支持OP列表
Ops
- affine_channel
- anchor_generator
- arg_max
- assign
- assign_value
- axpy
- batch_norm
- beam_search
- beam_search_decode
- bilinear_interp
- box_clip
- box_coder
- calib
- calib_once
- cast
- concat
- conv2d
- conv2d_transpose
- crop
- decode_bboxes
- density_prior_box
- depthwise_conv2d
- dropout
- elementwise_add
- elementwise_div
- elementwise_max
- elementwise_mul
- elementwise_sub
- equal
- exp
- expand
- fake_dequantize_max_abs
- fake_quantize_moving_average_abs_max
- fake_quantize_range_abs_max
- fc
- feed
- fetch
- fill_constant
- fill_constant_batch_size_like
- flatten
- flatten2
- floor
- fusion_elementwise_add_activation
- fusion_elementwise_div_activation
- fusion_elementwise_max_activation
- fusion_elementwise_mul_activation
- fusion_elementwise_sub_activation
- generate_proposals
- graph_op
- greater_equal
- greater_than
- gru
- gru_unit
- hard_sigmoid
- im2sequence
- increment
- io_copy
- io_copy_once
- is_empty
- layout
- layout_once
- leaky_relu
- less_equal
- less_than
- lod_reset
- log
- logical_and
- logical_not
- logical_or
- logical_xor
- lookup_table
- lrn
- matmul
- mean
- mul
- multiclass_nms
- nearest_interp
- negative
- norm
- notequal
- pad2d
- pool2d
- power
- prelu
- prior_box
- range
- read_from_array
- reduce_max
- reduce_mean
- relu
- relu6
- relu_clipped
- reshape
- reshape2
- roi_align
- scale
- sequence_expand
- sequence_expand_as
- sequence_pool
- sequence_softmax
- shape
- shuffle_channel
- sigmoid
- slice
- softmax
- split
- square
- squeeze
- squeeze2
- stack
- swish
- tanh
- top_k
- transpose
- transpose2
- uniform_random
- unsqueeze
- unsqueeze2
- while
- write_to_array
- yolo_box
Kernels
Host kernels
- feed
- fetch
- flatten
- flatten2
- multiclass_nms
- reshape
- reshape2
ARM kernels
- affine_channel
- anchor_generator
- arg_max
- assign
- assign_value
- axpy
- batch_norm
- beam_search
- beam_search_decode
- bilinear_interp
- box_clip
- box_coder
- cast
- concat
- conv2d
- conv2d_transpose
- crop
- decode_bboxes
- density_prior_box
- depthwise_conv2d
- dropout
- elementwise_add
- elementwise_div
- elementwise_max
- elementwise_mul
- elementwise_sub
- equal
- exp
- expand
- fc
- fill_constant
- floor
- fusion_elementwise_add_activation
- fusion_elementwise_div_activation
- fusion_elementwise_max_activation
- fusion_elementwise_mul_activation
- fusion_elementwise_sub_activation
- generate_proposals
- greater_equal
- greater_than
- gru
- gru_unit
- hard_sigmoid
- im2sequence
- increment
- is_empty
- leaky_relu
- less_equal
- less_than
- lod_reset
- log
- logical_and
- logical_not
- logical_or
- logical_xor
- lookup_table
- lrn
- matmul
- mul
- nearest_interp
- negative
- norm
- not_equal
- pad2d
- pool2d
- power
- prelu
- prior_box
- range
- read_from_array
- reduce_max
- reduce_mean
- relu
- relu6
- relu_clipped
- roi_align
- scale
- sequence_expand
- sequence_pool
- sequence_softmax
- shape
- shuffle_channel
- sigmoid
- slice
- softmax
- split
- squeeze
- squeeze2
- stack
- swish
- tanh
- top_k
- transpose
- transpose2
- unsqueeze
- unsqueeze2
- while
- write_to_array
- yolo_box
X86 kernels
- concat
- elementwise_add
- elementwise_sub
- fill_constant_batch_size_like
- gru
- matmul
- mul
- relu
- reshape
- reshape2
- scale
- sequence_expand_as
- sequence_pool
- shape
- slice
- softmax
- square
- squeeze
- squeeze2
OpenCL kernels
- conv2d
- depthwise_conv2d
- elementwise_add
- fc
- fusion_elementwise_add_activation
- io_copy
- io_copy_once
- mul
- pool2d
- relu