| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
|
|
| import atexit |
| import fcntl |
| import math |
| import os |
| import shutil |
| from contextlib import contextmanager, nullcontext |
| from enum import Enum |
| from threading import Lock |
| from typing import TYPE_CHECKING, List, Tuple |
|
|
| import torch |
| import torch_npu |
| from packaging.version import InvalidVersion, Version |
| from torch_npu.npu.streams import Event |
| from vllm.logger import logger |
|
|
| import vllm_ascend.envs as envs |
| from vllm_ascend.ascend_config import get_ascend_config |
|
|
| try: |
| |
| from torchair.scope import npu_stream_switch as _npu_stream_switch |
| from torchair.scope import npu_wait_tensor as _npu_wait_tensor |
| except ImportError: |
| from torchair.ops import NpuStreamSwitch as _npu_stream_switch |
| from torchair.ops import npu_wait_tensor as _npu_wait_tensor |
|
|
| if TYPE_CHECKING: |
| from vllm.config import VllmConfig |
| else: |
| VllmConfig = None |
|
|
| |
| |
| |
| |
| |
| MAX_CAPTURE_SIZE = 1920 |
|
|
| ASCEND_QUATIZATION_METHOD = "ascend" |
| SOC_VERSION_INFERENCE_SERIES = ["Ascend310P3"] |
|
|
| ACL_FORMAT_FRACTAL_ND = 2 |
| ACL_FORMAT_FRACTAL_NZ = 29 |
|
|
| _CUSTOM_OP_ENABLED = None |
| _IS_310P = None |
| _SLEEP_MODE_ENABLED = None |
| _CURRENT_STREAM = None |
|
|
|
|
| def is_310p(): |
| global _IS_310P |
| if _IS_310P is None: |
| from vllm_ascend import _build_info |
| _IS_310P = _build_info.__soc_version__.lower().startswith("ascend310p") |
| return _IS_310P |
|
|
|
|
| def sleep_mode_enabled(): |
| global _SLEEP_MODE_ENABLED |
| if _SLEEP_MODE_ENABLED is None: |
| from vllm_ascend import _build_info |
| _SLEEP_MODE_ENABLED = _build_info.__sleep_mode_enabled__ |
| return _SLEEP_MODE_ENABLED |
|
|
|
|
| def _round_up(x: int, align: int): |
| |
| |
| |
| |
| |
| |
| return (x + align - 1) // align * align |
|
|
|
|
| def _custom_pad(x, pad_dims): |
| |
| |
| |
| return torch.nn.functional.pad(x, pad_dims) |
|
|
|
|
| def _custom_reshape(x, target_shape): |
| |
| |
| |
| return x.reshape(target_shape) |
|
|
|
|
| def _custom_transpose(x, dim1, dim2): |
| |
| |
| |
| return x.transpose(dim1, dim2) |
|
|
|
|
| def nd_to_nz_2d(in_tensor: torch.Tensor) -> torch.Tensor: |
| |
| aux_dims = [1, 0, 0, 16] |
| |
| aux_dims[1] = _round_up(in_tensor.size(0), 16) |
| |
| aux_dims[2] = _round_up(in_tensor.size(1), 16) // 16 |
|
|
| |
|
|
| pad_dims = [0, 0, 0, 0] |
| |
| pad_dims[1] = _round_up(in_tensor.size(1), 16) - in_tensor.size(1) |
| |
| pad_dims[3] = _round_up(in_tensor.size(0), 16) - in_tensor.size(0) |
|
|
| |
|
|
| |
| return _custom_transpose( |
| _custom_reshape(_custom_pad(in_tensor, pad_dims), aux_dims), 1, |
| 2).contiguous() |
|
|
|
|
| def nd_to_nz_spec(mask_tensor: torch.Tensor) -> torch.Tensor: |
| num_tokens = mask_tensor.shape[0] |
| max_seq_len = mask_tensor.shape[1] |
|
|
| tokens_pad = (num_tokens + 15) // 16 * 16 |
| max_seq_len_pad = (max_seq_len + 15) // 16 * 16 |
|
|
| mask_tensor_pad = \ |
| torch.zeros((1, tokens_pad, max_seq_len_pad), dtype=mask_tensor.dtype, device=mask_tensor.device) |
| mask_tensor_pad[0][:num_tokens, :max_seq_len] = mask_tensor |
| mask = mask_tensor_pad.reshape( |
| (1, tokens_pad, max_seq_len_pad // 16, 16)).permute(0, 2, 1, 3) |
| return mask |
|
|
|
|
| def aligned_16(tensor: torch.Tensor): |
| """Aligned tensor for 310P""" |
|
|
| |
| n = tensor.size(0) |
|
|
| |
| n_aligned = ((n + 15) // 16) * 16 |
|
|
| |
| if n == n_aligned: |
| return tensor |
|
|
| |
| new_tensor = torch.zeros(n_aligned, |
| *tensor.shape[1:], |
| dtype=tensor.dtype, |
| device=tensor.device) |
|
|
| |
| new_tensor[:n] = tensor |
|
|
| return new_tensor |
|
|
|
|
