Skip to content

[https://nvbugs/5652062][fix] Rectify the checking rule for finishing a request #9516

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

[https://nvbugs/5652062][fix] Rectify the checking rule for finishing a request #9516

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
13 changes: 11 additions & 2 deletions tensorrt_llm/_torch/pyexecutor/sampler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1559,17 +1559,26 @@ def _are_end_id(self, requests: list[LlmRequest], tokens: torch.Tensor) -> torch
return tokens == end_ids_tensor

def _are_max_length(self, requests: list[LlmRequest]) -> torch.Tensor:
# Check if meeting the below conditions:
# either num_tokens - orig_prompt_len >= max_new_tokens
# or num_tokens >= max_seq_len
lengths_tensor = torch.tensor(
[
[
((req.get_num_tokens(BEAM) + num_tokens) - req.py_orig_prompt_len)
(req.get_num_tokens(BEAM) + num_tokens)
for num_tokens in range(1, self.max_tokens + 1)
]
for req in requests
]
)
max_lengths_tensor = torch.tensor(
[([min(req.py_max_new_tokens, self.max_seq_len)] * self.max_tokens) for req in requests]
[
(
[min(req.py_max_new_tokens + req.py_orig_prompt_len, self.max_seq_len)]
* self.max_tokens
)
for req in requests
]
)
return (
(lengths_tensor >= max_lengths_tensor)
Expand Down
Loading