biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Core] Deprecating block manager v1 and make block manager v2 default (#8704)

Browse Source 
Removing the block manager v1. This is the initial piece of prefix-caching-centric design. In order to achieve prefix-caching-centric design, we need to simplify the code path so that we only use v2 block manager (which has much higher performance on prefix caching).
This commit is contained in:
Kuntai Du
2024-10-17 11:38:15 -05:00
committed by GitHub
parent 5eda21e773
commit 81ede99ca4
45 changed files with 206 additions and 2109 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/spec_decode/e2e/test_integration_dist_tp2.py
View File

@@ -17,9 +17,6 @@ from .conftest import run_equality_correctness_test_tp
[[
# Skip cuda graph recording for fast test.
"--enforce-eager",
# Required for spec decode.
"--use-v2-block-manager",
"--tensor-parallel-size",
"2"
]])
@@ -74,9 +71,6 @@ def test_target_model_tp_gt_1(common_llm_kwargs, per_test_common_llm_kwargs,
[[
# Skip cuda graph recording for fast test.
"--enforce-eager",
# Required for spec decode.
"--use_v2_block_manager",
"--tensor_parallel_size",
"2",