_dims <= ne0
GGML_ASSERT(n_dims == ne0);
GGML_ASSERT(n_dims % 2 == 0);
// TODO: ext_factor != 0
GGML_ASSERT(ext_factor == 0);
// TODO: type == GGML_TYPE_F16
GGML_ASSERT(src0->type == GGML_TYPE_F32);

const float theta_scale = powf(freq_base, -2.0f / n_dims);

Expand Down Expand Up @@ -3194,7 +3187,217 @@ void ggml_cann_rope(ggml_backend_cann_context& ctx, ggml_tensor* dst) {
ggml_cann_create_tensor(cos_buffer, ACL_FLOAT, sizeof(float_t),
sin_reshape_ne, sin_reshape_nb, GGML_MAX_DIMS);
aclnn_cache_init(ctx, dst, acl_cos_reshape_tensor, acl_sin_reshape_tensor,
theta_scale, freq_scale, is_neox);
theta_scale, freq_scale, attn_factor, is_neox);

aclTensor* acl_src = ggml_cann_create_tensor(src0);
aclTensor* acl_dst = ggml_cann_create_tensor(dst);

#ifdef ASCEND_310P
// Special ROPE operation for 310P

// roll input
void* input_roll_buffer;
aclTensor* acl_minus_one_tensor;
void* minus_one_scale_buffer = nullptr;
ggml_cann_pool_alloc roll_allocator(ctx.pool(), ggml_nbytes(src0));
ggml_cann_pool_alloc minus_one_scale_allocator(
ctx.pool(), sizeof(float_t) * src0->ne[0]);
if (!is_neox) {
// roll input: [q0,q1,q2,q3,...] -> [q1,q0,q3,q2,...]
input_roll_buffer = roll_allocator.get();
int64_t input_roll_ne[4] = {2, src0->ne[1] * (src0->ne[0] / 2),
src0->ne[2], src0->ne[3]};
size_t input_roll_nb[GGML_MAX_DIMS];
input_roll_nb[0] = ggml_type_size(src0->type);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
input_roll_nb[i] = input_roll_nb[i - 1] * input_roll_ne[i - 1];
}
aclTensor* acl_input_roll_tensor = ggml_cann_create_tensor(
input_roll_buffer, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), input_roll_ne, input_roll_nb,
GGML_MAX_DIMS);
aclTensor* acl_input_tensor = ggml_cann_create_tensor(
src0->data, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), input_roll_ne, input_roll_nb,
GGML_MAX_DIMS);

int64_t shifts[] = {1};
int64_t dims[] = {3};
aclnn_roll(ctx, acl_input_tensor, acl_input_roll_tensor, shifts, dims);
ACL_CHECK(aclDestroyTensor(acl_input_roll_tensor));
ACL_CHECK(aclDestroyTensor(acl_input_tensor));

// init [-1, 1, -1, 1, ...]
minus_one_scale_buffer = minus_one_scale_allocator.get();

int64_t minus_one_ne[4] = {src0->ne[0], 1, 1, 1};
size_t minus_one_nb[GGML_MAX_DIMS];
minus_one_nb[0] = sizeof(float_t);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
minus_one_nb[i] = minus_one_nb[i - 1] * minus_one_ne[i - 1];
}
acl_minus_one_tensor = aclnn_values(
ctx, minus_one_scale_buffer, sizeof(float_t) * src0->ne[0],
minus_one_ne, GGML_MAX_DIMS, ACL_FLOAT, sizeof(float_t), 1);
int64_t dim = 3;
int64_t* index = new int64_t[src0->ne[0]];
for (int i = 0; i < src0->ne[0]; i++) {
index[i] = i / 2 * 2;
}
int64_t index_num = src0->ne[0];
float value = -1;
aclnn_index_fill_tensor(ctx, acl_minus_one_tensor, dim, index,
index_num, value);
} else {
// roll input: [q0,q1,q2,...] ->
// [q_half,q_half+1,...,q_end,q0,q1,...q_half-1]
input_roll_buffer = roll_allocator.get();
aclTensor* acl_input_roll_tensor = ggml_cann_create_tensor(
input_roll_buffer, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), src0->ne, src0->nb, GGML_MAX_DIMS);
aclTensor* acl_input_tensor = ggml_cann_create_tensor(src0);

int64_t shifts[] = {src0->ne[0] / 2};
int64_t dims[] = {3};
aclnn_roll(ctx, acl_input_tensor, acl_input_roll_tensor, shifts, dims);

