### What this PR does / why we need it?
1. MagicMTP (paper: "Block Verification Accelerates Speculative
Decoding") was introduced to consider the influence among multiple draft
tokens, improving the acceptance rate without compromising accuracy.
2. Added Triton and PyTorch implementations, and added E2E test cases.
### Does this PR introduce _any_ user-facing change?
MagicMTP will automatically take effect when the parameter
"num_speculative_tokens" >= 3.
- vLLM version: v0.13.0
- vLLM main:
7157596103
Signed-off-by: chenaoxuan <cax1165@163.com>
462 lines
18 KiB
Python
462 lines
18 KiB
Python
#
|
|
# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
|
|
# This file is a part of the vllm-ascend project.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
#
|
|
|
|
from vllm.triton_utils import tl, triton
|
|
|
|
from vllm_ascend.ops.triton.triton_utils import get_vectorcore_num
|
|
|
|
|
|
def cal_grid_and_block_size(batch_size: int):
|
|
vectorcore_num = get_vectorcore_num()
|
|
if batch_size <= vectorcore_num:
|
|
grid = batch_size
|
|
block_size = 1
|
|
else:
|
|
grid = vectorcore_num
|
|
block_size = triton.next_power_of_2(triton.cdiv(batch_size, grid))
|
|
return grid, block_size
|
|
|
|
|
|
@triton.jit(do_not_specialize=["max_spec_len"])
|
|
def bonus_renew_1(
|
|
bonus_token_ids_ptr,
|
|
position,
|
|
output_token_ids_ptr,
|
|
):
|
|
bonus_token_id = tl.load(bonus_token_ids_ptr + position)
|
|
tl.store(output_token_ids_ptr + position * 2 + 1, bonus_token_id)
|
|
|
|
|
|
@triton.jit(do_not_specialize=["max_spec_len"])
|
|
def rejection_greedy_sample_spec_len_1_triton(
|
|
output_token_ids_ptr, # [batch_size, 2]
|
|
draft_token_ids_ptr, # [num_tokens]
|
|
target_argmax_ptr, # [num_tokens]
|
|
bonus_token_ids_ptr,
|
|
vec_len,
|
|
BLOCK_SIZE: tl.constexpr,
|
|
):
|
|
block_idx = tl.program_id(0)
|
|
offset = block_idx * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
|
|
mask = offset < vec_len
|
|
|
|
draft_token_id = tl.load(draft_token_ids_ptr + offset, mask)
|
|
target_argmax_id = tl.load(target_argmax_ptr + offset, mask)
|
|
tl.store(output_token_ids_ptr + offset * 2, target_argmax_id, mask)
|
|
|
|
for pos in tl.range(0, BLOCK_SIZE):
|
|
draft_token_id1 = tl.get_element(draft_token_id, (pos, ))
|
|
target_argmax1 = tl.get_element(target_argmax_id, (pos, ))
|
|
position = block_idx * BLOCK_SIZE + pos
|
|
if draft_token_id1 == target_argmax1:
|
|
bonus_renew_1(
|
|
bonus_token_ids_ptr,
|
|
position,
|
|
output_token_ids_ptr,
|
|
)
|
|
|
|
|
|
@triton.jit(do_not_specialize=["max_spec_len"])
|
|
def bonus_renew(
|
|
bonus_token_ids_ptr,
|
|
position,
|
|
output_token_ids_ptr,
|
|
max_spec_len,
|
|
num_tokens1,
|
|
):
|
|
bonus_token_id = tl.load(bonus_token_ids_ptr + position)
|
|
tl.store(
|
|
output_token_ids_ptr + position * (max_spec_len + 1) + num_tokens1,
|
|
bonus_token_id)
|
|
|
|
|
|
@triton.jit(do_not_specialize=["max_spec_len"])
|
|
def rejection_greedy_sample_triton(
|
|
output_token_ids_ptr, # [batch_size, max_spec_len + 1]
|
|
cu_num_draft_tokens_ptr, # [batch_size]
|
|
draft_token_ids_ptr, # [num_tokens]
|
|
target_argmax_ptr, # [num_tokens]
|
|
bonus_token_ids_ptr, # [batch_size]
|
|
is_greedy_ptr, # [batch_size] or None
|
|
vec_len,
|
|
max_spec_len,
|
|
