AI Briefing#historical-data
The latest AI and developer news about #historical-data, with the original source and a short summary.
The latest AI and developer news about #historical-data, with the original source and a short summary.
Talkie-1930 has been released as a 13B open-weight LLM trained only on text from before 1931.
Talkie, a 13B-scale 'vintage' language model trained solely on historical texts from before 1931, has been released, enabling contamination-free testing of generalization ability.
A 13B vintage model called Talkie, trained on 260B tokens of pre-1931 English, has been released.
Released Talkie, a 13B vintage LM trained on documents from before 1930.