RU EN
Return to the leaderboard

GigaChat Pro

GIGACHAT Created at 01.11.2024 08:38
0.512
The overall result
90
Place in the rating
In the top by tasks:
7
ruModAr
The task is one of the main ones
Weak tasks:
526
RWSD
221
PARus
94
RCB
152
ruEthics
249
MultiQ
88
ruWorldTree
80
ruOpenBookQA
26
CheGeKa
151
ruMMLU
331
ruHateSpeech
216
ruDetox
170
ruHHH
154
ruTiE
156
ruHumanEval
112
MathLogicQA
203
ruMultiAr
190
SimpleAr
161
LCS
330
BPS
78
MaMuRAMu
139
ruCodeEval
+17
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.12 Accuracy
RCB 0.562 / 0.289 Accuracy F1 macro
USE 0.384 Grade norm
RWSD 0.392 Accuracy
PARus 0.798 Accuracy
ruTiE 0.708 Accuracy
MultiQ 0.341 / 0.182 F1 Exact match
CheGeKa 0.406 / 0.341 F1 Exact match
ruModAr 0.888 Exact match
MaMuRAMu 0.77 Accuracy
ruMultiAr 0.263 Exact match
ruCodeEval 0.056 / 0.068 / 0.073 Pass@k
MathLogicQA 0.47 Accuracy
ruWorldTree 0.956 / 0.956 Accuracy F1 macro
ruOpenBookQA 0.885 / 0.71 Accuracy F1 macro

Evaluation on open tasks:

Go to the ratings by subcategory

The table will scroll to the left

Task name Result Metric
BPS 0.806 Accuracy
ruMMLU 0.617 Accuracy
SimpleAr 0.967 Exact match
ruHumanEval 0.063 / 0.067 / 0.067 Pass@k
ruHHH 0.713
ruHateSpeech 0.596
ruDetox 0.174
ruEthics
Correct God Ethical
Virtue 0.353 0.33 0.372
Law 0.361 0.353 0.364
Moral 0.387 0.359 0.402
Justice 0.309 0.3 0.352
Utilitarianism 0.306 0.279 0.337

Information about the submission:

Mera version
v.1.2.0
Torch Version
2.4.0
The version of the codebase
db539c9
CUDA version
12.1
Precision of the model weights
-
Seed
1234
Butch
1
Transformers version
4.46.0.dev0
The number of GPUs and their type
5 x NVIDIA H100 80GB HBM3
Architecture
gigachat_llms

Team:

GIGACHAT

Name of the ML model:

GigaChat Pro

Link to the ML model:

https://developers.sber.ru/portal/products/gigachat-api

Model type:

Closed

API

SFT

MoE

Architecture description:

GigaChat Pro (version GigaChatPro:26.20) is a Large Language Model (LLM) that was fine-tuned on instruction corpus and has context length of 32k tokens. GigaChat Pro is Mixture of Experts model and has 13B active parameters. The version is available for users via API in preview mode https://developers.sber.ru/portal/products/gigachat-api

Description of the training:

-

Pretrain data:

-

License:

Proprietary model by Sber

Inference parameters

Generation Parameters:
simplear - do_sample=false;until=["\n"]; \nchegeka - do_sample=false;until=["\n"]; \nrudetox - do_sample=false;until=["\n"]; \nrumultiar - do_sample=false;until=["\n"]; \nuse - do_sample=false;until=["\n","."]; \nmultiq - do_sample=false;until=["\n"]; \nrumodar - do_sample=false;until=["\n"]; \nruhumaneval - do_sample=true;until=["\nclass","\ndef","\n#","\nif","\nprint"];temperature=0.6; \nrucodeeval - do_sample=true;until=["\nclass","\ndef","\n#","\nif","\nprint"];temperature=0.6;

System prompt:
Реши задачу по инструкции ниже. Не давай никаких объяснений и пояснений к своему ответу. Не пиши ничего лишнего. Пиши только то, что указано в инструкции. Если по инструкции нужно решить пример, то напиши только числовой ответ без хода решения и пояснений. Если по инструкции нужно вывести букву, цифру или слово, выведи только его. Если по инструкции нужно выбрать один из вариантов ответа и вывести букву или цифру, которая ему соответствует, то выведи только эту букву или цифру, не давай никаких пояснений, не добавляй знаки препинания, только 1 символ в ответе. Если по инструкции нужно дописать код функции на языке Python, пиши сразу код, соблюдая отступы так, будто ты продолжаешь функцию из инструкции, не давай пояснений, не пиши комментарии, используй только аргументы из сигнатуры функции в инструкции, не пробуй считывать данные через функцию input. Не извиняйся, не строй диалог. Выдавай только ответ и ничего больше.

