Skip to content

GitLab

Unverified Commit 72182747 authored by RunningLeon's avatar RunningLeon Committed by GitHub
Browse files

[Doc]: Fix deploy doc (#1512) 

* update title

* fix deploy docs in for readthedocs

* reformat table

* resolve comments
parent 7207397d
Hide whitespace changes
Inline Side-by-side
Showing with 114 additions and 107 deletions
docs/deployment/mmcv_ops_definition.md
View file @ 72182747
# Definition of custom operators in MMCV # MMCV Operators
To make custom operators in MMCV more standard, precise definitions of each operator are listed in this document.
<!-- TOC --> <!-- TOC -->
- [Definition of custom operators in MMCV](#definition-of-custom-operators-in-mmcv) - [MMCV Operators](#mmcv-operators)
- [MMCVBorderAlign](#mmcvborderalign) - [MMCVBorderAlign](#mmcvborderalign)
- [Description](#description) - [Description](#description)
- [Parameters](#parameters) - [Parameters](#parameters)
...@@ -80,25 +82,25 @@ ...@@ -80,25 +82,25 @@
- [Inputs](#inputs-12) - [Inputs](#inputs-12)
- [Outputs](#outputs-12) - [Outputs](#outputs-12)
- [Type Constraints](#type-constraints-12) - [Type Constraints](#type-constraints-12)
- [torch](#torch) - [grid_sampler*](#grid_sampler)
- [grid_sampler](#grid_sampler)
- [Description](#description-13) - [Description](#description-13)
- [Parameters](#parameters-13) - [Parameters](#parameters-13)
- [Inputs](#inputs-13) - [Inputs](#inputs-13)
- [Outputs](#outputs-13) - [Outputs](#outputs-13)
- [Type Constraints](#type-constraints-13) - [Type Constraints](#type-constraints-13)
- [cummax](#cummax) - [cummax*](#cummax)
- [Description](#description-14) - [Description](#description-14)
- [Parameters](#parameters-14) - [Parameters](#parameters-14)
- [Inputs](#inputs-14) - [Inputs](#inputs-14)
- [Outputs](#outputs-14) - [Outputs](#outputs-14)
- [Type Constraints](#type-constraints-14) - [Type Constraints](#type-constraints-14)
- [cummin](#cummin) - [cummin*](#cummin)
- [Description](#description-15) - [Description](#description-15)
- [Parameters](#parameters-15) - [Parameters](#parameters-15)
- [Inputs](#inputs-15) - [Inputs](#inputs-15)
- [Outputs](#outputs-15) - [Outputs](#outputs-15)
- [Type Constraints](#type-constraints-15) - [Type Constraints](#type-constraints-15)
- [Reminders](#reminders)
<!-- TOC --> <!-- TOC -->
## MMCVBorderAlign ## MMCVBorderAlign
...@@ -118,9 +120,9 @@ Read [BorderDet: Border Feature for Dense Object Detection](ttps://arxiv.org/abs ...@@ -118,9 +120,9 @@ Read [BorderDet: Border Feature for Dense Object Detection](ttps://arxiv.org/abs
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |-------|-------------|-------------------------------------------------------------------------------------|
| `int` | `pool_size` | number of positions sampled over the boxes' borders(e.g. top, bottom, left, right). | | `int` | `pool_size` | number of positions sampled over the boxes' borders(e.g. top, bottom, left, right). |
### Inputs ### Inputs
...@@ -152,11 +154,11 @@ Read [CARAFE: Content-Aware ReAssembly of FEatures](https://arxiv.org/abs/1905.0 ...@@ -152,11 +154,11 @@ Read [CARAFE: Content-Aware ReAssembly of FEatures](https://arxiv.org/abs/1905.0
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |---------|----------------|-----------------------------------------------|
| `int` | `kernel_size` | reassemble kernel size, should be odd integer| | `int` | `kernel_size` | reassemble kernel size, should be odd integer |
| `int` | `group_size` | reassemble group size | | `int` | `group_size` | reassemble group size |
| `float` | `scale_factor` | upsample ratio(>=1) | | `float` | `scale_factor` | upsample ratio(>=1) |
### Inputs ### Inputs
...@@ -187,8 +189,7 @@ Read [CCNet: Criss-Cross Attention for SemanticSegmentation](https://arxiv.org/p ...@@ -187,8 +189,7 @@ Read [CCNet: Criss-Cross Attention for SemanticSegmentation](https://arxiv.org/p
### Parameters ### Parameters
| Type | Parameter | Description | None
| ------- | --------------- | -------------------------------------------------------------- |
### Inputs ### Inputs
...@@ -219,8 +220,7 @@ Read [CCNet: Criss-Cross Attention for SemanticSegmentation](https://arxiv.org/p ...@@ -219,8 +220,7 @@ Read [CCNet: Criss-Cross Attention for SemanticSegmentation](https://arxiv.org/p
### Parameters ### Parameters
| Type | Parameter | Description | None
| ------- | --------------- | -------------------------------------------------------------- |
### Inputs ### Inputs
...@@ -251,9 +251,9 @@ Perform CornerPool on `input` features. Read [CornerNet -- Detecting Objects as ...@@ -251,9 +251,9 @@ Perform CornerPool on `input` features. Read [CornerNet -- Detecting Objects as
