search

awslabs / gluonts

Probabilistic time series modeling in Python
https://ts.gluon.ai
Apache License 2.0
4.64k stars 755 forks source link

Silence DeepAREstimator.train() messaging? #3146

Open aifaehnle opened 8 months ago

aifaehnle commented 8 months ago

Description

In running batches of DeepArEstimator.train(), the epoch output is growing to a point where it crashes Google Colab due to the ouput size. Please add a way to suppress these messages:

INFO:lightning.pytorch.callbacks.model_summary:
  | Name  | Type        | Params | In sizes                                                  | Out sizes  
----------------------------------------------------------------------------------------------------------------
0 | model | DeepARModel | 36.4 K | [[1, 1], [1, 1], [1, 45, 2], [1, 45], [1, 45], [1, 9, 2]] | [1, 100, 9]
----------------------------------------------------------------------------------------------------------------
36.4 K    Trainable params
0         Non-trainable params
36.4 K    Total params
0.146     Total estimated model params size (MB)
Epoch 2: 
 12/? [00:00<00:00, 15.76it/s, v_num=180, train_loss=0.721]
INFO: Epoch 0, global step 50: 'train_loss' reached 0.36485 (best 0.36485), saving model to '/content/lightning_logs/version_180/checkpoints/epoch=0-step=50.ckpt' as top 1
INFO:lightning.pytorch.utilities.rank_zero:Epoch 0, global step 50: 'train_loss' reached 0.36485 (best 0.36485), saving model to '/content/lightning_logs/version_180/checkpoints/epoch=0-step=50.ckpt' as top 1
INFO: Epoch 1, global step 100: 'train_loss' was not in top 1
INFO:lightning.pytorch.utilities.rank_zero:Epoch 1, global step 100: 'train_loss' was not in top 1