Skip to content

Model: Skip standalone load for drafters that load with their target #515

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
turboderp merged 1 commit into from
Nov 3, 2026
+6 −1
Merged

Model: Skip standalone load for drafters that load with their target #515

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
7 changes: 6 additions & 1 deletion backends/exllamav3/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1192,7 +1192,12 @@ def load_model_sync(self, progress_callback=None):
if value:
yield value

if self.use_draft_model:
if self.draft_model is not None and self.draft_model.caps.get("load_with_target"):
xlogger.info(
f"Draft model {self.draft_model_dir} loads with its target model, "
"skipping its standalone load"
)
elif self.use_draft_model:
for value in self.draft_model.load_gen(
reserve_per_device=(
None if (self.gpu_split or self.draft_gpu_split) else self.autosplit_reserve
Expand Down