EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Refactor] move the metadata from attention_v1 to util(ready for extract common_cp) & realize Ascendmetadata inherit from the parent class. (#5203)

Browse Source 
RFC: https://github.com/vllm-project/vllm-ascend/issues/4629

1. Remove the pcp-related code from attention_v1.
2. Establish the inheritance relationship of CommonAttentionMetadata.

TODO
1. extract common_cp
2. move cp metadata to common_cp.
3. remove commonAttentionMetadata for aclgraph.

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

---------

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>
Co-authored-by: weijinqian_v1 <weijinqian@huawei.com>
This commit is contained in:
weijinqian0
2025-12-23 00:10:52 +08:00
committed by GitHub
parent 3d9954eff0
commit 95e8a52156
9 changed files with 83 additions and 106 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/ut/attention/test_attention_v1.py
View File

@@ -117,7 +117,8 @@ class TestAscendAttentionMetadataBuilder(TestBase):
spec_attn_mask=None,
attn_state=AscendAttentionState.ChunkedPrefill,
num_computed_tokens_cpu=None,
seq_lens=None)
seq_lens=None,
max_seq_len=6)
mock_model = MagicMock()
self.builder.build(1, common_attn_metadata, mock_model)

15
tests/ut/attention/test_mla_v1.py
View File

@@ -606,7 +606,8 @@ class TestAscendMLAMetadataBuilderBuild(TestBase):
spec_attn_mask=None,
attn_state=AscendAttentionState.PrefillNoCache,
num_computed_tokens_cpu=None,
seq_lens=None)
seq_lens=None,
max_seq_len=6)
base_inputs = {
"num_actual_tokens": 10,
@@ -673,7 +674,8 @@ class TestAscendMLAMetadataBuilderBuild(TestBase):
spec_attn_mask=None,
attn_state=AscendAttentionState.ChunkedPrefill,
num_computed_tokens_cpu=None,
seq_lens=None)
seq_lens=None,
max_seq_len=6)
base_inputs = {
"num_actual_tokens": 15,
@@ -729,7 +731,8 @@ class TestAscendMLAMetadataBuilderBuild(TestBase):
spec_attn_mask=None,
attn_state=AscendAttentionState.DecodeOnly,
num_computed_tokens_cpu=None,
seq_lens=None)
seq_lens=None,
max_seq_len=6)
base_inputs = {
"num_actual_tokens": 3,
@@ -784,7 +787,8 @@ class TestAscendMLAMetadataBuilderBuild(TestBase):
spec_attn_mask=None,
attn_state=AscendAttentionState.DecodeOnly,
num_computed_tokens_cpu=None,
seq_lens=None)
seq_lens=None,
max_seq_len=6)
base_inputs = {
"num_actual_tokens": 3,
@@ -839,7 +843,8 @@ class TestAscendMLAMetadataBuilderBuild(TestBase):
spec_attn_mask=None,
attn_state=AscendAttentionState.PrefillNoCache,
num_computed_tokens_cpu=None,
seq_lens=None)
seq_lens=None,
max_seq_len=6)
builder = AscendMLAMetadataBuilder(kv_cache_spec=self.kv_cache_spec,
layer_names=["layer_0", "layer_1"],