Wyniki wyszukiwania - BazTech

Ograniczanie wyników

1 Computer Science

1 2025

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Bielik 7B v0.1: Polish language model–development, insights, andevaluation

Ociepa Krzysztof, Flis Łukasz, Wróbel Krzysztof, Gwoździej Adrian, Klinas Remigiusz

Computer Science

2025

T. 26 (4)

131--161

We introduce Bielik 7B v0.1 – a seven-billion-parameter generative text modelfor Polish language processing. Trained on curated Polish corpora, this model addresses key challenges in language model development through innovative techniques; these include Weighted Instruction Cross-Entropy Loss (which bal-ances the learning of different instruction types) and Adaptive Learning Rate(which dynamically adjusts the learning rate based on training progress). To evaluate performance, we created the Open PL LLM Leaderboard and Polish MT-Bench – novel frameworks assessing various NLP tasks and conversational abilities. Bielik 7B v0.1 demonstrates significant improvements, achieving a nine-percentage-point increase in its average score compared to Mistral-7B-v0.1 on the RAG Reader task. It also excels in the Polish MT-Bench –particularly in the Reasoning (6.15/10) and Role-playing (7.83/10) categories.This model represents a substantial advancement in Polish language AI, offering a powerful tool for diverse linguistic applications and setting new benchmarksin the field.