Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[BugFix][Core] Fix BlockManagerV2 when Encoder Input is None #9103

Merged
merged 60 commits into from
Oct 7, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
60 commits
Select commit Hold shift + click to select a range
5650b95
Merge pull request #1 from vllm-project/main
sroy745 May 29, 2024
8f36146
Merge branch 'vllm-project:main' into main
sroy745 Jun 3, 2024
9e75057
Merge branch 'vllm-project:main' into main
sroy745 Jun 3, 2024
db2c679
Merge branch 'vllm-project:main' into main
sroy745 Jun 7, 2024
8d7512c
Merge branch 'vllm-project:main' into main
sroy745 Jun 10, 2024
1473f74
Merge branch 'vllm-project:main' into main
sroy745 Jun 12, 2024
4013e1a
Merge branch 'vllm-project:main' into main
sroy745 Jun 14, 2024
2dbdd78
Merge branch 'vllm-project:main' into main
sroy745 Jun 17, 2024
b3575e9
Merge branch 'vllm-project:main' into main
sroy745 Jun 20, 2024
94b0d43
Merge branch 'vllm-project:main' into main
sroy745 Jun 24, 2024
fa8fedf
Merge branch 'vllm-project:main' into main
sroy745 Jun 27, 2024
6ed96b4
Merge branch 'vllm-project:main' into main
sroy745 Jun 27, 2024
b71c533
Merge branch 'vllm-project:main' into main
sroy745 Jun 28, 2024
57babef
Merge branch 'vllm-project:main' into main
sroy745 Jun 29, 2024
4b19bac
Merge branch 'vllm-project:main' into main
sroy745 Jul 1, 2024
eb7a1c4
Merge branch 'vllm-project:main' into main
sroy745 Jul 6, 2024
7e2c87e
Merge branch 'vllm-project:main' into main
sroy745 Jul 10, 2024
6212d5f
Merge branch 'vllm-project:main' into main
sroy745 Jul 15, 2024
5491438
Merge branch 'vllm-project:main' into main
sroy745 Jul 17, 2024
68e080a
Merge branch 'vllm-project:main' into main
sroy745 Jul 31, 2024
55e4332
Merge branch 'vllm-project:main' into main
sroy745 Aug 13, 2024
532eb48
Merge branch 'vllm-project:main' into main
sroy745 Aug 22, 2024
7cea056
Merge branch 'vllm-project:main' into main
sroy745 Aug 22, 2024
185e056
Merge branch 'vllm-project:main' into main
sroy745 Aug 24, 2024
e2be95f
Merge branch 'vllm-project:main' into main
sroy745 Aug 27, 2024
2ed5473
Merge branch 'vllm-project:main' into main
sroy745 Aug 28, 2024
efa4714
Merge branch 'vllm-project:main' into main
sroy745 Aug 29, 2024
fb87d34
Merge branch 'vllm-project:main' into main
sroy745 Aug 29, 2024
5419e49
Merge branch 'vllm-project:main' into main
sroy745 Aug 31, 2024
9ba12f8
Merge branch 'vllm-project:main' into main
sroy745 Sep 2, 2024
25cef3d
Merge branch 'vllm-project:main' into main
sroy745 Sep 3, 2024
9d4cd09
Merge branch 'vllm-project:main' into main
sroy745 Sep 4, 2024
c48cacb
Merge branch 'vllm-project:main' into main
sroy745 Sep 5, 2024
c42c399
Merge branch 'vllm-project:main' into main
sroy745 Sep 7, 2024
3d13e43
Merge branch 'vllm-project:main' into main
sroy745 Sep 9, 2024
7479775
Merge branch 'vllm-project:main' into main
sroy745 Sep 11, 2024
df9b966
Merge branch 'vllm-project:main' into main
sroy745 Sep 17, 2024
9a7ed92
Merge branch 'vllm-project:main' into main
sroy745 Sep 17, 2024
118e838
Merge branch 'vllm-project:main' into main
sroy745 Sep 19, 2024
e640c69
Merge branch 'vllm-project:main' into main
sroy745 Sep 20, 2024
89fb6cd
Merge branch 'vllm-project:main' into main
sroy745 Sep 23, 2024
5d886cc
Merge branch 'vllm-project:main' into main
sroy745 Sep 24, 2024
56f2065
Merge branch 'vllm-project:main' into main
sroy745 Sep 24, 2024
28e103e
Merge branch 'vllm-project:main' into main
sroy745 Sep 25, 2024
2fc1490
Merge branch 'vllm-project:main' into main
sroy745 Sep 25, 2024
8805750
Merge branch 'vllm-project:main' into main
sroy745 Sep 26, 2024
b30e5af
Merge branch 'vllm-project:main' into main
sroy745 Sep 28, 2024
92322f1
Merge branch 'vllm-project:main' into main
sroy745 Sep 30, 2024
85e9001
Merge branch 'vllm-project:main' into main
sroy745 Oct 1, 2024
cd4ff89
Merge branch 'vllm-project:main' into main
sroy745 Oct 1, 2024
0dd96ed
Merge branch 'vllm-project:main' into main
sroy745 Oct 1, 2024
9d4d969
Merge branch 'vllm-project:main' into main
sroy745 Oct 3, 2024
7d223b5
Merge branch 'vllm-project:main' into main
sroy745 Oct 5, 2024
f327d91
Merge branch 'vllm-project:main' into main
sroy745 Oct 5, 2024
d515360
Fix BlockManager V2 when the encoder input is None
sroy745 Oct 6, 2024
effc48d
Comments
sroy745 Oct 6, 2024
3296228
Comment
sroy745 Oct 6, 2024
2a484cb
Update vllm/core/block_manager_v2.py
sroy745 Oct 6, 2024
56b446d
Dummy
sroy745 Oct 6, 2024
27047a8
Dummy
sroy745 Oct 6, 2024
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 0 additions & 2 deletions vllm/core/block/block_table.py
Original file line number Diff line number Diff line change
Expand Up @@ -220,7 +220,6 @@ def free(self) -> None:
occupied by each block. After freeing all the blocks, the `_blocks` list
is set to `None`.
"""
assert self._is_allocated
for block in self.blocks:
self._allocator.free(block)
self._blocks.reset()
Expand All @@ -239,7 +238,6 @@ def physical_block_ids(self) -> List[int]:
List[int]: A list of physical block indices for the blocks in the
BlockTable.
"""
assert self._is_allocated
return self._blocks.ids()