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | ---------------------------------------------------------------- | |-------|-----------|------------------------------------------------------------------|
| `int` | `mode` | corner pool mode, (0: `top`, 1: `bottom`, 2: `left`, 3: `right`) | | `int` | `mode` | corner pool mode, (0: `top`, 1: `bottom`, 2: `left`, 3: `right`) |
### Inputs ### Inputs
...@@ -283,15 +283,15 @@ Read [Deformable Convolutional Networks](https://arxiv.org/pdf/1703.06211.pdf) f ...@@ -283,15 +283,15 @@ Read [Deformable Convolutional Networks](https://arxiv.org/pdf/1703.06211.pdf) f
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| -------------- | ------------------ | ------------------------------------------------------------------------------------- | |----------------|---------------------|-------------------------------------------------------------------------------------------------------------------|
| `list of ints` | `stride` | The stride of the convolving kernel, (sH, sW). Defaults to `(1, 1)`. | | `list of ints` | `stride` | The stride of the convolving kernel, (sH, sW). Defaults to `(1, 1)`. |
| `list of ints` | `padding` | Paddings on both sides of the input, (padH, padW). Defaults to `(0, 0)`. | | `list of ints` | `padding` | Paddings on both sides of the input, (padH, padW). Defaults to `(0, 0)`. |
| `list of ints` | `dilation` | The spacing between kernel elements (dH, dW). Defaults to `(1, 1)`. | | `list of ints` | `dilation` | The spacing between kernel elements (dH, dW). Defaults to `(1, 1)`. |
| `int` | `groups` | Split input into groups. `input_channel` should be divisible by the number of groups. Defaults to `1`.| | `int` | `groups` | Split input into groups. `input_channel` should be divisible by the number of groups. Defaults to `1`. |
| `int` | `deformable_groups` | Groups of deformable offset. Defaults to `1`. | | `int` | `deformable_groups` | Groups of deformable offset. Defaults to `1`. |
| `int` | `bias` | Whether to add a learnable bias to the output. `0` stands for `False` and `1` stands for `True`. Defaults to `0`. | | `int` | `bias` | Whether to add a learnable bias to the output. `0` stands for `False` and `1` stands for `True`. Defaults to `0`. |
| `int` | `im2col_step` | Groups of deformable offset. Defaults to `32`. | | `int` | `im2col_step` | Groups of deformable offset. Defaults to `32`. |
### Inputs ### Inputs
...@@ -323,11 +323,11 @@ Perform Modulated Deformable Convolution on input feature, read [Deformable Conv ...@@ -323,11 +323,11 @@ Perform Modulated Deformable Convolution on input feature, read [Deformable Conv
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| -------------- | ------------------ | ------------------------------------------------------------------------------------- | |----------------|---------------------|---------------------------------------------------------------------------------------|
| `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW) | | `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW) |
| `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW) | | `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW) |
| `list of ints` | `dilation` | The spacing between kernel elements. (dH, dW) | | `list of ints` | `dilation` | The spacing between kernel elements. (dH, dW) |
| `int` | `deformable_groups` | Groups of deformable offset. | | `int` | `deformable_groups` | Groups of deformable offset. |
| `int` | `groups` | Split input into groups. `input_channel` should be divisible by the number of groups. | | `int` | `groups` | Split input into groups. `input_channel` should be divisible by the number of groups. |
...@@ -365,13 +365,13 @@ Deformable roi pooling layer ...@@ -365,13 +365,13 @@ Deformable roi pooling layer
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |---------|------------------|---------------------------------------------------------------------------------------------------------------|
| `int` | `output_height` | height of output roi | | `int` | `output_height` | height of output roi |
| `int` | `output_width` | width of output roi | | `int` | `output_width` | width of output roi |
| `float` | `spatial_scale` | used to scale the input boxes | | `float` | `spatial_scale` | used to scale the input boxes |
| `int` | `sampling_ratio` | number of input samples to take for each output sample. `0` means to take samples densely for current models. | | `int` | `sampling_ratio` | number of input samples to take for each output sample. `0` means to take samples densely for current models. |
| `float` | `gamma` | gamma | | `float` | `gamma` | gamma |
