MCPcopy Create free account

hub / github.com/tile-ai/tilelang / functions

Functions7,254 in github.com/tile-ai/tilelang

Functionpytest_xdist_setupnodes
(config: Config, specs: Iterable[Any])
maint/scripts/pytest_cuda_scheduler.py:182
Functionq_multiply_shift
Execute a multiplication between two Q-numbers x and y followed by a right shift s. The mathematical expression is: out = round(x*y*2^-s)
tilelang/language/tir/op.py:3107
Functionq_multiply_shift_per_axis
Execute a multiplication between two Q-numbers x and y Parameters ---------- x : PrimExpr First Q-number. y : PrimExpr
tilelang/language/tir/op.py:3136
Functionqwq
(A: T.Tensor[(num_tokens,)])
testing/python/issue/test_tilelang_issue_1728.py:13
Functionqwq_kernel
()
testing/python/issue/test_tilelang_issue_1678.py:11
Functionraise_write_error
(*args, **kwargs)
testing/python/cache/test_tilelang_kernel_cache_atomic_save.py:362
Functionraise_write_error
(self, *args, **kwargs)
testing/python/autotune/test_tilelang_autotune_atomic_save.py:101
Functionrand_kernel
( A: T.Tensor((M,), "uint32"), B: T.Tensor((M,), "float32"), C: T.Tensor((M,), "float6
testing/python/language/test_tilelang_language_rand.py:15
Functionrasterization2DColumn
Map block indices to column-major swizzled rasterization coordinates.
tilelang/contrib/cutedsl/threadblock_swizzle.py:65
Functionrasterization2DRow
Map block indices to row-major swizzled rasterization coordinates.
tilelang/contrib/cutedsl/threadblock_swizzle.py:47
Methodraxis_order
(self)
tilelang/carver/roller/hint.py:238
Functionreads
The block buffer region reading statement. Parameters ---------- buffer_slices : List[Union[BufferRegion, BufferLoad]] The array
tilelang/language/ast/ir.py:392
Functionrealize
Create a realization. Parameters ---------- buffer_slice : BufferRegion The region of buffer access. storage_scope : str
tilelang/language/ast/ir.py:952
Functionrecompute_w_u_fwd_kernel
( q, k, qg, kg, v, beta, w, u, A, gk, cu_seqlens, chunk_indice
examples/kda/FLA_KDA/fla_wy_fast.py:28
Functionrecord_exception
(message, *args, **kwargs)
testing/python/cache/test_tilelang_kernel_cache_atomic_save.py:365
Functionrecord_exception
(message, *args, **kwargs)
testing/python/autotune/test_tilelang_autotune_atomic_save.py:104
Functionreduce_absmax
Perform reduce absolute max on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out
tilelang/language/reduce_op.py:232
Functionreduce_abssum
Perform reduce absolute sum on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out
tilelang/language/reduce_op.py:216
Functionreduce_bitand
Perform reduce bitwise-and on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out (
tilelang/language/reduce_op.py:253
Functionreduce_bitor
Perform reduce bitwise-or on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out (t
tilelang/language/reduce_op.py:269
Functionreduce_bitxor
Perform reduce bitwise-xor on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out (
tilelang/language/reduce_op.py:285
Functionreduce_func
( A: T.Tensor((M, N), T.int32), B: T.Tensor((M), T.int32), Output: T.Tensor((M), T.int
testing/python/math/test_math_bitwise_reduce.py:24
Functionreduce_max
Perform reduce max on input buffer, store the result to output buffer Parameters ---------- buffer : Buffer The input buffer.
