Nixtla/neuralforecast

Issues

• [Auto] options to control randomness and number of epochs

  #1204 opened 3 days ago by yarnabrina
  0

• [Auto] Training of `Auto*` models always fail with `Expected a parent` error on `macOS`

  #1203 opened 4 days ago by yarnabrina
  0

• [BaseMultivariate] OutOfMemoryError on GPU

  #1202 opened 5 days ago by Antoine-Schwartz
  0

• HINT Predictions Not Created with nf.fit() and nf.predict() but with nf.cross_validation()

  #1125 opened 3 months ago by breadwall
  2

• [<Library component: Models|Core|etc...>]

  #1200 opened 7 days ago by kevinleahy-switchdin
  0

• `maximum size for tensor at dimension` for NBEATS model

  #1193 opened 17 days ago by j-adamczyk
  2

• [Probabilistic forecasting with TFT] Quantiles predictions values

  #1199 opened 8 days ago by ZineebF
  0

• [ NeuralForecast: TFT|TemporalCovariateEncoder| nn.lstm ]

  #1185 opened 8 days ago by Otsutsukii
  1

• [<Library component: Model|Core|etc...>] static exogenous variables not found in input dataset

  #1194 opened 17 days ago by saad1912
  1

• Add TimeXer?

  #1197 opened 9 days ago by jinhangjiang
  0

• [ACNet] Add the ACNet model

  #1120 opened 12 days ago by WenjuanOlympus
  1

• Neural Forecast: Tweedie Distribution Loss Broken (NBEATSx)

  #1181 opened a month ago by jrodenberg0
  2

• Add more LLMs / Zero-shot / time-series-foundation-models

  #1121 opened 3 months ago by CoteDave
  1

• [Core] PermissionDeniedError with TensorBoard Logging When Running NeuralForecast in Parallel

  #1187 opened a month ago by Yibei990826
  5

• TFT model has no attribute "attention_weights"

  #1156 opened 24 days ago by joanna-baran-allegro
  2

• [Model Request] Add TimeXer

  #1111 opened a month ago by krishnapriya-18
  4

• step_size in cross-validation

  #1161 opened a month ago by HAUNGTOFU
  1

• TFT Model: Trainer.__init__() got an unexpected keyword argument 'grn_activation'

  #1182 opened a month ago by skuzmier
  1

• [<Library component: Models|Core|etc...>]

  #1180 opened a month ago by HAUNGTOFU
  0

• [TFT] TFT.feature_importances() returns error when there is no futr_exog_list

  #1169 opened a month ago by jeanboy44
  1

• Nbeatsx RuntimeError on Example

  #1143 opened a month ago by Songloading
  5

• [Loss] Tweedie DistributionLoss not working

  #1159 opened a month ago by Antoine-Schwartz
  1

• [TiDE] Exception("No windows available for training")

  #1172 opened a month ago by rjguedes8
  2

• [BUG] KeyError: 'autormok' on loading

  #1166 opened 2 months ago by elephaint
  0

• Core - Transfer Learning functionality with cross validation

  #1157 opened 2 months ago by DaneLyttinen
  1

• What is the purpose of unique_id during prediction? Different unique_id values produce the same prediction results.

  #1162 opened 2 months ago by sunggc
  1

• [Recurrent Models] do not support step_size > 1

  #1145 opened 2 months ago by JoaquinDF-UniLU
  3

• Combine Direct and Recursive Forecasting Strategies allowing to produce longer prediction then training horizon

  #1155 opened 2 months ago by anthonygiorgio97
  1

• Getting Access to Slack Channel

  #1153 opened 2 months ago by Povinelli
  4

• neuralforecast - KAN - support for exogenous features

  #1132 opened 2 months ago by marcopeix
  0

• TimeMixer model issue: tensor shape mismatch error in MultiScaleSeasonMixing.forward(self, season_list)

  #1131 opened 2 months ago by carusyte
  0

• NeuralForecast: TFT support prediction for any start date without retraining

  #1147 opened 2 months ago by sophiag16
  3

• Distributed training of AutoNHITS on spark is very slow compared to local training

  #1142 opened 2 months ago by kiranklc
  4

• NeuralForecast |TimeMixer Model - exogeneous variables support

  #1137 opened 3 months ago by MLfreakPy
  3

• AutoTFT can't be trained on multiple GPUs

  #1110 opened 3 months ago by jmoralez
  0

• PatchTST prediction is very slow.

  #1106 opened 3 months ago by Povinelli
  17

• Weighting scheme for TFT (to reduce the impact of a period, like the COVID one, in the training phase)

  #1130 opened 3 months ago by DG2DATA
  0

• AttributeError: 'TimeMixer' object has no attribute 'pred_len'

  #1127 opened 3 months ago by carusyte
  2

• How refit on train+val set?

  #1124 opened 3 months ago by EdgedSquirrels
  2

• Hyperparameter optimization for multivariate forecasting

  #1119 opened 3 months ago by masadshoaib
  1

• Do the models in the current library support multivariate forecasting?

  #1122 opened 3 months ago by YSAA1
  1

• [Model Request] - Koopman Neural Forecaster for Time Series with Temporal Distribution Shifts

  #1123 opened 3 months ago by dafdaf1234444
  0

• Hyper-param optimization giving suboptimal results compared to non-optimized models

  #1102 opened 3 months ago by masadshoaib
  2

• Does neuralforecast can use TensorBoard to show the histograms for network weights?

  #1118 opened 3 months ago by ZhikangLai
  0

• Can we increase support for cross entropy loss？

  #1116 opened 3 months ago by chensohg
  0

• Early Stop

  #1108 opened 3 months ago by stanley-tsgs
  2

• [Model Request] Add CATS: official code included

  #1112 opened 3 months ago by krishnapriya-18
  0

• Early Stop

  #1105 opened 3 months ago by stanley-tsgs
  3

• All Multivariate Models (Except SOFTS) cannot be trained on multiple GPUs

  #1100 opened 3 months ago by Teculos
  5

• How can I close the message output during training and prediction?

  #1101 opened 3 months ago by ZhikangLai
  2