### Inputs ### Inputs
...@@ -404,10 +404,10 @@ Read [Pixel Recurrent Neural Networks](https://arxiv.org/abs/1601.06759) for mor ...@@ -404,10 +404,10 @@ Read [Pixel Recurrent Neural Networks](https://arxiv.org/abs/1601.06759) for mor
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |----------------|-----------|----------------------------------------------------------------------------------|
| `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW). **Only support stride=1 in mmcv** | | `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW). **Only support stride=1 in mmcv** |
| `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW). Defaults to `(0, 0)`. | | `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW). Defaults to `(0, 0)`. |
### Inputs ### Inputs
...@@ -443,10 +443,10 @@ Read [PSANet: Point-wise Spatial Attention Network for Scene Parsing](https://hs ...@@ -443,10 +443,10 @@ Read [PSANet: Point-wise Spatial Attention Network for Scene Parsing](https://hs
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |----------------|-------------|----------------------------------------------|
| `int` | `psa_type` | `0` means collect and `1` means `distribute` | | `int` | `psa_type` | `0` means collect and `1` means `distribute` |
| `list of ints` | `mask_size` | The size of mask | | `list of ints` | `mask_size` | The size of mask |
### Inputs ### Inputs
...@@ -477,7 +477,7 @@ Note this definition is slightly different with [onnx: NonMaxSuppression](https: ...@@ -477,7 +477,7 @@ Note this definition is slightly different with [onnx: NonMaxSuppression](https:
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | ---------------------------- | ------------------------------------------------------------------------------------------------------------------------------------ | |---------|------------------------------|--------------------------------------------------------------------------------------------------------------------------------------|
| `int` | `center_point_box` | 0 - the box data is supplied as [y1, x1, y2, x2], 1-the box data is supplied as [x_center, y_center, width, height]. | | `int` | `center_point_box` | 0 - the box data is supplied as [y1, x1, y2, x2], 1-the box data is supplied as [x_center, y_center, width, height]. |
| `int` | `max_output_boxes_per_class` | The maximum number of boxes to be selected per batch per class. Default to 0, number of output boxes equal to number of input boxes. | | `int` | `max_output_boxes_per_class` | The maximum number of boxes to be selected per batch per class. Default to 0, number of output boxes equal to number of input boxes. |
| `float` | `iou_threshold` | The threshold for deciding whether boxes overlap too much with respect to IoU. Value range [0, 1]. Default to 0. | | `float` | `iou_threshold` | The threshold for deciding whether boxes overlap too much with respect to IoU. Value range [0, 1]. Default to 0. |
...@@ -515,7 +515,7 @@ Perform RoIAlign on output feature, used in bbox_head of most two-stage detector ...@@ -515,7 +515,7 @@ Perform RoIAlign on output feature, used in bbox_head of most two-stage detector
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | ---------------- | ------------------------------------------------------------------------------------------------------------- | |---------|------------------|---------------------------------------------------------------------------------------------------------------|
| `int` | `output_height` | height of output roi | | `int` | `output_height` | height of output roi |
| `int` | `output_width` | width of output roi | | `int` | `output_width` | width of output roi |
| `float` | `spatial_scale` | used to scale the input boxes | | `float` | `spatial_scale` | used to scale the input boxes |
...@@ -543,7 +543,6 @@ Perform RoIAlign on output feature, used in bbox_head of most two-stage detector ...@@ -543,7 +543,6 @@ Perform RoIAlign on output feature, used in bbox_head of most two-stage detector
- T:tensor(float32) - T:tensor(float32)
## MMCVRoIAlignRotated ## MMCVRoIAlignRotated
### Description ### Description
...@@ -552,15 +551,15 @@ Perform RoI align pooling for rotated proposals ...@@ -552,15 +551,15 @@ Perform RoI align pooling for rotated proposals
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |---------|------------------|---------------------------------------------------------------------------------------------------------------|
| `int` | `output_height` | height of output roi | | `int` | `output_height` | height of output roi |
| `int` | `output_width` | width of output roi | | `int` | `output_width` | width of output roi |
| `float` | `spatial_scale` | used to scale the input boxes | | `float` | `spatial_scale` | used to scale the input boxes |
