Skip to content

vllm.v1.attention.backends.cpu_attn

TorchSDPABackend

Source code in vllm/v1/attention/backends/cpu_attn.py 
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
class TorchSDPABackend:
    accept_output_buffer: bool = False

    @staticmethod
    def get_name() -> str:
        return "TORCH_SDPA_VLLM_V1"

    @staticmethod
    def get_impl_cls() -> type["TorchSDPABackendImpl"]:
        return TorchSDPABackendImpl

    @staticmethod
    def get_metadata_cls() -> type["AttentionMetadata"]:
        return TorchSDPAMetadata

    @staticmethod
    def get_state_cls() -> type["CommonAttentionState"]:
        return CommonAttentionState

    @staticmethod
    def get_builder_cls() -> type["TorchSDPAMetadataBuilderV1"]:
        return TorchSDPAMetadataBuilderV1

    @staticmethod
    def get_kv_cache_shape(
        num_blocks: int,
        block_size: int,
        num_kv_heads: int,
        head_size: int,
    ) -> tuple[int, ...]:
        return PagedAttention.get_kv_cache_shape(num_blocks, block_size,
                                                 num_kv_heads, head_size)

    @staticmethod
    def use_cascade_attention(*args, **kwargs) -> bool:
        return False

accept_output_buffer class-attribute instance-attribute 

accept_output_buffer: bool = False

get_builder_cls staticmethod 

get_builder_cls() -> type[TorchSDPAMetadataBuilderV1]
Source code in vllm/v1/attention/backends/cpu_attn.py 
39
40
41
@staticmethod
def get_builder_cls() -> type["TorchSDPAMetadataBuilderV1"]:
    return TorchSDPAMetadataBuilderV1

get_impl_cls staticmethod 

get_impl_cls() -> type[TorchSDPABackendImpl]
Source code in vllm/v1/attention/backends/cpu_attn.py 
27
28
29
@staticmethod
def get_impl_cls() -> type["TorchSDPABackendImpl"]:
    return TorchSDPABackendImpl

get_kv_cache_shape staticmethod 

get_kv_cache_shape(
    num_blocks: int,
    block_size: int,
    num_kv_heads: int,
    head_size: int,
) -> tuple[int, ...]
Source code in vllm/v1/attention/backends/cpu_attn.py 
43
44
45
46
47
48
49
50
51
@staticmethod
def get_kv_cache_shape(
    num_blocks: int,
    block_size: int,
    num_kv_heads: int,
    head_size: int,
) -> tuple[int, ...]:
    return PagedAttention.get_kv_cache_shape(num_blocks, block_size,
                                             num_kv_heads, head_size)

get_metadata_cls staticmethod 

get_metadata_cls() -> type[AttentionMetadata]
Source code in vllm/v1/attention/backends/cpu_attn.py 
31
32
33
@staticmethod
def get_metadata_cls() -> type["AttentionMetadata"]:
    return TorchSDPAMetadata

get_name staticmethod 

get_name() -> str
Source code in vllm/v1/attention/backends/cpu_attn.py 
23
24
25
@staticmethod
def get_name() -> str:
    return "TORCH_SDPA_VLLM_V1"

get_state_cls staticmethod 

get_state_cls() -> type[CommonAttentionState]
Source code in vllm/v1/attention/backends/cpu_attn.py 
35
36
37
@staticmethod
def get_state_cls() -> type["CommonAttentionState"]:
    return CommonAttentionState

use_cascade_attention staticmethod 

use_cascade_attention(*args, **kwargs) -> bool
Source code in vllm/v1/attention/backends/cpu_attn.py 
53
54
55
@staticmethod
def use_cascade_attention(*args, **kwargs) -> bool:
    return False

TorchSDPAMetadataBuilderV1

Bases: AttentionMetadataBuilder[TorchSDPAMetadata]

Source code in vllm/v1/attention/backends/cpu_attn.py 
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
class TorchSDPAMetadataBuilderV1(AttentionMetadataBuilder[TorchSDPAMetadata]):

    def __init__(self, runner: CPUModelRunner, kv_cache_spec: AttentionSpec,
                 block_table: BlockTable) -> None:
        self.runner = runner
        self.block_table = block_table

        # For reorder
        self.reorder_prompt_req_index_list = np.empty(self.runner.max_num_reqs,
                                                      dtype=np.int64)
        self.reorder_decode_req_index_list = np.empty(self.runner.max_num_reqs,
                                                      dtype=np.int64)
        self.num_prompt_req: int = 0

        self.seq_start_loc_cpu = torch.zeros(
            runner.max_num_reqs + 1,
            dtype=torch.int32,
            device="cpu",
        )
        self.seq_start_loc_np = self.seq_start_loc_cpu.numpy()

    def reorder_batch(self, input_batch: InputBatch,
                      scheduler_output: SchedulerOutput) -> bool:
        prompt_list_idx = 0
        decode_list_idx = 0
        for req_index in range(input_batch.num_reqs):
            if input_batch.num_computed_tokens_cpu[
                    req_index] < input_batch.num_prompt_tokens[req_index]:
                # prompt stage
                self.reorder_prompt_req_index_list[prompt_list_idx] = req_index
                prompt_list_idx += 1
            else:
                # decode stage
                self.reorder_decode_req_index_list[decode_list_idx] = req_index
                decode_list_idx += 1
        assert decode_list_idx + prompt_list_idx == input_batch.num_reqs

