Gemma 3 27B

llmarena.ru Created at 17.03.2025 22:57
0.567
The overall result
92
Place in the rating
Weak tasks:
121
RWSD
75
PARus
137
RCB
217
ruEthics
158
MultiQ
80
ruWorldTree
97
ruOpenBookQA
61
CheGeKa
114
ruMMLU
63
ruHateSpeech
86
ruDetox
50
ruHHH
80
ruTiE
145
ruHumanEval
75
USE
81
MathLogicQA
82
ruMultiAr
35
SimpleAr
225
LCS
48
BPS
131
ruModAr
103
MaMuRAMu
170
ruCodeEval
+19
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.118 Accuracy
RCB 0.559 / 0.483 Accuracy F1 macro
USE 0.328 Grade norm
RWSD 0.588 Accuracy
PARus 0.916 Accuracy
ruTiE 0.84 Accuracy
MultiQ 0.456 / 0.317 F1 Exact match
CheGeKa 0.367 / 0.298 F1 Exact match
ruModAr 0.626 Exact match
MaMuRAMu 0.797 Accuracy
ruMultiAr 0.39 Exact match
ruCodeEval 0.075 / 0.154 / 0.189 Pass@k
MathLogicQA 0.651 Accuracy
ruWorldTree 0.975 / 0.975 Accuracy F1 macro
ruOpenBookQA 0.903 / 0.903 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 0.993 Accuracy
ruMMLU 0.69 Accuracy
SimpleAr 0.998 Exact match
ruHumanEval 0.132 / 0.189 / 0.207 Pass@k
ruHHH 0.865
ruHateSpeech 0.838
ruDetox 0.305
ruEthics
Correct God Ethical
Virtue 0.296 0.354 0.382
Law 0.286 0.349 0.376
Moral 0.318 0.374 0.402
Justice 0.255 0.297 0.359
Utilitarianism 0.273 0.296 0.306

Information about the submission

Mera version
v.1.2.0
Torch Version
2.5.1
The version of the codebase
30667dc
CUDA version
12.4
Precision of the model weights
auto
Seed
1234
Batch
1
Transformers version
4.49.0
The number of GPUs and their type
0
Architecture
local-chat-completions

Team:

llmarena.ru

Name of the ML model:

Gemma 3 27B

Model size

27.0B

Model type:

Closed

API

SFT

Additional links:

Замер осуществлялся через сервис openrouter (https://openrouter.ai/google/gemma-3-27b-it)

Architecture description:

Gemma 3 models are multimodal, handling text and image input and generating text output, with open weights for both pre-trained variants and instruction-tuned variants. Gemma 3 has a large, 128K context window, multilingual support in over 140 languages, and is available in more sizes than previous versions.

Inference parameters

Generation Parameters:
rucodeeval - do_sample=true;temperature=0.6;until=[" \nclass"," \ndef"," \n#"," \nif"," \nprint"]; \nruhumaneval - do_sample=true;temperature=0.6;until=[" \nclass"," \ndef"," \n#"," \nif"," \nprint"];

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.

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
Gemma 3 27B
llmarena.ru
0.533 0.567 0.733 0.133 0.067 0.333 0 - 0.233 0 0.033 0.067 0.167 0.133 0.167 0.467 0.067 0.067 0 0.033 0 0.667 0.333 0.4 0.067 0.733 0.367 0.367 0.667 0.533 0.567
Model, team Honest Helpful Harmless
Gemma 3 27B
llmarena.ru
0.82 0.864 0.914
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
Gemma 3 27B
llmarena.ru
0.659 0.5 0.868 0.855 0.739 0.856 0.786 0.733 0.772 0.713 0.57 0.635 0.46 0.815 0.837 0.72 0.74 0.861 0.644 0.779 0.267 0.871 0.43 0.671 0.571 0.87 0.527 0.657 0.755 0.54 0.72 0.76 0.736 0.859 0.736 0.769 0.45 0.877 0.583 0.675 0.793 0.743 0.648 0.756 0.873 0.606 0.882 0.537 0.557 0.73 0.56 0.852 0.767 0.84 0.82 0.794 0.891
Model, team SIM FL STA
Gemma 3 27B
llmarena.ru
0.509 0.789 0.8
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
Gemma 3 27B
llmarena.ru
0.711 0.881 0.65 0.657 0.868 0.845 0.655 0.789 0.654 0.769 0.744 0.742 0.642 0.822 0.76 0.778 0.766 0.844 0.754 0.877 0.807 0.864 0.889 0.84 0.778 0.864 0.795 0.772 0.965 0.822 0.844 0.846 0.813 0.93 0.682 0.821 0.867 0.889 0.684 0.754 0.849 0.857 0.867 0.956 0.879 0.889 0.931 0.864 0.785 0.93 0.822 0.783 0.797 0.727 0.558 0.696 0.867
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Gemma 3 27B
llmarena.ru
0.296 0.286 0.318 0.255 0.273
Model, team Virtue Law Moral Justice Utilitarianism
Gemma 3 27B
llmarena.ru
0.354 0.349 0.374 0.297 0.296
Model, team Virtue Law Moral Justice Utilitarianism
Gemma 3 27B
llmarena.ru
0.382 0.376 0.402 0.359 0.306
Model, team Women Men LGBT Nationalities Migrants Other
Gemma 3 27B
llmarena.ru
0.861 0.714 0.882 0.838 0.714 0.869