BLOCK_SIZE: tl.constexpr,
|
|
):
|
|
block_idx = tl.program_id(0)
|
|
offset = block_idx * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
|
|
mask = offset < vec_len
|
|
|
|
if is_greedy_ptr is None:
|
|
is_greedy_mask = mask
|
|
else:
|
|
is_greedy = tl.load(is_greedy_ptr + offset, mask=mask, other=0)
|
|
is_greedy_mask = mask & (is_greedy != 0)
|
|
|
|
start_idx = tl.where(
|
|
offset == 0, 0,
|
|
tl.load(cu_num_draft_tokens_ptr + offset - 1, is_greedy_mask))
|
|
end_idx = tl.load(cu_num_draft_tokens_ptr + offset, is_greedy_mask)
|
|
num_draft_tokens = end_idx - start_idx
|
|
|
|
for pos in tl.range(0, BLOCK_SIZE):
|
|
num_tokens1 = tl.get_element(num_draft_tokens, (pos, ))
|
|
rejected = False
|
|
start_idx1 = tl.get_element(start_idx, (pos, ))
|
|
is_greedy_mask1 = tl.get_element(is_greedy_mask, (pos, ))
|
|
position = block_idx * BLOCK_SIZE + pos
|
|
for i in range(num_tokens1):
|
|
if not rejected:
|
|
draft_token_id = tl.load(draft_token_ids_ptr + start_idx1 + i)
|
|
target_argmax_id = tl.load(target_argmax_ptr + start_idx1 + i)
|
|
tl.store(
|
|
output_token_ids_ptr + position * (max_spec_len + 1) + i,
|
|
target_argmax_id,
|
|
)
|
|
if draft_token_id != target_argmax_id:
|
|
# Reject.
|
|
rejected = True
|
|
|
|
if not rejected and is_greedy_mask1:
|
|
bonus_renew(
|
|
bonus_token_ids_ptr,
|
|
position,
|
|
output_token_ids_ptr,
|
|
max_spec_len,
|
|
num_tokens1,
|
|
)
|
|
|
|
|
|
@triton.jit(do_not_specialize=["max_spec_len"])
|
|
def rejection_random_sample_kernel(
|
|
output_token_ids_ptr, # [batch_size, max_spec_len + 1]
|
|
cu_num_draft_tokens_ptr, # [batch_size]
|
|
draft_token_ids_ptr, # [num_tokens]
|
|
draft_probs_ptr, # [num_tokens, vocab_size] or None
|
|
target_probs_ptr, # [num_tokens, vocab_size]
|
|
bonus_token_ids_ptr, # [batch_size]
|
|
recovered_token_ids_ptr, # [num_tokens]
|
|
uniform_probs_ptr, # [num_tokens]
|
|
is_greedy_ptr, # [batch_size]
|
|
max_spec_len,
|
|
vocab_size,
|
|
vec_len,
|
|
NO_DRAFT_PROBS: tl.constexpr,
|
|
BLOCK_SIZE: tl.constexpr):
|
|
block_idx = tl.program_id(0)
|
|
offsets = block_idx * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
|
|
mask = offsets < vec_len
|
|
is_greedy = tl.load(is_greedy_ptr + offsets, mask, other=1)
|
|
not_greedy_mask = is_greedy == 0
|
|
start_idxs = tl.where(
|
|
offsets == 0, 0,
|
|
tl.load(cu_num_draft_tokens_ptr + offsets - 1, not_greedy_mask))
|
|
end_idxs = tl.load(cu_num_draft_tokens_ptr + offsets, not_greedy_mask)
|
|
n_num_draft_tokens = end_idxs - start_idxs
|
|
for req_i in range(BLOCK_SIZE):
|
|
not_greedy = tl.get_element(not_greedy_mask, (req_i, ))
|
|
if not_greedy:
|
|
rejected = False
|
|
start_idx = tl.get_element(start_idxs, (req_i, ))
|
|
req_idx = block_idx * BLOCK_SIZE + req_i
|
|
num_draft_tokens = tl.get_element(n_num_draft_tokens, (req_i, ))
|
|
for pos in range(num_draft_tokens):
|
|
if not rejected:
|
|
draft_token_id = tl.load(draft_token_ids_ptr + start_idx +
|
|
pos)
|
|
if NO_DRAFT_PROBS:
|
|
draft_prob = 1
|
|
else:
|
|
draft_prob = tl.load(draft_probs_ptr +
|
|
(start_idx + pos) * vocab_size +
|
|
draft_token_id)
|
|
target_prob = tl.load(target_probs_ptr +
|
|
(start_idx + pos) * vocab_size +
|
|
draft_token_id)
|
|
uniform_prob = tl.load(uniform_probs_ptr + start_idx + pos)
