sparse_swa.py v1/attention/backends/mla/sparse_swa.py
attention.py models/deepseek_v4_1/attention.py
flashinfer_sparse.py models/deepseek_v4_1/nvidia/flashinfer_sparse.py
engram.py models/deepseek_v4_1/common/engram.py
weight_utils.py model_executor/model_loader/weight_utils.py
model_state.py models/deepseek_v4_1/nvidia/model_state.py
sparse_attn_indexer.py model_executor/layers/sparse_attn_indexer.py
kv_cache_interface.py v1/kv_cache_interface.py
kv_cache_utils.py v1/core/kv_cache_utils.py
kv_cache_coordinator.py v1/core/kv_cache_coordinator.py
gpu_block_table.py v1/worker/gpu/block_table.py
gpu_model_runner.py v1/worker/gpu/model_runner.py
gpu_attn_utils.py v1/worker/gpu/attn_utils.py
v1_attention_backend.py v1/attention/backend.py
compressor.py models/deepseek_v4_1/compressor.py
compressor_utils.py v1/attention/backends/mla/compressor_utils.py
mla_indexer.py v1/attention/backends/mla/indexer.py
candidate_blocks.py model_executor/kernels/attention/dsa/candidate_blocks.py
v41_sparse_mla.py models/deepseek_v4_1/sparse_mla.py
worker_cp_utils.py v1/worker/cp_utils.py
dflash_speculator.py v1/worker/gpu/spec_decode/dflash/speculator.py
fused_moe_b12x.py model_executor/layers/fused_moe/b12x.py
communication_op.py distributed/communication_op.py