Sökning: "Hämtningsförstärkta språkmodeller"

Hittade 1 uppsats innehållade orden Hämtningsförstärkta språkmodeller.

  1. 1. The Impact of the Retrieval Text Set for Text Sentiment Classification With the Retrieval-Augmented Language Model REALM

    Master-uppsats, KTH/Matematik (Inst.)

    Författare :Oscar Blommegård; [2023]
    Nyckelord :The Impact of the Retrieval Text Set for Text Sentiment Classification With the Retrieval-Augmented Language Model REALM; Hämtningsförstärkta språkmodeller; Natural Language Processing; Transformers; Djupinlärning; Textklassificering;

    Sammanfattning : Large Language Models (LLMs) have demonstrated impressive results across various language technology tasks. By training on large corpora of diverse text collections from the internet, these models learn to process text effectively, allowing them to acquire comprehensive world knowledge. LÄS MER