关于GCU、沐曦GPGPU、MLU、0卡V100资源4月7日恢复上架的公告>>> 关于共建具身智能开源数据集的倡议>>> 关于云脑任务中统一路径访问方式的公告>>> 关于将启智集群GPU资源迁移至智算集群的公告>>>
gbstack
Public Activity Repositories Datasets Starred Repositories
0
 Following
0
 Followers
0
 Achievement Badge
1
Loading Heatmap…

gbstack pushed to mask_rcnn_slim at gbstack/PaddleDetection

  • b002d9866f change learning rate, schedule, warmup and weight decay
  • ff88185d8b add mask rcnn pruning configuration file

2 years ago

gbstack pushed to multi-scale-test at gbstack/PaddleDetection

  • 5322e072e0 do not import nms method from static.ppdet.utils.post_process
  • 7ecbb8c4e3 use isinstance for sequence type checking. add empty checking for samples sequence
  • 4e31d8c718 update documentation
  • a9d99ca7d2 add image files for testing. use relative path for input images
  • d229c30a25 load pretrained weights from URL

2 years ago

gbstack pushed to ppyolo-dali at gbstack/PaddleDetection

  • f9d9bfe0f1 use DALI for data preprocessing if dataset is DALICOCODataSet
  • 334ec682ab Read resize option from DALI pipeline. Set use_dali for all batch_transforms. Apply batch transforms after loading data from DALI pipeline. Convert image loaded from DALI pipeline to paddle Tensor.
  • a4518efe3a allow using dali for NormalizeBox, RandomCrop, RandomExpand, Resize, RandomFlip, RandomDistort, NormalizeImage and Permute operations
  • 93d1337330 extract Compose and BatchCompose into separate file to prevent circular import
  • adaaa6549c add use_dali parameter for BatchRandomResize and Gt2YoloTarget. allow setting resize size from outer code

2 years ago

gbstack pushed to release/2.2 at gbstack/PaddleDetection

2 years ago

gbstack pushed to paa-on-develop at gbstack/PaddleDetection

  • e213a71c06 add PAAFPN
  • c4d2ba18dd change GMM covariance_type to 'full'. use all 0 value for bg class instead of using a separate bg class when providing label for focal loss. add comparing debugging code (comparing between current implementation and original one). calculate anchor score and reassign for one batch everytime, instead of combining batches and process all. use giou loss for bbox regression. set loss weight.
  • 3f901e6253 use GroupNorm for cls_convs and reg_convs
  • 0e43717faf fixed calling cast on empty tensor will fail
  • b345ca5fc6 allow using GroupNorm for norm_type parameter of ConvNormLayer

2 years ago

gbstack created repository gbstack/PaddleDetection

2 years ago

Please read the following content carefully:

Dear OpenI User

Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.

For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》

Agree and continue
Disagree, exit