You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

RELEASE.md 2.0 kB

5 years ago
5 years ago
123456789101112131415161718192021222324252627282930313233
  1. ## MindInsight
  2. # Release 0.2.0-alpha
  3. ## Major Features and Improvements
  4. * Parameter distribution graph (Histogram).
  5. Now you can use [`HistogramSummary`](https://www.mindspore.cn/api/zh-CN/master/api/python/mindspore/mindspore.ops.operations.html#mindspore.ops.operations.HistogramSummary) and MindInsight to record and visualize distribution info of tensors. See our [tutorial](https://www.mindspore.cn/tutorial/zh-CN/master/advanced_use/visualization_tutorials.html) for details.
  6. * Lineage support Custom information
  7. * GPU support
  8. * Model and dataset tracking linkage support
  9. ## Bugfixes
  10. * Fix graph bug when node name is empty.([#I1EQ5H](https://gitee.com/mindspore/mindinsight/issues/I1EQ5H?from=project-issue))
  11. * Fix sha256 checksum missing bug.([#I1EHVK](https://gitee.com/mindspore/mindinsight/issues/I1EHVK?from=project-issue))
  12. * Fix start/stop command exit-code incorrect.([#I1EY3F](https://gitee.com/mindspore/mindinsight/issues/I1EY3F?from=project-issue))
  13. * Reduce cyclomatic complexity of list_summary_directories.
  14. ([#I1DANB](https://gitee.com/mindspore/mindinsight/issues/I1DANB?from=project-issue))
  15. * Fix unsafe functions and duplication files and redundant codes.([#I1DDGF]https://gitee.com/mindspore/mindinsight/issues/I1DDGF?from=project-issue))
  16. ## Thanks to our Contributors
  17. Thanks goes to these wonderful people:
  18. Ye Huang, Weifeng Huang, Zhenzhong Kou, Pengting Luo, Hongzhang Li, Yongxiong Liang, Gongchang Ou, Hui Pan, Luyu Qiu, Junyan Qin, Kai Wen, Weining Wang, Yifan Xia, Yunshu Zhang, Ting Zhao
  19. Contributions of any kind are welcome!
  20. # Release 0.1.0-alpha
  21. * Training process observation
  22. * Provides and displays training process information, including computational graphs and training process indicators.
  23. * Training result tracing
  24. * Provides functions of tracing and visualizing model training parameter information, including filtering and sorting of training data, model accuracy and training hyperparameters.

MindInsight为MindSpore提供了简单易用的调优调试能力。在训练过程中,可以将标量、张量、图像、计算图、模型超参、训练耗时等数据记录到文件中,通过MindInsight可视化页面进行查看及分析。