PR #25319 [CPU] GQA supports attention scores output - SemanticDiff

Commits

Allow GQA to output attention scores

derdeljan-msft committed 167 days ago
Fix docs

derdeljan-msft committed 167 days ago
Update attribute comments and docs

derdeljan-msft committed 166 days ago
Fix docs pipeline

derdeljan-msft committed 166 days ago
Update parameter ordering

derdeljan-msft committed 163 days ago
Update shape of output QK

derdeljan-msft committed 163 days ago
Fix shape inference

derdeljan-msft committed 163 days ago
more shape guards

derdeljan-msft committed 162 days ago
Fix qk_output attr default value

derdeljan-msft committed 162 days ago
fix docs

derdeljan-msft committed 162 days ago
fix PR comments

derdeljan-msft committed 162 days ago
fix docs

derdeljan-msft committed 161 days ago

Loading