Loading search index…
No recent searches
No results for "Query here"
← Back to Glossary
A base model fine-tuned with instruction-following data and/or RLHF to make it useful for conversation and following directions. Contrasted with a raw base/pretrained model which just predicts next tokens.