Skip to content

Onboarding Minimax with DP/CP in MSA - #1329

Draft
kdulla wants to merge 19 commits into
quic:mainfrom
kdulla:minimax_v2
Draft

kdulla wants to merge 19 commits into
quic:mainfrom
kdulla:minimax_v2

Conversation

@kdulla

@kdulla kdulla commented Sep 11, 2026

Copy link
Copy Markdown
Contributor

Onboarded Minimax, with new attention structure of MSA - includes indexer and sparse attention.

Added DP/CP support in the model for optimized MSA, including in indexer and sparse attention.

Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
@kdulla kdulla self-assigned this Sep 11, 2026
@kdulla
kdulla marked this pull request as draft September 11, 2026 14:52
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>
…e embeddings

Signed-off-by: Kushal Dulla <kdulla@qti.qualcomm.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant