karylab_agents/vllm
1
0
Fork 0
forked from Karylab-cklius/vllm
Code Pull Requests 2 Actions 1 Packages Activity

using PAD_SLOT_ID

Browse Source 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
yewentao256
2026-02-22 14:57:28 +00:00
parent fb9a6b8da1
commit 9f07a15403
1 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vllm/v1/worker/gpu/model_runner.py
+3 -2
View File
@@ -23,6 +23,7 @@ from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.sequence import IntermediateTensors
from vllm.utils.mem_utils import DeviceMemoryProfiler, format_gib
from vllm.utils.torch_utils import STR_DTYPE_TO_TORCH_DTYPE
from vllm.v1.attention.backends.utils import PAD_SLOT_ID
from vllm.v1.core.sched.output import GrammarOutput, SchedulerOutput
from vllm.v1.kv_cache_interface import KVCacheConfig
from vllm.v1.outputs import DraftTokenIds, ModelRunnerOutput
@@ -726,9 +727,9 @@ class GPUModelRunner(LoRAModelRunnerMixin):
attn_num_reqs = min(num_tokens_after_padding, self.max_num_reqs)
attn_num_tokens = num_tokens_after_padding
if attn_num_reqs > num_reqs:
# fill -1 for the padded blocks
# Fill PAD_SLOT_ID for padded block-table rows.
for input_block_table in self.block_tables.input_block_tables:
input_block_table[num_reqs:attn_num_reqs].fill_(-1)
input_block_table[num_reqs:attn_num_reqs].fill_(PAD_SLOT_ID)
attn_block_tables = tuple(
input_block_table[:attn_num_reqs]
for input_block_table in self.block_tables.input_block_tables