EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

[feature] Ascend NPU graph support (#9399)

Browse Source 
Co-authored-by: ronnie_zheng <zl19940307@163.com>
Co-authored-by: yezhifeng (D) <y00897525@china.huawei.com>
Co-authored-by: anon189Ty <Stari_Falcon@outlook.com>
Co-authored-by: Maksim <makcum888e@mail.ru>
Co-authored-by: ssshinigami <44640852+ssshinigami@users.noreply.github.com>
This commit is contained in:
VDV1985
2025-08-21 07:13:27 +03:00
committed by GitHub
parent 7cd2ee06d7
commit 2c4b4b786b
9 changed files with 470 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
python/sglang/srt/distributed/parallel_state.py
View File

@@ -55,7 +55,7 @@ _is_npu = is_npu()
@dataclass
class GraphCaptureContext:
stream: torch.cuda.Stream
stream: torch.cuda.Stream if not _is_npu else torch.npu.Stream
TensorMetadata = namedtuple("TensorMetadata", ["device", "dtype", "size"])
@@ -252,8 +252,11 @@ class GroupCoordinator:
if is_cuda_alike():
self.device = torch.device(f"cuda:{local_rank}")
elif _is_npu:
self.device = torch.device(f"npu:{local_rank}")
else:
self.device = torch.device("cpu")
self.device_module = torch.get_device_module(self.device)
self.use_pynccl = use_pynccl
self.use_pymscclpp = use_pymscclpp
@@ -402,7 +405,7 @@ class GroupCoordinator:
self, graph_capture_context: Optional[GraphCaptureContext] = None
):
if graph_capture_context is None:
stream = torch.cuda.Stream()
stream = self.device_module.Stream()
graph_capture_context = GraphCaptureContext(stream)
else:
stream = graph_capture_context.stream
@@ -413,11 +416,11 @@ class GroupCoordinator:
# ensure all initialization operations complete before attempting to
# capture the graph on another stream
curr_stream = torch.cuda.current_stream()
curr_stream = self.device_module.current_stream()
if curr_stream != stream:
stream.wait_stream(curr_stream)
with torch.cuda.stream(stream), maybe_ca_context:
with self.device_module.stream(stream), maybe_ca_context:
# In graph mode, we have to be very careful about the collective
# operations. The current status is:
# allreduce \ Mode | Eager | Graph |
@@ -1641,6 +1644,8 @@ def cleanup_dist_env_and_memory(shutdown_ray: bool = False):
)
elif hasattr(torch, "xpu") and torch.xpu.is_available():
torch.xpu.empty_cache()
elif hasattr(torch, "npu") and torch.npu.is_available():
torch.npu.empty_cache()
def in_the_same_node_as(pg: ProcessGroup, source_rank: int = 0) -> List[bool]:

