feat: Support speculative decoding with Hopper XQA - #3269
Merged
Conversation
lowsfer
force-pushed
the
user/yaoy/hopper-xqa-spec-dec
branch
from
April 3, 2025 10:36
a577b6f to
cb709e3
Compare
juney-nvidia
changed the title
lowsfer
force-pushed
the
user/yaoy/hopper-xqa-spec-dec
branch
from
April 7, 2025 01:32
cb709e3 to
247079b
Compare
ming-wei
approved these changes
Apr 7, 2025
lowsfer
force-pushed
the
user/yaoy/hopper-xqa-spec-dec
branch
from
April 7, 2025 06:33
247079b to
27b48bb
Compare
Member
Author
|
/bot run --disable-fail-fast |
Collaborator
|
PR_Github #1273 [ run ] triggered by Bot |
symphonylyh
approved these changes
Apr 7, 2025
Collaborator
|
PR_Github #1273 [ run ] completed with state |
lowsfer
enabled auto-merge (squash)
April 7, 2025 08:37
lowsfer
disabled auto-merge
April 7, 2025 08:37
lowsfer
enabled auto-merge (squash)
April 7, 2025 08:38
lowsfer
force-pushed
the
user/yaoy/hopper-xqa-spec-dec
branch
from
April 7, 2025 08:38
27b48bb to
755f1be
Compare
Member
Author
|
/bot reuse-pipeline --comment "rebase" |
Collaborator
|
PR_Github #1301 [ reuse-pipeline ] triggered by Bot |
Collaborator
|
PR_Github #1301 [ reuse-pipeline ] completed with state |
Signed-off-by: Yao Yao <lowsfer@users.noreply.github.com>
lowsfer
force-pushed
the
user/yaoy/hopper-xqa-spec-dec
branch
from
April 7, 2025 08:59
755f1be to
db38872
Compare
Member
Author
|
/bot reuse-pipeline --comment "rebase" |
Collaborator
|
PR_Github #1308 [ reuse-pipeline ] triggered by Bot |
Collaborator
|
PR_Github #1308 [ reuse-pipeline ] completed with state |
sarattha
pushed a commit
to sarattha/TensorRT-LLM
that referenced
this pull request
Apr 9, 2025
Signed-off-by: Yao Yao <lowsfer@users.noreply.github.com> Signed-off-by: sarattha <j.sarattha@gmail.com>
wu1du2
pushed a commit
to wu1du2/TensorRT-LLM
that referenced
this pull request
May 11, 2025
Signed-off-by: Yao Yao <lowsfer@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Before this change, speculative decoding is supported with Ampere XQA on Hopper. This change make Hopper XQA also support speculative decoding for better perf on Hopper GPUs.