Skip to content
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

Jump to bottom

Fix instruct models UI issue #78

Merged
merged 2 commits into from
Jul 23, 2024
Merged

Fix instruct models UI issue #78

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4247afe
feat(tgi): allow top_k = 0 and top_p = 1 when do_sample = True
tengomucho Jul 23, 2024
e0690aa
chore: update version to 0.1.4
tengomucho Jul 23, 2024
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
5 changes: 0 additions & 5 deletions optimum/tpu/generation/logits_process.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,8 +48,6 @@ def from_config(cls, generation_config: GenerationConfig) -> "FusedLogitsWarper"
Returns:
a `FusedLogitsWarper` or None if neither top-k nor top-p are configured.
"""
if generation_config.do_sample and generation_config.top_k == 0 and generation_config.top_p == 1.0:
raise ValueError("Multinomial sampling requires at least top-k or top-p to be specified.")
return cls(generation_config.temperature, generation_config.top_k, generation_config.top_p)

def __call__(self, logits: torch.FloatTensor) -> Tuple[torch.FloatTensor, torch.LongTensor]:
Expand All @@ -59,9 +57,6 @@ def __call__(self, logits: torch.FloatTensor) -> Tuple[torch.FloatTensor, torch.
do_top_k = self.top_k > 0 and self.top_k < logits.shape[-1]
do_top_p = self.top_p < 1.0 and self.top_p > 0.0

if not do_top_k and not do_top_p:
return logits, None

if do_top_k:
sorted_logits, sorted_indices = torch.topk(logits, self.top_k)
else:
Expand Down
2 changes: 1 addition & 1 deletion optimum/tpu/version.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,5 +15,5 @@
from pkg_resources import parse_version


__version__ = "0.1.3"
__version__ = "0.1.4"
VERSION = parse_version(__version__)
2 changes: 1 addition & 1 deletion text-generation-inference/server/text_generation_server/version.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
from pkg_resources import parse_version


__version__ = "0.1.3"
__version__ = "0.1.4"
VERSION = parse_version(__version__)
Loading