154
python/sglang/srt/layers/attention/ascend_backend.py
View File

@@ -1,7 +1,7 @@
from __future__ import annotations
from dataclasses import dataclass
from typing import TYPE_CHECKING, Optional
from typing import TYPE_CHECKING, List, Optional
import torch
import torch_npu
@@ -27,6 +27,7 @@ class ForwardMetadata:
# seq len inputs
extend_seq_lens_cpu_int: Optional[torch.Tensor] = None
seq_lens_cpu_int: Optional[torch.Tensor] = None
seq_lens_cpu_list: Optional[List[int]] = None
class AscendAttnBackend(AttentionBackend):
@@ -51,7 +52,7 @@ class AscendAttnBackend(AttentionBackend):
def __init__(self, model_runner: ModelRunner):
super().__init__()
self.forward_metadata = ForwardMetadata()
self.forward_metadata = None
self.device = model_runner.device
self.gen_attention_mask(128, model_runner.dtype)
self.page_size = model_runner.page_size
@@ -60,9 +61,15 @@ class AscendAttnBackend(AttentionBackend):
self.kv_lora_rank = model_runner.model_config.kv_lora_rank
self.qk_rope_head_dim = model_runner.model_config.qk_rope_head_dim
self.native_attn = TorchNativeAttnBackend(model_runner)
self.graph_metadata = {}
self.max_context_len = model_runner.model_config.context_len
self.req_to_token = model_runner.req_to_token_pool.req_to_token
self.graph_mode = False
def init_forward_metadata(self, forward_batch: ForwardBatch):
"""Init the metadata for a forward pass."""
self.forward_metadata = ForwardMetadata()
self.forward_metadata.block_tables = (
forward_batch.req_to_token_pool.req_to_token[
forward_batch.req_pool_indices, : forward_batch.seq_lens.max()
@@ -75,6 +82,63 @@ class AscendAttnBackend(AttentionBackend):
)
self.forward_metadata.seq_lens_cpu_int = forward_batch.seq_lens_cpu.int()
self.graph_mode = False
def init_cuda_graph_state(self, max_bs: int, max_num_tokens: int):
self.graph_metadata = {
"block_tables": torch.empty(
(max_bs, self.max_context_len // self.page_size),
dtype=torch.int32,
device=self.device,
),
}
def init_forward_metadata_capture_cuda_graph(
self,
bs: int,
num_tokens: int,
req_pool_indices: torch.Tensor,
seq_lens: torch.Tensor,
encoder_lens: Optional[torch.Tensor],
forward_mode: ForwardMode,
spec_info: Optional[Union[EagleDraftInput, EagleVerifyInput]],
):
metadata = ForwardMetadata()
metadata.block_tables = self.graph_metadata["block_tables"][:bs, :]
metadata.seq_lens_cpu_list = seq_lens.cpu().int().tolist()
self.graph_metadata[bs] = metadata
self.forward_metadata = metadata
self.graph_mode = True
def init_forward_metadata_replay_cuda_graph(
self,
bs: int,
req_pool_indices: torch.Tensor,
seq_lens: torch.Tensor,
seq_lens_sum: int,
encoder_lens: Optional[torch.Tensor],
forward_mode: ForwardMode,
spec_info: Optional[Union[EagleDraftInput, EagleVerifyInput]],
seq_lens_cpu: Optional[torch.Tensor],
):
metadata = self.graph_metadata[bs]
max_len = seq_lens_cpu[:bs].max().item()
max_seq_pages = (max_len + self.page_size - 1) // self.page_size
metadata.block_tables[:bs, :max_seq_pages].copy_(
self.req_to_token[req_pool_indices[:bs], :max_len][:, :: self.page_size]
// self.page_size
)
