# # Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved. # # 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. # This file is a part of the vllm-ascend project. # from dataclasses import dataclass from enum import Enum from typing import ClassVar, List, Optional, Tuple, Type import numpy as np import torch import torch.distributed as dist import torch.nn as nn import torch.nn.functional as F import torch_npu from vllm.attention.backends.abstract import (AttentionBackend, AttentionImpl, AttentionLayer, AttentionType) from vllm.config import VllmConfig from vllm.distributed import (get_dcp_group, get_decode_context_model_parallel_rank, get_decode_context_model_parallel_world_size) from vllm.forward_context import ForwardContext, get_forward_context from vllm.utils import cdiv from vllm.v1.attention.backends.utils import AttentionCGSupport from vllm.v1.core.sched.output import SchedulerOutput from vllm.v1.kv_cache_interface import AttentionSpec from vllm_ascend.attention.utils import (AscendCommonAttentionMetadata, split_decodes_and_prefills) from vllm_ascend.compilation.acl_graph import (get_graph_params, update_graph_params_workspaces) from vllm_ascend.ops.attention import vanilla_chunked_prefill from vllm_ascend.utils import (ACL_FORMAT_FRACTAL_NZ, aligned_16, is_310p, nd_to_nz_2d, nd_to_nz_spec, prefill_context_parallel_enable, version_check, weak_ref_tensors) # isort: off if prefill_context_parallel_enable(): from vllm.distributed import (get_pcp_group, get_prefill_context_model_parallel_rank, get_prefill_context_model_parallel_world_size ) # isort: on class AscendAttentionBackend(AttentionBackend): accept_output_buffer: bool = True @staticmethod def get_name() -> str: return "ASCEND" @staticmethod def get_impl_cls() -> Type["AscendAttentionBackendImpl"]: return AscendAttentionBackendImpl @staticmethod def get_metadata_cls() -> Type["AscendMetadata"]: return AscendMetadata @staticmethod def get_builder_cls() -> type["AscendAttentionMetadataBuilder"]: return AscendAttentionMetadataBuilder @staticmethod def get_kv_cache_shape( num_blocks: int, block_size: int, num_kv_heads: int, head_size: int, ) -> Tuple[int, ...]: if is_310p(): return (2, num_blocks, num_kv_heads * head_size // 16, block_size, 16) return (2, num_blocks, block_size, num_kv_heads, head_size) @staticmethod def get_bsh_kv_cache_shape( num_blocks: int, block_size: int, num_kv_heads: int, head_size: int, ) -> Tuple[int, ...]: return (2, num_blocks, block_size, num_kv_heads * head_size) @staticmethod def swap_blocks( src_kv_cache: List[torch.Tensor], dst_kv_cache: List[torch.Tensor], src_to_dst: torch.Tensor, ) -> None: src_key_cache, src_value_cache = src_kv_cache[0], src_kv_cache[1] dst_key_cache, dst_value_cache = dst_kv_cache[0], dst_kv_cache[1] src_indices = src_to_dst[:, 0] dst_indices = src_to_dst[:, 1] dst_key_cache[dst_indices] = src_key_cache[src_indices].to( dst_key_cache.device) dst_value_cache[dst_indices] = src_value_cache[src_indices].to( dst_key_cache.device) @staticmethod def copy_blocks( kv_caches: List[torch.Tensor], src_to_dists: torch.Tensor, ) -> None: src_indices = src_to_dists[:, 0] dst_indices = src_to_dists[:, 1] for kv_cache in kv_caches: key_caches = kv_cache[0] value_caches = kv_cache[1] key_caches[dst_indices] = key_caches[src_indices] value_caches[dst_indices] = value_caches[src_indices] @staticmethod def get_supported_block_size() -> list[int]: return [64] class AscendAttentionState(Enum): PrefillNoCache = 0 PrefillCacheHit = 1 DecodeOnly = 2 ChunkedPrefill = 3 SpecDecoding = 4 @dataclass class AscendPCPMetadata: q_head_idx: torch.Tensor = None q_tail_idx: torch.Tensor = None kv_with_q_head_nomask_idx: torch.Tensor = None kv_with_q_head_mask_idx: torch.Tensor = None kv_with_q_tail_nomask_idx: torch.Tensor = None kv_with_q_tail_mask_idx: torch.Tensor = None attn_mask_seqlens: torch.Tensor = None head_attn_nomask_seqlens: torch.Tensor = None tail_attn_nomask_seqlens: torch.Tensor = None q_full_idx: torch.Tensor = None pcp_prefill_mask: torch.Tensor = None @dataclass class AscendMetadataForPrefill: """ Prefill Specific Metadata for Ascend""" pcp_metadata: Optional[AscendPCPMetadata] = None pcp_allgather_restore_idx: Optional[List[int]] = None @dataclass class AscendMetadataForDecode: """ Decode Specific Metadata for Ascend""" num_computed_tokens_of_pcp_dcp: Optional[list[Optional[list[Optional[ list[int]]]]]] = None @dataclass class AscendMetadata: # **************************** Basic Properties ************************** # attn_mask: Optional[torch.Tensor] = None # Current state of this attention run. attn_state: AscendAttentionState = AscendAttentionState.ChunkedPrefill # Number of tokens excluding padding. num_actual_tokens_pcp_padded: int = 0 num_actual_tokens: int = 0 num_decode_tokens: int = 0 num_prefills: int = 0 num_decodes: int = 0 # The sequence length per sequence. Sequence length means the computed # tokens + new tokens (is None if it is a decoding). # (batch_size,) # TODO(Angazenn): The following parameters are quite redundant and # contains similar information (such as seq_lens seq_lens_list). We # should simplified these parameters once attention schema in vLLM-Ascend # is unified. seq_lens: torch.Tensor = None seq_lens_list: List[int] = None # type: ignore actual_seq_lengths_q: List[int] = None # type: ignore query_start_loc: torch.Tensor = None query_lens: torch.Tensor = None # Maximum query length in the batch (None for decoding). max_query_len: Optional[int] = None # ********************** KV Cache Related Properties ********************* # # Block addresses per sequence (Seq id -> list of physical block). # (batch_size, max_blocks_per_seq) block_tables: torch.Tensor = None # The indices of the token slots that input tokens will be stored into. # E.g., if `slot_mapping` is [35, 2, 17] and the block size is 16, the # three tokens are stored in the 3rd slot in block 2, 2nd slot in block 0, # and 1st slot in block 1, respectively. # (num_tokens,) slot_mapping: torch.Tensor = None prefill: Optional[AscendMetadataForPrefill] = None decode_meta: Optional[AscendMetadataForDecode] = None class AscendAttentionMetadataBuilder: # Does this backend/builder support ACL Graphs for attention (default: no). aclgraph_support: ClassVar[AttentionCGSupport] = \ AttentionCGSupport.UNIFORM_SINGLE_TOKEN_DECODE # Does this backend/builder reorder the batch? # If not, set this to None. Otherwise set it to the query # length that will be pulled into the front of the batch. reorder_batch_threshold: ClassVar[int] = 1 def __init__( self, kv_cache_spec: AttentionSpec, layer_names: list[str], vllm_config: VllmConfig, device: torch.device, ): self.vllm_config = vllm_config self.model_config = vllm_config.model_config self.device = device self.max_num_blocks_per_req = cdiv( self.model_config.max_model_len, AscendAttentionBackend.get_supported_block_size()[0]) def reorder_batch(self, input_batch, scheduler_output: "SchedulerOutput") -> bool: return False def build( self, common_prefix_len: int, common_attn_metadata: AscendCommonAttentionMetadata, model: Optional[nn.Module] = None, ): num_reqs = common_attn_metadata.num_reqs num_actual_tokens = common_attn_metadata.num_actual_tokens query_start_loc_cpu = common_attn_metadata.query_start_loc_cpu[: num_reqs + 1] decode_threshold = 1 num_decodes, num_prefills, num_decode_tokens, num_prefill_tokens = \ split_decodes_and_prefills(common_attn_metadata, decode_threshold=decode_threshold) assert num_decodes + num_prefills == num_reqs assert num_decode_tokens + num_prefill_tokens == num_actual_tokens block_table = common_attn_metadata.block_table_tensor query_lens = query_start_loc_cpu[1:] - query_start_loc_cpu[:-1] seq_lens = common_attn_metadata.seq_lens_cpu[:num_reqs] long_seq_metadata = common_attn_metadata.prefill_context_parallel_metadata num_actual_tokens_pcp_padded = long_seq_metadata.num_actual_tokens_pcp_padded if long_seq_metadata else None if num_actual_tokens_pcp_padded is None: num_actual_tokens_pcp_padded = num_actual_tokens slot_mapping = common_attn_metadata.slot_mapping[: num_actual_tokens_pcp_padded] # slot_mapping = common_attn_metadata.slot_mapping[:num_actual_tokens] attn_mask = common_attn_metadata.attn_mask attn_state = common_attn_metadata.attn_state query_start_loc_cpu = common_attn_metadata.query_start_loc_cpu[: num_reqs + 1] if attn_state == AscendAttentionState.DecodeOnly and \ common_attn_metadata.num_input_tokens > num_actual_tokens: padded_num_tokens = common_attn_metadata.num_input_tokens - num_actual_tokens seq_lens = torch.cat([ seq_lens, torch.ones(padded_num_tokens, dtype=seq_lens.dtype, device=seq_lens.device) ]) block_table_padding = torch.zeros( (padded_num_tokens, ) + block_table.shape[1:], dtype=block_table.dtype, device=block_table.device) block_table = torch.cat([block_table, block_table_padding], dim=0) query_start_loc_cpu = torch.cat([ query_start_loc_cpu, torch.arange(query_start_loc_cpu[-1] + 1, query_start_loc_cpu[-1] + padded_num_tokens, dtype=query_start_loc_cpu.dtype, device=query_start_loc_cpu.device) ]) query_start_loc = query_start_loc_cpu.to(self.device, non_blocking=True) if is_310p(): if attn_state == AscendAttentionState.PrefillNoCache: mask_nz = nd_to_nz_2d(attn_mask) attn_mask = torch_npu.npu_format_cast(mask_nz.contiguous(), ACL_FORMAT_FRACTAL_NZ) elif