v1.1.8

Updated mapmaking helpers and updated defaults to GPT 5.5 from GPT 5.4.

v1.1.7

Improve run ID and attribute batching compatibility

v1.1.6

Updated model defaults to gpt-5.4-mini and allowed priority service tier passthrough.

v1.1.5

Added gabriel.poll function for synthetic surveying.

v1.1.4

Improved printouts and high parallelization infrastructure.

v1.1.3

Added passthrough support for get_all_responses parameters.

v1.1.2

Updated model defaults to GPT-5.4 and added image detail passthrough support.

v1.1.1

Added the ability to pass custom embedding_fn and get_all_embeddings_fn. Increased save robustness for embeddings and csvs.

v1.1.0

Added tail failure robustness to get_all_responses.

v1.0.9

Bug fixes for very large runs (robustness in saving large csvs, timeout initialization from checkpoint saves).