        # Update prompt requests number
        self.num_prompt_req = prompt_list_idx

        reorder_req_num = 0
        for req_index in range(decode_list_idx):
            if self.reorder_decode_req_index_list[req_index] < prompt_list_idx:
                reorder_req_num += 1
            else:
                break

        if reorder_req_num == 0:
            return False

        reorder_prompt_list = (
            self.reorder_prompt_req_index_list[:prompt_list_idx]
            [-reorder_req_num:])
        reorder_decode_list = (
            self.reorder_decode_req_index_list[:decode_list_idx]
            [:reorder_req_num])
        assert reorder_decode_list.size == reorder_prompt_list.size

        for idx in range(reorder_req_num):
            prompt_req_index = reorder_prompt_list[idx].item()
            decode_req_index = reorder_decode_list[idx].item()
            input_batch.swap_states(prompt_req_index, decode_req_index)

        return True

    def build(self, common_prefix_len: int,
              common_attn_metadata: CommonAttentionMetadata):
        num_reqs = common_attn_metadata.num_reqs
        num_actual_tokens = common_attn_metadata.num_actual_tokens
        max_query_len = common_attn_metadata.max_query_len

        runner = self.runner
        block_table = self.block_table
        seq_lens_np = runner.seq_lens_np[:num_reqs]
        num_prompt_req = self.num_prompt_req
        max_prefill_seq_len = seq_lens_np[:num_prompt_req].max().item(
        ) if num_prompt_req > 0 else 0
        max_decode_seq_len = seq_lens_np[num_prompt_req:num_reqs].max().item(
        ) if num_prompt_req < num_reqs else 0
        self.seq_start_loc_np[0] = 0
        np.cumsum(seq_lens_np, out=self.seq_start_loc_np[1:num_reqs + 1])
        num_prefill_tokens = runner.query_start_loc_np[num_prompt_req].item()
        num_decode_tokens = runner.query_start_loc_np[num_reqs].item(
        ) - num_prefill_tokens
        slot_mapping = block_table.slot_mapping_cpu[:num_actual_tokens].long()
        block_table_tensor = block_table.get_device_tensor()
        attn_metadata = TorchSDPAMetadata(
            num_prefills=num_prompt_req,
            num_prefill_tokens=num_prefill_tokens,
            num_decode_tokens=num_decode_tokens,
            slot_mapping=slot_mapping,
            seq_lens_tensor=runner.
            seq_lens_cpu[num_prompt_req:num_reqs],  # decode
            max_decode_seq_len=max_decode_seq_len,  # decode
            block_tables=block_table_tensor[num_prompt_req:num_reqs],  # decode
            chunked_prefill=True,
            max_query_len=max_query_len,
            max_kv_len=max_prefill_seq_len,
            prefill_query_start_loc=runner.
            query_start_loc_cpu[:num_prompt_req + 1],  # prefill
            kv_start_loc=self.seq_start_loc_cpu[:num_prompt_req +
                                                1],  # prefill
            prefill_block_tables=block_table_tensor[:
                                                    num_prompt_req],  # prefill
            query_start_loc=runner.query_start_loc_cpu[:num_reqs +
                                                       1],  # for logits index
            multi_modal_placeholder_index_maps=None,
            enable_kv_scales_calculation=False,
        )

        return attn_metadata

block_table instance-attribute 

block_table = block_table

num_prompt_req instance-attribute 

num_prompt_req: int = 0

reorder_decode_req_index_list instance-attribute 

reorder_decode_req_index_list = empty(
    max_num_reqs, dtype=int64
)

reorder_prompt_req_index_list instance-attribute 

reorder_prompt_req_index_list = empty(
    max_num_reqs, dtype=int64
)

runner instance-attribute 

runner = runner

seq_start_loc_cpu instance-attribute 

seq_start_loc_cpu = zeros(
    max_num_reqs + 1, dtype=int32, device="cpu"
)

seq_start_loc_np instance-attribute 

seq_start_loc_np = numpy()

__init__ 

__init__(
    runner: CPUModelRunner,
    kv_cache_spec: AttentionSpec,
    block_table: BlockTable,
) -> None
Source code in vllm/v1/attention/backends/cpu_attn.py 
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
def __init__(self, runner: CPUModelRunner, kv_cache_spec: AttentionSpec,
             block_table: BlockTable) -> None:
    self.runner = runner
    self.block_table = block_table