metadata.block_tables[:bs, max_seq_pages:].fill_(0)
metadata.block_tables[bs:, :].fill_(0)
self.forward_metadata = metadata
self.graph_mode = True
def get_cuda_graph_seq_len_fill_value(self):
return 1
@@ -167,28 +231,74 @@ class AscendAttnBackend(AttentionBackend):
layer, forward_batch.out_cache_loc, k, v
)
if not self.use_mla:
k_cache = forward_batch.token_to_kv_pool.get_key_buffer(layer.layer_id)
v_cache = forward_batch.token_to_kv_pool.get_value_buffer(layer.layer_id)
if self.graph_mode:
k_cache = forward_batch.token_to_kv_pool.get_key_buffer(
layer.layer_id
).view(-1, self.page_size, layer.tp_k_head_num * layer.qk_head_dim)
v_cache = forward_batch.token_to_kv_pool.get_value_buffer(
layer.layer_id
).view(-1, self.page_size, layer.tp_v_head_num * layer.v_head_dim)
query = q.view(-1, 1, layer.tp_q_head_num * layer.qk_head_dim)
num_tokens = query.shape[0]
workspace = (
torch_npu._npu_fused_infer_attention_score_get_max_workspace(
query,
k_cache,
v_cache,
block_table=self.forward_metadata.block_tables,
block_size=self.page_size,
num_heads=layer.tp_q_head_num,
num_key_value_heads=layer.tp_k_head_num,
input_layout="BSH",
scale=layer.scaling,
actual_seq_lengths_kv=self.forward_metadata.seq_lens_cpu_list,
)
)
output = torch.empty(
(num_tokens, 1, layer.tp_q_head_num * layer.v_head_dim),
dtype=q.dtype,
device=q.device,
)
softmax_lse = torch.empty(1, dtype=q.dtype, device=q.device)
torch_npu.npu_fused_infer_attention_score.out(
query,
k_cache,
v_cache,
block_table=self.forward_metadata.block_tables,
block_size=self.page_size,
num_heads=layer.tp_q_head_num,
num_key_value_heads=layer.tp_k_head_num,
input_layout="BSH",
scale=layer.scaling,
actual_seq_lengths_kv=self.forward_metadata.seq_lens_cpu_list,
workspace=workspace,
out=[output, softmax_lse],
)
else:
k_cache = forward_batch.token_to_kv_pool.get_key_buffer(layer.layer_id)
v_cache = forward_batch.token_to_kv_pool.get_value_buffer(
layer.layer_id
)
query = q.view(-1, layer.tp_q_head_num, layer.qk_head_dim)
num_tokens = query.shape[0]
output = torch.empty(
(num_tokens, layer.tp_q_head_num, layer.v_head_dim),
dtype=query.dtype,
device=query.device,
)
query = q.view(-1, layer.tp_q_head_num, layer.qk_head_dim)
num_tokens = query.shape[0]
output = torch.empty(
(num_tokens, layer.tp_q_head_num, layer.v_head_dim),
dtype=query.dtype,
device=query.device,
)
torch_npu._npu_paged_attention(
query=query,
key_cache=k_cache,
value_cache=v_cache,
num_heads=layer.tp_q_head_num,
num_kv_heads=layer.tp_k_head_num,
scale_value=layer.scaling,
block_table=self.forward_metadata.block_tables,
context_lens=self.forward_metadata.seq_lens_cpu_int,
out=output,
)
torch_npu._npu_paged_attention(
query=query,
key_cache=k_cache,
value_cache=v_cache,
num_heads=layer.tp_q_head_num,
num_kv_heads=layer.tp_k_head_num,
scale_value=layer.scaling,
block_table=self.forward_metadata.block_tables,
context_lens=self.forward_metadata.seq_lens_cpu_int,
out=output,
)
return output.view(num_tokens, layer.tp_q_head_num * layer.v_head_dim)
else:
query = q.view(-1, layer.tp_q_head_num, layer.head_dim)