| `int` | `sampling_ratio` | number of input samples to take for each output sample. `0` means to take samples densely for current models. | | `int` | `sampling_ratio` | number of input samples to take for each output sample. `0` means to take samples densely for current models. |
| `str` | `mode` | pooling mode in each bin. `avg` or `max` | | `str` | `mode` | pooling mode in each bin. `avg` or `max` |
| `int` | `aligned` | If `aligned=0`, use the legacy implementation in MMDetection. Else, align the results more perfectly. | | `int` | `aligned` | If `aligned=0`, use the legacy implementation in MMDetection. Else, align the results more perfectly. |
| `int` | `clockwise` | If `aligned=0`, use the legacy implementation in MMDetection. Else, align the results more perfectly. | | `int` | `clockwise` | If `aligned=0`, use the legacy implementation in MMDetection. Else, align the results more perfectly. |
### Inputs ### Inputs
...@@ -581,9 +580,7 @@ Perform RoI align pooling for rotated proposals ...@@ -581,9 +580,7 @@ Perform RoI align pooling for rotated proposals
- T:tensor(float32) - T:tensor(float32)
# torch ## grid_sampler*
## grid_sampler
### Description ### Description
...@@ -594,7 +591,7 @@ Check [torch.nn.functional.grid_sample](https://pytorch.org/docs/stable/generate ...@@ -594,7 +591,7 @@ Check [torch.nn.functional.grid_sample](https://pytorch.org/docs/stable/generate
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ----- | -------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | |-------|----------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| `int` | `interpolation_mode` | Interpolation mode to calculate output values. (0: `bilinear` , 1: `nearest`) | | `int` | `interpolation_mode` | Interpolation mode to calculate output values. (0: `bilinear` , 1: `nearest`) |
| `int` | `padding_mode` | Padding mode for outside grid values. (0: `zeros`, 1: `border`, 2: `reflection`) | | `int` | `padding_mode` | Padding mode for outside grid values. (0: `zeros`, 1: `border`, 2: `reflection`) |
| `int` | `align_corners` | If `align_corners=1`, the extrema (`-1` and `1`) are considered as referring to the center points of the input's corner pixels. If `align_corners=0`, they are instead considered as referring to the corner points of the input's corner pixels, making the sampling more resolution agnostic. | | `int` | `align_corners` | If `align_corners=1`, the extrema (`-1` and `1`) are considered as referring to the center points of the input's corner pixels. If `align_corners=0`, they are instead considered as referring to the corner points of the input's corner pixels, making the sampling more resolution agnostic. |
...@@ -619,7 +616,7 @@ Check [torch.nn.functional.grid_sample](https://pytorch.org/docs/stable/generate ...@@ -619,7 +616,7 @@ Check [torch.nn.functional.grid_sample](https://pytorch.org/docs/stable/generate
- T:tensor(float32, Linear) - T:tensor(float32, Linear)
## cummax ## cummax*
### Description ### Description
...@@ -627,9 +624,9 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative maximum e ...@@ -627,9 +624,9 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative maximum e
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | ---------------------------------------------------------------- | |-------|-----------|----------------------------------------|
| `int` | `dim` | the dimension to do the operation over | | `int` | `dim` | the dimension to do the operation over |
### Inputs ### Inputs
...@@ -651,7 +648,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative maximum e ...@@ -651,7 +648,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative maximum e
- T:tensor(float32) - T:tensor(float32)
## cummin ## cummin*
### Description ### Description
...@@ -659,9 +656,9 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative minimum e ...@@ -659,9 +656,9 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative minimum e
### Parameters ### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | ---------------------------------------------------------------- | |-------|-----------|----------------------------------------|
| `int` | `dim` | the dimension to do the operation over | | `int` | `dim` | the dimension to do the operation over |
### Inputs ### Inputs
...@@ -682,3 +679,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative minimum e ...@@ -682,3 +679,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative minimum e
### Type Constraints ### Type Constraints
- T:tensor(float32) - T:tensor(float32)
## Reminders
- Operators endwith `*` are defined in Torch and are included here for the conversion to ONNX.
