Staff Engineer & Solutions Architect
-
maracatu-80m
Open-weight Brazilian Portuguese LLM trained from scratch on 1.6B tokens. 87.8M params, Llama-style with GQA. Validation perplexity 21.34. Apache 2.0. Base model.
46 Pulls 1 Tag Updated 6 months ago
-
maracatu-20m
17M-param Brazilian Portuguese LLM, trained from scratch on Wikipedia PT. Base model, Apache 2.0.
30 Pulls 1 Tag Updated 6 months ago