| def maybe_converting_weight_acl_format(model, format=ACL_FORMAT_FRACTAL_NZ): |
| |
| |
| |
| |
| |
| |
| from vllm.model_executor.layers.fused_moe.layer import FusedMoE |
|
|
| use_torchair = get_ascend_config().torchair_graph_config.enabled |
| if not is_310p() or not use_torchair: |
| return |
| for module in model.modules(): |
| if isinstance(module, FusedMoE): |
| if torch_npu.get_npu_format(module.w13_weight.data) == format: |
| return |
| module.w13_weight.data = torch_npu.npu_format_cast( |
| module.w13_weight.data, format) |
| module.w2_weight.data = torch_npu.npu_format_cast( |
| module.w2_weight.data, format) |
|
|
|
|
| def try_register_lib(lib_name: str, lib_info: str = ""): |
| import importlib |
| import importlib.util |
| try: |
| module_spec = importlib.util.find_spec(lib_name) |
| if module_spec is not None: |
| importlib.import_module(lib_name) |
| if lib_info: |
| logger.info(lib_info) |
| except Exception: |
| pass |
|
|
|
|
| def enable_custom_op(): |
| """ |
| Enable lazy init for vllm_ascend_C to avoid early initialization of CANN's RTS component. |
| Ensure that ASCEND_RT_VISIBLE_DEVICES can be dynamically modified before torch.npu.set_device(). |
| """ |
| global _CUSTOM_OP_ENABLED |
| if _CUSTOM_OP_ENABLED is not None: |
| return _CUSTOM_OP_ENABLED |
| try: |
| |
| import vllm_ascend.vllm_ascend_C |
| _CUSTOM_OP_ENABLED = True |
| except ImportError: |
| _CUSTOM_OP_ENABLED = False |
| logger.warning( |
| "Warning: Failed to register custom ops, all custom ops will be disabled" |
| ) |
| return _CUSTOM_OP_ENABLED |
|
|
|
|
| def find_hccl_library() -> str: |
| """ |
| We either use the library file specified by the `HCCL_SO_PATH` |
| environment variable, or we find the library file brought by PyTorch. |
| After importing `torch`, `libhccl.so` can be |
| found by `ctypes` automatically. |
| """ |
| so_file = envs.HCCL_SO_PATH |
|
|
| |
| if so_file: |
| logger.info("Found hccl from environment variable HCCL_SO_PATH=%s", |
| so_file) |
| else: |
| if torch.version.cann is not None: |
| so_file = "libhccl.so" |
| else: |
| raise ValueError("HCCL only supports Ascend NPU backends.") |
| logger.info("Found hccl from library %s", so_file) |
| return so_file |
|
|
|
|
| def current_stream() -> torch.npu.Stream: |
| """ |
| replace `torch.npu.current_stream()` with `vllm.utils.current_stream()`. |
| it turns out that `torch.npu.current_stream()` is quite expensive, |
| as it will construct a new stream object at each call. |
| here we patch `torch.npu.set_stream` to keep track of the current stream |
| directly, so that we can avoid calling `torch.npu.current_stream()`. |
| |
| """ |
| global _CURRENT_STREAM |
| if _CURRENT_STREAM is None: |
| |
| |
| _CURRENT_STREAM = torch.npu.current_stream() |
| return _CURRENT_STREAM |
|
|
|
|
| def adapt_patch(is_global_patch: bool = False): |
| if is_global_patch: |
| from vllm_ascend.patch import platform |
| else: |
| from vllm_ascend.patch import worker |
|
|
|
|
| def vllm_version_is(target_vllm_version: str): |
| if envs.VLLM_VERSION is not None: |
| vllm_version = envs.VLLM_VERSION |
| else: |
| import vllm |
| vllm_version = vllm.__version__ |
| try: |
| return Version(vllm_version) == Version(target_vllm_version) |
| except InvalidVersion: |
| raise ValueError( |
| f"Invalid vllm version {vllm_version} found. A dev version of vllm " |
| "is installed probably. Set the environment variable VLLM_VERSION " |
| "to control it by hand. And please make sure the value follows the " |
| "format of x.y.z.") |
|
|
|
|
| def update_aclgraph_sizes(vllm_config: VllmConfig) -> None: |
| """Update ACL graph capture sizes based on hardware limitations""" |
| |
| compilation_config = vllm_config.compilation_config |
