↓ 1 callersMethod_upad_input(self, query_layer, key_layer, value_layer, attention_mask, query_length)
examples/bitnet-1.58b/modeling_bitnet.py:701
↓ 1 callersFunctionassert_bitnet_158_int8xint2_decode_correctness(M, N, K, in_dtype, out_dtype, accum_dtype, fast_decoding=True)
examples/bitnet-1.58b/kernel_benchmark/tilelang_bitnet_158_int8xint2_decode.py:237
↓ 1 callersFunctionassert_bitnet_158_int8xint2_prefill_correctness(M, N, K, in_dtype, out_dtype, accum_dtype, fast_decoding=True)
examples/bitnet-1.58b/kernel_benchmark/tilelang_bitnet_158_int8xint2_prefill.py:360
↓ 1 callersFunctionassert_gemm_codegen(
M,
N,
K,
block_M,
block_N,
block_K,
dtype=T.float16,
accum_dtype=T.float32,
testing/python/webgpu/test_webgpu_codegen.py:34
↓ 1 callersFunctionassert_matmul_codegen(M=1024, N=1024, K=1024, block_M=128, block_N=128, block_K=32)
testing/python/cpu/test_tilelang_cpu_gemm.py:48
↓ 1 callersFunctionassert_tl_matmul_correctness(M, N, K, k_pack=1, a_transposed=False, b_transposed=True)
examples/gemm_fp8/example_tilelang_gemm_amd_fp8_preshuffle.py:169
↓ 1 callersFunctionassert_tl_matmul_correctness(M, N, K, in_dtype, out_dtype, accum_dtype=T.float32, a_transposed=False, b_transposed=True, k_pack=1)
testing/python/amd/test_tilelang_gemm_mfma_intrinsic.py:165
↓ 1 callersFunctionassert_tl_matmul_with_ladder_weight_only_transform_block_reduce_int4_correctness(
M,
N,
K,
in_dtype,
out_dtype,
accum_dtype,
transform_b,
)
examples/dequantize_gemm/example_dequant_gemm_fine_grained.py:357
↓ 1 callersFunctionatomic_add_complicated_parallel_program(K, M, N, block_M, block_N, dtype=T.float32)
testing/python/language/test_tilelang_language_atomic.py:242
↓ 1 callersFunctionatomic_max_program(K, M, N, block_M, block_N, dtype=T.float32)
testing/python/language/test_tilelang_language_atomic.py:418
↓ 1 callersFunctionatomic_memory_order_program(K, M, N, block_M, block_N, dtype=T.float32)
testing/python/language/test_tilelang_language_atomic.py:37
↓ 1 callersFunctionatomic_min_program(K, M, N, block_M, block_N, dtype=T.float32)
testing/python/language/test_tilelang_language_atomic.py:451
↓ 1 callersFunctionauto_tma_store_copy(M, N, block_M, block_N, dtype, threads)
testing/python/language/test_tilelang_language_tma_store.py:103
↓ 1 callersFunctionbench_tilelang(M, N, K, block_M, block_N, block_K, warmup, repeats)
benchmark/matmul_metal/benchmark_matmul_metal.py:65
↓ 1 callersFunctionbenchmark(M=512, N=8192, K=8192, seed=0, warmup=25, rep=100)
examples/deepseek_v4/fp8_fp4_gemm_1d1d_sm100.py:373
↓ 1 callersFunctionbenchmark_fwd(
BATCH: int = 1,
H: int = 8,
N_CTX: int = 512,
N_KV: int = 1024,
D_HEAD: int = 128,
T
examples/deepseek_v4/sparse_attn_fwd_sm90.py:225