Instructions to use huggingface/time-series-transformer-tourism-monthly with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use huggingface/time-series-transformer-tourism-monthly with Transformers:
# Load model directly from transformers import AutoTokenizer, TimeSeriesTransformerForPrediction tokenizer = AutoTokenizer.from_pretrained("huggingface/time-series-transformer-tourism-monthly") model = TimeSeriesTransformerForPrediction.from_pretrained("huggingface/time-series-transformer-tourism-monthly", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Meaning of label_length parameter in AutoformerConfig
#15
by yuvalarbel - opened
Hi!
Could you explain a bit further the meaning of the label_length config parameter in the Autoformer?
The docs state:label_length (int, optional, defaults to 10) β Start token length of the Autoformer decoder, which is used for direct multi-step prediction (i.e. non-autoregressive generation).
Still, I'm not sure how this parameter should be used, or under what circumstances different values should be used.
Thanks!