37
python/sglang/srt/model_executor/cuda_graph_runner.py
View File

@@ -240,6 +240,8 @@ class CudaGraphRunner:
def __init__(self, model_runner: ModelRunner):
# Parse args
self.model_runner = model_runner
self.device = model_runner.device
self.device_module = torch.get_device_module(self.device)
self.graphs = {}
self.output_buffers = {}
self.enable_torch_compile = model_runner.server_args.enable_torch_compile
@@ -305,13 +307,15 @@ class CudaGraphRunner:
self.model_runner.lora_manager.init_cuda_graph_batch_info(self.max_bs)
# Graph inputs
with torch.device("cuda"):
with torch.device(self.device):
self.input_ids = torch.zeros((self.max_num_token,), dtype=torch.int64)
self.req_pool_indices = torch.zeros((self.max_bs,), dtype=torch.int32)
self.seq_lens = torch.full(
(self.max_bs,), self.seq_len_fill_value, dtype=torch.int32
)
self.out_cache_loc = torch.zeros((self.max_num_token,), dtype=torch.int64)
self.out_cache_loc = torch.zeros(
(self.max_num_token,), dtype=self._cache_loc_dtype()
)
self.positions = torch.zeros((self.max_num_token,), dtype=torch.int64)
self.mrope_positions = torch.zeros((3, self.max_bs), dtype=torch.int64)
self.num_token_non_padded = torch.zeros((1,), dtype=torch.int32)
@@ -366,12 +370,12 @@ class CudaGraphRunner:
* self.num_tokens_per_bs
),
dtype=torch.bool,
device="cuda",
device=self.device,
)
self.next_token_logits_buffer = torch.zeros(
(self.max_num_token, self.model_runner.model_config.vocab_size),
dtype=torch.float,
device="cuda",
device=self.device,
)
# Capture
@@ -383,6 +387,9 @@ class CudaGraphRunner:
f"Capture cuda graph failed: {e}\n{CUDA_GRAPH_CAPTURE_FAILED_MSG}"
)
def _cache_loc_dtype(self):
return torch.int64
def can_run(self, forward_batch: ForwardBatch):
if self.require_mlp_tp_gather:
cuda_graph_bs = (
@@ -502,8 +509,16 @@ class CudaGraphRunner:
)
logger.info(log_message)
def _capture_graph(self, graph, pool, stream, run_once_fn):
with self.device_module.graph(graph, pool=pool, stream=stream):
out = run_once_fn()
return out
def _create_device_graph(self):
return torch.cuda.CUDAGraph()
def capture_one_batch_size(self, bs: int, forward: Callable):
graph = torch.cuda.CUDAGraph()
graph = self._create_device_graph()
stream = self.stream
num_tokens = bs * self.num_tokens_per_bs
@@ -643,19 +658,17 @@ class CudaGraphRunner:
return logits_output_or_pp_proxy_tensors
for _ in range(2):
torch.cuda.synchronize()
self.device_module.synchronize()
self.model_runner.tp_group.barrier()
run_once()
if get_global_graph_memory_pool() is None:
set_global_graph_memory_pool(torch.cuda.graph_pool_handle())
set_global_graph_memory_pool(self.device_module.graph_pool_handle())
# Set graph pool id globally to be able to use symmetric memory
set_graph_pool_id(get_global_graph_memory_pool())
with torch.cuda.graph(
graph, pool=get_global_graph_memory_pool(), stream=stream
):
out = run_once()
out = self._capture_graph(
graph, get_global_graph_memory_pool(), stream, run_once
)
return graph, out

26
python/sglang/srt/model_executor/model_runner.py
View File

@@ -91,6 +91,7 @@ from sglang.srt.mem_cache.memory_pool import (
)
from sglang.srt.model_executor.cuda_graph_runner import CudaGraphRunner
from sglang.srt.model_executor.forward_batch_info import ForwardBatch, PPProxyTensors
from sglang.srt.model_executor.npu_graph_runner import NPUGraphRunner
from sglang.srt.model_loader import get_model
from sglang.srt.model_loader.loader import DefaultModelLoader, get_model_loader
from sglang.srt.model_loader.utils import set_default_torch_dtype
@@ -341,9 +342,12 @@ class ModelRunner:
if self.device == "cuda":
self.init_cublas()
self.init_attention_backend()
self.init_cuda_graphs()
self.init_device_graphs()
elif self.device == "npu":
self.init_attention_backend()
self.init_device_graphs()
else:
self.cuda_graph_runner = None
self.graph_runner = None
self.cuda_graph_mem_usage = 0
self.init_attention_backend()
@@ -917,7 +921,8 @@ class ModelRunner:
)
# We need to get device after patch otherwise the device would be wrong
infered_device = torch.cuda.current_device()
self.device_module = torch.get_device_module(self.device)
infered_device = self.device_module.current_device()
named_tensors = [
(name, _unwrap_tensor(tensor, tp_rank=self.tp_rank, device=infered_device))
@@ -1592,9 +1597,9 @@ class ModelRunner:
.cuda()
)
def init_cuda_graphs(self):
def init_device_graphs(self):
"""Capture cuda graphs."""
self.cuda_graph_runner = None
self.graph_runner = None
self.cuda_graph_mem_usage = 0
if not self.is_generation:
@@ -1609,8 +1614,9 @@ class ModelRunner:
logger.info(
f"Capture cuda graph begin. This can take up to several minutes. avail mem={before_mem:.2f} GB"
)
self.cuda_graph_runner = CudaGraphRunner(self)
self.graph_runner = (
CudaGraphRunner(self) if not _is_npu else NPUGraphRunner(self)
)
after_mem = get_available_gpu_memory(self.device, self.gpu_id)
self.cuda_graph_mem_usage = before_mem - after_mem
logger.info(
@@ -1762,11 +1768,11 @@ class ModelRunner:
) -> Tuple[Union[LogitsProcessorOutput, PPProxyTensors], bool]:
can_run_cuda_graph = bool(
forward_batch.forward_mode.is_cuda_graph()
and self.cuda_graph_runner
and self.cuda_graph_runner.can_run(forward_batch)
and self.graph_runner
and self.graph_runner.can_run(forward_batch)
)
if can_run_cuda_graph:
ret = self.cuda_graph_runner.replay(
ret = self.graph_runner.replay(
forward_batch,
skip_attn_backend_init=skip_attn_backend_init,
pp_proxy_tensors=pp_proxy_tensors,

