Text Industrial
Leaderboard Tasks How MERA works
Return to the leaderboard

T-lite-it-1.0

T-Tech Created at 10.12.2024 08:07
0.552
The overall result
67
Place in the rating
Weak tasks:
226
RWSD
99
PARus
59
RCB
214
ruEthics
66
MultiQ
71
ruWorldTree
59
ruOpenBookQA
94
ruMMLU
154
ruHateSpeech
218
ruDetox
110
ruHHH
85
ruTiE
202
ruHumanEval
226
USE
38
MathLogicQA
65
ruMultiAr
43
SimpleAr
83
LCS
289
BPS
179
ruModAr
81
MaMuRAMu
126
ruCodeEval
+18
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.144 Accuracy
RCB 0.571 / 0.533 Accuracy F1 macro
USE 0.147 Grade norm
RWSD 0.535 Accuracy
PARus 0.894 Accuracy
ruTiE 0.786 Accuracy
MultiQ 0.523 / 0.398 F1 Exact match
CheGeKa 0.502 / 0.413 F1 Exact match
ruModAr 0.493 Exact match
MaMuRAMu 0.775 Accuracy
ruMultiAr 0.346 Exact match
ruCodeEval 0.082 / 0.168 / 0.226 Pass@k
MathLogicQA 0.662 Accuracy
ruWorldTree 0.964 / 0.964 Accuracy F1 macro
ruOpenBookQA 0.905 / 0.905 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 0.887 Accuracy
ruMMLU 0.664 Accuracy
SimpleAr 0.994 Exact match
ruHumanEval 0.013 / 0.018 / 0.024 Pass@k
ruHHH 0.781
ruHateSpeech 0.762
ruDetox 0.175
ruEthics
Correct God Ethical
Virtue 0.279 0.265 0.286
Law 0.257 0.244 0.264
Moral 0.291 0.277 0.297
Justice 0.246 0.234 0.254
Utilitarianism 0.215 0.234 0.265

Information about the submission:

Mera version
v.1.2.0
Torch Version
2.5.1
The version of the codebase
44ddcb3
CUDA version
12.4
Precision of the model weights
bfloat16
Seed
1234
Butch
1
Transformers version
4.46.3
The number of GPUs and their type
4 x NVIDIA H100 80GB HBM3
Architecture
vllm

Team:

T-Tech

Name of the ML model:

T-lite-it-1.0

Link to the ML model:

https://huggingface.co/t-tech/T-lite-it-1.0

Model size

7.0B

Model type:

Opened

SFT

Architecture description:

7B open source russian model

Description of the training:

Based on Qwen2.5-7b-it post-trained to increase performance on russian language tasks

Pretrain data:

Diverse high-quality datamix of pre-train and synthetic data

License:

Apache 2.0

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;

The size of the context:
simplear, chegeka, rudetox, rumultiar, use, multiq, rumodar, ruhumaneval, rucodeeval - 4096

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.

Description of the template:
{% for message in messages %} \n {{- '<|im_start|>' + message.role + ' \n' + message.content + '<|im_end|>' + ' \n' -}} \n{% endfor %} \n{%- if add_generation_prompt %} \n {{- '<|im_start|>assistant \n' -}} \n{%- endif %}

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
T-lite-it-1.0
T-Tech 		0.067 0.567 0.667 0.267 0.1 0.233 0 - 0 0 0 0 0.1 0.1 0.033 0.117 0 0 0 0 0 0.033 0.267 0.1 0.067 0.183 0.033 0.133 0.367 0.367 0.533
Model, team Honest Helpful Harmless
T-lite-it-1.0
T-Tech 		0.689 0.797 0.862
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
T-lite-it-1.0
T-Tech 		0.615 0.548 0.822 0.838 0.719 0.831 0.816 0.701 0.728 0.704 0.579 0.54 0.4 0.704 0.788 0.708 0.75 0.785 0.433 0.756 0.36 0.836 0.5 0.63 0.598 0.77 0.452 0.62 0.702 0.48 0.83 0.785 0.724 0.808 0.728 0.739 0.51 0.855 0.556 0.626 0.848 0.702 0.669 0.7 0.858 0.676 0.804 0.493 0.457 0.695 0.59 0.819 0.759 0.777 0.83 0.77 0.813
Model, team SIM FL STA
T-lite-it-1.0
T-Tech 		0.293 0.78 0.79
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
T-lite-it-1.0
T-Tech 		0.667 0.871 0.7 0.722 0.803 0.69 0.724 0.737 0.712 0.769 0.795 0.75 0.5 0.814 0.708 0.765 0.71 0.733 0.614 0.825 0.614 0.898 0.867 0.84 0.756 0.818 0.808 0.649 0.895 0.867 0.867 0.821 0.732 0.86 0.727 0.804 0.889 0.8 0.667 0.723 0.824 0.841 0.8 0.844 0.897 0.889 0.879 0.864 0.815 0.895 0.844 0.826 0.81 0.727 0.698 0.708 0.867
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
T-lite-it-1.0
T-Tech 		0.279 0.257 0.291 0.246 0.215
Model, team Virtue Law Moral Justice Utilitarianism
T-lite-it-1.0
T-Tech 		0.265 0.244 0.277 0.234 0.234
Model, team Virtue Law Moral Justice Utilitarianism
T-lite-it-1.0
T-Tech 		0.286 0.264 0.297 0.254 0.265
Model, team Women Men LGBT Nationalities Migrants Other
T-lite-it-1.0
T-Tech 		0.759 0.571 0.941 0.811 0.714 0.803

MERA

Useful links

© 2025 MERA. All rights reserved.
User agreement Privacy policy

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