woutermans 69b284b8b7 Fix PR integration: token prep slots, input ordering and output lifetime

2026-10-07 22:39:37 +02:00

..

rrlite_dec1_4x4.hip

Token-lane block for the RDNA4 native-FP8 layers: enc1, dec1 and the enc3 tube

2026-10-06 22:10:41 +02:00

rrlite_dec2_4x4.hip

Merge PR #26: RDNA4 token-lane Swin and hardware FP8 texture tails

2026-10-07 22:26:30 +02:00

rrlite_enc1_4x4.hip

Token-lane block for the RDNA4 native-FP8 layers: enc1, dec1 and the enc3 tube

2026-10-06 22:10:41 +02:00

rrlite_enc2_4x4.hip

Token lanes for enc2 and dec2 too

2026-10-06 23:00:03 +02:00

rrlite_enc3_tube_4x4.hip

Merge PR #26: RDNA4 token-lane Swin and hardware FP8 texture tails

2026-10-07 22:26:30 +02:00

swin_block.h

Merge PR #26: RDNA4 token-lane Swin and hardware FP8 texture tails

2026-10-07 22:26:30 +02:00

swin_common.h

Merge PR #26: RDNA4 token-lane Swin and hardware FP8 texture tails

2026-10-07 22:26:30 +02:00

swin_tok.h

Fix PR integration: token prep slots, input ordering and output lifetime

2026-10-07 22:39:37 +02:00