Skip to content

[feat] Implement model-agnostic one-engine eagle3 #4778

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
merged 4 commits into from
Jun 13, 2025
Merged

[feat] Implement model-agnostic one-engine eagle3 #4778

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
36197eb
[feat] Implement model-agnostic one-engine eagle3
nv-yilinf May 29, 2025
33b16a0
Generalize for future SpecDec implementations
nv-yilinf Jun 4, 2025
2cd7f79
Fix UT
nv-yilinf Jun 9, 2025
51c4c90
Fix UT
nv-yilinf Jun 13, 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
2 changes: 2 additions & 0 deletions tensorrt_llm/_torch/models/modeling_auto.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,8 @@ def from_config(
# Hack to detect eagle3 checkpoints. TODO: should we provide
# our own checkpoints with the correct arch? It would let us
# avoid nasty stuff like this.
model_arch = model_arch.replace("Eagle3",
"") # Strip the appended EAGLE3
if hasattr(config.pretrained_config, "draft_vocab_size"):
model_arch = "EAGLE3" + model_arch

Expand Down
Loading