Skip to content
Merged
Changes from 1 commit
Commits
Show all changes
48 commits
Select commit Hold shift + click to select a range
e8e96b1
allow shared prefix question and system prompt variance and calculate…
kaushikmitr Dec 8, 2025
fa0dd37
fix bug
kaushikmitr Dec 9, 2025
6baa036
fix shared prefix bug
kaushikmitr Dec 9, 2025
e724dc9
fix unit tests
kaushikmitr Jan 29, 2026
a3fb970
fix test errors
kaushikmitr Jan 29, 2026
eb2d9e2
fix completion.py to_payload method
kaushikmitr Jan 29, 2026
7fc8965
fix completion.py to_payload method 2
kaushikmitr Jan 29, 2026
179dbda
fix lint error
kaushikmitr Jan 29, 2026
c4a89be
fix lint error 2
kaushikmitr Jan 29, 2026
e63d45a
fix lint error 3
kaushikmitr Jan 29, 2026
f317737
fix lint error 4
kaushikmitr Jan 29, 2026
a6bbd5b
make slo headers configurable
kaushikmitr Jan 31, 2026
b796162
make slo headers configurable 2
kaushikmitr Jan 31, 2026
ffc0836
make slo headers configurable 3
kaushikmitr Jan 31, 2026
69e417f
make slo headers configurable 4
kaushikmitr Jan 31, 2026
58ec725
make slo headers configurable 5
kaushikmitr Jan 31, 2026
c645f49
make slo headers configurable 6
kaushikmitr Jan 31, 2026
e8af6ba
make slo headers configurable 7
kaushikmitr Jan 31, 2026
c16a494
make slo headers configurable 8
kaushikmitr Jan 31, 2026
175306f
make slo headers configurable 9
kaushikmitr Jan 31, 2026
e9e42df
precalculate ntpot in RequestLifecycleMetric
kaushikmitr Jan 31, 2026
8db50d5
update docs
kaushikmitr Feb 1, 2026
32dd309
update docs 1
kaushikmitr Feb 1, 2026
69616b5
update docs 3
kaushikmitr Feb 1, 2026
0ce006a
update docs 4
kaushikmitr Feb 1, 2026
91172d3
calculate tpot, ntpot, ttft, and slo metrics in post processing and g…
kaushikmitr Feb 7, 2026
c5db12e
calculate tpot, ntpot, ttft, and slo metrics in post processing and g…
kaushikmitr Feb 7, 2026
a12a7e6
fix rebase errors
kaushikmitr Feb 7, 2026
5240ac9
fix rebase errors 1
kaushikmitr Feb 7, 2026
369b4cd
fix rebase errors 2
kaushikmitr Feb 7, 2026
f1ad486
fix rebase errors 3
kaushikmitr Feb 7, 2026
800b490
fix rebase errors 4
kaushikmitr Feb 7, 2026
bfb15b9
fix rebase errors 5
kaushikmitr Feb 7, 2026
3788ed5
fix rebase errors 6
kaushikmitr Feb 7, 2026
dd850c3
fix rebase errors 7
kaushikmitr Feb 7, 2026
74db16d
fix rebase errors 8
kaushikmitr Feb 7, 2026
f507467
fix rebase errors 9
kaushikmitr Feb 7, 2026
cd10f95
fix rebase errors 10
kaushikmitr Feb 7, 2026
d7acb61
fix rebase errors 11
kaushikmitr Feb 7, 2026
93d4c80
fix rebase errors 12
kaushikmitr Feb 7, 2026
74927a2
fix rebase errors 13
kaushikmitr Feb 7, 2026
ffe7635
fix rebase errors 14
kaushikmitr Feb 8, 2026
fc35bbe
fix rebase errors 15
kaushikmitr Feb 8, 2026
c8be278
fix rebase errors 16
kaushikmitr Feb 8, 2026
0ba9ace
fix rebase errors 17
kaushikmitr Feb 8, 2026
e7023e8
restore shared prefix defaults
kaushikmitr Feb 11, 2026
094ebc7
restore shared prefix defaults 1
kaushikmitr Feb 12, 2026
89e4cb3
use dist class for shared prefix
kaushikmitr Feb 12, 2026
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
Prev Previous commit
Next Next commit
restore shared prefix defaults 1
  • Loading branch information
kaushikmitr committed Feb 12, 2026
commit 094ebc726634037a24cf3a32c575392105e2e4ca
16 changes: 8 additions & 8 deletions inference_perf/config.py
Original file line number Diff line number Diff line change
Expand Up @@ -85,24 +85,24 @@ class SharedPrefix(BaseModel):
system_prompt_len: int = 100
question_len: int = 50
question_len_std: float = 0 # Variation in question length within a group
Comment thread
jjk-g marked this conversation as resolved.
Outdated
question_len_min: Optional[int] = None
question_len_max: Optional[int] = None
question_len_min: int = -1 # Set dynamically by validator if not provided
question_len_max: int = -1 # Set dynamically by validator if not provided

output_len: int = 50
output_len_std: float = 0 # Variation in output length within a group
output_len_min: Optional[int] = None
output_len_max: Optional[int] = None
output_len_min: int = -1 # Set dynamically by validator if not provided
output_len_max: int = -1 # Set dynamically by validator if not provided
enable_multi_turn_chat: bool = False

@model_validator(mode="after")
def set_dynamic_defaults(self) -> "SharedPrefix":
if self.question_len_min is None:
if self.question_len_min == -1:
self.question_len_min = self.question_len if self.question_len_std == 0 else 1
if self.question_len_max is None:
if self.question_len_max == -1:
self.question_len_max = self.question_len if self.question_len_std == 0 else 32768
if self.output_len_min is None:
if self.output_len_min == -1:
self.output_len_min = self.output_len if self.output_len_std == 0 else 1
if self.output_len_max is None:
if self.output_len_max == -1:
self.output_len_max = self.output_len if self.output_len_std == 0 else 32768
return self

Expand Down