Skip to content

[None][fix] Remove overlap scheduler adjustment for max sequence length in create_py_executor function #9229

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

Draft
Funatiq wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

[None][fix] Remove overlap scheduler adjustment for max sequence length in create_py_executor function #9229

Show file tree
Hide file tree
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
3 changes: 1 addition & 2 deletions tensorrt_llm/_torch/pyexecutor/model_engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -423,8 +423,7 @@ def __init__(
# This way it can also be used for CUDA graphs.
if self.use_beam_search:
self.cache_indirection_attention = torch.zeros(
(self.batch_size, self.max_beam_width, self.max_seq_len +
(0 if self._disable_overlap_scheduler else 1)),
(self.batch_size, self.max_beam_width, self.max_seq_len),
device="cuda",
dtype=torch.int32)
else:
Expand Down
4 changes: 0 additions & 4 deletions tensorrt_llm/_torch/pyexecutor/py_executor_creator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -403,10 +403,6 @@ def drafting_loop_wrapper(model):
# PyTorchModelEngine modifies these fields, update them
model_engine_max_seq_len = model_engine.max_seq_len
net_max_seq_len = model_engine_max_seq_len
if not llm_args.disable_overlap_scheduler:
model_engine_max_seq_len = model_engine.max_seq_len + 1
if spec_config is not None:
model_engine_max_seq_len += spec_config.max_total_draft_tokens

if spec_config is not None:
model_engine_max_seq_len += get_num_extra_kv_tokens(spec_config)
Expand Down
Loading