| original_sizes, compilation_config.cudagraph_capture_sizes = \ |
| compilation_config.cudagraph_capture_sizes, None |
|
|
| |
| num_hidden_layers = vllm_config.model_config.hf_config.num_hidden_layers |
| parallel_config = vllm_config.parallel_config |
|
|
| |
| parallel_factor = 1 + sum(size > 1 for size in [ |
| parallel_config.data_parallel_size_local, |
| parallel_config.tensor_parallel_size, |
| parallel_config.expert_parallel_size, |
| parallel_config.expert_tensor_parallel_size, |
| ]) |
|
|
| |
| max_num_batch_sizes = math.floor(MAX_CAPTURE_SIZE / |
| (num_hidden_layers + 1) / parallel_factor) |
| logger.info("Calculated maximum supported batch sizes for ACL graph: %s", |
| max_num_batch_sizes) |
|
|
| |
| if max_num_batch_sizes < len(original_sizes): |
| |
| step = (len(original_sizes) - 1) / (max_num_batch_sizes - 1) |
| indices = [round(i * step) for i in range(max_num_batch_sizes)] |
|
|
| |
| indices[0], indices[-1] = 0, len(original_sizes) - 1 |
|
|
| sampled_sizes = [original_sizes[i] for i in indices] |
| compilation_config.init_with_cudagraph_sizes(sampled_sizes) |
|
|
| logger.info( |
| "Adjusted ACL graph batch sizes for %s model (layers: %d): %d → %d sizes", |
| vllm_config.model_config.architectures[0], |
| num_hidden_layers, |
| len(original_sizes), |
| len(compilation_config. |
| cudagraph_capture_sizes |
| )) |
| else: |
| |
| compilation_config.cudagraph_capture_sizes = original_sizes |
| logger.info( |
| "No adjustment needed for ACL graph batch sizes: %s model (layers: %d) with %d sizes", |
| vllm_config.model_config.architectures[0], num_hidden_layers, |
| len(original_sizes)) |
|
|
|
|
| |
| def dispose_tensor(x: torch.Tensor): |
| x.set_(torch.empty((0, ), device=x.device, dtype=x.dtype)) |
|
|
|
|
| class ProfileExecuteDuration: |
| _instance = None |
| _observations: List[Tuple[str, Event, Event]] = [] |
| _lock = Lock() |
|
|
| def __new__(cls): |
| with cls._lock: |
| if cls._instance is None: |
| cls._instance = super().__new__(cls) |
| atexit.register(cls._instance.destroy) |
| return cls._instance |
|
|
| def destroy(self): |
| with self._lock: |
| self._observations.clear() |
|
|
| @contextmanager |
| def capture_async(self, duration_tag: str): |
| if not envs.VLLM_ASCEND_MODEL_EXECUTE_TIME_OBSERVE: |
| yield |
| return |
|
|
| observe_start = Event(enable_timing=True) |
| observe_start.record() |
| try: |
| yield |
| finally: |
| observe_end = Event(enable_timing=True) |
| observe_end.record() |
| with self._lock: |
| self._observations.append( |
| (duration_tag, observe_start, observe_end)) |
|
|
| def pop_captured_sync(self) -> dict: |
| """Pop and synchronize all events in the observation list""" |
| durations: dict[str, float] = {} |
| if not envs.VLLM_ASCEND_MODEL_EXECUTE_TIME_OBSERVE: |
| return durations |
|
|
| while self._observations: |
| with self._lock: |
| tag, observe_start, observe_end = self._observations.pop() |
| observe_end.synchronize() |
| durations[tag] = observe_start.elapsed_time(observe_end) |
|
|
| return durations |
|
|
|
|
| |
| def npu_stream_switch(tag: str, priority: int, *, enabled: bool = True): |
| return _npu_stream_switch(tag, priority) if enabled else nullcontext() |
|
|
|
|
| |
| def npu_wait_tensor(self: torch.Tensor, |
| dependency: torch.Tensor, |
| *, |
| enabled: bool = True): |
| return _npu_wait_tensor(self, dependency) if enabled else self |
|
|
|
|
| |
| def npu_prefetch(input: torch.Tensor, |
| dependency: torch.Tensor, |
| max_size: int = 0, |
| *, |
| enabled: bool = True): |
| if not enabled: |
| return |
| input_size = input.element_size() * input.numel() |
| if max_size <= 0 or max_size > input_size: |
| max_size = input_size |
| torch_npu.npu_prefetch(input, dependency, max_size) |