tilelang/language/reduce_op.py:139
Functionreduce_min
Perform reduce min on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out (tirx.Buf
tilelang/language/reduce_op.py:168
Functionreduce_ref
(Q, K, V, glse, Output_partial, causal)
examples/flash_decoding/example_mha_inference.py:168
Functionreduce_ref
(Q, Q_pe, KV, K_pe, glse, Output_partial)
examples/deepseek_mla/torch_refs.py:69
Functionreduce_sum
Perform reduce sum on input buffer, store the result to output buffer. Args: buffer (tirx.Buffer): The input buffer out (tirx.Buf
tilelang/language/reduce_op.py:190
Functionref_program
(A, B)
maint/gemm/correctness_evaluation_tcgen05.py:76
Functionref_program
(A, B)
maint/gemm/correctness_evaluation_tcgen05_2cta.py:83
Functionref_program
(A, B)
maint/gemm/correctness_evaluation_sm70.py:75
Functionref_program
(A, B)
maint/gemm/correctness_evaluation_tcgen05_fp4_unpacked.py:324
Functionref_program
(A, B)
maint/gemm/correctness_evaluation.py:76
Functionref_program
Argument: cb: (batch, nchunks, ngroups, chunk_size, chunk_size) x: (batch, seqlen, nheads, headdim) dt: (batch, nheads, n
benchmark/mamba2/benchmark_mamba_chunk_scan.py:24
Functionref_program
A reference matrix multiplication program, used to compare performance. Parameters ---------- A : numpy.ndarray The matrix w
benchmark/matmul/benchmark_matmul_sp.py:16
Functionref_program
A reference matrix multiplication program, used to compare performance. Parameters ---------- A : numpy.ndarray The matrix w
benchmark/matmul/benchmark_matmul.py:14
Functionref_program
A reference matrix multiplication program, used to compare performance. Parameters ---------- A : numpy.ndarray The matrix w
benchmark/matmul_fp8/benchmark_matmul.py:14
Functionref_program
(A, B)
examples/gemv/example_gemv.py:10
Functionref_program
Compute the matrix product of A and the transpose of B. A and B are expected to be 2-D tensors where A has shape (M, K) and B has shape (N,
examples/gemm/example_gemm_autotune.py:13
Functionref_program
(A, B)
examples/gemm/example_gemm_persistent.py:86
Functionref_program
(A, B)
examples/gemm/example_gemm_intrinsics.py:157
Functionref_program
Compute the matrix product of A and the transpose of B. A and B are expected to be 2-D tensors where A has shape (M, K) and B has shape (N,
examples/gemm/example_gemm_advanced_autotune.py:21
Functionref_program
Argument: B: (batch, seqlen, ngroups, headdim) x: (batch, seqlen, nheads, headdim) dt: (batch, nheads, nchunks, chunk_siz
examples/linear_attention/example_mamba_chunk_state.py:17
Functionref_program
Argument: cb: (batch, nchunks, ngroups, chunk_size, chunk_size) x: (batch, seqlen, nheads, headdim) dt: (batch, nheads, n
examples/linear_attention/example_mamba_chunk_scan.py:17
Functionref_program
(A, qB)
examples/dequantize_gemm/example_dequant_gemm_fine_grained.py:125
Functionref_program
(A, qB)
examples/dequantize_gemm/example_dequant_gemm_fp4_hopper.py:223
Functionref_program
(A, qB)
examples/dequantize_gemm/example_dequant_gemm_w4a8.py:77
Functionref_program
(Q, K, V, is_causal, groups=1)
examples/amd/example_amd_flash_attn_fwd.py:37
Functionref_program
(Q, K, V, causal)
examples/flash_decoding/example_mha_inference.py:158
Functionref_program
(Q, K, V, is_causal)
examples/flash_attention/example_mha_fwd_bshd.py:97
Functionref_program
(Q, K, V, is_causal, groups=1)
examples/flash_attention/example_gqa_fwd_bshd.py:144
Functionref_program
(Q, K, V, is_causal)
examples/flash_attention/example_mha_fwd_bhsd.py:105
Functionref_program
(x)
examples/norm/rms_norm.py:65
Functionref_program
(q, q_pe, kv, k_pe, glse, Output_partial)
examples/warp_specialize/example_warp_specialize_flashmla.py:316
Functionref_program
CPU reference for forward only; backward ref omitted for brevity.
examples/flash_attention_sm100/gqa_bwd_bshd.py:267
Functionref_program
CPU reference forward (for validation); backward ref not implemented.
examples/flash_attention_sm100/mha_bwd_bshd.py:237
Functionref_program
(q, q_pe, kv, k_pe)
examples/deepseek_mla/example_mla_decode.py:184
Functionref_program
(q, q_pe, kv, k_pe, glse, Output_partial)
examples/deepseek_mla/example_mla_decode_ws.py:512
Functionref_program
(q, q_pe, kv, k_pe, glse, Output_partial)
examples/deepseek_mla/example_mla_decode_persistent.py:139
Functionref_program
(q, q_pe, kv, k_pe)
examples/deepseek_mla/experimental/example_mla_decode_kv_fp8.py:91
Functionref_program
(A, B)
testing/python/components/test_tilelang_pass_config_disable_warp_specialized.py:91
Functionref_program
(A)
testing/python/language/test_tilelang_language_reshape.py:32
Functionref_program
(A, B)
testing/python/language/test_tilelang_language_tma_store.py:94
Functionref_program
(A)
testing/python/language/test_tilelang_language_view.py:38
Functionref_program
(A, B)
testing/python/language/test_tilelang_language_tma_copy.py:94
Functionref_program
(A, B)
testing/python/amd/test_tilelang_test_amd.py:87
Functionref_program
(A, B)
testing/python/amd/test_tilelang_gfx950_copy_async.py:210
Functionref_program
(A, B)
testing/python/analysis/test_tilelang_nested_loop_checker.py:177
Functionref_program
A reference matrix multiplication program, used to compare performance. Parameters ---------- A : torch.Tensor The matrix wi
testing/python/autotune/test_tilelang_autotune.py:8
Functionref_program
(A, B)
testing/python/kernel/test_tilelang_kernel_gemm.py:87
Functionref_program
(A, B)
testing/python/kernel/test_tilelang_kernel_gemm_batched.py:95
Functionref_program_simple
Reference: simple nibble-by-nibble dequantization then matmul.
