Public Member Functions
void	mat_mul_transposed (const struct matmul_params *params)

void	mat_mul_accelerator_transposed_fastover_column (const struct matmul_params *params)

void	mat_mul_accelerator_transposed_fastover_column_bias (const struct matmul_params *params)

void	mat_mul_accelerator_untransposed_fastover_column (const struct matmul_params *params)

void	naive_mat_mul_int8 (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_32unroll_over_column (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll_nobias (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll_nobias_batch (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll_nobias_ofp32 (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll_nobias_ofp32_batch (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll_bfp32_ofp32 (const struct matmul_params *params)

void	mat_mul_accelerator_int8_fast_2x2_32unroll_bfp32_ofp32_over_column (const struct matmul_params *params)

void	mat_mul_accelerator_int4_fast (const struct matmul_params *params)

void	mat_mul_accelerator_int4_fast_no_offset (const struct matmul_params *params)

void	mat_mul_accelerator_int8_int4_fast_no_offset (struct matmul_params *params)

void	gemv_accelerator_int8_int4_fast_no_offset (struct matmul_params *params)

void	gemm_accelerator_int8_int4_fast_no_offset (struct matmul_params *params)

void	gemm_accelerator_int8_int4_fast_no_offset_v2 (struct matmul_params *params)

void	cblas_gemm_accelerator_no_offset (struct matmul_params *params)

void	naive_mat_mul_int4 (const struct matmul_params *params)

void	naive_mat_mul_int4_with_offset (const struct matmul_params *params)

void	naive_mat_mul_fp16_int4 (const struct matmul_params *params)

void	mat_mul_cuda (const struct matmul_params *params)

void	gemm_forward_cuda (const struct matmul_params *params, int split_k_iters)

void	gemm_forward_cuda_8splits (const struct matmul_params params, float16_t split_8_buffer)

void	gemm_forward_cuda_half (const struct matmul_params *params, int split_k_iters)

void	gemm_forward_cuda_half_test (const struct matmul_params *params, int split_k_iters)

void	gemv_forward_cuda (const struct matmul_params *params)

The documentation for this class was generated from the following files:

kernels/matmul.h
kernels/avx/matmul_avx_fp32.cc
kernels/avx/matmul_avx_int4.cc
kernels/avx/matmul_avx_int8.cc
kernels/avx/matmul_avx_int8_int4.cc
kernels/cuda/matmul_ref_fp32.cc
kernels/cuda/matmul_ref_int8.cc
kernels/matmul_imp.cc
kernels/matmul_int4.cc
kernels/matmul_int8.cc
kernels/metal/matmul_metal_int4.cc
kernels/metal/matmul_ref_fp32.cc
kernels/metal/matmul_ref_int8.cc
kernels/neon/matmul_neon_fp32.cc
kernels/neon/matmul_neon_int4.cc
kernels/neon/matmul_neon_int4_offset.cc
kernels/neon/matmul_neon_int8_int4.cc
kernels/neon/matmul_ref_int8.cc
kernels/ref/matmul_ref_fp32.cc
kernels/ref/matmul_ref_int4.cc
kernels/ref/matmul_ref_int8.cc

Public Member Functions