attn_state == AscendAttentionState.ChunkedPrefill: mask_nz = nd_to_nz_spec(attn_mask) attn_mask = torch_npu.npu_format_cast(mask_nz.contiguous(), ACL_FORMAT_FRACTAL_NZ) prefill_metadata = None if num_prefills > 0: pcp_metadata = None common_long_seq_metadata = common_attn_metadata.prefill_context_parallel_metadata if common_long_seq_metadata is not None: pcp_metadata = AscendPCPMetadata( q_head_idx=common_long_seq_metadata.q_head_idx_tensor, q_tail_idx=common_long_seq_metadata.q_tail_idx_tensor, kv_with_q_head_nomask_idx=common_long_seq_metadata. kv_with_q_head_nomask_idx_tensor, kv_with_q_head_mask_idx=common_long_seq_metadata. kv_with_q_head_mask_idx_tensor, kv_with_q_tail_nomask_idx=common_long_seq_metadata. kv_with_q_tail_nomask_idx_tensor, kv_with_q_tail_mask_idx=common_long_seq_metadata. kv_with_q_tail_mask_idx_tensor, attn_mask_seqlens=common_long_seq_metadata. attn_mask_seqlens, head_attn_nomask_seqlens=common_long_seq_metadata. head_attn_nomask_seqlens, tail_attn_nomask_seqlens=common_long_seq_metadata. tail_attn_nomask_seqlens, q_full_idx=common_long_seq_metadata.q_full_idx, pcp_prefill_mask=common_long_seq_metadata.pcp_prefill_mask) prefill_metadata = AscendMetadataForPrefill( pcp_metadata=pcp_metadata, pcp_allgather_restore_idx=common_long_seq_metadata. pcp_allgather_restore_idx if common_long_seq_metadata is not None else None) decode_metadata = None if num_decodes > 0: common_long_seq_metadata = common_attn_metadata.prefill_context_parallel_metadata if common_long_seq_metadata is not None: num_computed_tokens_of_pcp_dcp = common_long_seq_metadata.num_computed_tokens_of_pcp_dcp num_computed_tokens_of_pcp_dcp = np.array( num_computed_tokens_of_pcp_dcp) decode_metadata = AscendMetadataForDecode( num_computed_tokens_of_pcp_dcp= num_computed_tokens_of_pcp_dcp) attn_metadata = AscendMetadata( num_actual_tokens=num_actual_tokens, num_decode_tokens=num_decode_tokens, num_actual_tokens_pcp_padded=num_actual_tokens_pcp_padded, block_tables=block_table, query_start_loc=query_start_loc, query_lens=query_lens, seq_lens=seq_lens, seq_lens_list=seq_lens.tolist(), max_query_len=common_attn_metadata.max_query_len, actual_seq_lengths_q=query_start_loc_cpu[1:].tolist(), slot_mapping=slot_mapping, attn_mask=attn_mask, attn_state=attn_state, num_prefills=num_prefills, num_decodes=num_decodes, prefill=prefill_metadata, decode_meta=decode_metadata) return attn_metadata def build_for_graph_capture( self, common_attn_metadata: AscendCommonAttentionMetadata, attn_state: AscendAttentionState = AscendAttentionState.DecodeOnly, model: Optional[nn.Module] = None, ): if attn_state == AscendAttentionState.DecodeOnly: attn_metadata = self.build( common_prefix_len=0, common_attn_metadata=common_attn_metadata, ) else: raise NotImplementedError( "Currently we only support building dummy metadata for DecodeOnly state" ) attn_metadata.attn_state = attn_state return attn_metadata class AscendAttentionBackendImpl(AttentionImpl): def __init__( self, num_heads: int, head_size: int, scale: float, num_kv_heads: int, alibi_slopes: Optional[List[float]], sliding_window: Optional[int], kv_cache_dtype: str, logits_soft_cap: Optional[float], attn_type: str, kv_sharing_target_layer_name: Optional[str], **kwargs, ) -> None: self.num_heads = num_heads self.head_size = head_size self.scale = float(scale) self.num_kv_heads = num_heads if num_kv_heads is None else num_kv_heads self.hidden_size = self.num_heads * self.head_size self.kv_cache_dtype = kv_cache_dtype self.sliding_window = sliding_window if alibi_slopes is not None: alibi_slopes = torch.tensor(alibi_slopes, dtype=torch.float32, device="npu") self.alibi_slopes = alibi_slopes self.attn_type = attn_type assert self.num_heads % self.num_kv_heads == 0 self.num_queries_per_kv = self.num_heads // self.num_kv_heads self.key_cache = None self.value_cache = None self.torch_npu_check = version_check() self.pcp_size = get_prefill_context_model_parallel_world_size( ) if prefill_context_parallel_enable() else 1 self.pcp_rank = get_prefill_context_model_parallel_rank( ) if self.pcp_size > 1 else 0 self.pcp_group = get_pcp_group( ).device_group if self.pcp_size > 1 else None self.dcp_size = get_decode_context_model_parallel_world_size() self.dcp_rank = get_decode_context_model_parallel_rank( ) if self.dcp_size > 1 else 0 self.dcp_group = get_dcp_group( ).device_group if self.dcp_size > 1 else None def _forward_prefill_no_cache( self, query: torch.Tensor, key: torch.Tensor, value: torch.Tensor, attn_metadata: AscendMetadata, output: Optional[torch.Tensor] = None, num_tokens=0, ) -> torch.Tensor: assert attn_metadata is not None assert