Skip to content

[#9147][feat] AutoDeploy: Draft Target Speculative Decoding #9275

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 4 commits into
base: main
Choose a base branch
from
+744 −195
Open

[#9147][feat] AutoDeploy: Draft Target Speculative Decoding #9275

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
e1c4a68
DraftTarget speculative decoding support in AutoDeploy. Target model …
govind-ramnarayan Nov 11, 2025
54dc4d3
fixed some unit tests
govind-ramnarayan Nov 20, 2025
26099fa
fixed integration test model path and went over _util.py to make sure…
govind-ramnarayan Nov 20, 2025
c66fca2
fix bug with extra arg to kv cache manager
govind-ramnarayan Nov 20, 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
5 changes: 5 additions & 0 deletions tensorrt_llm/_torch/auto_deploy/llm_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -185,6 +185,11 @@ class AutoDeployConfig(DynamicYamlMixInForSettings, BaseSettings):
),
)

draft_checkpoint_loader: Optional[object] = Field(
default=None,
description="The checkpoint loader to use for the draft model when using speculative decoding with two models.",
)

### SEQUENCE INTERFACE CONFIG ##################################################################
max_input_len: int = Field(default=1024, description="The maximum input length.")
max_num_tokens: Optional[int] = Field(default=None, description="The maximum number of tokens.")
Expand Down
Loading