|
|
# NOTE(woosuk): While the draft probability should never be 0,
|
|
# we check it to avoid NaNs. If it happens to be 0, we reject.
|
|
if draft_prob > 0 and target_prob / draft_prob >= uniform_prob:
|
|
# Accept.
|
|
token_id = draft_token_id
|
|
else:
|
|
# Reject. Use recovered token.
|
|
rejected = True
|
|
token_id = tl.load(recovered_token_ids_ptr +
|
|
start_idx + pos)
|
|
tl.store(
|
|
output_token_ids_ptr + req_idx * (max_spec_len + 1) +
|
|
pos, token_id)
|
|
if not rejected:
|
|
# If all tokens are accepted, append the bonus token.
|
|
bonus_token_id = tl.load(bonus_token_ids_ptr + req_idx)
|
|
tl.store(
|
|
output_token_ids_ptr + req_idx * (max_spec_len + 1) +
|
|
num_draft_tokens,
|
|
bonus_token_id,
|
|
)
|
|
|
|
|
|
@triton.jit(do_not_specialize=["replace_from", "replace_to"])
|
|
def expand_kernel(
|
|
output_ptr, # [num_tokens]
|
|
input_ptr, # [batch_size]
|
|
cu_num_tokens_ptr, # [batch_size]
|
|
replace_from,
|
|
replace_to,
|
|
vec_len,
|
|
MAX_NUM_TOKENS: tl.constexpr,
|
|
BLOCK_SIZE: tl.constexpr,
|
|
):
|
|
req_idx = tl.program_id(0)
|
|
offset = req_idx * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
|
|
len_mask = offset < vec_len
|
|
|
|
start_idx = tl.where(offset == 0, 0,
|
|
tl.load(cu_num_tokens_ptr + offset - 1, len_mask))
|
|
end_idx = tl.load(cu_num_tokens_ptr + offset, len_mask)
|
|
num_tokens = end_idx - start_idx
|
|
|
|
src_val = tl.load(input_ptr + offset, len_mask)
|
|
src_val = tl.where(src_val == replace_from, replace_to, src_val)
|
|
|
|
for i in tl.range(0, BLOCK_SIZE):
|
|
num_tokens1 = tl.get_element(num_tokens, (i, ))
|
|
start_idx1 = tl.get_element(start_idx, (i, ))
|
|
src_val1 = tl.get_element(src_val, (i, ))
|
|
offset1 = tl.arange(0, MAX_NUM_TOKENS)
|
|
tl.store(output_ptr + start_idx1 + offset1,
|
|
src_val1,
|
|
mask=offset1 < num_tokens1)
|
|
|
|
|
|
@triton.jit
|
|
def sample_recovered_tokens_kernel(
|
|
output_token_ids_ptr, # [num_tokens]
|
|
cu_num_draft_tokens_ptr, # [batch_size]
|
|
draft_token_ids_ptr, # [num_tokens]
|
|
draft_probs_ptr, # [num_tokens, vocab_size] or None
|
|
target_probs_ptr, # [num_tokens, vocab_size]
|
|
q_ptr, # [batch_size, vocab_size]
|
|
vocab_size,
|
|
PADDED_VOCAB_SIZE: tl.constexpr,
|
|
NO_DRAFT_PROBS: tl.constexpr,
|
|
SUB_BLOCK: tl.constexpr,
|
|
):
|
|
req_idx = tl.program_id(0)
|
|
start_idx = 0 if req_idx == 0 else tl.load(cu_num_draft_tokens_ptr +
|
|
req_idx - 1)
|
|
end_idx = tl.load(cu_num_draft_tokens_ptr + req_idx)
|
|
num_draft_tokens = end_idx - start_idx
|
|
|
|
# Early exit for out-of-range positions.
|
|
pos = tl.program_id(1)
|
|
if pos >= num_draft_tokens:
|
|
return
|
|
|
|
loop = (vocab_size + SUB_BLOCK - 1) // SUB_BLOCK
|
|
global_recovered_id = -1
|
|
global_max_p = -1.0
|
|
if NO_DRAFT_PROBS:
|
|
draft_token_id = tl.load(draft_token_ids_ptr + start_idx + pos)
|
|
orig_prob = tl.load(target_probs_ptr + (start_idx + pos) * vocab_size +
|
|
draft_token_id)
