A collection of resources for evaluation of LLM capabilities in the Estonian language.
AI & ML interests
Natural Language Processing
Recent Activity
Organization Card
We are the research group of natural language processing at the Institute of Computer Science, University of Tartu. Our areas of focus include machine translation, speech synthesis, NLP for Estonian and others.
Llama-2-based LLMs fine-tuned for grammatical error correction. This collection also includes AEG (Artificial Error Generation) models.
-
To Err Is Human, but Llamas Can Learn It Too
Paper • 2403.05493 • Published • 6 -
tartuNLP/Llamma-2-7b-ukr-p1-llama-errors-p2-GEC
Text Generation • 7B • Updated • 5 • 2 -
tartuNLP/Llammas-base-p1-llama-errors-p2-GEC
Text Generation • 7B • Updated • 6 • 3 -
tartuNLP/leo-hessianai-7b-p1-llama-errors-p2-GEC
Text Generation • 7B • Updated • 3
A collection of resources for evaluation of LLM capabilities in the Estonian language.
Llama-2-based LLMs fine-tuned for grammatical error correction. This collection also includes AEG (Artificial Error Generation) models.
-
To Err Is Human, but Llamas Can Learn It Too
Paper • 2403.05493 • Published • 6 -
tartuNLP/Llamma-2-7b-ukr-p1-llama-errors-p2-GEC
Text Generation • 7B • Updated • 5 • 2 -
tartuNLP/Llammas-base-p1-llama-errors-p2-GEC
Text Generation • 7B • Updated • 6 • 3 -
tartuNLP/leo-hessianai-7b-p1-llama-errors-p2-GEC
Text Generation • 7B • Updated • 3
models
62

tartuNLP/XTTS-v2-est
Updated
•
18

tartuNLP/Llama-3.1-8B-est-gec-highrec-july-2025
Text Generation
•
8B
•
Updated
•
9

tartuNLP/Llama-3.1-8B-est-gec-july-2025
Text Generation
•
8B
•
Updated
•
167

tartuNLP/Llama-3.1-8B-est-ged-july-2025
Text Generation
•
8B
•
Updated
•
13

tartuNLP/Llama-3.1-8B-est-gee-july-2025
Text Generation
•
8B
•
Updated
•
13

tartuNLP/Qwen2.5-3B-Instruct-hsb-dsb
Text Generation
•
3B
•
Updated
•
21

tartuNLP/Llammas
Text Generation
•
7B
•
Updated
•
9
•
5

tartuNLP/Llammas-base-AEG
Text Generation
•
7B
•
Updated
•
2

tartuNLP/Llammas-base-p1-GPT-4o-human-error-mix-paragraph-GEC
Text Generation
•
7B
•
Updated
•
79

tartuNLP/Llammas-base-p1-GPT-4o-human-error-pseudo-m2
Text Generation
•
7B
•
Updated
•
3
datasets
21
tartuNLP/ifeval_et
Viewer
•
Updated
•
541
•
194
tartuNLP/winogrande_et
Viewer
•
Updated
•
4.83k
•
174
tartuNLP/grammar_et_multiple_choice
Updated
tartuNLP/finno-ugric-train
Updated
•
51
•
1
tartuNLP/finno-ugric-benchmark
Updated
•
79
tartuNLP/liv4ever
Updated
•
203
•
2
tartuNLP/smugri-mt-bench
Viewer
•
Updated
•
500
•
95
•
2
tartuNLP/flores-smugri-pairs
Viewer
•
Updated
•
16.5k
•
52
tartuNLP/est_gec_ged_gee_2025
Viewer
•
Updated
•
299k
•
100
tartuNLP/magpie-gemma-3-12b-it-100k-et
Viewer
•
Updated
•
101k
•
89