IAttention FP8#4209

Draft

narendasan wants to merge 2 commits intomainfrom

narendasan/quantization_fixes

Commits on Apr 23, 2026

feat: Adding the scale normalize flag in IAttention layer using a pass to annotate target nodes
narendasan
committed
fix add a pass to insert softmax quantization node for decomposed attention
narendasan
committed