|
|
# Temporarily zero out the probability of the draft token.
|
|
# This is essentially the same as target_prob - draft_prob, except that
|
|
# n-gram does not have draft_prob. We regard it as 1.
|
|
tl.store(
|
|
target_probs_ptr + (start_idx + pos) * vocab_size + draft_token_id,
|
|
0)
|
|
for loop_i in range(loop):
|
|
vocab_start = loop_i * SUB_BLOCK
|
|
vocab_offset = vocab_start + tl.arange(0, SUB_BLOCK)
|
|
prob = tl.load(target_probs_ptr + (start_idx + pos) * vocab_size +
|
|
vocab_offset,
|
|
mask=vocab_offset < vocab_size,
|
|
other=0)
|
|
q = tl.load(q_ptr + req_idx * vocab_size + vocab_offset,
|
|
mask=vocab_offset < vocab_size,
|
|
other=float("-inf"))
|
|
new_p = prob / q
|
|
recovered_id = tl.argmax(new_p, axis=-1)
|
|
max_p = tl.get_element(new_p, (recovered_id, ))
|
|
if max_p > global_max_p:
|
|
global_max_p = max_p
|
|
global_recovered_id = vocab_start + recovered_id
|
|
else:
|
|
for loop_i in range(loop):
|
|
vocab_start = loop_i * SUB_BLOCK
|
|
vocab_offset = vocab_start + tl.arange(0, SUB_BLOCK)
|
|
draft_prob = tl.load(draft_probs_ptr +
|
|
(start_idx + pos) * vocab_size + vocab_offset,
|
|
mask=vocab_offset < vocab_size,
|
|
other=0)
|
|
target_prob = tl.load(target_probs_ptr +
|
|
(start_idx + pos) * vocab_size +
|
|
vocab_offset,
|
|
mask=vocab_offset < vocab_size,
|
|
other=0)
|
|
prob = tl.maximum(target_prob - draft_prob, 0)
|
|
# NOTE(woosuk): We don't need `prob = prob / tl.sum(prob)` here because
|
|
# `tl.argmax` will select the maximum value.
|
|
|
|
q = tl.load(q_ptr + req_idx * vocab_size + vocab_offset,
|
|
mask=vocab_offset < vocab_size,
|
|
other=float("-inf"))
|
|
new_p = prob / q
|
|
recovered_id = tl.argmax(new_p, axis=-1)
|
|
max_p = tl.get_element(new_p, (recovered_id, ))
|
|
if max_p > global_max_p:
|
|
global_max_p = max_p
|
|
global_recovered_id = vocab_start + recovered_id
|
|
|
|
tl.store(output_token_ids_ptr + start_idx + pos, global_recovered_id)
|
|
|
|
if NO_DRAFT_PROBS:
|
|
# Restore the original probability.
|
|
tl.store(
|
|
target_probs_ptr + (start_idx + pos) * vocab_size + draft_token_id,
|
|
orig_prob)
|
|
|
|
|
|
def rejection_greedy_sample_with_triton(output_token_ids, num_draft_tokens,
|
|
cu_num_draft_tokens, draft_token_ids,
|
|
target_argmax, bonus_token_ids,
|
|
is_greedy, max_spec_len, grid,
|
|
block_size):
|
|
vec_len = output_token_ids.shape[0]
|
|
|
|
if min(num_draft_tokens) == 1 and max(
|
|
num_draft_tokens) == 1 and is_greedy is None:
|
|
rejection_greedy_sample_spec_len_1_triton[(grid, )](
|
|
output_token_ids,
|
|
draft_token_ids,
|
|
target_argmax,
|
|
bonus_token_ids,
|
|
vec_len,
|
|
BLOCK_SIZE=block_size,
|
|
)
|
|
else:
|
|
rejection_greedy_sample_triton[(grid, )](
|
|
output_token_ids,
|
|
cu_num_draft_tokens,
|
|
draft_token_ids,
|
|
target_argmax,
|
|
bonus_token_ids,
|
|
is_greedy,
|
|
vec_len,
|
|
max_spec_len,
|
|
BLOCK_SIZE=block_size,
|
|
)
|
|
|
|
|
|
def expand_triton(batch_size, expanded_x, x, cu_num_tokens, replace_from,
|
|
replace_to, max_num_tokens):
|
|
vec_len = batch_size
|
|
grid, block_size = cal_grid_and_block_size(batch_size)
|
|
|
|
expand_kernel[(grid, )](
|
|
expanded_x,
|
|
x,
|
|
cu_num_tokens,
|
|
replace_from,
|
|
replace_to,
|
|
vec_len,
|
|
MAX_NUM_TOKENS=max_num_tokens, # To avoid recompilation.