|
|
|
|
| |
| class FusedMoEState(Enum): |
| AllGather = 0 |
| All2All = 1 |
| MC2 = 2 |
| AllGatherEP = 3 |
| NaiveMulticast = 4 |
|
|
|
|
| |
| |
| |
| def get_rm_router_logits_state(ep_size: int, dp_size: int, |
| is_deepseek_v3_r1: bool): |
| |
| |
| if dp_size > 1: |
| if (envs.VLLM_ENABLE_FUSED_EXPERTS_ALLGATHER_EP and ep_size > 1 |
| and is_deepseek_v3_r1): |
| return True |
| elif ep_size == 1 and is_deepseek_v3_r1: |
| return True |
| return False |
|
|
|
|
| |
| |
| |
| def get_all_reduce_merge_state(ep_size: int, is_deepseek_v3_r1: bool): |
| |
| |
| if (envs.VLLM_ENABLE_FUSED_EXPERTS_ALLGATHER_EP and ep_size > 1 |
| and is_deepseek_v3_r1): |
| return True |
| elif ep_size == 1 and is_deepseek_v3_r1: |
| return True |
| return False |
|
|
|
|
| |
| def get_fused_moe_state(ep_size: int, with_prefill: bool, |
| is_deepseek_v3_r1: bool): |
| |
| |
| if (envs.VLLM_ENABLE_FUSED_EXPERTS_ALLGATHER_EP and ep_size > 1 |
| and is_deepseek_v3_r1 and not with_prefill): |
| return FusedMoEState.AllGatherEP |
| elif ep_size == 1: |
| if with_prefill: |
| return FusedMoEState.NaiveMulticast |
| else: |
| return FusedMoEState.AllGather |
| |
| elif ep_size < 16 or with_prefill: |
| return FusedMoEState.All2All |
| else: |
| return FusedMoEState.MC2 |
|
|
|
|
| KV_CACHE_BYTES_CACHE_PATH_NAME = ".kv_cache_bytes" |
| KV_CACHE_BYTES_CACHE_FILE_NAME = "kv_cache_bytes" |
| TORCHAIR_CACHE_PATH_NAME = ".torchair_cache" |
| TORCHAIR_CACHE_DIR = os.getenv( |
| 'TORCHAIR_CACHE_HOME', os.path.join(os.getcwd(), TORCHAIR_CACHE_PATH_NAME)) |
|
|
|
|
| def get_torchair_current_work_dir(file_name=None): |
| if file_name is None: |
| return TORCHAIR_CACHE_DIR |
| return os.path.join(TORCHAIR_CACHE_DIR, file_name) |
|
|
|
|
| def check_torchair_cache_exist(): |
| res = False |
| torch_air_abs_path = get_torchair_current_work_dir() |
| if os.path.exists(torch_air_abs_path): |
| file_list = os.listdir(torch_air_abs_path) |
| if len(file_list) != 0: |
| res = True |
| return res |
|
|
|
|
| def check_kv_cache_bytes_cache_exist(): |
| res = False |
| kv_cache_bytes_cache_abs_path = get_torchair_current_work_dir( |
| KV_CACHE_BYTES_CACHE_PATH_NAME) |
| if os.path.exists(kv_cache_bytes_cache_abs_path): |
| file_list = os.listdir(kv_cache_bytes_cache_abs_path) |
| if len(file_list) != 0: |
| res = True |
| return res |
|
|
|
|
| def read_kv_cache_bytes_from_file(rank) -> int: |
| kv_cache_bytes = -1 |
| kv_cache_bytes_cache_abs_path = get_torchair_current_work_dir( |
| KV_CACHE_BYTES_CACHE_PATH_NAME) |
| kv_cache_bytes_file = os.path.join( |
| kv_cache_bytes_cache_abs_path, |
| f"{rank}_{KV_CACHE_BYTES_CACHE_FILE_NAME}") |
| with open(kv_cache_bytes_file, "r", encoding="utf-8") as f: |
| with file_lock(f, fcntl.LOCK_SH): |
| kv_cache_bytes = int(f.readline()) |
| return kv_cache_bytes |
|
|
|
|
| @contextmanager |
| def file_lock(file_descriptor, lock_type): |
| fcntl.flock(file_descriptor, lock_type) |
| try: |
| yield |
| finally: |
| fcntl.flock(file_descriptor, fcntl.LOCK_UN) |
|
|
|
|
| def write_kv_cache_bytes_to_file(rank, kv_cache_bytes): |
| kv_cache_bytes_cache_abs_path = get_torchair_current_work_dir( |
| KV_CACHE_BYTES_CACHE_PATH_NAME) |
| os.makedirs(kv_cache_bytes_cache_abs_path, exist_ok=True) |
| kv_cache_bytes_file = os.path.join( |
| kv_cache_bytes_cache_abs_path, |
| f"{rank}_{KV_CACHE_BYTES_CACHE_FILE_NAME}") |
| with open(kv_cache_bytes_file, "w", encoding="utf-8") as f: |
| with file_lock(f, fcntl.LOCK_EX): |
| f.write(f"{kv_cache_bytes}") |
|
|
|
|
| def delete_torchair_cache_file(): |
| torch_air_abs_path = get_torchair_current_work_dir() |
| if os.path.exists(torch_air_abs_path): |
| shutil.rmtree(torch_air_abs_path) |
|
|