Large language model training method and device, code generation method and device and storage medium
The invention discloses a large language model training method, a code generation method, equipment and a storage medium, and the method comprises the steps: carrying out the full-parameter fine tuning of a preset language model based on a first training sample set, and obtaining an initial large la...
Gespeichert in:
Hauptverfasser: | , , , , |
---|---|
Format: | Patent |
Sprache: | chi ; eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The invention discloses a large language model training method, a code generation method, equipment and a storage medium, and the method comprises the steps: carrying out the full-parameter fine tuning of a preset language model based on a first training sample set, and obtaining an initial large language model; selecting a seed data set in the first training sample set, and constructing a second training data set based on the seed data set; and performing all-parameter fine tuning on the initial large language model based on the second training sample set to obtain a trained large language model. According to the method, the preset language model is trained by adopting the first training sample set, then the second training sample set carrying more information is generated based on the instruction data in the first training sample set, and the initial large language model is finely adjusted through the second training sample set; in this way, the large language model can better understand the human intention |
---|