googleapis / googleapis/python-aiplatform
aiplatofrm.log_model() to represent itself under Model section instead of Parameters
- 主要语言
- Python
- 星标
- 905
- 派生
- 465
- 平均合并
- 1 天 13 小时
- 30 天内合并 PR
- 44
描述
Current `aiplatform.log_model()` method allows for model registry into experiment tracking. However, instead of representing model under separate MODEL section, it just adds it under specified name under ARTIFACTS, that I think is not optimal: such as in case of multiple and many artifacts the ease of finding the MODEL parameter is not optimal (lost in the list of multiple artifacts that are itemized under Artifact ID). Instead I suggest the model to be placed in a separate MODEL/MODELS section similar to PARAMETERS and ARTIFACTS; surely MODEL being an artifact, but having it under separate section improves readibility.
贡献指南
调研方向
首先跟踪 aiplatform.log_model() 如何在实验跟踪中记录模型,以及 PARAMETERS 和 ARTIFACTS 部分是如何表示的。将当前的 artifact 列表与所请求的 MODEL/MODELS 部分进行比较,然后确定受影响的实现和测试。当已记录的模型可以在其专属部分中被发现,同时不丢失与 artifact 的关系时,即表示完成。
由索引模型根据 Issue 内容生成。
评估
- 技术栈
- google-cloud, python
- 领域
- cloud, machine-learning
- Issue 类型
- 功能
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 活跃度
- 停滞
- 描述清晰度
- 需要澄清
- 新手友好度
- 30/100