xc-llm-kunlun

Files

Xinyu Dong bf9369f733 Migrate XTorch operations to Kunlun operations (accelerating iteration) (#177 )

Signed-off-by: dongxinyu03 <dongxinyu03@baidu.com>

2026-02-12 18:13:00 +08:00

__init__.py

2025-12-10 17:51:24 +08:00

chunk_delta_h.py

2025-12-10 17:51:24 +08:00

chunk_o.py

2025-12-10 17:51:24 +08:00

chunk_scaled_dot_kkt.py

2025-12-10 17:51:24 +08:00

chunk.py

2026-01-05 10:24:51 +08:00

cumsum.py

2025-12-10 17:51:24 +08:00

fused_recurrent.py

2026-02-12 18:13:00 +08:00

index.py

2025-12-10 17:51:24 +08:00

l2norm.py

2026-02-12 18:13:00 +08:00

layernorm_guard.py

2025-12-10 17:51:24 +08:00

op.py

2025-12-10 17:51:24 +08:00

solve_tril.py

2025-12-10 17:51:24 +08:00

torch_fla.py

2025-12-10 17:51:24 +08:00

utils.py

2025-12-10 17:51:24 +08:00

wy_fast.py

2025-12-10 17:51:24 +08:00