Skip to content

P
Paddle-Lite

PaddlePaddle / Paddle-Lite

通知 337
Star 4
Fork 1
P
Paddle-Lite
未验证 提交 8bc88cff 编写于 作者: J juncaipeng 提交者： GitHub
浏览文件
操作

Add support op and kernel (#2606) 

* add op, test=develop
上级 c9f45332
隐藏空白更改
内联 并排
Showing with 240 addition and 0 deletion
_all_pages/develop/support_operation_list.md
浏览文件 @ 8bc88cff
...@@ -10,6 +10,7 @@ title: 支持OP列表 ...@@ -10,6 +10,7 @@ title: 支持OP列表
- arg_max - arg_max
- assign - assign
- assign_value - assign_value
- attention_padding_mask
- axpy - axpy
- batch_norm - batch_norm
- beam_search - beam_search
...@@ -20,13 +21,16 @@ title: 支持OP列表 ...@@ -20,13 +21,16 @@ title: 支持OP列表
- calib - calib
- calib_once - calib_once
- cast - cast
- collect_fpn_proposals
- concat - concat
- conditional_block
- conv2d - conv2d
- conv2d_transpose - conv2d_transpose
- crop - crop
- decode_bboxes - decode_bboxes
- density_prior_box - density_prior_box
- depthwise_conv2d - depthwise_conv2d
- distribute_fpn_proposals
- dropout - dropout
- elementwise_add - elementwise_add
- elementwise_div - elementwise_div
...@@ -36,7 +40,9 @@ title: 支持OP列表 ...@@ -36,7 +40,9 @@ title: 支持OP列表
- equal - equal
- exp - exp
- expand - expand
- fake_channel_wise_dequantize_max_abs
- fake_dequantize_max_abs - fake_dequantize_max_abs
- fake_quantize_dequantize_moving_average_abs_max
- fake_quantize_moving_average_abs_max - fake_quantize_moving_average_abs_max
- fake_quantize_range_abs_max - fake_quantize_range_abs_max
- fc - fc
...@@ -52,6 +58,7 @@ title: 支持OP列表 ...@@ -52,6 +58,7 @@ title: 支持OP列表
- fusion_elementwise_max_activation - fusion_elementwise_max_activation
- fusion_elementwise_mul_activation - fusion_elementwise_mul_activation
- fusion_elementwise_sub_activation - fusion_elementwise_sub_activation
- gather
- generate_proposals - generate_proposals
- graph_op - graph_op
- greater_equal - greater_equal
...@@ -61,9 +68,11 @@ title: 支持OP列表 ...@@ -61,9 +68,11 @@ title: 支持OP列表
- hard_sigmoid - hard_sigmoid
- im2sequence - im2sequence
- increment - increment
- instance_norm
- io_copy - io_copy
- io_copy_once - io_copy_once
- is_empty - is_empty
- layer_norm
- layout - layout
- layout_once - layout_once
- leaky_relu - leaky_relu
...@@ -76,9 +85,12 @@ title: 支持OP列表 ...@@ -76,9 +85,12 @@ title: 支持OP列表
- logical_or - logical_or
- logical_xor - logical_xor
- lookup_table - lookup_table
- lookup_table_v2
- lrn - lrn
- match_matrix_tensor
- matmul - matmul
- mean - mean
- merge_lod_tensor
- mul - mul
- multiclass_nms - multiclass_nms
- nearest_interp - nearest_interp
...@@ -94,23 +106,43 @@ title: 支持OP列表 ...@@ -94,23 +106,43 @@ title: 支持OP列表
- read_from_array - read_from_array
- reduce_max - reduce_max
- reduce_mean - reduce_mean
- reduce_prod
- reduce_sum
- relu - relu
- relu6 - relu6
- relu_clipped - relu_clipped
- reshape - reshape
- reshape2 - reshape2
- roi_align - roi_align
- rsqrt
- scale - scale
- search_aligned_mat_mul
- search_attention_padding_mask
- search_fc
- search_grnn
- search_group_padding
- search_seq_arithmetic
- search_seq_depadding
- search_seq_fc
- search_seq_softmax
- sequence_arithmetic
- sequence_concat
- sequence_expand - sequence_expand
- sequence_expand_as - sequence_expand_as
- sequence_pool - sequence_pool