attn_metadata.attn_mask is not None mask = attn_metadata.attn_mask if is_310p(): # align q k v output tensors query = aligned_16(query) key = aligned_16(key) value = aligned_16(value) output = aligned_16(output) # do reformat in case of broadcasted tensors mask = mask.repeat(attn_metadata.seq_lens.size(0), 1, 1, 1) mask = torch_npu.npu_format_cast(mask.contiguous(), ACL_FORMAT_FRACTAL_NZ) torch_npu._npu_flash_attention(query=query, key=key, value=value, mask=mask, seq_len=attn_metadata.seq_lens, scale_value=self.scale, num_heads=self.num_heads, num_kv_heads=self.num_kv_heads, out=output) assert output is not None return output[:num_tokens] def _forward_prefill_cache_hit( self, query: torch.Tensor, attn_metadata: AscendMetadata, output: Optional[torch.Tensor] = None, ) -> torch.Tensor: assert attn_metadata is not None assert attn_metadata.attn_mask is not None compress_mask = attn_metadata.attn_mask batch_size = attn_metadata.query_lens.shape[0] block_table = attn_metadata.block_tables[:batch_size, :] torch_npu._npu_flash_attention_qlens( query=query, key_cache=self.key_cache, value_cache=self.value_cache, block_table=block_table, mask=compress_mask, seq_len=attn_metadata.query_lens, context_lens=attn_metadata.seq_lens, num_kv_heads=self.num_kv_heads, num_heads=self.num_heads, scale_value=self.scale, out=output) return output def _forward_decode_only( self, query: torch.Tensor, attn_metadata: AscendMetadata, output: Optional[torch.Tensor] = None, ) -> torch.Tensor: if is_310p(): # seq_lens_tensor needs to be transferred to the device for 310P. attn_metadata.seq_lens = \ attn_metadata.seq_lens.to(device=query.device) if self.sliding_window is not None and attn_metadata.seq_lens.shape[ 0] == query.size(0): batch_size = attn_metadata.seq_lens.shape[0] block_size = 128 query = query.view(batch_size, 1, self.num_heads * self.head_size) key = self.key_cache value = self.value_cache if self.key_cache is not None and self.value_cache is not None: block_size = self.key_cache.shape[1] key = self.key_cache.flatten(2, 3).contiguous() value = self.value_cache.flatten(2, 3).contiguous() output, _ = torch_npu.npu_fused_infer_attention_score( query, key, value, num_heads=self.num_heads, num_key_value_heads=self.num_kv_heads, input_layout="BSH", block_size=block_size, pre_tokens=self.sliding_window, scale=self.scale, block_table=attn_metadata.block_tables, actual_seq_lengths=[1] * len(attn_metadata.seq_lens), actual_seq_lengths_kv=attn_metadata.seq_lens) output = output.view(batch_size, self.num_heads, self.head_size) else: graph_params = get_graph_params() forward_context: ForwardContext = get_forward_context() num_tokens = query.shape[0] if forward_context.capturing: if self.torch_npu_check: # Get workspace from cache or calculate it if not present. workspace = graph_params.workspaces.get(num_tokens) if workspace is None: workspace = torch_npu._npu_paged_attention_get_workspace( query=query, key_cache=self.key_cache, value_cache=self.value_cache, num_kv_heads=self.num_kv_heads, num_heads=self.num_heads, scale_value=self.scale, block_table=attn_metadata.block_tables, context_lens=attn_metadata.seq_lens, out=output) update_graph_params_workspaces( num_tokens, weak_ref_tensors(workspace)) # Handle graph capturing mode stream = torch_npu.npu.current_stream() event = torch.npu.ExternalEvent() event.wait(stream) event.reset(stream) graph_params.events[num_tokens].append(event) graph_params.attn_params[num_tokens].append(( weak_ref_tensors(query), weak_ref_tensors(self.key_cache), weak_ref_tensors(self.value_cache), self.num_kv_heads, self.num_heads, self.scale, attn_metadata.block_tables, attn_metadata.seq_lens, weak_ref_tensors(output), )) torch.npu.graph_task_group_begin(stream) if self.torch_npu_check: torch_npu._npu_paged_attention( query=query, key_cache=self.key_cache, value_cache=self.value_cache, num_kv_heads=self.num_kv_heads, num_heads=self.num_heads, scale_value=self.scale, block_table=attn_metadata.block_tables, context_lens=attn_metadata.seq_lens, out=output, workspace=workspace) else: torch_npu._npu_paged_attention( query=query, key_cache=self.key_cache, value_cache=self.value_cache, num_kv_heads=self.num_kv_heads, num_heads=self.num_heads, scale_value=self.scale, block_table=attn_metadata.block_tables, context_lens=attn_metadata.seq_lens, out=output) handle = torch.npu.graph_task_group_end(stream) graph_params.handles[num_tokens].append(handle) else: torch_npu._npu_paged_attention( query=query, key_cache=self.key_cache, value_cache=self.value_cache, num_kv_heads=self.num_kv_heads, num_heads=self.num_heads, scale_value=self.scale, block_table=attn_metadata.block_tables, context_lens=attn_metadata.seq_lens, out=output) return