docs/deployment/onnx.md
View file @ 72182747
## Introduction of onnx module in MMCV (Experimental) ## Introduction of mmcv.onnx module
### register_extra_symbolics ### register_extra_symbolics
...@@ -14,6 +14,10 @@ opset_version = 11 ...@@ -14,6 +14,10 @@ opset_version = 11
register_extra_symbolics(opset_version) register_extra_symbolics(opset_version)
``` ```
#### Reminder
- *Please note that this feature is experimental and may change in the future.*
#### FAQs #### FAQs
- None - None
docs/deployment/onnxruntime_custom_ops.md
View file @ 72182747
## Onnxruntime Custom Ops ## ONNX Runtime Custom Ops
<!-- TOC --> <!-- TOC -->
- [Onnxruntime Custom Ops](#onnxruntime-custom-ops) - [ONNX Runtime Custom Ops](#onnx-runtime-custom-ops)
- [SoftNMS](#softnms) - [SoftNMS](#softnms)
- [Description](#description) - [Description](#description)
- [Parameters](#parameters) - [Parameters](#parameters)
...@@ -69,7 +69,7 @@ Perform soft NMS on `boxes` with `scores`. Read [Soft-NMS -- Improving Object De ...@@ -69,7 +69,7 @@ Perform soft NMS on `boxes` with `scores`. Read [Soft-NMS -- Improving Object De
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | -------------------------------------------------------------- | |---------|-----------------|----------------------------------------------------------------|
| `float` | `iou_threshold` | IoU threshold for NMS | | `float` | `iou_threshold` | IoU threshold for NMS |
| `float` | `sigma` | hyperparameter for gaussian method | | `float` | `sigma` | hyperparameter for gaussian method |
| `float` | `min_score` | score filter threshold | | `float` | `min_score` | score filter threshold |
...@@ -107,7 +107,7 @@ Perform RoIAlign on output feature, used in bbox_head of most two-stage detector ...@@ -107,7 +107,7 @@ Perform RoIAlign on output feature, used in bbox_head of most two-stage detector
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | ---------------- | ------------------------------------------------------------------------------------------------------------- | |---------|------------------|---------------------------------------------------------------------------------------------------------------|
| `int` | `output_height` | height of output roi | | `int` | `output_height` | height of output roi |
| `int` | `output_width` | width of output roi | | `int` | `output_width` | width of output roi |
| `float` | `spatial_scale` | used to scale the input boxes | | `float` | `spatial_scale` | used to scale the input boxes |
...@@ -144,7 +144,7 @@ Filter out boxes has high IoU overlap with previously selected boxes. ...@@ -144,7 +144,7 @@ Filter out boxes has high IoU overlap with previously selected boxes.
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ------- | --------------- | ---------------------------------------------------------------------------------------------------------------- | |---------|-----------------|------------------------------------------------------------------------------------------------------------------|
| `float` | `iou_threshold` | The threshold for deciding whether boxes overlap too much with respect to IoU. Value range [0, 1]. Default to 0. | | `float` | `iou_threshold` | The threshold for deciding whether boxes overlap too much with respect to IoU. Value range [0, 1]. Default to 0. |
| `int` | `offset` | 0 or 1, boxes' width or height is (x2 - x1 + offset). | | `int` | `offset` | 0 or 1, boxes' width or height is (x2 - x1 + offset). |
...@@ -177,7 +177,7 @@ Perform sample from `input` with pixel locations from `grid`. ...@@ -177,7 +177,7 @@ Perform sample from `input` with pixel locations from `grid`.
