xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-05-23 05:24:25 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge 1c13c28087de72432b092836fa28694f0efbe801 into 254f6b986720c92ddf97fbb1a6a6465da8e87e29

Browse Source
This commit is contained in:
Rei. 2025-12-25 00:06:56 +00:00 committed by GitHub
commit de99c56dc5
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/v1/worker/gpu_model_runner.py
View File

@ -1708,7 +1708,11 @@ class GPUModelRunner(
_get_block_table_and_slot_mapping(kv_cache_gid) _get_block_table_and_slot_mapping(kv_cache_gid)
) )
if self.speculative_config and spec_decode_common_attn_metadata is None: if (
self.speculative_config
and spec_decode_common_attn_metadata is None
and hasattr(self, "drafter")
):
if isinstance(self.drafter, EagleProposer): if isinstance(self.drafter, EagleProposer):
if self.drafter.attn_layer_names[0] in kv_cache_group.layer_names: if self.drafter.attn_layer_names[0] in kv_cache_group.layer_names:
spec_decode_common_attn_metadata = cm spec_decode_common_attn_metadata = cm