We provide a script to convert the model to [ONNX](https://github.com/onnx/onnx) format. The converted model could be visualized by tools like [Netron](https://github.com/lutzroeder/netron). Besides, we also support comparing the output results between Pytorch and ONNX model.
**Note**: This tool is still experimental. For now, some customized operators are not supported, and we only support a subset of detection and recognition algorithms.
| DBNet | [dbnet_r18_fpnc_1200e_icdar2015.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/dbnet/dbnet_r18_fpnc_1200e_icdar2015.py) | Y | N | |
| PSENet | [psenet_r50_fpnf_600e_ctw1500.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/psenet/psenet_r50_fpnf_600e_ctw1500.py) | Y | Y | |
| PSENet | [psenet_r50_fpnf_600e_icdar2015.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/psenet/psenet_r50_fpnf_600e_icdar2015.py) | Y | Y | |
| PANet | [panet_r18_fpem_ffm_600e_ctw1500.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/panet/panet_r18_fpem_ffm_600e_ctw1500.py) | Y | Y | |
| PANet | [panet_r18_fpem_ffm_600e_icdar2015.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/panet/panet_r18_fpem_ffm_600e_icdar2015.py) | Y | Y | |
| CRNN | [crnn_academic_dataset.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textrecog/crnn/crnn_academic_dataset.py) | Y | Y | CRNN only accepts input with height 32 |
We also provide a script to convert [ONNX](https://github.com/onnx/onnx) model to [TensorRT](https://github.com/NVIDIA/TensorRT) format. Besides, we support comparing the output results between ONNX and TensorRT model.
| `model_config` | str | The path to a model config file. |
| `model_type` | 'recog', 'det' | The model type of the config file. |
| `image_path` | str | The path to input image file. |
| `onnx_file` | str | The path to input ONNX file. |
| `--trt-file` | str | The path of output TensorRT model. Defaults to `tmp.trt`. |
| `--max-shape` | int * 4 | Maximum shape of model input. |
| `--min-shape` | int * 4 | Minimum shape of model input. |
| `--workspace-size`| int | Max workspace size in GiB. Defaults to 1. |
| `--fp16`| bool | Determines whether to export TensorRT with fp16 mode. Defaults to `False`. |
| `--verify`| bool | Determines whether to verify the correctness of an exported model. Defaults to `False`. |
| `--show`| bool | Determines whether to show the output of ONNX and TensorRT. Defaults to `False`. |
| `--verbose`| bool | Determines whether to verbose logging messages while creating TensorRT engine. Defaults to `False`. |
**Note**: This tool is still experimental. For now, some customized operators are not supported, and we only support a subset of detection and recognition algorithms.
### List of supported models exportable to TensorRT
| DBNet | [dbnet_r18_fpnc_1200e_icdar2015.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/dbnet/dbnet_r18_fpnc_1200e_icdar2015.py) | Y | N | |
| PSENet | [psenet_r50_fpnf_600e_ctw1500.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/psenet/psenet_r50_fpnf_600e_ctw1500.py) | Y | Y | |
| PSENet | [psenet_r50_fpnf_600e_icdar2015.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/psenet/psenet_r50_fpnf_600e_icdar2015.py) | Y | Y | |
| PANet | [panet_r18_fpem_ffm_600e_ctw1500.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/panet/panet_r18_fpem_ffm_600e_ctw1500.py) | Y | Y | |
| PANet | [panet_r18_fpem_ffm_600e_icdar2015.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textdet/panet/panet_r18_fpem_ffm_600e_icdar2015.py) | Y | Y | |
| CRNN | [crnn_academic_dataset.py](https://github.com/open-mmlab/mmocr/blob/main/configs/textrecog/crnn/crnn_academic_dataset.py) | Y | Y | CRNN only accepts input with height 32 |
**Notes**:
- *All models above are tested with Pytorch==1.8.1, onnxruntime==1.7.0 and tensorrt==7.2.1.6*
To evaluate ONNX and TensorRT models, ONNX, ONNXRuntime and TensorRT should be installed first. Install `mmcv-full` with ONNXRuntime custom ops and TensorRT plugins follow [ONNXRuntime in mmcv](https://mmcv.readthedocs.io/en/latest/onnxruntime_op.html) and [TensorRT plugin in mmcv](https://github.com/open-mmlab/mmcv/blob/master/docs/tensorrt_plugin.md).
- TensorRT upsampling operation is a little different from PyTorch. For DBNet and PANet, we suggest replacing upsampling operations with the nearest mode to operations with bilinear mode. [Here](https://github.com/open-mmlab/mmocr/blob/50a25e718a028c8b9d96f497e241767dbe9617d1/mmocr/models/textdet/necks/fpem_ffm.py#L33) for PANet, [here](https://github.com/open-mmlab/mmocr/blob/50a25e718a028c8b9d96f497e241767dbe9617d1/mmocr/models/textdet/necks/fpn_cat.py#L111) and [here](https://github.com/open-mmlab/mmocr/blob/50a25e718a028c8b9d96f497e241767dbe9617d1/mmocr/models/textdet/necks/fpn_cat.py#L121) for DBNet. As is shown in the above table, networks with tag * mean the upsampling mode is changed.
- Note that changing upsampling mode reduces less performance compared with using the nearest mode. However, the weights of networks are trained through the nearest mode. To pursue the best performance, using bilinear mode for both training and TensorRT deployment is recommended.
- All ONNX and TensorRT models are evaluated with dynamic shapes on the datasets, and images are preprocessed according to the original config file.
- This tool is still experimental, and we only support a subset of detection and recognition algorithms for now.