output def _forward_v1_style( self, query: torch.Tensor, attn_metadata: AscendMetadata, output: Optional[torch.Tensor] = None, ) -> torch.Tensor: # Use chunked prefill for head size 192 scenario, like deepseek # paged_attention_splitfuse maybe crash at such scenario. # TODO: vanilla path will be removed after the kernel support # head_size 192 scenario. if self.head_size == 192: cu_seqlen_q = [0] + attn_metadata.query_lens.tolist() cu_seqlen_k = [0] + attn_metadata.seq_lens.tolist() cu_seqlen_q = torch.tensor(cu_seqlen_q, device=query.device) cu_seqlen_k = torch.tensor(cu_seqlen_k, device=query.device) cu_seqlen_q = torch.cumsum(cu_seqlen_q, dim=0) cu_seqlen_k = torch.cumsum(cu_seqlen_k, dim=0) max_seqlen_q = torch.max(attn_metadata.query_lens) max_seqlen_k = torch.max(attn_metadata.seq_lens) vanilla_chunked_prefill(output, query, self.key_cache, self.value_cache, attn_metadata.block_tables, cu_seqlen_q, cu_seqlen_k, max_seqlen_q, max_seqlen_k, self.scale, None, True) return output # Use paged attention. assert attn_metadata is not None assert attn_metadata.attn_mask is not None if is_310p(): # Do reformat in case of broadcasted tensors. attn_metadata.attn_mask = \ torch_npu.npu_format_cast(attn_metadata.attn_mask.contiguous(), ACL_FORMAT_FRACTAL_NZ) attn_metadata.seq_lens = \ attn_metadata.seq_lens.to(device=query.device) if torch.version.cann.startswith("8.3"): # TODO:The npu_fused_infer_attention_score op is planned to # be utilized in a wider range in upcoming versions. num_block, block_size, _, _ = self.key_cache.shape # type: ignore key = self.key_cache.view( # type: ignore num_block, block_size, -1) value = self.value_cache.view( # type: ignore num_block, block_size, -1) output, _ = torch_npu.npu_fused_infer_attention_score( query=query, key=key, value=value, atten_mask=attn_metadata.attn_mask, block_table=attn_metadata.block_tables, input_layout="TND", block_size=block_size, actual_seq_lengths=attn_metadata.actual_seq_lengths_q, actual_seq_lengths_kv=attn_metadata.seq_lens_list, num_key_value_heads=self.num_kv_heads, num_heads=self.num_heads, scale=self.scale, sparse_mode=3, ) else: torch_npu._npu_paged_attention_splitfuse( query=query, key_cache=self.key_cache, value_cache=self.value_cache, mask=attn_metadata.attn_mask, block_table=attn_metadata.block_tables, seq_len=attn_metadata.query_lens, context_lens=attn_metadata.seq_lens, num_kv_heads=self.num_kv_heads, num_heads=self.num_heads, scale_value=self.scale, out=output) return output def _pack_tnd_2_bsnd(self, tensor_tnd: torch.Tensor, lengths: List[int]) -> torch.Tensor: max_len = max(lengths) splits = torch.split(tensor_tnd, lengths, dim=0) padded = [] for s in splits: pad_len = max_len - s.shape[0] s_pad = F.pad(s, (0, 0, 0, 0, 0, pad_len)) padded.append(s_pad) tensor_bsnd = torch.stack(padded, dim=0) return tensor_bsnd def _unpack_bsnd_2_tnd(self, tensor_bsnd: torch.Tensor, lengths: List[int]) -> torch.Tensor: slices = [] for i, length in enumerate(lengths): slices.append(tensor_bsnd[i, :length]) tensor_tnd = torch.cat(slices, dim=0) return tensor_tnd def _attention_with_nomask_and_mask(self, q: torch.Tensor, q_seqlens: List[int], k_nomask: torch.Tensor, v_nomask: torch.Tensor, kv_seqlens_nomask: List[int], k_mask: torch.Tensor, v_mask: torch.Tensor, kv_seqlens_mask: List[int], mask: torch.Tensor) -> torch.Tensor: q = self._pack_tnd_2_bsnd(q, q_seqlens) # nomask Attention if k_nomask is not None: attn_out_nomask, attn_lse_nomask = torch.ops.npu.npu_fused_infer_attention_score( q, self._pack_tnd_2_bsnd(k_nomask, kv_seqlens_nomask), self._pack_tnd_2_bsnd(v_nomask, kv_seqlens_nomask), num_heads=self.num_heads, num_key_value_heads=self.num_kv_heads, input_layout="BSND", atten_mask=None, scale=self.scale, sparse_mode=0, antiquant_mode=0, antiquant_scale=None, softmax_lse_flag=True, actual_seq_lengths_kv=kv_seqlens_nomask, actual_seq_lengths=q_seqlens) attn_out_nomask = self._unpack_bsnd_2_tnd(attn_out_nomask, q_seqlens) # (B, N, Q_S, 1) -> (B, Q_S, N, 1) -> (T, N, 1) attn_lse_nomask = self._unpack_bsnd_2_tnd( attn_lse_nomask.permute([0, 2, 1, 3]), q_seqlens) # mask Attention attn_out_mask, attn_lse_mask = torch.ops.npu.npu_fused_infer_attention_score( q, self._pack_tnd_2_bsnd(k_mask, kv_seqlens_mask), self._pack_tnd_2_bsnd(v_mask, kv_seqlens_mask), num_heads=self.num_heads, num_key_value_heads=self.num_kv_heads, input_layout="BSND", atten_mask=mask, scale=self.scale, sparse_mode=0, antiquant_mode=0, antiquant_scale=None, softmax_lse_flag=True, actual_seq_lengths_kv=kv_seqlens_mask, actual_seq_lengths=q_seqlens) attn_out_mask = self._unpack_bsnd_2_tnd(attn_out_mask, q_seqlens) attn_lse_mask = self._unpack_bsnd_2_tnd( attn_lse_mask.permute([0, 2, 1, 3]), q_seqlens) # update output = attn_out_mask if k_nomask is not None: output, _ = self._update_out_and_lse( torch.stack([attn_out_nomask, attn_out_mask], dim=0), torch.stack([attn_lse_nomask, attn_lse_mask], dim=0)) return output def _forward_prefill_cp(self, query: torch.Tensor, key: torch.Tensor, value: torch.Tensor, attn_metadata: AscendMetadata) -> torch.Tensor: assert attn_metadata is not None assert attn_metadata.prefill is not None assert attn_metadata.prefill.pcp_metadata is not None # Use precomputed indices from the metadata (already converted to tensors and on device) q_head_idx = attn_metadata.prefill.pcp_metadata.q_head_idx q_tail_idx = attn_metadata.prefill.pcp_metadata.q_tail_idx kv_with_q_head_nomask_idx = attn_metadata.prefill.pcp_metadata.kv_with_q_head_nomask_idx kv_with_q_head_mask_idx = attn_metadata.prefill.pcp_metadata.kv_with_q_head_mask_idx kv_with_q_tail_nomask_idx = attn_metadata.prefill.pcp_metadata.kv_with_q_tail_nomask_idx kv_with_q_tail_mask_idx = attn_metadata.prefill.pcp_metadata.kv_with_q_tail_mask_idx attn_mask_seqlens = attn_metadata.prefill.pcp_metadata.attn_mask_seqlens head_attn_nomask_seqlens = attn_metadata.prefill.pcp_metadata.head_attn_nomask_seqlens tail_attn_nomask_seqlens = attn_metadata.prefill.pcp_metadata.tail_attn_nomask_seqlens mask = attn_metadata.prefill.pcp_metadata.pcp_prefill_mask # 1. Attention calculation in the first half of Q in load balancing output_head = self._attention_with_nomask_and_mask( q=torch.index_select(query, 0, q_head_idx), q_seqlens=attn_mask_seqlens[0].tolist(), k_nomask=torch.index_select(key, 0, kv_with_q_head_nomask_idx) if self.pcp_rank > 0 else None, v_nomask=torch.index_select(value, 0, kv_with_q_head_nomask_idx) if self.pcp_rank > 0 else None, kv_seqlens_nomask=head_attn_nomask_seqlens[1].tolist(), k_mask=torch.index_select(key, 0, kv_with_q_head_mask_idx), v_mask=torch.index_select(value, 0, kv_with_q_head_mask_idx), kv_seqlens_mask=attn_mask_seqlens[0].tolist(), mask=mask) # 2. the Attention calculation in the latter half of Q in load balancing # pcp_rank0: Q3*KV0~KV2 + Q3*KV3 # pcp_rank1: Q2*KV0~KV1 + Q2*KV2 output_tail = self._attention_with_nomask_and_mask( q=torch.index_select(query, 0, q_tail_idx), q_seqlens=attn_mask_seqlens[0].tolist(), k_nomask=torch.index_select(key, 0, kv_with_q_tail_nomask_idx), v_nomask=torch.index_select(value, 0, kv_with_q_tail_nomask_idx), kv_seqlens_nomask=tail_attn_nomask_seqlens[1].tolist(), k_mask=torch.index_select(key, 0, kv_with_q_tail_mask_idx), v_mask=torch.index_select(value, 0, kv_with_q_tail_mask_idx), kv_seqlens_mask=attn_mask_seqlens[0].tolist(), mask=mask) # 3. Combine the output of the first half and second half. q_full_idx = attn_metadata.prefill.pcp_metadata.q_full_idx output = torch.index_select( torch.cat([output_head, output_tail], dim=0), 0, q_full_idx) return output def _update_out_and_lse(self, out_list: torch.Tensor, lse_list: torch.Tensor) -> torch.Tensor: """LSE_final = log(sum(exp(LSE_i))), O_final = sum(exp(LSE_i - LSE_final) * O_i) Args: out_list: shape = [N, batch_size, num_heads, head_size] lse_list: shape = [N, batch_size, num_heads, 1] Returns: out_final: shape = [batch_size, num_heads, head_size] lse_final: shape = [batch_size, num_heads, 1] """ lse_final = torch.logsumexp(lse_list, dim=0, keepdim=False) out_final = torch.sum(torch.exp(lse_list - lse_final) * out_list, dim=0) return out_final, lse_final def _forward_decode_pcp_dcp(self, query: torch.Tensor, attn_metadata: AscendMetadata) -> torch.Tensor: assert self.key_cache is not None assert self.value_cache is not None if self.dcp_size > 1: query = get_dcp_group().all_gather(query, 1) num_heads = self.num_heads * self.dcp_size else: num_heads = self.num_heads # 1. Compute out&lse by "npu_fused_infer_attention_score" q_nope = query.view(query.shape[0], 1, query.shape[1], query.shape[2]) # [b,num_heads,head_size] -> [b,1,num_heads,head_size] k_nope = self.key_cache.view(self.key_cache.shape[0], self.key_cache.shape[1], -1) value = self.value_cache.view(self.key_cache.shape[0], self.key_cache.shape[1], -1) common_kwargs = { 'num_heads': num_heads, 'num_key_value_heads': self.num_kv_heads, 'input_layout': "BSND", 'atten_mask': None, 'scale': self.scale, 'antiquant_mode': 0, 'antiquant_scale': None, 'softmax_lse_flag': True, 'block_table': attn_metadata.block_tables, 'block_size': self.key_cache.shape[1], "actual_seq_lengths_kv": attn_metadata.decode_meta. num_computed_tokens_of_pcp_dcp[:, self.pcp_rank, self.dcp_rank], } graph_params = get_graph_params() forward_context: ForwardContext = get_forward_context() num_tokens = query.shape[0] if forward_context.capturing: stream = torch_npu.npu.current_stream() event = torch.npu.ExternalEvent() event.wait(stream) event.reset(stream) graph_params.events[num_tokens].append(event) workspace = graph_params.workspaces.get(num_tokens) if workspace is None: workspace = torch_npu._npu_fused_infer_attention_score_get_max_workspace( q_nope, k_nope, value, **common_kwargs) update_graph_params_workspaces(num_tokens, weak_ref_tensors(workspace)) attn_out = torch.empty_like(q_nope) attn_lse = torch.empty((num_tokens, num_heads, 1, 1), dtype=torch.float, device=q_nope.device) graph_params.attn_params[num_tokens].append( (weak_ref_tensors(q_nope), weak_ref_tensors(k_nope), weak_ref_tensors(value), self.num_heads, self.num_kv_heads, self.scale, attn_metadata.block_tables, self.key_cache.shape[1], attn_metadata.decode_meta. num_computed_tokens_of_pcp_dcp[:, self.pcp_rank, self.dcp_rank], weak_ref_tensors(attn_out), weak_ref_tensors(attn_lse), self.pcp_rank, self.dcp_rank, self.dcp_size)) torch.npu.graph_task_group_begin(stream) torch_npu.npu_fused_infer_attention_score.out( q_nope, k_nope, value, **common_kwargs, workspace=workspace, out=[attn_out, attn_lse]) handle = torch.npu.graph_task_group_end(stream) graph_params.handles[num_tokens].append(handle) else: attn_out, attn_lse = torch_npu.npu_fused_infer_attention_score( q_nope, k_nope, value, **common_kwargs) attn_out = attn_out.view(attn_out.shape[0], attn_out.shape[2], attn_out.shape[3]) attn_lse = attn_lse.view(attn_lse.shape[0], attn_lse.shape[1], 1) if self.dcp_size > 1: # Concat out&lse: [bs,num_heads,v_head_dim] + [bs,num_heads,1] -> [bs,num_heads,v_head_dim+1] attn_out_lse = torch.cat([attn_out, attn_lse], dim=-1) # permute: [bs, num_heads, v_head_dim+1] -> [num_heads, v_head_dim+1, bs] attn_out_lse = attn_out_lse.permute([1, 2, 0]).contiguous() attn_out_lse_all2all = torch.empty_like(attn_out_lse) dist.all_to_all_single(attn_out_lse_all2all, attn_out_lse, group=self.dcp_group) # permute: [num_heads, v_head_dim+1, bs] -> [bs, num_heads, v_head_dim+1] attn_out_lse_all2all = attn_out_lse_all2all.permute([2, 0, 1]) attn_out_lse_split_on_seq = list( torch.chunk(attn_out_lse_all2all, self.dcp_size, dim=1)) attn_out_lse_split_dcp = torch.stack( attn_out_lse_split_on_seq, dim=0) # [dcp, batch_size, num_heads, head_size+1] # Update out&lse attn_out_split_dcp, attn_lse_split_dcp = torch.split( attn_out_lse_split_dcp, [self.head_size, 1], dim=-1) attn_out, attn_lse = self._update_out_and_lse( attn_out_split_dcp, attn_lse_split_dcp) if self.pcp_size > 1: # 2. Concat out&lse: [bs,num_heads,head_size] + [bs,num_heads,1] -> [bs,num_heads,head_size+1] attn_out_lse = torch.cat([attn_out, attn_lse], dim=-1) # 3. AllGather out&lse within CP group attn_out_lse_list = [ torch.empty_like(attn_out_lse) for _ in range(self.pcp_size) ] dist.all_gather(attn_out_lse_list, attn_out_lse, group=self.pcp_group) # 4. Update out&lse attn_out_lse_allgather = torch.stack( attn_out_lse_list, dim=0) # [pcp, batch_size, num_heads, head_size+1] attn_out_allgather, attn_lse_allgather = torch.split( attn_out_lse_allgather, [self.head_size, 1], dim=-1) attn_out, _ = self._update_out_and_lse(attn_out_allgather, attn_lse_allgather) return attn_out def _forward_pcp_dcp(self, query: torch.Tensor, key: torch.Tensor, value: torch.Tensor, attn_metadata: AscendMetadata, output: torch.Tensor) -> torch.Tensor: assert attn_metadata is not None has_decode = attn_metadata.num_decodes > 0 has_prefill = attn_metadata.num_prefills > 0 num_decode_tokens = attn_metadata.num_decode_tokens if has_decode: decode_query = query[:num_decode_tokens] output_decode = self._forward_decode_pcp_dcp( decode_query, attn_metadata) output[:num_decode_tokens] = output_decode if has_prefill: prefill_query = query[num_decode_tokens:] key = key[self.pcp_size * num_decode_tokens:] value = value[self.pcp_size * num_decode_tokens:] if self.pcp_size > 1: output_prefill = self._forward_prefill_cp( prefill_query, key, value, attn_metadata) else: max_prefill_seq_len = attn_metadata.seq_lens[ attn_metadata.num_decode_tokens:].max().item() if attn_metadata.attn_mask is not None: attn_metadata.attn_mask = attn_metadata.attn_mask[: max_prefill_seq_len, : max_prefill_seq_len] else: ValueError("Attn_metadata.attn_mask is required") seq_lens_back = attn_metadata.seq_lens attn_metadata.seq_lens = attn_metadata.seq_lens[ attn_metadata.num_decode_tokens:] output_prefill = self._forward_prefill_no_cache( prefill_query, key, value, attn_metadata, output[num_decode_tokens:], prefill_query.shape[0]) attn_metadata.seq_lens = seq_lens_back