ACL_CHECK(aclDestroyTensor(acl_input_roll_tensor));
ACL_CHECK(aclDestroyTensor(acl_input_tensor));
// init [-1, -1, -1, 1, 1，1，...]
minus_one_scale_buffer = minus_one_scale_allocator.get();
int64_t minus_one_ne[4] = {src0->ne[0], 1, 1, 1};
size_t minus_one_nb[GGML_MAX_DIMS];
minus_one_nb[0] = sizeof(float_t);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
minus_one_nb[i] = minus_one_nb[i - 1] * minus_one_ne[i - 1];
}
acl_minus_one_tensor = aclnn_values(
ctx, minus_one_scale_buffer, sizeof(float_t) * src0->ne[0],
minus_one_ne, GGML_MAX_DIMS, ACL_FLOAT, sizeof(float_t), 1);
// -1 * first half
int64_t first_half_ne[4] = {src0->ne[0] / 2, 1, 1, 1};
size_t first_half_nb[GGML_MAX_DIMS];
first_half_nb[0] = sizeof(float_t);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
first_half_nb[i] = first_half_nb[i - 1] * first_half_ne[i - 1];
}
aclTensor* acl_first_half_tensor = ggml_cann_create_tensor(
minus_one_scale_buffer, ACL_FLOAT, sizeof(float_t), first_half_ne,
first_half_nb, GGML_MAX_DIMS);
bool inplace = true;
float scale = -1;
aclnn_muls(ctx, acl_first_half_tensor, scale, nullptr, inplace);
ACL_CHECK(aclDestroyTensor(acl_first_half_tensor));
}

// TODO: n_dims < ne0
GGML_ASSERT(n_dims == src0->ne[0]);

// input * scale
ggml_cann_pool_alloc roll_mul_scale_allocator(ctx.pool(),
ggml_nbytes(src0));
void* input_roll_mul_scale_buffer = roll_mul_scale_allocator.get();
size_t input_nb[GGML_MAX_DIMS];
input_nb[0] = ggml_type_size(src0->type);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
input_nb[i] = input_nb[i - 1] * src0->ne[i - 1];
}
aclTensor* acl_input_roll_mul_scale_tensor = ggml_cann_create_tensor(
input_roll_mul_scale_buffer, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), src0->ne, input_nb, GGML_MAX_DIMS);
aclTensor* acl_input_roll_reshape_tensor = ggml_cann_create_tensor(
input_roll_buffer, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), src0->ne, input_nb, GGML_MAX_DIMS);

aclnn_mul(ctx, acl_input_roll_reshape_tensor, acl_minus_one_tensor,
acl_input_roll_mul_scale_tensor);

// output
void* output_fp32_buffer;
if (src0->type == GGML_TYPE_F32) {
aclnn_inplace_mul(ctx, acl_src, acl_cos_reshape_tensor);
aclnn_inplace_mul(ctx, acl_input_roll_mul_scale_tensor,
acl_sin_reshape_tensor);
aclnn_add(ctx, acl_src, acl_input_roll_mul_scale_tensor, acl_dst);
// TODO: ne0 != n_dims in mode2
} else if (src0->type == GGML_TYPE_F16) {
size_t input_fp32_nb[GGML_MAX_DIMS];
input_fp32_nb[0] = sizeof(float_t);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
input_fp32_nb[i] = input_fp32_nb[i - 1] * dst->ne[i - 1];
}
ggml_cann_pool_alloc fp32_allocator1(
ctx.pool(), ggml_nelements(dst) * sizeof(float_t));
void* input_fp32_buffer1 = fp32_allocator1.get();
aclTensor* input_fp32_tensor1 = ggml_cann_create_tensor(
input_fp32_buffer1, ACL_FLOAT, sizeof(float_t), dst->ne,
input_fp32_nb, GGML_MAX_DIMS);
ggml_cann_pool_alloc fp32_allocator2(
ctx.pool(), ggml_nelements(dst) * sizeof(float_t));
void* input_fp32_buffer2 = fp32_allocator2.get();
aclTensor* input_fp32_tensor2 = ggml_cann_create_tensor(
input_fp32_buffer2, ACL_FLOAT, sizeof(float_t), dst->ne,
input_fp32_nb, GGML_MAX_DIMS);

ggml_cann_pool_alloc fp32_allocator(
ctx.pool(), ggml_nelements(dst) * sizeof(float_t));
output_fp32_buffer = fp32_allocator.get();
aclTensor* output_fp32_tensor = ggml_cann_create_tensor(
output_fp32_buffer, ACL_FLOAT, sizeof(float_t), dst->ne,
input_fp32_nb, GGML_MAX_DIMS);
aclnn_mul(ctx, acl_src, acl_cos_reshape_tensor, input_fp32_tensor1);
aclnn_mul(ctx, acl_input_roll_mul_scale_tensor, acl_sin_reshape_tensor,
input_fp32_tensor2);
aclnn_add(ctx, input_fp32_tensor1, input_fp32_tensor2,
output_fp32_tensor);
aclnn_cast(ctx, output_fp32_tensor, acl_dst, ACL_FLOAT16);

ACL_CHECK(aclDestroyTensor(input_fp32_tensor1));
ACL_CHECK(aclDestroyTensor(input_fp32_tensor2));
ACL_CHECK(aclDestroyTensor(output_fp32_tensor));
ACL_CHECK(aclDestroyTensor(acl_sin_reshape_tensor));
ACL_CHECK(aclDestroyTensor(acl_minus_one_tensor));
ACL_CHECK(aclDestroyTensor(acl_input_roll_mul_scale_tensor));
ACL_CHECK(aclDestroyTensor(acl_input_roll_reshape_tensor));
ACL_CHECK(aclDestroyTensor(acl_src));
}
return;
#endif

