周周周 d957ccd46d seq_lens related tensor shape -> [max_num_seqs] (#6535)

2026-03-02 11:18:30 +08:00

..

__init__.py

[Optimization] Support FA2/FA3/FA4 with attn_mask_q (#6354)

2026-02-05 14:39:00 +08:00

append_attention.py

[Feature]Supports SWA based on appendattn (#6547)

2026-03-01 19:02:08 +08:00

flash_mask_attention.py

seq_lens related tensor shape -> [max_num_seqs] (#6535)

2026-03-02 11:18:30 +08:00

get_attn_mask_q.py

[Optimization] Support FA2/FA3/FA4 with attn_mask_q (#6354)

2026-02-05 14:39:00 +08:00

get_block_shape_and_split_kv_block.py

[Others]get_block_shape_and_split_kv_block clean code (#5123)

2025-11-20 16:40:04 +08:00

gqa_rope_write_cache.py

FA3 support qwen3 (#5441)

2025-12-09 16:16:16 +08:00

init_kv_signal_per_query.py

[PD Disaggregation][XPU] Add XPU support for PD disaggregation (#5113)

2025-11-21 14:09:01 +08:00

init_signal_layerwise.py

[PD Disaggregation][XPU] Add XPU support for PD disaggregation (#5113)

2025-11-21 14:09:01 +08:00

open_shm_and_get_meta_signal.py

[PD Disaggregation][XPU] Add XPU support for PD disaggregation (#5113)

2025-11-21 14:09:01 +08:00

pre_cache_len_concat.py

[Others] Remove useless code (#5404)

2025-12-08 13:59:46 +08:00