output[num_decode_tokens:] = output_prefill return output def forward( self, layer: AttentionLayer, query: torch.Tensor, key: torch.Tensor, value: torch.Tensor, kv_cache: Tuple[torch.Tensor], attn_metadata: AscendMetadata, output: Optional[torch.Tensor] = None, output_scale: Optional[torch.Tensor] = None, output_block_scale: Optional[torch.Tensor] = None, ) -> torch.Tensor: """Forward pass with Ascend attention. Args: query: shape = [num_tokens, num_heads, head_size] key: shape = [num_tokens, num_kv_heads, head_size] value: shape = [num_tokens, num_kv_heads, head_size] kv_cache: shape = [2, num_blocks, block_size, num_kv_heads, head_size] attn_metadata: Metadata for attention. Returns: shape = [num_tokens, num_heads * head_size] """ assert output is not None, "Output tensor must be provided." if output_scale is not None or output_block_scale is not None: raise NotImplementedError( "fused output quantization is not yet supported" " for AscendAttentionBackendImpl") num_tokens = query.shape[0] if attn_metadata is None: return output # NOTE: Currently, we have various attention paths for different # scenarios, and not all of them are in-place operations. Therefore, # we need to create a separate tensor to hold the attention result. # In the future, we may consolidate them into fewer paths, which will # hopefully allow us to use in-place operation by default. intermediate_output: torch.Tensor assert layer._k_scale_float == 1.0 and layer._v_scale_float == 1.0 attn_type = self.attn_type if attn_type != AttentionType.DECODER and attn_type != AttentionType.ENCODER_ONLY: raise NotImplementedError("Encoder/decoder cross-attention " "are not implemented for " "PallasAttentionBackendImpl") num_decode_tokens = attn_metadata.num_decode_tokens has_decode = attn_metadata.num_decodes > 0 has_prefill = attn_metadata.num_prefills > 0 if len(kv_cache) > 1: if self.key_cache is None: self.key_cache, self.value_cache = kv_cache[0], kv_cache[1] if has_decode: slot_mapping = attn_metadata.slot_mapping[:num_decode_tokens * self.pcp_size: self.pcp_size] \ if self.pcp_size * self.dcp_size > 1 else attn_metadata.slot_mapping[:num_decode_tokens] torch_npu._npu_reshape_and_cache( key=key[:num_decode_tokens], value=value[:num_decode_tokens], key_cache=self.key_cache, value_cache=self.value_cache, slot_indices=slot_mapping) if has_prefill: if self.pcp_size > 1: kv = torch.cat([key, value], dim=-1) all_kv = get_pcp_group().all_gather(kv, dim=0) pcp_allgather_restore_idx = attn_metadata.prefill.pcp_allgather_restore_idx if attn_metadata.prefill else None all_kv = torch.index_select(all_kv, 0, pcp_allgather_restore_idx) key, value = all_kv.split([self.head_size, self.head_size], dim=-1) torch_npu._npu_reshape_and_cache( key=key[self.pcp_size * num_decode_tokens:attn_metadata. num_actual_tokens_pcp_padded], value=value[self.pcp_size * num_decode_tokens:attn_metadata. num_actual_tokens_pcp_padded], key_cache=self.key_cache, value_cache=self.value_cache, slot_indices=attn_metadata. slot_mapping[self.pcp_size * num_decode_tokens:attn_metadata. num_actual_tokens_pcp_padded]) if self.pcp_size * self.dcp_size > 1: intermediate_output = self._forward_pcp_dcp( query, key, value, attn_metadata, output) elif attn_type == AttentionType.ENCODER_ONLY: # TODO(zzzwwjj): Deal with this `cum_seq_len` more elegantly. cum_seq_len = attn_metadata.query_start_loc[1:].tolist() intermediate_output = torch_npu.npu_fusion_attention( query, key, value, head_num=self.num_heads, input_layout="TND", scale=self.scale, sparse_mode=4, atten_mask=attn_metadata.attn_mask, pre_tockens=attn_metadata.max_query_len, next_tockens=attn_metadata.max_query_len, actual_seq_qlen=cum_seq_len, actual_seq_kvlen=cum_seq_len, )[0] # V0-Style scheduler situation. elif attn_metadata.attn_state == AscendAttentionState.PrefillNoCache: intermediate_output = self._forward_prefill_no_cache( query, key, value, attn_metadata, output, num_tokens) elif attn_metadata.attn_state == \ AscendAttentionState.PrefillCacheHit: intermediate_output = self._forward_prefill_cache_hit( query, attn_metadata, output) elif attn_metadata.attn_state == AscendAttentionState.DecodeOnly: intermediate_output = self._forward_decode_only( query, attn_metadata, output) # Normal V1 situation. else: if torch.version.cann.startswith("8.3"): # npu_fused_infer_attention_score does not support cases # where query.shape[0] != attn_metadata.query_start_loc[-1]. # Thus we need unpad it here. num_tokens = attn_metadata.query_start_loc[-1] query = query[:num_tokens] intermediate_output = self._forward_v1_style( query, attn_metadata, output) output[:num_tokens] = intermediate_output[:num_tokens] return output