Tag: llms

New Model LIMA Claims To Provide GPT-4-Like Quality After Being Fine-Tuned On Only 1000 Curated Examples

Thus far, the conventional wisdom appears to have been that larger models and larger fine-tuning datasets would lead to more sophisticated models, but this is increasingly being put to the test. A new model named LIMA claims to provide performance comparable to GPT-4 after being fine-tuned on only a set of 1000 curated examples. LIMA, […]