-
Notifications
You must be signed in to change notification settings - Fork 655
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[timeseries] add some basic block and deepAR model #2027
Conversation
Codecov ReportBase: 72.08% // Head: 71.36% // Decreases project coverage by
Additional details and impacted files@@ Coverage Diff @@
## master #2027 +/- ##
============================================
- Coverage 72.08% 71.36% -0.73%
- Complexity 5126 6275 +1149
============================================
Files 473 624 +151
Lines 21970 27795 +5825
Branches 2351 2997 +646
============================================
+ Hits 15838 19835 +3997
- Misses 4925 6500 +1575
- Partials 1207 1460 +253
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
ae48241
to
c873fc3
Compare
* and limitations under the License. | ||
*/ | ||
|
||
package ai.djl.timeseries.block; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please add unit test
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Got it.
) * feature: add TimeSeriesDataset and training transform * feature: some basic block and deepar model * feature: add train example * feature: add m5-demo and air passengers demo Co-authored-by: Carkham <1302112560@qq.com> Co-authored-by: Frank Liu <frankfliu2000@gmail.com> Co-authored-by: KexinFeng <fengx463@umn.edu>
Description
This PR adds two new sections for creating trainable models.
The first part is embedding and scaler, which are responsible for encoding the input features and scaling the target values.
The second part is that we implemented a deepAR network model with the above-mentioned blcok, which refers to the implementation in gluonTS.
Since the model predicts a probability distribution, we need to override the different forward methods when training and inference with different block.