-
### bug描述 Describe the Bug
#保存模型的代码
DeepAR.save("./DeepAR",network_model=True,dygraph_to_static=True)
#使用paddle inference 推理
from paddlets.models.model_loader import load
from paddlets.datasets…
-
- Is there a "right way" to handle missing values for this model?
- How should I pad when time series have non-uniform lengths in a batch?
-
Hi,
I have noticed that the CPU utilization when using deployed MOJO models for forecasting is low (in my case, I am using a DRF model and CPU utilization is below 10%). Is there a way to speed up …
-
Now the unit tests for the time series forecasting are quite formal and the low-quality predictions can still be fine in the test (because the benchmarks are too simple or the pass criteria are too we…
-
Moving this point flagged up by @sultanorazbayev to the bug tracker for investigation.
The issue is with the line `NaiveForecaster(strategy='drift'))]), 'pipeline_y__forecaster__strategy': 'last'}`…
-
this can be done as activity: giving each team dataset, its description and research question. and ask whether it is ok to use ML:
-
- PyTorch-Forecasting version: 0.10.2
- PyTorch version: 1.11.0
- Python version: 3.9.12
- Operating System: Windows 7
### Expected behavior
I executed code like in the tutorial [Demand fore…
-
**Describe the bug**
When fitting N models one for each forecasting horizon using **make_reduction** with **strategy="direct"** and **pooling="global"**, the dataset that is used to train the model…
-
Just create this dataset:
```python
import numpy as np
import pandas as pd
multi_target_test_data = pd.DataFrame(
dict(
target1=np.random.rand(30),
target2=np.random.ran…
-
**Combine neural nets with statistical models**
Some papers/competitions have shown that combining statistical models with neural networks can increase accuracy. An example is the winner of the M4-co…