// src0 == GGML_TYPE_F16
// TODO: optimization this `if` code
if (src0->type == GGML_TYPE_F16) {
ggml_cann_pool_alloc sin_final_allocator(
ctx.pool(), src0->ne[0] * src0->ne[2] * ggml_type_size(src0->type));
ggml_cann_pool_alloc cos_final_allocator(
ctx.pool(), src0->ne[0] * src0->ne[2] * ggml_type_size(src0->type));
void* sin_final_buffer = sin_final_allocator.get();
void* cos_final_buffer = cos_final_allocator.get();

int64_t sin_final_ne[4] = {src0->ne[0], 1, src0->ne[2], 1};
size_t sin_final_nb[GGML_MAX_DIMS];
sin_final_nb[0] = ggml_type_size(src0->type);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
sin_final_nb[i] = sin_final_nb[i - 1] * sin_final_ne[i - 1];
}
aclTensor* acl_sin_final_tensor = ggml_cann_create_tensor(
sin_final_buffer, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), sin_final_ne, sin_final_nb,
GGML_MAX_DIMS);
aclTensor* acl_cos_final_tensor = ggml_cann_create_tensor(
cos_final_buffer, ggml_cann_type_mapping(src0->type),
ggml_type_size(src0->type), sin_final_ne, sin_final_nb,
GGML_MAX_DIMS);

aclnn_cast(ctx, acl_sin_reshape_tensor, acl_sin_final_tensor,
ggml_cann_type_mapping(src0->type));
aclnn_cast(ctx, acl_cos_reshape_tensor, acl_cos_final_tensor,
ggml_cann_type_mapping(src0->type));
ACL_CHECK(aclDestroyTensor(acl_cos_reshape_tensor));
ACL_CHECK(aclDestroyTensor(acl_sin_reshape_tensor));
acl_sin_reshape_tensor = acl_sin_final_tensor;
acl_cos_reshape_tensor = acl_cos_final_tensor;
}

uint64_t workspaceSize = 0;
aclOpExecutor* executor;
Expand All @@ -3206,10 +3409,8 @@ void ggml_cann_rope(ggml_backend_cann_context& ctx, ggml_tensor* dst) {
acl_mode = 1;
}

aclTensor* acl_x = ggml_cann_create_tensor(src0);
aclTensor* acl_dst = ggml_cann_create_tensor(dst);
ACL_CHECK(aclnnRotaryPositionEmbeddingGetWorkspaceSize(
acl_x, acl_cos_reshape_tensor, acl_sin_reshape_tensor, acl_mode,
acl_src, acl_cos_reshape_tensor, acl_sin_reshape_tensor, acl_mode,
acl_dst, &workspaceSize, &executor));
if (workspaceSize > 0) {
ggml_cann_pool_alloc workspace_allocator(ctx.pool(), workspaceSize);
Expand All @@ -3219,7 +3420,7 @@ void ggml_cann_rope(ggml_backend_cann_context& ctx, ggml_tensor* dst) {
ACL_CHECK(aclnnRotaryPositionEmbedding(workspaceAddr, workspaceSize,
executor, ctx.stream()));

ACL_CHECK(aclDestroyTensor(acl_x));
ACL_CHECK(aclDestroyTensor(acl_src));
ACL_CHECK(aclDestroyTensor(acl_cos_reshape_tensor));
ACL_CHECK(aclDestroyTensor(acl_sin_reshape_tensor));
ACL_CHECK(aclDestroyTensor(acl_dst));
Expand Down
13 changes: 1 addition & 12 deletions ggml/src/ggml-cann/ggml-cann.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1739,7 +1739,6 @@ static bool ggml_backend_cann_supports_op(ggml_backend_dev_t dev,
case GGML_OP_ROPE: {
// TODO: with ops-test v == 1
float * ext_factor = (float*)((int32_t*)op->op_params + 7);
float * attn_factor = (float*)((int32_t*)op->op_params + 8);
// TODO: n_dims <= ne0
if (op->src[0]->ne[0] != op->op_params[1]) {
return false;
Expand All @@ -1748,17 +1747,7 @@ static bool ggml_backend_cann_supports_op(ggml_backend_dev_t dev,
if (*ext_factor != 0) {
return false;
}
// TODO: attn_factor != 1
if (*attn_factor != 1) {
return false;
}
//TODO: type == GGML_TYPE_F16
switch (op->src[0]->type) {
case GGML_TYPE_F32:
return true;
default:
return false;
}
return true;
}
case GGML_OP_UPSCALE: {
// aclnnUpsampleNearest2dGetWorkspaceSize not support
Expand Down

0 comments on commit 938f608

Please sign in to comment.