amitesh863 / nbeats_forecast Goto Github PK

View Code? Open in Web Editor NEW

82.0 82.0 18.0 60 KB

N-Beats library implementation

License: MIT License

Python 100.00%

nbeats_forecast's People

Contributors

Stargazers

Watchers

Forkers

jwc92 katerega sususushi zeroonekevin taogeanton2 xhfei1224 joewillbe panda-puff wittshan robbyan koushikl0l dal3006 lavender912 dainguyenvan clarenceke jiawei0322 gaoxiaojun surajitdb

nbeats_forecast's Issues

Funkiness with CHECKPOINT_NAME

When training a new model, "checkpoint_file_name" kwarg needs to be set (or model.CHECKPOINT_NAME set directly). Leaving it at default value can produce host of size mismatch errors that don't give information as to what's going on. Suggest raising error in that case instead of allowing "RuntimeError" via torch.

TypeError: init() got an unexpected keyword argument 'thetas_dims'

model = NBeats(data=data, period_to_forecast=194, train_percent=0.75)

This code used to work but now it won't. Is this beacause of using wrong version of packages?

Where is the NBeatsNet ?

As declared in nbeats_forecast.py , from nbeats_pytorch.model import NBeatsNet

But I can't find the model.py to see the code of NBeatsNet.

data_generator() method is wrong

In data_generator, you define a split() method to split data into multiple batches, but the while condition len(arr) > size is incorrect because arr = (x_full, y_full) which means len(arr) is equal to 2 forever. I found this bug when I debug into your code. Besides, I can't find the definition of the model, could you please show me your complete code in github.

Positional arguments error

initiating model produces the following error:

init() takes from 4 to 6 positional arguments but 7 were given

can you please look into it?

Cuda doesn't work

using mode="cuda"

gives






| N-Beats
| --  Stack Generic (#0) (share_weights_in_stack=False)
     | -- GenericBlock(units=128, thetas_dim=7, backcast_length=36, forecast_length=12, share_thetas=False) at @2157678533888
     | -- GenericBlock(units=128, thetas_dim=7, backcast_length=36, forecast_length=12, share_thetas=False) at @2157678534032
     | -- GenericBlock(units=128, thetas_dim=7, backcast_length=36, forecast_length=12, share_thetas=False) at @2157678534128
| --  Stack Generic (#1) (share_weights_in_stack=False)
     | -- GenericBlock(units=128, thetas_dim=8, backcast_length=36, forecast_length=12, share_thetas=False) at @2157678533072
     | -- GenericBlock(units=128, thetas_dim=8, backcast_length=36, forecast_length=12, share_thetas=False) at @2157532770560
     | -- GenericBlock(units=128, thetas_dim=8, backcast_length=36, forecast_length=12, share_thetas=False) at @2157532720288
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
c:\Users\UserName\development\pythonDev\Stonks\tryingNBEATSforcast.ipynb Cell 4' in <cell line: 3>()
      1 model = NBeats(data=train, period_to_forecast=12,mode="cuda")
      2 # model.fit(optimiser=)
----> 3 model.fit(optimiser=Adam(model.parameters, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0.01, amsgrad=False),plot=False)
      4 forecast = model.predict()

File c:\Users\UserName\AppData\Local\Programs\Python\Python38\lib\site-packages\nbeats_forecast\nbeats_forecast.py:213, in NBeats.fit(self, epoch, optimiser, plot, verbose)
    211 test_losses = []
    212 for i in range(self.epoch):
--> 213     self.eval_test(test_losses, x_test, y_test)
    214     self.train_100_grad_steps(train_data, test_losses)
    215 if self.check_save:

File c:\Users\UserName\AppData\Local\Programs\Python\Python38\lib\site-packages\nbeats_forecast\nbeats_forecast.py:167, in NBeats.eval_test(self, test_losses, x_test, y_test)
    165 self.net.eval()
    166 _, forecast = self.net(torch.tensor(x_test, dtype=torch.float))
--> 167 test_losses.append(F.mse_loss(forecast, torch.tensor(y_test, dtype=torch.float)).item())
    168 p = forecast.detach().numpy()
    169 if self.plot:

File c:\Users\UserName\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\functional.py:3262, in mse_loss(input, target, size_average, reduce, reduction)
   3259     reduction = _Reduction.legacy_get_string(size_average, reduce)
   3261 expanded_input, expanded_target = torch.broadcast_tensors(input, target)
-> 3262 return torch._C._nn.mse_loss(expanded_input, expanded_target, _Reduction.get_enum(reduction))

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu!

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.

Jobs

Jooble