Expand information

Ratings by subcategory

Metric: Grade Norm
Model, team 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 8_0 8_1 8_2 8_3 8_4
GigaChat Pro
GIGACHAT 		0.867 0.467 0.733 0.2 0.333 0.5 0.2 - 0.267 0.1 0.167 0.133 0.667 0.2 0.167 0.6 0.1 0.133 0.2 0.1 0.233 0.567 0.3 0.1 0.067 0.683 0.433 0.267 0.5 0.333 0.8
Model, team Honest Helpful Harmless
GigaChat Pro
GIGACHAT 		0.705 0.695 0.741
Model, team Anatomy Virology Astronomy Marketing Nutrition Sociology Management Philosophy Prehistory Human aging Econometrics Formal logic Global facts Jurisprudence Miscellaneous Moral disputes Business ethics Biology (college) Physics (college) Human Sexuality Moral scenarios World religions Abstract algebra Medicine (college) Machine learning Medical genetics Professional law PR Security studies Chemistry (школьная) Computer security International law Logical fallacies Politics Clinical knowledge Conceptual_physics Math (college) Biology (high school) Physics (high school) Chemistry (high school) Geography (high school) Professional medicine Electrical engineering Elementary mathematics Psychology (high school) Statistics (high school) History (high school) Math (high school) Professional accounting Professional psychology Computer science (college) World history (high school) Macroeconomics Microeconomics Computer science (high school) European history Government and politics
GigaChat Pro
GIGACHAT 		0.593 0.524 0.783 0.838 0.725 0.841 0.748 0.704 0.728 0.65 0.474 0.476 0.35 0.713 0.787 0.61 0.66 0.75 0.433 0.725 0.351 0.801 0.41 0.59 0.402 0.76 0.402 0.602 0.694 0.48 0.74 0.802 0.663 0.818 0.687 0.607 0.37 0.787 0.43 0.493 0.833 0.688 0.669 0.509 0.82 0.505 0.745 0.37 0.454 0.625 0.53 0.734 0.692 0.748 0.75 0.776 0.834
Model, team SIM FL STA
GigaChat Pro
GIGACHAT 		0.331 0.739 0.734
Model, team Anatomy Virology Astronomy Marketing Nutrition Sociology Managment Philosophy Pre-History Gerontology Econometrics Formal logic Global facts Jurisprudence Miscellaneous Moral disputes Business ethics Bilology (college) Physics (college) Human sexuality Moral scenarios World religions Abstract algebra Medicine (college) Machine Learning Genetics Professional law PR Security Chemistry (college) Computer security International law Logical fallacies Politics Clinical knowledge Conceptual physics Math (college) Biology (high school) Physics (high school) Chemistry (high school) Geography (high school) Professional medicine Electrical Engineering Elementary mathematics Psychology (high school) Statistics (high school) History (high school) Math (high school) Professional Accounting Professional psychology Computer science (college) World history (high school) Macroeconomics Microeconomics Computer science (high school) Europe History Government and politics
GigaChat Pro
GIGACHAT 		0.711 0.881 0.633 0.639 0.842 0.759 0.741 0.719 0.788 0.754 0.756 0.742 0.483 0.814 0.743 0.728 0.71 0.778 0.614 0.807 0.421 0.881 0.822 0.846 0.667 0.788 0.821 0.754 0.842 0.822 0.844 0.846 0.723 0.895 0.667 0.786 0.822 0.867 0.561 0.677 0.873 0.857 0.844 0.733 0.862 0.822 0.897 0.795 0.892 0.877 0.778 0.884 0.848 0.714 0.419 0.76 0.9
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
GigaChat Pro
GIGACHAT 		0.353 0.361 0.387 0.309 0.306
Model, team Virtue Law Moral Justice Utilitarianism
GigaChat Pro
GIGACHAT 		0.33 0.353 0.359 0.3 0.279
Model, team Virtue Law Moral Justice Utilitarianism
GigaChat Pro
GIGACHAT 		0.372 0.364 0.402 0.352 0.337
Model, team Women Men LGBT Nationalities Migrants Other
GigaChat Pro
GIGACHAT 		0.778 0.543 0 0.162 0.286 0.77

Пользовательское соглашение

We use cookies to store data. By continuing to use the site, you consent to work with these files