EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

[CI][Router] Fix bench_one_batch_server for pd router test (#7731)

Browse Source 
Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
This commit is contained in:
Shangming Cai
2025-07-03 14:18:24 +08:00
committed by GitHub
parent 84f2e4a0f8
commit 2ff572e28c
2 changed files with 6 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
.github/workflows/pr-test-pd-router.yml vendored
View File

@@ -114,7 +114,7 @@ jobs:
run: |
echo "Installing SGLang with all extras..."
python3 -m pip --no-cache-dir install -e "python[all]" --break-system-packages
python3 -m pip --no-cache-dir install mooncake-transfer-engine==0.3.4.post1
python3 -m pip --no-cache-dir install mooncake-transfer-engine==0.3.4.post2
- name: Build and install sgl-router
run: |

7
python/sglang/bench_one_batch_server.py
View File

@@ -253,8 +253,11 @@ def run_benchmark(server_args: ServerArgs, bench_args: BenchArgs):
else:
proc, base_url = launch_server_process(server_args)
server_info = requests.get(base_url + "/get_server_info")
tokenizer_path = server_info.json()["tokenizer_path"]
server_info = requests.get(base_url + "/get_server_info").json()
if "tokenizer_path" in server_info:
tokenizer_path = server_info["tokenizer_path"]
elif "prefill" in server_info:
tokenizer_path = server_info["prefill"][0]["tokenizer_path"]
tokenizer = get_tokenizer(tokenizer_path)
# warmup