Skip to content

NewRLInfra-3003 add over-context-len prompt tokens into stat #8111

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

Open
rain7996 wants to merge 1 commit into
base: develop
Choose a base branch
from
Open

NewRLInfra-3003 add over-context-len prompt tokens into stat #8111

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
15 changes: 15 additions & 0 deletions benchmarks/benchmark_serving.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -294,6 +294,21 @@ def calculate_metrics(
stacklevel=2,
)

# === 零输出(prefill-only)请求补计 ===
# prompt 超长时服务端(allow_auto_truncate)会把输入截到 max_req_input_len 并把 max_new_tokens
# 压到 0：prefill 真实算过，但回包没有任何 content，客户端判成 "No generated text found!"。
# 这类请求的输入 token 只补进 ITPS(input/total token throughput)分子，不参与
# completed / TTFT / TPOT / OTPS 等其他指标。
prefill_only_reqs = 0
prefill_only_input = 0
for o in outputs:
if o.prompt_tokens > 0 and not o.output_tokens:
prefill_only_reqs += 1
prefill_only_input += o.prompt_tokens
if prefill_only_reqs:
total_input += prefill_only_input
print(f"零输出(prefill-only)请求: {prefill_only_reqs} 条, " f"补入 ITPS 的输入 token: {prefill_only_input}")

# === 解码速度过滤：TPOT < 1ms 的请求视为不可信(引擎批量flush伪象) ===
MIN_TPOT_S = 0.001 # 1ms
reliable_s_decodes = []
Expand Down
Loading