Study data
收藏资源简介:
将命令式深度学习程序迁移到图形执行的挑战:一项实证研究 文件描述 文件描述 commit_categorizations.csv 我们数据集中提交的分类。 commits.csv 我们数据集中的提交信息 datasets.csv 包含我们数据集的名称和描述。 issue_categorizations.csv 从我们的数据集中选择的问题的分类。 issues.csv 我们数据集中问题的信息。 pipeline_stages.csv DL 管道阶段及其各自的描述。 question_categories.csv 问题类别及其各自的描述。 issue_causes.csv 问题原因及其各自的描述。 problem_fixes.csv 问题修复及其各自的描述。 problem_symptoms.csv 问题症状及其各自的描述。 study_subjects_commits.csv 提交的项目数据。 study_subjects_issues.csv 问题的项目数据。列说明 commit_categorizations.csv 列说明 tf.function 相关修复?当找到与 tf.function 相关的错误修复时为 TRUE,否则为 FALSE。如果为 FALSE,则后续列值将为空。发现问题修复的阶段 DL 管道阶段。 issue_categorizations.csv 列说明 tf.function 相关问题?当发现与 tf.function 相关的错误时为 TRUE,否则为 FALSE。如果为 FALSE,则后续列值将为空。发现问题的阶段 DL 管道阶段。 GH_id GitHub 问题唯一标识符。 issues.csv 列 描述 GH_id GitHub 问题唯一标识符。
Challenges in Migrating Imperative Deep Learning Programs to Graph Execution: An Empirical Study File Descriptions commit_categorizations.csv: Categorizations of commits in our dataset. commits.csv: Commit information in our dataset. datasets.csv: Contains the names and descriptions of our dataset. issue_categorizations.csv: Categorizations of selected issues from our dataset. issues.csv: Information of issues in our dataset. pipeline_stages.csv: DL pipeline stages and their respective descriptions. question_categories.csv: Question categories and their respective descriptions. issue_causes.csv: Issue causes and their respective descriptions. problem_fixes.csv: Problem fixes and their respective descriptions. problem_symptoms.csv: Problem symptoms and their respective descriptions. study_subjects_commits.csv: Project commit data. study_subjects_issues.csv: Project issue data. Column Descriptions For commit_categorizations.csv: tf.function-related fix? TRUE if an error fix related to tf.function is found, otherwise FALSE. If FALSE, the values of subsequent columns will be empty. Stage where the problem fix was discovered: DL pipeline stage. For issue_categorizations.csv: tf.function-related issue? TRUE if an error related to tf.function is found, otherwise FALSE. If FALSE, the values of subsequent columns will be empty. Stage where the issue was discovered: DL pipeline stage. GH_id: Unique GitHub issue identifier. For issues.csv: GH_id: Unique GitHub issue identifier.