94
python/sglang/srt/model_executor/npu_graph_runner.py Normal file
View File

@@ -0,0 +1,94 @@
# Copyright 2023-2024 SGLang Team
# 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.
# ==============================================================================
"""Run the model with npu graph and torch.compile."""
from __future__ import annotations
import logging
import threading
from typing import TYPE_CHECKING, Optional, Union
import torch
from sglang.srt.model_executor.cuda_graph_runner import CudaGraphRunner
logger = logging.getLogger(__name__)
if TYPE_CHECKING:
from sglang.srt.model_executor.model_runner import ModelRunner
from sglang.srt.layers.logits_processor import LogitsProcessorOutput
from sglang.srt.model_executor.forward_batch_info import ForwardBatch, PPProxyTensors
class NPUGraphRunner(CudaGraphRunner):
"""A NPUGraphRunner runs the forward pass of a model with npu graph and torch.compile."""
def __init__(self, model_runner: ModelRunner):
super().__init__(model_runner)
def _create_device_graph(self):
return torch.npu.NPUGraph()
def _capture_graph(self, graph, pool, stream, run_once_fn):
with torch.npu.graph(
graph,
pool=pool,
stream=stream,
auto_dispatch_capture=True,
):
out = run_once_fn()
return out
def _update_inputs(self, seq_lens):
self.graphs[self.bs].update(
cpu_update_input=[{"actual_seq_lengths_kv": seq_lens}]
)
def _cache_loc_dtype(self):
return torch.int32
def replay(
self,
forward_batch: ForwardBatch,
skip_attn_backend_init: bool = False,
pp_proxy_tensors: Optional[PPProxyTensors] = None,
) -> Union[LogitsProcessorOutput, PPProxyTensors]:
if not skip_attn_backend_init:
self.replay_prepare(forward_batch, pp_proxy_tensors)
else:
# In speculative decoding, these two fields are still needed.
self.input_ids[: self.raw_num_token].copy_(forward_batch.input_ids)
self.positions[: self.raw_num_token].copy_(forward_batch.positions)
# Replay
seq_lens = forward_batch.seq_lens.cpu().tolist() + [0] * (self.bs - self.raw_bs)
thread = threading.Thread(target=self._update_inputs, args=(seq_lens,))
thread.start()
self.graphs[self.bs].replay()
thread.join()
output = self.output_buffers[self.bs]
if isinstance(output, LogitsProcessorOutput):
return LogitsProcessorOutput(
next_token_logits=output.next_token_logits[: self.raw_num_token],
hidden_states=(
output.hidden_states[: self.raw_num_token]
if output.hidden_states is not None
else None
),
)
else:
assert isinstance(output, PPProxyTensors)
return PPProxyTensors({k: v[: self.bs] for k, v in output.tensors.items()})