|
|
BLOCK_SIZE=block_size,
|
|
)
|
|
|
|
|
|
@triton.jit(do_not_specialize=["max_spec_len"])
|
|
def rejection_random_sample_block_verify_kernel(
|
|
output_token_ids_ptr, # [batch_size, max_spec_len + 1]
|
|
cu_num_draft_tokens_ptr, # [batch_size]
|
|
draft_token_ids_ptr, # [num_tokens]
|
|
draft_probs_ptr, # [num_tokens, vocab_size] or None
|
|
target_probs_ptr, # [num_tokens, vocab_size]
|
|
bonus_token_ids_ptr, # [batch_size]
|
|
recovered_token_ids_ptr, # [num_tokens]
|
|
uniform_probs_ptr, # [num_tokens]
|
|
is_greedy_ptr, # [batch_size]
|
|
max_spec_len,
|
|
vocab_size,
|
|
vec_len,
|
|
NO_DRAFT_PROBS: tl.constexpr,
|
|
BLOCK_SIZE: tl.constexpr):
|
|
block_idx = tl.program_id(0)
|
|
offsets = block_idx * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
|
|
mask = offsets < vec_len
|
|
is_greedy = tl.load(is_greedy_ptr + offsets, mask, other=1)
|
|
not_greedy_mask = is_greedy == 0
|
|
start_idxs = tl.where(
|
|
offsets == 0, 0,
|
|
tl.load(cu_num_draft_tokens_ptr + offsets - 1, not_greedy_mask))
|
|
end_idxs = tl.load(cu_num_draft_tokens_ptr + offsets, not_greedy_mask)
|
|
n_num_draft_tokens = end_idxs - start_idxs
|
|
for req_i in range(BLOCK_SIZE):
|
|
not_greedy = tl.get_element(not_greedy_mask, (req_i, ))
|
|
if not_greedy:
|
|
|
|
rejected = False
|
|
pi = 1.0
|
|
uniform_prob = 1.0
|
|
last_accepted_token_pos = -1
|
|
start_idx = tl.get_element(start_idxs, (req_i, ))
|
|
req_idx = block_idx * BLOCK_SIZE + req_i
|
|
num_draft_tokens = tl.get_element(n_num_draft_tokens, (req_i, ))
|
|
|
|
for pos in range(num_draft_tokens):
|
|
draft_token_id = tl.load(draft_token_ids_ptr + start_idx + pos)
|
|
target_prob = tl.load(target_probs_ptr +
|
|
(start_idx + pos) * vocab_size +
|
|
draft_token_id)
|
|
tmp_uniform_prob = tl.load(uniform_probs_ptr + start_idx + pos)
|
|
uniform_prob = uniform_prob * tmp_uniform_prob
|
|
|
|
if NO_DRAFT_PROBS:
|
|
draft_prob = 1
|
|
else:
|
|
draft_prob = tl.load(draft_probs_ptr +
|
|
(start_idx + pos) * vocab_size +
|
|
draft_token_id)
|
|
|
|
pi = min(pi * target_prob / draft_prob, 1.0)
|
|
if draft_prob > 0 and pi >= uniform_prob:
|
|
last_accepted_token_pos = pos
|
|
rejected = False
|
|
else:
|
|
rejected = True
|
|
|
|
if last_accepted_token_pos > -1:
|
|
for pos in range(last_accepted_token_pos + 1):
|
|
token_id = tl.load(draft_token_ids_ptr + start_idx + pos)
|
|
tl.store(
|
|
output_token_ids_ptr + req_idx * (max_spec_len + 1) +
|
|
pos, token_id)
|
|
|
|
if rejected:
|
|
recovered_token_id = tl.load(recovered_token_ids_ptr +
|
|
start_idx +
|
|
last_accepted_token_pos + 1)
|
|
tl.store(
|
|
output_token_ids_ptr + req_idx * (max_spec_len + 1) +
|
|
last_accepted_token_pos + 1, recovered_token_id)
|
|
else:
|
|
bonus_token_id = tl.load(bonus_token_ids_ptr + req_idx)
|
|
tl.store(
|
|
output_token_ids_ptr + req_idx * (max_spec_len + 1) +
|
|
num_draft_tokens, bonus_token_id)
|