↓ 1 callersFunctionflashattn(batch_size, groups, UQ, UKV, heads, dim, is_causal, block_M=64, block_N=64, num_stages=1, threads=128)
examples/flash_attention/example_gqa_fwd_varlen.py:17
↓ 1 callersFunctionflashattn(batch, heads, kv_head_num, seqlen_kv, dim, pe_dim, block_N, block_H, num_split)
examples/warp_specialize/example_warp_specialize_flashmla.py:11
↓ 1 callersFunctionflashattn(batch, heads, kv_head_num, seqlen_kv, dim, pe_dim, block_N, block_H, num_split, softmax_scale)
examples/deepseek_mla/example_mla_decode_ws.py:28
↓ 1 callersFunctionflashattn(batch, heads, kv_head_num, seqlen_kv, dim, pe_dim, block_N, block_H, num_split)
examples/deepseek_mla/example_mla_decode_persistent.py:17
↓ 1 callersFunctionflashattn(batch, heads, kv_head_num, seqlen_kv, dim, pe_dim, block_N, block_H)
examples/deepseek_mla/experimental/example_mla_decode_kv_fp8.py:16
↓ 1 callersFunctionflashattn_bwd(
batch,
heads,
seq_len,
dim,
is_causal,
groups,
block_M: int,
block_N: int,
examples/amd/example_amd_flash_attn_bwd.py:309
↓ 1 callersFunctionflashattn_bwd(
batch_size,
groups,
UQ,
UKV,
N_CTX,
heads,
max_seq_len,
dim,
is_causal,
examples/attention_sink/example_gqa_sink_bwd_varlen.py:301
↓ 1 callersFunctionflashattn_bwd_atomic_add(batch, heads, seq_len, dim_qk, dim_v, is_causal, block_M, block_N, threads=256, num_stages=2, groups=1)
examples/flash_attention/example_gqa_bwd_tma_reduce.py:171
↓ 1 callersFunctionflashattn_bwd_atomic_add(batch, heads, seq_len, dim_qk, dim_v, is_causal, block_M, block_N, threads=256, num_stages=2, groups=1)
examples/flash_attention/example_gqa_bwd.py:156
↓ 1 callersFunctionflashattn_bwd_atomic_add(
batch, total_q, total_kv, N_CTX, heads, max_seq_len, dim_qk, dim_v, is_causal, block_M, block_N, threads
examples/flash_attention/example_gqa_bwd_tma_reduce_varlen.py:236
↓ 1 callersFunctionflashattn_bwd_dsink(batch_size, heads, N_CTX, max_seq_len, block=256, dtype: T.dtype = T.float16)
examples/attention_sink/example_gqa_sink_bwd_varlen.py:464
↓ 1 callersFunctionflashattn_bwd_postprocess(batch, heads, head_kv, seq_len, dim_qk, dim_v)
examples/flash_attention/example_gqa_bwd_tma_reduce.py:133
↓ 1 callersFunctionflashattn_bwd_postprocess(total_q, total_kv, heads, head_kv, dim_qk, dim_v)
examples/flash_attention/example_gqa_bwd_tma_reduce_varlen.py:198
↓ 1 callersFunctionflashattn_bwd_preprocess(batch_size, heads, UQ, N_CTX, max_seq_len, dim, dtype: T.dtype = T.float16)
examples/attention_sink/example_gqa_sink_bwd_varlen.py:223
↓ 1 callersFunctionflashattn_bwd_split_novarlen(batch, heads, seq_len, dim_qk, dim_v, is_causal, block_M, block_N, threads=256, num_stages=2, groups=1)
examples/flash_attention/example_gqa_bwd_tma_reduce.py:258
↓ 1 callersFunctionflashattn_fwd(batch, heads, seq_len, dim, is_causal, block_M, block_N)
examples/flash_attention/example_mha_bwd_bhsd.py:15
↓ 1 callersFunctionflashattn_fwd(batch, heads, seq_len, dim_qk, dim_v, is_causal, block_M, block_N, groups=1)
examples/flash_attention/example_gqa_bwd_tma_reduce.py:15
↓ 1 callersFunctionflashattn_fwd(batch, heads, seq_len, dim_qk, dim_v, is_causal, block_M, block_N, groups=1)
examples/flash_attention/example_gqa_bwd.py:14
↓ 1 callersFunctionflashattn_fwd(batch, heads, seq_len, dim, is_causal, block_M, block_N)
examples/flash_attention/example_mha_bwd_bshd.py:15
↓ 1 callersFunctionflashattn_fwd(batch, total_q, total_kv, N_CTX, heads, max_seq_len, dim_qk, dim_v, is_causal, block_M, block_N, groups=1)
examples/flash_attention/example_gqa_bwd_tma_reduce_varlen.py:29
↓ 1 callersFunctionflashattn_fwd(
batch_size,
groups,
UQ,
UKV,
N_CTX,
heads,
max_seq_len,
dim,
is_causal,
examples/attention_sink/example_gqa_sink_bwd_varlen.py:29
↓ 1 callersFunctionflashattn_sink(
batch_size,
groups,
UQ,
UKV,
heads,
dim,
is_causal,
window_size=None, # Non
examples/attention_sink/example_gqa_sink_fwd_varlen.py:24
↓ 1 callersFunctionfp8_matmul(M, N, K, block_M, block_N, block_K, num_stages, num_threads, k_pack, gemm_type)
examples/gemm_fp8/example_tilelang_gemm_amd.py:57
↓ 1 callersFunctionfp8_native_block_mean_pooling(
K,
KScale,
dim: int = 128,
pooling_block_size: int = 128,
block_N: int = 64,
num_sta
examples/dsa_hisa/fp8_block_mean_pooling.py:12
↓ 1 callersFunctionfp8_native_block_sparse_mqa_attn_return_logits(
IndexQ,
IndexK,
IndexKScale,
TopKBlockIndex,
Weights,
CuSeqLenKS,
CuSeqLenKE,
examples/dsa_hisa/block_sparse_mqa_fp8.py:14
↓ 1 callersMethodfrom_ptr(
cls, pointer_var: Var, shape: ShapeType, dtype: DType = "float32", strides: tuple[PrimExpr, ...]
tilelang/language/proxy.py:242
↓ 1 callersFunctiongemm_ptr(
A: T.ptr,
B: T.ptr,
C: T.ptr,
M: int,
N: int,
K: int,
testing/python/language/test_tilelang_language_eager_jit.py:49
↓ 1 callersFunctiongemv_simt(
M: int,
N: int,
K: int,
in_dtype: str,
out_dtype: str,
accum_dtype: str,
trans_A
testing/python/kernel/test_tilelang_kernel_gemv_simt.py:14