List of all items
Macros
- avx2_do
- avx2_fma_do
- avx512_do
- avx_do
- avx_f16c_do
- base_do
- embedding_idx_specialization
- embedding_specialization
- fused_8bit_rowwise_embedding_idx_specialization
Functions
- adagrad_fp16_update_prefetch
- adagrad_fp16_update_prefetch__avx2_fma
- adagrad_fp16_update_prefetch_base
- adagrad_update
- adagrad_update__avx2_fma
- adagrad_update_base
- adagrad_update_base_inlined
- adagrad_update_no_prefetch
- adagrad_update_prefetch
- adagrad_update_prefetch__avx2_fma
- adagrad_update_prefetch_base
- adagrad_update_prefetch_inlined
- cvtsh_ss
- cvtss_sh
- decompress_and_dequantize__avx2
- decompress_and_dequantize__base
- embedding_lookup
- embedding_lookup_generic_slow
- embedding_lookup_generic_slow_idx
- embedding_lookup_idx
- embedding_lookup_idx_int32_t_float_float_avx2_fma
- embedding_lookup_idx_int32_t_float_float_false_avx2_fma
- embedding_lookup_idx_int32_t_float_float_true_avx2_fma
- embedding_lookup_idx_int32_t_half_float_avx2_fma
- embedding_lookup_idx_int32_t_half_float_false_avx2_fma
- embedding_lookup_idx_int32_t_half_float_true_avx2_fma
- embedding_lookup_idx_int32_t_uint8_t_float_avx2_fma
- embedding_lookup_idx_int32_t_uint8_t_float_false_avx2_fma
- embedding_lookup_idx_int32_t_uint8_t_float_true_avx2_fma
- embedding_lookup_idx_int64_t_float_float_avx2_fma
- embedding_lookup_idx_int64_t_float_float_false_avx2_fma
- embedding_lookup_idx_int64_t_float_float_true_avx2_fma
- embedding_lookup_idx_int64_t_half_float_avx2_fma
- embedding_lookup_idx_int64_t_half_float_false_avx2_fma
- embedding_lookup_idx_int64_t_half_float_true_avx2_fma
- embedding_lookup_idx_int64_t_uint8_t_float_avx2_fma
- embedding_lookup_idx_int64_t_uint8_t_float_false_avx2_fma
- embedding_lookup_idx_int64_t_uint8_t_float_true_avx2_fma
- embedding_lookup_int32_t_float_float_avx2_fma
- embedding_lookup_int32_t_float_float_false_avx2_fma
- embedding_lookup_int32_t_float_float_true_avx2_fma
- embedding_lookup_int32_t_half_float_avx2_fma
- embedding_lookup_int32_t_half_float_false_avx2_fma
- embedding_lookup_int32_t_half_float_true_avx2_fma
- embedding_lookup_int32_t_uint8_t_float_avx2_fma
- embedding_lookup_int32_t_uint8_t_float_false_avx2_fma
- embedding_lookup_int32_t_uint8_t_float_true_avx2_fma
- embedding_lookup_int64_t_float_float_avx2_fma
- embedding_lookup_int64_t_float_float_false_avx2_fma
- embedding_lookup_int64_t_float_float_true_avx2_fma
- embedding_lookup_int64_t_half_float_avx2_fma
- embedding_lookup_int64_t_half_float_false_avx2_fma
- embedding_lookup_int64_t_half_float_true_avx2_fma
- embedding_lookup_int64_t_uint8_t_float_avx2_fma
- embedding_lookup_int64_t_uint8_t_float_false_avx2_fma
- embedding_lookup_int64_t_uint8_t_float_true_avx2_fma
- float_to_fused_8bit_rowwise_quantized
- float_to_fused_8bit_rowwise_quantized_base
- float_to_fusedn_bit_rowwise_quantized_sb_half
- float_to_fusedn_bit_rowwise_quantized_sb_half_base
- fused_8bit_rowwise_embedding_lookup
- fused_8bit_rowwise_embedding_lookup_generic_slow
- fused_8bit_rowwise_embedding_lookup_generic_slow_idx
- fused_8bit_rowwise_embedding_lookup_idx
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_float_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_float_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_float_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_half_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_half_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_half_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_uint8_t_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_uint8_t_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int32_t_uint8_t_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_float_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_float_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_float_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_half_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_half_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_half_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_uint8_t_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_uint8_t_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_idx_int64_t_uint8_t_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_float_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_float_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_float_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_half_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_half_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_half_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_uint8_t_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_uint8_t_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int32_t_uint8_t_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_float_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_float_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_float_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_half_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_half_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_half_float_true_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_uint8_t_float_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_uint8_t_float_false_avx2_fma
- fused_8bit_rowwise_embedding_lookup_int64_t_uint8_t_float_true_avx2_fma
- fused_8bit_rowwise_quantized_to_float
- fused_8bit_rowwise_quantized_to_float_base
- fusedn_bit_rowwise_quantized_sb_half_to_float
- fusedn_bit_rowwise_quantized_sb_half_to_float_base
- lstm_unit_gradient_impl_avx2_fmaf32
- lstm_unit_gradient_implf32
- lstm_unit_impl_avx2_fmaf32
- lstm_unit_implf32
- quantize_and_compress__avx2
- quantize_and_compress__base
- typed_axpy
- typed_axpy__avx2_fma
- typed_axpy__avx_f_16c
- typed_axpy__base
- typed_axpy_halffloat__avx2_fma
- typed_axpy_halffloat__avx_f_16c
- typed_axpy_halffloat__base
- typed_axpy_uint8_float__avx2_fma
- typed_axpy_uint8_float__base
- typed_axpyf16f32
- typed_axpyf32f32
- typed_axpyu8f32