- sequence_reshape
- sequence_reverse
- sequence_softmax - sequence_softmax
- sequence_topk_avg_pooling
- shape - shape
- shuffle_channel - shuffle_channel
- sigmoid - sigmoid
- slice - slice
- softmax - softmax
- softsign
- split - split
- split_lod_tensor
- sqrt
- square - square
- squeeze - squeeze
- squeeze2 - squeeze2
...@@ -123,6 +155,7 @@ title: 支持OP列表 ...@@ -123,6 +155,7 @@ title: 支持OP列表
- uniform_random - uniform_random
- unsqueeze - unsqueeze
- unsqueeze2 - unsqueeze2
- var_conv_2d
- while - while
- write_to_array - write_to_array
- yolo_box - yolo_box
...@@ -154,13 +187,16 @@ title: 支持OP列表 ...@@ -154,13 +187,16 @@ title: 支持OP列表
- box_clip - box_clip
- box_coder - box_coder
- cast - cast
- collect_fpn_proposals
- concat - concat
- conditional_block
- conv2d - conv2d
- conv2d_transpose - conv2d_transpose
- crop - crop
- decode_bboxes - decode_bboxes
- density_prior_box - density_prior_box
- depthwise_conv2d - depthwise_conv2d
- distribute_fpn_proposals
- dropout - dropout
- elementwise_add - elementwise_add
- elementwise_div - elementwise_div
...@@ -172,12 +208,14 @@ title: 支持OP列表 ...@@ -172,12 +208,14 @@ title: 支持OP列表
- expand - expand
- fc - fc
- fill_constant - fill_constant
- fill_constant_batch_size_like
- floor - floor
- fusion_elementwise_add_activation - fusion_elementwise_add_activation
- fusion_elementwise_div_activation - fusion_elementwise_div_activation
- fusion_elementwise_max_activation - fusion_elementwise_max_activation
- fusion_elementwise_mul_activation - fusion_elementwise_mul_activation
- fusion_elementwise_sub_activation - fusion_elementwise_sub_activation
- gather
- generate_proposals - generate_proposals
- greater_equal - greater_equal
- greater_than - greater_than
...@@ -186,7 +224,11 @@ title: 支持OP列表 ...@@ -186,7 +224,11 @@ title: 支持OP列表
- hard_sigmoid - hard_sigmoid
- im2sequence - im2sequence
- increment - increment
- instance_norm
- is_empty - is_empty
- layer_norm
- layout
- layout_once
- leaky_relu - leaky_relu
- less_equal - less_equal
- less_than - less_than
...@@ -197,8 +239,10 @@ title: 支持OP列表 ...@@ -197,8 +239,10 @@ title: 支持OP列表
- logical_or - logical_or
- logical_xor - logical_xor
- lookup_table - lookup_table
- lookup_table_v2
- lrn - lrn
- matmul - matmul
- merge_lod_tensor
- mul - mul
- nearest_interp - nearest_interp
- negative - negative
...@@ -213,10 +257,12 @@ title: 支持OP列表 ...@@ -213,10 +257,12 @@ title: 支持OP列表
- read_from_array - read_from_array
- reduce_max - reduce_max
- reduce_mean - reduce_mean
- reduce_prod
- relu - relu
- relu6 - relu6
- relu_clipped - relu_clipped
- roi_align - roi_align
- rsqrt
- scale - scale
- sequence_expand - sequence_expand
- sequence_pool - sequence_pool
...@@ -227,6 +273,7 @@ title: 支持OP列表 ...@@ -227,6 +273,7 @@ title: 支持OP列表
- slice - slice
- softmax - softmax
- split - split
- split_lod_tensor
- squeeze - squeeze
- squeeze2 - squeeze2
- stack - stack
...@@ -241,26 +288,97 @@ title: 支持OP列表 ...@@ -241,26 +288,97 @@ title: 支持OP列表
- write_to_array - write_to_array
- yolo_box - yolo_box
### X86 kernels ### X86 kernels
- batch_norm
- cast
- concat - concat
- conv2d
- depthwise_conv2d
- dropout
- elementwise_add - elementwise_add
- elementwise_sub - elementwise_sub
- fc
