-
##### System information (version)
- OpenVINO Source=> Runtime - Intel Archive
- OpenVINO Version=> Version 2023.0
- Operating System / Platform => L4T ( Linux for Tegra - based on Ubuntu 18.0…
-
To align KerasCV with the KerasNLP API, we're migrating our backbone models to subclass keras_cv.models.Backbone.
This allows us to have a more structured, modular API for high-level task models, inc…
-
## Description
### What steps will reproduce the problem?
1. Import PyTorch
2. Run code, works fine
3. Run code again
4. Error occurs
This is an issue with Spyder and nothing else, as I can do …
-
(ppocr) F:\python_project\PaddleOCR>python tools/train.py -c configs\det\my_det_mv3_db.yml
[2023/09/18 11:46:43] ppocr INFO: Architecture :
[2023/09/18 11:46:43] ppocr INFO: Backbone :
[2023/…
-
[2022/09/13 15:24:00] ppocr INFO: valid dataloader has 31 iters
Traceback (most recent call last):
File "tools/train.py", line 202, in
main(config, device, logger, vdl_writer)
File "tools…
-
请问下ocr版本release/2.1 release/2.2 release/2.3区别在哪呢？他们是分支？？直接就发布2.3？
为什么两段式算法中识别部分有些模型差别很大？
哪能去看代码提交记录吗，想看看当时是为什么那么修改
2.3版本训练backbone为MobileNetV3的db算法识别模型
一样的预训练模型 一样的配置参数 识别验证集准确率只有40%
2.1 可以达…
-
-
### 🐛 Describe the bug
I am trying to perform object detection using ssdlite with mobilenetv3 as backbone. I would like to use the pre-trained model present on your repository and retrain a part of…
-
I tried to validate the pretrained MobileNet V2 and V3 models available at keras.applications.MobileNetV2() and keras.applications.MobileNetV3(). To my surprise, both yielded lower-than-expected T…
-
- 系统环境/System Environment：npu_x86
- 版本号/Version：Paddle： PaddleOCR amp
- 运行指令/Command Code：python3 tools/train.py -c configs/det/det_mv3_db.yml \
-o Global.pretrained_model=./pretrain_models…