docs: add PyTorch Profiler support details to TensorBoard documentation #4615
Conversation
📝 WalkthroughWalkthroughThis pull request updates the TensorBoard documentation by reformatting the JSON configuration block for the training section and adding a new section titled "PyTorch Profiler With TensorBoard." The new section provides installation instructions for the necessary packages, configuration guidelines, and details on visualizing profiling data within TensorBoard. The command for launching TensorBoard has also been updated for clarity. Changes
Sequence Diagram(s)sequenceDiagram
participant User
participant Profiler
participant TensorBoard
User->>Profiler: Executes training with profiler enabled
Profiler->>TensorBoard: Generates and exports profiling data
User->>TensorBoard: Launches TensorBoard to view profiling data
TensorBoard-->>User: Displays profiling visualizations
Suggested labels
Suggested reviewers
Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (2)
doc/train/tensorboard.md (2)
29-51: JSON Configuration Block ImprovementThe JSON example has several duplicate
_commentkeys. In strict JSON, duplicate keys can lead to undefined behavior where only the last instance is preserved. Since this example is intended for documentation purposes, consider either using uniquely named comment keys (e.g.,_comment_1,_comment_2) or moving explanatory comments outside the JSON block to avoid confusion for readers who may copy and use the snippet.
64-73: New PyTorch Profiler Section ReviewThe newly added PyTorch Profiler With TensorBoard section is well integrated and provides concise steps on installing
tensorboardandtorch-tb-profiler, setting"enable_profiler": true, and running a short training task. It might be helpful to include a brief caution or note regarding any potential performance overhead when profiling is enabled. This will prepare users for any increase in execution time during profiling.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
doc/train/tensorboard.md(1 hunks)
🔇 Additional comments (1)
doc/train/tensorboard.md (1)
57-57: Clarify Log Directory InstructionThis additional note—"TensorBoard requires a logdir to read logs from..."—is clear. For further clarity, you might briefly mention what a user can expect from running
tensorboard --help(for example, configuration details or option listings). Overall, this addition improves the guidance.
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## devel #4615 +/- ##
=======================================
Coverage 84.58% 84.58%
=======================================
Files 680 680
Lines 64544 64544
Branches 3538 3540 +2
=======================================
Hits 54596 54596
Misses 8807 8807
Partials 1141 1141 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
…on (deepmodeling#4615) This PR adds instructions on how to profile with the PyTorch backend. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Documentation** - Added a new section detailing the integration of PyTorch Profiler with TensorBoard. - Provided clear instructions on package installation, configuration adjustments, and how to visualize profiling data. - Enhanced the readability of commands and the overall formatting of the training documentation. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Han Wang <92130845+wanghan-iapcm@users.noreply.github.com> (cherry picked from commit 80d445b)
This PR adds instructions on how to profile with the PyTorch backend.
Summary by CodeRabbit