Таблица скроллится влево
Задача | Результат | Метрика |
---|---|---|
LCS | 0.084 | Accuracy |
RCB | 0.543 / 0.452 | Avg. F1 / Accuracy |
USE | 0.284 | Grade Norm |
RWSD | 0.627 | Accuracy |
PARus | 0.848 | Accuracy |
ruTiE | 0.726 | Accuracy |
MultiQ | 0.193 / 0.071 | F1-score/EM |
CheGeKa | 0.063 / 0 | F1 / EM |
ruModAr | 0.77 | EM |
ruMultiAr | 0.216 | EM |
MathLogicQA | 0.45 | Accuracy |
ruWorldTree | 0.897 / 0.897 | Avg. F1 / Accuracy |
ruOpenBookQA | 0.823 / 0.822 | Avg. F1 / Accuracy |
Таблица скроллится влево
Задача | Результат | Метрика | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
BPS | 0.412 | Accuracy | ||||||||||||||||||||||||
ruMMLU | 0.783 | Accuracy | ||||||||||||||||||||||||
SimpleAr | 0.9 | EM | ||||||||||||||||||||||||
ruHumanEval | 0.018 / 0.088 / 0.177 | pass@k | ||||||||||||||||||||||||
ruHHH |
0.753
|
Accuracy | ||||||||||||||||||||||||
ruHateSpeech |
0.774
|
Accuracy | ||||||||||||||||||||||||
ruDetox |
|
Общая средняя оценка (J) Оценка сохранения смысла (SIM) Оценка натуральности (FL) Точность переноса стиля (STA) |
||||||||||||||||||||||||
ruEthics |
Результаты таблицы:
[[-0.336, -0.332
, -0.351, -0.31
, -0.237], |
5 MCC |
GIGACHAT
GigaChat Lite
GigaChat Lite (version `GigaChat:4.0.26.8`) is a Large Language Model (LLM) with 7B parameters that was fine-tuned on instruction corpus and has context length of 8192 tokens. The version is available for users via API since 13.07.
-
-
-
-
Code version v.1.1.0. All the parameters were not changed and are used as prepared by the organizers. Details: - 2 x NVIDIA A100 + accelerate - dtype float16 - Pytorch 2.3.1 + CUDA 12.1 - Transformers 4.42.3 - Context length 8192