rllm.preprocessing.embed_text_column¶
- rllm.preprocessing.embed_text_column(col_series: Series, config: TextEmbedderConfig) Tensor[source]¶
Embed a text column into dense vector representations. The function supports both one-shot and mini-batch embedding, depending on configuration.
- Parameters:
col_series (Series) – Input text column.
config (TextEmbedderConfig) – Embedding configuration.
- Returns:
Embedded features with shape \((N, D)\) and dtype
torch.float32.- Return type:
Tensor