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ----- | -------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | |-------|----------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| `int` | `interpolation_mode` | Interpolation mode to calculate output values. (0: `bilinear` , 1: `nearest`) | | `int` | `interpolation_mode` | Interpolation mode to calculate output values. (0: `bilinear` , 1: `nearest`) |
| `int` | `padding_mode` | Padding mode for outside grid values. (0: `zeros`, 1: `border`, 2: `reflection`) | | `int` | `padding_mode` | Padding mode for outside grid values. (0: `zeros`, 1: `border`, 2: `reflection`) |
| `int` | `align_corners` | If `align_corners=1`, the extrema (`-1` and `1`) are considered as referring to the center points of the input's corner pixels. If `align_corners=0`, they are instead considered as referring to the corner points of the input's corner pixels, making the sampling more resolution agnostic. | | `int` | `align_corners` | If `align_corners=1`, the extrema (`-1` and `1`) are considered as referring to the center points of the input's corner pixels. If `align_corners=0`, they are instead considered as referring to the corner points of the input's corner pixels, making the sampling more resolution agnostic. |
...@@ -211,7 +211,7 @@ Perform CornerPool on `input` features. Read [CornerNet -- Detecting Objects as ...@@ -211,7 +211,7 @@ Perform CornerPool on `input` features. Read [CornerNet -- Detecting Objects as
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ----- | --------- | ---------------------------------------------------------------- | |-------|-----------|------------------------------------------------------------------|
| `int` | `mode` | corner pool mode, (0: `top`, 1: `bottom`, 2: `left`, 3: `right`) | | `int` | `mode` | corner pool mode, (0: `top`, 1: `bottom`, 2: `left`, 3: `right`) |
#### Inputs #### Inputs
...@@ -241,7 +241,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative maximum e ...@@ -241,7 +241,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative maximum e
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ----- | --------- | -------------------------------------- | |-------|-----------|----------------------------------------|
| `int` | `dim` | the dimension to do the operation over | | `int` | `dim` | the dimension to do the operation over |
#### Inputs #### Inputs
...@@ -273,7 +273,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative minimum e ...@@ -273,7 +273,7 @@ Returns a tuple (`values`, `indices`) where `values` is the cumulative minimum e
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| ----- | --------- | -------------------------------------- | |-------|-----------|----------------------------------------|
| `int` | `dim` | the dimension to do the operation over | | `int` | `dim` | the dimension to do the operation over |
#### Inputs #### Inputs
...@@ -305,7 +305,7 @@ Perform Modulated Deformable Convolution on input feature, read [Deformable Conv ...@@ -305,7 +305,7 @@ Perform Modulated Deformable Convolution on input feature, read [Deformable Conv
#### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| -------------- | ------------------- | ------------------------------------------------------------------------------------- | |----------------|---------------------|---------------------------------------------------------------------------------------|
| `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW) | | `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW) |
| `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW) | | `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW) |
| `list of ints` | `dilation` | The spacing between kernel elements. (dH, dW) | | `list of ints` | `dilation` | The spacing between kernel elements. (dH, dW) |
...@@ -338,16 +338,16 @@ Perform Modulated Deformable Convolution on input feature, read [Deformable Conv ...@@ -338,16 +338,16 @@ Perform Modulated Deformable Convolution on input feature, read [Deformable Conv
- T:tensor(float32, Linear) - T:tensor(float32, Linear)
## MMCVDeformConv2d ### MMCVDeformConv2d
### Description #### Description
Perform Deformable Convolution on input feature, read [Deformable Convolutional Network](https://arxiv.org/abs/1703.06211) for detail. Perform Deformable Convolution on input feature, read [Deformable Convolutional Network](https://arxiv.org/abs/1703.06211) for detail.
### Parameters #### Parameters
| Type | Parameter | Description | | Type | Parameter | Description |
