frozenleaves/vllm
1
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2025-10-20 23:03:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Misc][Minor] Fix CPU block num log in CPUExecutor. (#4088)

Browse Source
This commit is contained in:
Li, Jiang
2024-04-15 23:35:55 +08:00
committed by GitHub
parent e11e200736
commit 0003e9154b
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/executor/cpu_executor.py
View File

@ -74,7 +74,10 @@ class CPUExecutor(ExecutorBase):
# NOTE: We log here to avoid multiple logs when number of workers is # NOTE: We log here to avoid multiple logs when number of workers is
# greater than one. We could log in the engine, but not all executors # greater than one. We could log in the engine, but not all executors
# have GPUs. # have GPUs.
logger.info(f"# CPU blocks: {num_cpu_blocks}") # NOTE: `cpu block` for CPU backend is located on CPU memory but is
# referred as `gpu block`. Because we want to reuse the existing block
# management procedure.
logger.info(f"# CPU blocks: {num_gpu_blocks}")
self.driver_worker.initialize_cache(num_gpu_blocks, num_cpu_blocks) self.driver_worker.initialize_cache(num_gpu_blocks, num_cpu_blocks)
def execute_model(self, def execute_model(self,