xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-06-05 11:54:26 +08:00
Code Issues Packages Projects Releases Wiki Activity

Increase #blocks

Browse Source
This commit is contained in:
Woosuk Kwon 2024-04-24 08:56:58 +00:00
parent f42b4c27d8
commit 5323969fcf
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/worker/tpu_worker.py
View File

@ -64,7 +64,7 @@ class TPUWorker(LoraNotSupportedWorkerBase):
self.model_runner.load_model() self.model_runner.load_model()
def determine_num_available_blocks(self) -> Tuple[int, int]: def determine_num_available_blocks(self) -> Tuple[int, int]:
num_tpu_blocks = 100 num_tpu_blocks = 1000
return num_tpu_blocks, 0 return num_tpu_blocks, 0
def initialize_cache( def initialize_cache(