| -------------- | ------------------ | --------------------------------------------------------------------------------------------------------------------------------- | |----------------|--------------------|-----------------------------------------------------------------------------------------------------------------------------------|
| `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW) | | `list of ints` | `stride` | The stride of the convolving kernel. (sH, sW) |
| `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW) | | `list of ints` | `padding` | Paddings on both sides of the input. (padH, padW) |
| `list of ints` | `dilation` | The spacing between kernel elements. (dH, dW) | | `list of ints` | `dilation` | The spacing between kernel elements. (dH, dW) |
...@@ -355,7 +355,7 @@ Perform Deformable Convolution on input feature, read [Deformable Convolutional ...@@ -355,7 +355,7 @@ Perform Deformable Convolution on input feature, read [Deformable Convolutional
| `int` | `group` | Split input into groups. `input_channel` should be divisible by the number of groups. | | `int` | `group` | Split input into groups. `input_channel` should be divisible by the number of groups. |
| `int` | `im2col_step` | DeformableConv2d use im2col to compute convolution. im2col_step is used to split input and offset, reduce memory usage of column. | | `int` | `im2col_step` | DeformableConv2d use im2col to compute convolution. im2col_step is used to split input and offset, reduce memory usage of column. |
### Inputs #### Inputs
<dl> <dl>
<dt><tt>inputs[0]</tt>: T</dt> <dt><tt>inputs[0]</tt>: T</dt>
...@@ -366,13 +366,13 @@ Perform Deformable Convolution on input feature, read [Deformable Convolutional ...@@ -366,13 +366,13 @@ Perform Deformable Convolution on input feature, read [Deformable Convolutional
<dd>Input weight; 4-D tensor of shape (output_channel, input_channel, kH, kW).</dd> <dd>Input weight; 4-D tensor of shape (output_channel, input_channel, kH, kW).</dd>
</dl> </dl>
### Outputs #### Outputs
<dl> <dl>
<dt><tt>outputs[0]</tt>: T</dt> <dt><tt>outputs[0]</tt>: T</dt>
<dd>Output feature; 4-D tensor of shape (N, output_channel, outH, outW).</dd> <dd>Output feature; 4-D tensor of shape (N, output_channel, outH, outW).</dd>
</dl> </dl>
### Type Constraints #### Type Constraints
- T:tensor(float32, Linear) - T:tensor(float32, Linear)
docs/deployment/onnxruntime_op.md
View file @ 72182747
## Custom operators for ONNX Runtime in MMCV ## ONNX Runtime Deployment
### Introduction of ONNX Runtime ### Introduction of ONNX Runtime
...@@ -15,15 +15,15 @@ ...@@ -15,15 +15,15 @@
### List of operators for ONNX Runtime supported in MMCV ### List of operators for ONNX Runtime supported in MMCV
| Operator | CPU | GPU | MMCV Releases | | Operator | CPU | GPU | MMCV Releases |
| :----------------------------------------------------: | :---: | :---: | :-----------: | |:-------------------------------------------------------|:---:|:---:|:-------------:|
| [SoftNMS](onnxruntime_custom_ops.md#softnms) | Y | N | 1.2.3 | | [SoftNMS](onnxruntime_custom_ops.md#softnms) | Y | N | 1.2.3 |
| [RoIAlign](onnxruntime_custom_ops.md#roialign) | Y | N | 1.2.5 | | [RoIAlign](onnxruntime_custom_ops.md#roialign) | Y | N | 1.2.5 |
| [NMS](onnxruntime_custom_ops.md#nms) | Y | N | 1.2.7 | | [NMS](onnxruntime_custom_ops.md#nms) | Y | N | 1.2.7 |
| [grid_sampler](onnxruntime_custom_ops.md#grid_sampler) | Y | N | 1.3.1 | | [grid_sampler](onnxruntime_custom_ops.md#grid_sampler) | Y | N | 1.3.1 |
| [CornerPool](onnxruntime_custom_ops.md#cornerpool) | Y | N | 1.3.4 | | [CornerPool](onnxruntime_custom_ops.md#cornerpool) | Y | N | 1.3.4 |
| [cummax](onnxruntime_custom_ops.md#cummax) | Y | N | master | | [cummax](onnxruntime_custom_ops.md#cummax) | Y | N | 1.3.4 |
| [cummin](onnxruntime_custom_ops.md#cummin) | Y | N | master | | [cummin](onnxruntime_custom_ops.md#cummin) | Y | N | 1.3.4 |
### How to build custom operators for ONNX Runtime ### How to build custom operators for ONNX Runtime
...@@ -88,6 +88,8 @@ onnx_results = sess.run(None, {'input' : input_data}) ...@@ -88,6 +88,8 @@ onnx_results = sess.run(None, {'input' : input_data})
#### Reminder #### Reminder
- *Please note that this feature is experimental and may change in the future. Strongly suggest users always try with the latest master branch.*
- The custom operator is not included in [supported operator list](https://github.com/microsoft/onnxruntime/blob/master/docs/OperatorKernels.md) in ONNX Runtime. - The custom operator is not included in [supported operator list](https://github.com/microsoft/onnxruntime/blob/master/docs/OperatorKernels.md) in ONNX Runtime.
- The custom operator should be able to be exported to ONNX. - The custom operator should be able to be exported to ONNX.
......
