talkie-1930-13b-it is a 13B vintage language model. It is an instruction-tuned post-train of talkie-1930-13b-base, which was trained on 260B tokens of pre-1931 English-language text.

talkie-1930-13b-it was finetuned using a novel dataset of instruction-response pairs extracted from pre-1931 reference works, including etiquette manuals, encyclopedias, and letter-writing manuals. The model then underwent reinforcement learning (online DPO with an LLM-as-a-judge) to improve instruction-following ability.

Read more about talkie in our report.

Reference code to run talkie is available on GitHub.