- fill_constant_batch_size_like - fill_constant_batch_size_like
- gather
- gelu
- gru - gru
- layer_norm
- match_matrix_tensor
- matmul - matmul
- mul - mul
- pool2d
- reduce_sum
- relu - relu
- reshape - reshape
- reshape2 - reshape2
- scale - scale
- search_aligned_mat_mul
- search_attention_padding_mask
- search_fc
- search_grnn
- search_group_padding
- search_seq_arithmetic
- search_seq_depadding
- search_seq_fc
- search_seq_softmax
- sequence_arithmetic
- sequence_concat
- sequence_expand_as - sequence_expand_as
- sequence_pool - sequence_pool
- sequence_reverse
- sequence_topk_avg_pooling
- shape - shape
- slice - slice
- softmax - softmax
- softsign
- square - square
- squeeze - squeeze
- squeeze2 - squeeze2
- stack
- tanh
- transpose
- transpose2
- var_conv_2d
### CUDA kernels
- attention_padding_mask
- bilinear_interp
- calib
- concat
- conv
- dropout
- elementwise_add
- fusion_elementwise_add_activation
- fusion_elementwise_mul_activation
- elementwise_mul
- feed
- io_copy
- layout
- layout_once
- leaky_relu
- lookup_table
- match_matrix_tensor
- mul
- nearest_interp
- pool2d
- relu
- scale
- search_aligned_mat_mul
- search_fc
- search_grnn
- search_group_padding
- search_seq_depadding
- search_seq_fc
- sequence_arithmetic
- sequence_concat
- sequence_pool
- sequence_reverse
- sequence_topk_avg_pooling
- softmax
- transpose
- var_conv_2d
- yolo_box
### OpenCL kernels ### OpenCL kernels
- conv2d - conv2d
...@@ -268,6 +386,8 @@ title: 支持OP列表 ...@@ -268,6 +386,8 @@ title: 支持OP列表
- elementwise_add - elementwise_add
- fc - fc
- fusion_elementwise_add_activation - fusion_elementwise_add_activation
- layout
- layout_once
- io_copy - io_copy
- io_copy_once - io_copy_once
- mul - mul
......
_all_pages/v2.1.0/support_operation_list.md
浏览文件 @ 8bc88cff
...@@ -10,6 +10,7 @@ title: 支持OP列表 ...@@ -10,6 +10,7 @@ title: 支持OP列表
- arg_max - arg_max
- assign - assign
- assign_value - assign_value
- attention_padding_mask
- axpy - axpy
- batch_norm - batch_norm
- beam_search - beam_search
...@@ -20,13 +21,16 @@ title: 支持OP列表 ...@@ -20,13 +21,16 @@ title: 支持OP列表
- calib - calib
- calib_once - calib_once
- cast - cast
- collect_fpn_proposals
- concat - concat
- conditional_block
- conv2d - conv2d
- conv2d_transpose - conv2d_transpose
- crop - crop
- decode_bboxes - decode_bboxes
- density_prior_box - density_prior_box
- depthwise_conv2d - depthwise_conv2d
- distribute_fpn_proposals
- dropout - dropout
- elementwise_add - elementwise_add
- elementwise_div - elementwise_div
...@@ -36,7 +40,9 @@ title: 支持OP列表 ...@@ -36,7 +40,9 @@ title: 支持OP列表
- equal - equal
- exp - exp
- expand - expand
- fake_channel_wise_dequantize_max_abs
- fake_dequantize_max_abs - fake_dequantize_max_abs
- fake_quantize_dequantize_moving_average_abs_max
- fake_quantize_moving_average_abs_max - fake_quantize_moving_average_abs_max
- fake_quantize_range_abs_max - fake_quantize_range_abs_max
- fc - fc
...@@ -52,6 +58,7 @@ title: 支持OP列表 ...@@ -52,6 +58,7 @@ title: 支持OP列表
- fusion_elementwise_max_activation - fusion_elementwise_max_activation
- fusion_elementwise_mul_activation - fusion_elementwise_mul_activation
- fusion_elementwise_sub_activation - fusion_elementwise_sub_activation
- gather