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_cdna4.py:262
Functionref_program_simple
Compute a BF16 matrix product A · B^T from a quantized B with simple (non-twiddling) dequantization. Converts the quantized tensor `qB` to f
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_hopper.py:426
Functionref_program_simple
Compute a reference BF16 matrix multiply using a simple (non-twiddled) dequantization of qB. Converts the quantized tensor `qB` to full-prec
examples/dequantize_gemm/example_dequant_gemm_bf16_fp4_hopper.py:370
Functionref_program_simple_with_bias
(A, qB, Scale, Bias)
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_cdna4.py:277
Functionref_program_simple_with_bias
Compute a BF16 matrix product A · B^T from a quantized B with simple (non-twiddling) dequantization. Converts the quantized tensor `qB` to f
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_hopper.py:450
Functionref_program_twiddling
Reference: dequantize via bit-twiddling then matmul.
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_cdna4.py:254
Functionref_program_twiddling
Compute A @ B^T where B is reconstructed from bit-twiddled 4-bit quantized data and per-block scales, returning bfloat16 results. Converts t
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_hopper.py:381
Functionref_program_twiddling
Compute reference BF16 matrix multiply using bit-twiddled FP4 quantized B. Converts qB (a bit-twiddled, packed FP4 representation of matrix
examples/dequantize_gemm/example_dequant_gemm_bf16_fp4_hopper.py:349
Functionref_program_twiddling_with_bias
(A, qB, Scale, Bias)
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_cdna4.py:270
Functionref_program_twiddling_with_bias
Compute A @ B^T where B is reconstructed from bit-twiddled 4-bit quantized data and per-block scales, returning bfloat16 results. Converts t
examples/dequantize_gemm/example_dequant_gemm_bf16_mxfp4_hopper.py:403
Functionregister_c_postproc_callback
Decorator for registering C host post-processing callback function. Can be used with or without parentheses: @register_c_postproc_callbac
tilelang/engine/callback.py:121
Functionregister_cuda_postproc_callback
Decorator for registering CUDA post-processing callback function. Can be used with or without parentheses: @register_cuda_postproc_callba
tilelang/engine/callback.py:55
Functionregister_hip_postproc_callback
Decorator for registering HIP post-processing callback function. Can be used with or without parentheses: @register_hip_postproc_callback
tilelang/engine/callback.py:88
Functionregister_metal_postproc_callback
Decorator for registering Metal post-processing callback function. Can be used with or without parentheses: @register_metal_postproc_call
tilelang/engine/callback.py:154
Functionregression
Run entrypoints in the caller module and print a markdown table. This is invoked by many example scripts.
tilelang/testing/perf_regression.py:72
Functionregression_block_sparse_attn_tilelang
()
examples/seer_attention/regression_block_sparse_attn_tilelang.py:5
Functionregression_example_blocksparse_gemm
()
examples/blocksparse_gemm/regression_example_blocksparse_gemm.py:5
Functionregression_example_convolution
()
examples/convolution/regression_example_convolution.py:6
Functionregression_example_convolution_autotune
()
examples/convolution/regression_example_convolution.py:10
Functionregression_example_dequant_gemm_bf16_fp4_hopper
()
examples/dequantize_gemm/regression_example_dequantize_gemm.py:17
Functionregression_example_dequant_gemm_bf16_mxfp4_hopper
()
examples/dequantize_gemm/regression_example_dequantize_gemm.py:21
Functionregression_example_dequant_gemm_fp4_hopper
()
examples/dequantize_gemm/regression_example_dequantize_gemm.py:13
Functionregression_example_dequant_gemm_w4a8
()
examples/dequantize_gemm/regression_example_dequantize_gemm.py:25
Functionregression_example_dequant_gemv_fp16xint4
()
examples/dequantize_gemm/regression_example_dequantize_gemm.py:9
Functionregression_example_dynamic
()
examples/dynamic_shape/regression_example_dynamic.py:5
Functionregression_example_elementwise_add
()
examples/elementwise/regression_example_elementwise.py:5
Functionregression_example_fusedmoe_tilelang
()
examples/fusedmoe/regression_example_fusedmoe.py:5
Functionregression_example_gemm
()
examples/gemm/regression_example_gemm.py:10
Functionregression_example_gemm_intrinsics
()
examples/gemm/regression_example_gemm.py:6
Functionregression_example_gemv
()
examples/gemv/regression_example_gemv.py:5
← previousnext →5,401–5,500 of 7,254, ranked by callers