def get_unseen_token_ids(self, sequence_token_ids: List[int]) -> List[int]:
Expand Down
4 changes: 3 additions & 1 deletion vllm/core/block_manager_v2.py
Original file line number Diff line number Diff line change
Expand Up @@ -151,7 +151,9 @@ def _allocate_sequence(self, seq: Sequence) -> BlockTable:
block_allocator=self.block_allocator,
max_block_sliding_window=self.max_block_sliding_window,
)
block_table.allocate(seq.get_token_ids())
if seq.get_token_ids():
# Add blocks to the block table only if the sequence is non empty.
block_table.allocate(seq.get_token_ids())

return block_table

Expand Down
5 changes: 0 additions & 5 deletions vllm/engine/arg_utils.py
Original file line number Diff line number Diff line change
Expand Up @@ -903,11 +903,6 @@ def create_engine_config(self) -> EngineConfig:
"--enable-prefix-caching is currently not "
"supported for multimodal models and has been disabled.")
self.enable_prefix_caching = False
if model_config.is_encoder_decoder_model:
logger.warning(
"Block Manager v2 does not support encoder-decoder models"
" currently. Using Block Manager v1 as fallback.")
self.use_v2_block_manager = False

cache_config = CacheConfig(
block_size=self.block_size if self.device != "neuron" else
Expand Down
Loading