-
通过paddle_lite_opt参数--valid_targets=huawei_ascend_npu,arm导出nb模型,在Ascend NPU arm64中出现No device is specified
错误信息
"
[F 12/16 7:53:31.217 ...le-Lite/lite/kernels/nnadapter/engine.cc:281 Engine] Check …
L-lei updated
2 years ago
-
就在PaddleDetection的首页里面描述：
PP-YOLO在COCO数据集精度45.9%，Tesla V100预测速度72.9FPS，精度速度均优于YOLOv4
PP-YOLO v2是对PP-YOLO模型的进一步优化，在COCO数据集精度49.5%，Tesla V100预测速度68.9FPS
-
## Steps to Reproduce
**Expected results:**
An instance of SharedPreferences should be returned.
**Actual results:**
Initialization fails with: `Unhandled Exception: Null check opera…
-
# 【PaddlePaddle Hackathon】任务总览
- **NEWS：** 本次黑客松活动，线上部分已结束，欢迎大家继续认领&完成感兴趣的任务，可以@TCChenlong review相关PR；此外，欢迎大家参与报名线下的 Coding Party ，报名表见：[2021飞桨黑客松 48H Coding Party 报名表](https://www.wjx.top/vj/e0QwY…
-
开发者您好，我因为使用Win10系统，所以没法使用主流的目标检测框架（detectron2或者mmdetection等），考虑使用paddledetection完成我的研究。paddle框架下的模型训练很正常，检测精度也很高，但为什么检测速度会比pytorch框架的模型或代码差那么多呢？
我使用
1、[YOLOv5](https://github.com/ultralytics/yolov5)…
-
(PP-YOLO) wtc@wtc-All-Series:~/PaddleDetection$ python tools/train.py -c configs/ppyolo/ppyolo_r50vd_dcn_voc.yml
```
/home/wtc/anaconda3/envs/PP-YOLO/lib/python3.8/site-packages/paddle/tensor/crea…
-
Hi,
we have tested the pretreined model on COCO dataset and we received only ~11 FPS using Tesla V100. We used the pretrained COCO model, which performance is reported as 72.9 FPS. We used this model…
-
请问PaddleDetection飞桨目标检测开发套件公布的代码中不包含YOLOv4、PP-YOLOv2吗？官方有发布YOLOv4和PP-YOLOv2的配置文件吗
-
Hi,
I am using ppyolo for inference. But when I set batch_size more than 1 in ppyolo_reader.yml, an error occurs.
My command is
`python tools/infer_renqun.py -c config s/ppyolo/ppyolo.yml -…
-
我将inference_product.yaml钟的det_inference_model_dir替换为ppyolo_r50vd_dcn_1x_coco模型，则会报错：
InvalidArgumentError: The 2-th dimension of input[0] and input[1] is expected to be equal.But received input[0…