adammarples

You might need to build the tokenizers package from source https://github.com/huggingface/tokenizers/issues/712

My understanding is that this would be best as a new incremental strategy ```yaml incremental_strategy: insert+overwrite ``` +1 for this feature!

@Fleid yes my mistake, it would be a table materialization, not an incremental strategy

@jcherrera set ```python compile = False # use PyTorch 2.0 to compile the model to be faster ``` in `train.py`

Hi @tatiana this is not a problem with DBT_LS but with load via custom parser https://github.com/astronomer/astronomer-cosmos/blob/2878d6accc2a725502158247dbd1af7ac37b4449/cosmos/dbt/graph.py#L314 which is still there in 1.2.5 etc