    # For reorder
    self.reorder_prompt_req_index_list = np.empty(self.runner.max_num_reqs,
                                                  dtype=np.int64)
    self.reorder_decode_req_index_list = np.empty(self.runner.max_num_reqs,
                                                  dtype=np.int64)
    self.num_prompt_req: int = 0

    self.seq_start_loc_cpu = torch.zeros(
        runner.max_num_reqs + 1,
        dtype=torch.int32,
        device="cpu",
    )
    self.seq_start_loc_np = self.seq_start_loc_cpu.numpy()

build 

build(
    common_prefix_len: int,
    common_attn_metadata: CommonAttentionMetadata,
)
Source code in vllm/v1/attention/backends/cpu_attn.py 
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
def build(self, common_prefix_len: int,
          common_attn_metadata: CommonAttentionMetadata):
    num_reqs = common_attn_metadata.num_reqs
    num_actual_tokens = common_attn_metadata.num_actual_tokens
    max_query_len = common_attn_metadata.max_query_len

    runner = self.runner
    block_table = self.block_table
    seq_lens_np = runner.seq_lens_np[:num_reqs]
    num_prompt_req = self.num_prompt_req
    max_prefill_seq_len = seq_lens_np[:num_prompt_req].max().item(
    ) if num_prompt_req > 0 else 0
    max_decode_seq_len = seq_lens_np[num_prompt_req:num_reqs].max().item(
    ) if num_prompt_req < num_reqs else 0
    self.seq_start_loc_np[0] = 0
    np.cumsum(seq_lens_np, out=self.seq_start_loc_np[1:num_reqs + 1])
    num_prefill_tokens = runner.query_start_loc_np[num_prompt_req].item()
    num_decode_tokens = runner.query_start_loc_np[num_reqs].item(
    ) - num_prefill_tokens
    slot_mapping = block_table.slot_mapping_cpu[:num_actual_tokens].long()
    block_table_tensor = block_table.get_device_tensor()
    attn_metadata = TorchSDPAMetadata(
        num_prefills=num_prompt_req,
        num_prefill_tokens=num_prefill_tokens,
        num_decode_tokens=num_decode_tokens,
        slot_mapping=slot_mapping,
        seq_lens_tensor=runner.
        seq_lens_cpu[num_prompt_req:num_reqs],  # decode
        max_decode_seq_len=max_decode_seq_len,  # decode
        block_tables=block_table_tensor[num_prompt_req:num_reqs],  # decode
        chunked_prefill=True,
        max_query_len=max_query_len,
        max_kv_len=max_prefill_seq_len,
        prefill_query_start_loc=runner.
        query_start_loc_cpu[:num_prompt_req + 1],  # prefill
        kv_start_loc=self.seq_start_loc_cpu[:num_prompt_req +
                                            1],  # prefill
        prefill_block_tables=block_table_tensor[:
                                                num_prompt_req],  # prefill
        query_start_loc=runner.query_start_loc_cpu[:num_reqs +
                                                   1],  # for logits index
        multi_modal_placeholder_index_maps=None,
        enable_kv_scales_calculation=False,
    )

    return attn_metadata

reorder_batch 

reorder_batch(
    input_batch: InputBatch,
    scheduler_output: SchedulerOutput,
) -> bool
Source code in vllm/v1/attention/backends/cpu_attn.py 
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
def reorder_batch(self, input_batch: InputBatch,
                  scheduler_output: SchedulerOutput) -> bool:
    prompt_list_idx = 0
    decode_list_idx = 0
    for req_index in range(input_batch.num_reqs):
        if input_batch.num_computed_tokens_cpu[
                req_index] < input_batch.num_prompt_tokens[req_index]:
            # prompt stage
            self.reorder_prompt_req_index_list[prompt_list_idx] = req_index
            prompt_list_idx += 1
        else:
            # decode stage
            self.reorder_decode_req_index_list[decode_list_idx] = req_index
            decode_list_idx += 1
    assert decode_list_idx + prompt_list_idx == input_batch.num_reqs

    # Update prompt requests number
    self.num_prompt_req = prompt_list_idx

    reorder_req_num = 0
    for req_index in range(decode_list_idx):
        if self.reorder_decode_req_index_list[req_index] < prompt_list_idx:
            reorder_req_num += 1
        else:
            break

    if reorder_req_num == 0:
        return False

    reorder_prompt_list = (
        self.reorder_prompt_req_index_list[:prompt_list_idx]
        [-reorder_req_num:])
    reorder_decode_list = (
        self.reorder_decode_req_index_list[:decode_list_idx]
        [:reorder_req_num])
    assert reorder_decode_list.size == reorder_prompt_list.size

    for idx in range(reorder_req_num):
        prompt_req_index = reorder_prompt_list[idx].item()
        decode_req_index = reorder_decode_list[idx].item()
        input_batch.swap_states(prompt_req_index, decode_req_index)

    return True