- generate_proposals - generate_proposals
- graph_op - graph_op
- greater_equal - greater_equal
...@@ -61,9 +68,11 @@ title: 支持OP列表 ...@@ -61,9 +68,11 @@ title: 支持OP列表
- hard_sigmoid - hard_sigmoid
- im2sequence - im2sequence
- increment - increment
- instance_norm
- io_copy - io_copy
- io_copy_once - io_copy_once
- is_empty - is_empty
- layer_norm
- layout - layout
- layout_once - layout_once
- leaky_relu - leaky_relu
...@@ -76,9 +85,12 @@ title: 支持OP列表 ...@@ -76,9 +85,12 @@ title: 支持OP列表
- logical_or - logical_or
- logical_xor - logical_xor
- lookup_table - lookup_table
- lookup_table_v2
- lrn - lrn
- match_matrix_tensor
- matmul - matmul
- mean - mean
- merge_lod_tensor
- mul - mul
- multiclass_nms - multiclass_nms
- nearest_interp - nearest_interp
...@@ -94,23 +106,43 @@ title: 支持OP列表 ...@@ -94,23 +106,43 @@ title: 支持OP列表
- read_from_array - read_from_array
- reduce_max - reduce_max
- reduce_mean - reduce_mean
- reduce_prod
- reduce_sum
- relu - relu
- relu6 - relu6
- relu_clipped - relu_clipped
- reshape - reshape
- reshape2 - reshape2
- roi_align - roi_align
- rsqrt
- scale - scale
- search_aligned_mat_mul
- search_attention_padding_mask
- search_fc
- search_grnn
- search_group_padding
- search_seq_arithmetic
- search_seq_depadding
- search_seq_fc
- search_seq_softmax
- sequence_arithmetic
- sequence_concat
- sequence_expand - sequence_expand
- sequence_expand_as - sequence_expand_as
- sequence_pool - sequence_pool
- sequence_reshape
- sequence_reverse
- sequence_softmax - sequence_softmax
- sequence_topk_avg_pooling
- shape - shape
- shuffle_channel - shuffle_channel
- sigmoid - sigmoid
- slice - slice
- softmax - softmax
- softsign
- split - split
- split_lod_tensor
- sqrt
- square - square
- squeeze - squeeze
- squeeze2 - squeeze2
...@@ -123,6 +155,7 @@ title: 支持OP列表 ...@@ -123,6 +155,7 @@ title: 支持OP列表
- uniform_random - uniform_random
- unsqueeze - unsqueeze
- unsqueeze2 - unsqueeze2
- var_conv_2d
- while - while
- write_to_array - write_to_array
- yolo_box - yolo_box
...@@ -154,13 +187,16 @@ title: 支持OP列表 ...@@ -154,13 +187,16 @@ title: 支持OP列表
- box_clip - box_clip
- box_coder - box_coder
- cast - cast
- collect_fpn_proposals
- concat - concat
- conditional_block
- conv2d - conv2d
- conv2d_transpose - conv2d_transpose
- crop - crop
- decode_bboxes - decode_bboxes
- density_prior_box - density_prior_box
- depthwise_conv2d - depthwise_conv2d
- distribute_fpn_proposals
- dropout - dropout
- elementwise_add - elementwise_add
- elementwise_div - elementwise_div
...@@ -172,12 +208,14 @@ title: 支持OP列表 ...@@ -172,12 +208,14 @@ title: 支持OP列表
- expand - expand
- fc - fc
- fill_constant - fill_constant
- fill_constant_batch_size_like
- floor - floor
- fusion_elementwise_add_activation - fusion_elementwise_add_activation
- fusion_elementwise_div_activation - fusion_elementwise_div_activation
- fusion_elementwise_max_activation - fusion_elementwise_max_activation
- fusion_elementwise_mul_activation - fusion_elementwise_mul_activation
- fusion_elementwise_sub_activation - fusion_elementwise_sub_activation
- gather
- generate_proposals - generate_proposals
- greater_equal - greater_equal
- greater_than - greater_than
...@@ -186,7 +224,11 @@ title: 支持OP列表 ...@@ -186,7 +224,11 @@ title: 支持OP列表
