Skip to content

[CB] add warning when exceeding 32K context length #385

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

Merged
yannicks1 merged 2 commits into from
Aug 15, 2025
Merged

[CB] add warning when exceeding 32K context length #385

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3653154
add warning when exceeding 32K context length
yannicks1 Aug 15, 2025
9907a59
Update vllm_spyre/platform.py
yannicks1 Aug 15, 2025
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
7 changes: 7 additions & 0 deletions vllm_spyre/platform.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -200,6 +200,13 @@ def check_and_update_config(cls, vllm_config: VllmConfig) -> None:
# set env vars for torch_sendnn to consume
os.environ["VLLM_DT_MAX_CONTEXT_LEN"] = str(
vllm_config.model_config.max_model_len)
if (envs_spyre.VLLM_SPYRE_USE_CB
and vllm_config.model_config.max_model_len > 32 * 1024):
logger.warning(
'Max context length too big. Currently only 32K (32768) ' \
Comment thread
yannicks1 marked this conversation as resolved.
Outdated
'context length is supported on Spyre for continuous ' \
'batching. Results might be off!'
)
# min value 2 needed for VLLM_DT_MAX_BATCH_SIZE (compiler constraint)
# Note that we can still have decodes of batch size 1 as the env var
# only concerns the max batch size.
Expand Down