--emb_model

Switch

--add_emb_feat

Description

Adds a Huggingface Transformer feature table.

Argument and Default Value

No argument. Boolean switch, default: False.

Details

Alias for --bert_model (--add_bert, --add_embedding); all three flags run the same code path, which extracts contextual embedding features for every message using a Huggingface Transformer model and aggregates them into a feature table, one row per group. The model defaults to bert-base-uncased (--emb_model) but any Huggingface model name works, e.g. roberta-base or distilbert-base-uncased.

Other Switches

Required Switches:

• -d, -t , -g

Optional Switches:

• --emb_layers

• --emb_msg_aggregation

• fwflag_emb_layer_aggregation

Example Commands

Creates a default BERT feature table.

dlatkInterface.py -d dla_tutorial -t msgs -c user_id --add_emb_feat --emb_model bert-base-uncased