- hard_sigmoid - hard_sigmoid
- im2sequence - im2sequence
- increment - increment
- instance_norm
- is_empty - is_empty
- layer_norm
- layout
- layout_once
- leaky_relu - leaky_relu
- less_equal - less_equal
- less_than - less_than
...@@ -197,8 +239,10 @@ title: 支持OP列表 ...@@ -197,8 +239,10 @@ title: 支持OP列表
- logical_or - logical_or
- logical_xor - logical_xor
- lookup_table - lookup_table
- lookup_table_v2
- lrn - lrn
- matmul - matmul
- merge_lod_tensor
- mul - mul
- nearest_interp - nearest_interp
- negative - negative
...@@ -213,10 +257,12 @@ title: 支持OP列表 ...@@ -213,10 +257,12 @@ title: 支持OP列表
- read_from_array - read_from_array
- reduce_max - reduce_max
- reduce_mean - reduce_mean
- reduce_prod
- relu - relu
- relu6 - relu6
- relu_clipped - relu_clipped
- roi_align - roi_align
- rsqrt
- scale - scale
- sequence_expand - sequence_expand
- sequence_pool - sequence_pool
...@@ -227,6 +273,7 @@ title: 支持OP列表 ...@@ -227,6 +273,7 @@ title: 支持OP列表
- slice - slice
- softmax - softmax
- split - split
- split_lod_tensor
- squeeze - squeeze
- squeeze2 - squeeze2
- stack - stack
...@@ -241,26 +288,97 @@ title: 支持OP列表 ...@@ -241,26 +288,97 @@ title: 支持OP列表
- write_to_array - write_to_array
- yolo_box - yolo_box
### X86 kernels ### X86 kernels
- batch_norm
- cast
- concat - concat
- conv2d
- depthwise_conv2d
- dropout
- elementwise_add - elementwise_add
- elementwise_sub - elementwise_sub
- fc
- fill_constant_batch_size_like - fill_constant_batch_size_like
- gather
- gelu
- gru - gru
- layer_norm
- match_matrix_tensor
- matmul - matmul
- mul - mul
- pool2d
- reduce_sum
- relu - relu
- reshape - reshape
- reshape2 - reshape2
- scale - scale
- search_aligned_mat_mul
- search_attention_padding_mask
- search_fc
- search_grnn
- search_group_padding
- search_seq_arithmetic
- search_seq_depadding
- search_seq_fc
- search_seq_softmax
- sequence_arithmetic
- sequence_concat
- sequence_expand_as - sequence_expand_as
- sequence_pool - sequence_pool
- sequence_reverse
- sequence_topk_avg_pooling
- shape - shape
- slice - slice
- softmax - softmax
- softsign
- square - square
- squeeze - squeeze
- squeeze2 - squeeze2
- stack
- tanh
- transpose
- transpose2
- var_conv_2d
### CUDA kernels
- attention_padding_mask
- bilinear_interp
- calib
- concat
- conv
- dropout
- elementwise_add
- fusion_elementwise_add_activation
- fusion_elementwise_mul_activation
- elementwise_mul
- feed
- io_copy
- layout
- layout_once
- leaky_relu
- lookup_table
- match_matrix_tensor
- mul
- nearest_interp
- pool2d
- relu
- scale
- search_aligned_mat_mul
- search_fc
- search_grnn
- search_group_padding
- search_seq_depadding
- search_seq_fc
- sequence_arithmetic
- sequence_concat
- sequence_pool
- sequence_reverse
- sequence_topk_avg_pooling
- softmax
- transpose
- var_conv_2d
- yolo_box
### OpenCL kernels ### OpenCL kernels
- conv2d - conv2d
...@@ -268,6 +386,8 @@ title: 支持OP列表 ...@@ -268,6 +386,8 @@ title: 支持OP列表
- elementwise_add - elementwise_add
- fc - fc
- fusion_elementwise_add_activation - fusion_elementwise_add_activation
- layout
- layout_once
- io_copy - io_copy
- io_copy_once - io_copy_once
- mul - mul
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或