docs/deployment/tensorrt_plugin.md
View file @ 72182747
## TensorRT Plugins for custom operators in MMCV (Experimental) ## TensorRT Deployment
<!-- TOC --> <!-- TOC -->
- [TensorRT Plugins for custom operators in MMCV (Experimental)](#tensorrt-plugins-for-custom-operators-in-mmcv-experimental) - [TensorRT Deployment](#tensorrt-deployment)
- [Introduction](#introduction) - [Introduction](#introduction)
- [List of TensorRT plugins supported in MMCV](#list-of-tensorrt-plugins-supported-in-mmcv) - [List of TensorRT plugins supported in MMCV](#list-of-tensorrt-plugins-supported-in-mmcv)
- [How to build TensorRT plugins in MMCV](#how-to-build-tensorrt-plugins-in-mmcv) - [How to build TensorRT plugins in MMCV](#how-to-build-tensorrt-plugins-in-mmcv)
...@@ -24,17 +24,17 @@ To ease the deployment of trained models with custom operators from `mmcv.ops` u ...@@ -24,17 +24,17 @@ To ease the deployment of trained models with custom operators from `mmcv.ops` u
### List of TensorRT plugins supported in MMCV ### List of TensorRT plugins supported in MMCV
| ONNX Operator | TensorRT Plugin | MMCV Releases | | ONNX Operator | TensorRT Plugin | MMCV Releases |
| :-----------------------: | :-----------------------------------------------------------------------------: | :-----------: | |:--------------------------|:--------------------------------------------------------------------------------|:-------------:|
| MMCVRoiAlign | [MMCVRoiAlign](./tensorrt_custom_ops.md#mmcvroialign) | 1.2.6 | | MMCVRoiAlign | [MMCVRoiAlign](./tensorrt_custom_ops.md#mmcvroialign) | 1.2.6 |
| ScatterND | [ScatterND](./tensorrt_custom_ops.md#scatternd) | 1.2.6 | | ScatterND | [ScatterND](./tensorrt_custom_ops.md#scatternd) | 1.2.6 |
| NonMaxSuppression | [NonMaxSuppression](./tensorrt_custom_ops.md#nonmaxsuppression) | 1.3.0 | | NonMaxSuppression | [NonMaxSuppression](./tensorrt_custom_ops.md#nonmaxsuppression) | 1.3.0 |
| MMCVDeformConv2d | [MMCVDeformConv2d](./tensorrt_custom_ops.md#mmcvdeformconv2d) | 1.3.0 | | MMCVDeformConv2d | [MMCVDeformConv2d](./tensorrt_custom_ops.md#mmcvdeformconv2d) | 1.3.0 |
| grid_sampler | [grid_sampler](./tensorrt_custom_ops.md#grid-sampler) | 1.3.1 | | grid_sampler | [grid_sampler](./tensorrt_custom_ops.md#grid-sampler) | 1.3.1 |
| cummax | [cummax](./tensorrt_custom_ops.md#cummax) | 1.3.5 | | cummax | [cummax](./tensorrt_custom_ops.md#cummax) | 1.3.5 |
| cummin | [cummin](./tensorrt_custom_ops.md#cummin) | 1.3.5 | | cummin | [cummin](./tensorrt_custom_ops.md#cummin) | 1.3.5 |
| MMCVInstanceNormalization | [MMCVInstanceNormalization](./tensorrt_custom_ops.md#mmcvinstancenormalization) | 1.3.5 | | MMCVInstanceNormalization | [MMCVInstanceNormalization](./tensorrt_custom_ops.md#mmcvinstancenormalization) | 1.3.5 |
| MMCVModulatedDeformConv2d | [MMCVModulatedDeformConv2d](./tensorrt_custom_ops.md#mmcvmodulateddeformconv2d) | master | | MMCVModulatedDeformConv2d | [MMCVModulatedDeformConv2d](./tensorrt_custom_ops.md#mmcvmodulateddeformconv2d) | 1.3.8 |
Notes Notes
...@@ -163,6 +163,8 @@ Below are the main steps: ...@@ -163,6 +163,8 @@ Below are the main steps:
#### Reminders #### Reminders
- *Please note that this feature is experimental and may change in the future. Strongly suggest users always try with the latest master branch.*
- Some of the [custom ops](https://mmcv.readthedocs.io/en/latest/ops.html) in `mmcv` have their cuda implementations, which could be referred. - Some of the [custom ops](https://mmcv.readthedocs.io/en/latest/ops.html) in `mmcv` have their cuda implementations, which could be referred.
### Known Issues ### Known Issues
......
docs/index.rst
View file @ 72182747
...@@ -29,12 +29,10 @@ You can switch between Chinese and English documents in the lower-left corner of ...@@ -29,12 +29,10 @@ You can switch between Chinese and English documents in the lower-left corner of
:maxdepth: 2 :maxdepth: 2
:caption: Deployment :caption: Deployment
deployment/mmcv_ops_definition.md
deployment/onnx.md deployment/onnx.md
deployment/onnxruntime_op.md deployment/onnxruntime_op.md
deployment/onnxruntime_custom_ops.md
deployment/tensorrt_plugin.md deployment/tensorrt_plugin.md
deployment/tensorrt_custom_ops.md
deployment/mmcv_ops_definition.md
.. toctree:: .. toctree::
:maxdepth: 2 :maxdepth: 2
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment