Comments (3)
A bit embarrassing, but there was a bug in Cached
: #3060
from gluonts.
@shchur thanks, great reproducer!
What puzzles me is that extrapolating the validation data loader constructor gives me the right length (i.e. number of validation batches) every time I go through them:
from gluonts.itertools import Cached
module = estimator.create_lightning_module()
transformation = estimator.create_transformation()
transformed_validation_data = transformation.apply(val, is_train=True)
transformed_validation_data = Cached(transformed_validation_data)
validation_data_loader = estimator.create_validation_data_loader(
transformed_validation_data,
module,
)
print(len(list(validation_data_loader)))
print(len(list(validation_data_loader)))
print(len(list(validation_data_loader)))
print(len(list(validation_data_loader)))
yields
313
313
313
313
from gluonts.
Not sure, it looks like some strange interplay with Lightning. I tried dropping a debugger here, and both len(validation_data_loader)
and len(list(validation_data_loader))
are equal to the expected length. However, the trainer only takes 1 batch from the validation loader.
Could we just replace the Cached
wrapper with list
?
if cache_data:
transformed_validation_data = list(transformed_validation_data)
from gluonts.
Related Issues (20)
- AssertionError when reproduction of Dlinear model HOT 18
- best way for non-periodicity data? HOT 5
- Update dependencies HOT 2
- Dependency clashes (in kaggle) HOT 1
- Convert ListDataset to Multivariate dataset with gluonts.dataset.multivariate_grouper HOT 2
- GluonTSDataError: Input for field "target" does not have the requireddimension (field: target, ndim observed: 2, expected ndim: 1) HOT 5
- Which datasets provided in the library are multivariate? HOT 1
- GPVAREstimator - AssertionError: HOT 3
- Numerical overflow in StdScaler HOT 1
- Training is extremely slow on Gluonts [Torch] HOT 1
- Unify QuantileOutput and DistributionOutput
- Unify the design of seq2seq PyTorch models
- Add support for all covariate types to all PyTorch models
- Change of behaviour from lightning logger HOT 2
- BUG No module named 'gluonts.torch.model.i_transformer' HOT 1
- The annotation of the parameter device of PyTorchPredictor is incomplete HOT 2
- Dataframe index is not uniformly spaced. If your dataframe contains data from multiple series in the same column ("long" format), consider constructing the dataset with `PandasDataset.from_long_dataframe` instead. HOT 1
- get_aggregate_metrics() returns `inf` for MSIS and MASE when just a single target series in the dataset is invariant across time
- Consolidate R methods HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from gluonts.