Skip to content

[Bugfix][TPU] Fix V1 TPU worker for sliding window #16059

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.

Sign up for GitHub

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

Merged
merged 1 commit into from
Apr 4, 2025
Merged

[Bugfix][TPU] Fix V1 TPU worker for sliding window #16059

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
7 changes: 4 additions & 3 deletions vllm/v1/worker/tpu_worker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@
from vllm.model_executor import set_random_seed
from vllm.utils import STR_DTYPE_TO_TORCH_DTYPE
from vllm.v1.core.sched.output import SchedulerOutput
from vllm.v1.kv_cache_interface import (FullAttentionSpec, KVCacheConfig,
from vllm.v1.kv_cache_interface import (AttentionSpec, KVCacheConfig,
KVCacheSpec)
from vllm.v1.outputs import ModelRunnerOutput
from vllm.v1.utils import bind_kv_cache
Expand Down Expand Up @@ -137,7 +137,7 @@ def determine_available_memory(self) -> int:
kv_caches: dict[str, torch.Tensor] = {}
kv_cache_spec = self.model_runner.get_kv_cache_spec()
for layer_name, layer_spec in kv_cache_spec.items():
if isinstance(layer_spec, FullAttentionSpec):
if isinstance(layer_spec, AttentionSpec):
dtype = layer_spec.dtype

# Use an empty tensor instead of `None`` to force Dynamo to pass
Expand All @@ -147,7 +147,8 @@ def determine_available_memory(self) -> int:
device=self.device)
kv_caches[layer_name] = tpu_kv_cache
else:
raise NotImplementedError
raise NotImplementedError(
f"Unsupported KV cache spec '{type(layer_spec)}'")

runner_kv_caches: list[torch.Tensor] = []
bind_kv_cache(
Expand Down