Qwen3-32B

MERA Created at 04.03.2026 20:25
0.731
The overall result
21
Place in the rating
In the top by tasks:
2
ruEthics
Weak tasks:
27
RWSD
221
PARus
95
RCB
90
MultiQ
23
ruWorldTree
135
CheGeKa
24
ruMMLU
75
ruDetox
85
ruHHH
21
ruHumanEval
37
USE
27
ruMultiAr
39
SimpleAr
100
BPS
25
ruModAr
36
MaMuRAMu
26
ruCodeEval
+13
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.568 Accuracy
RCB 0.568 / 0.401 Accuracy F1 macro
USE 0.421 Grade norm
RWSD 0.738 Accuracy
PARus 0.832 Accuracy
ruTiE 0.909 Accuracy
MultiQ 0.533 / 0.352 F1 Exact match
CheGeKa 0.248 / 0.192 F1 Exact match
ruModAr 0.975 Exact match
MaMuRAMu 0.851 Accuracy
ruMultiAr 0.945 Exact match
ruCodeEval 0.577 / 0.798 / 0.841 Pass@k
MathLogicQA 0.992 Accuracy
ruWorldTree 0.99 / 0.793 Accuracy F1 macro
ruOpenBookQA 0.96 / 0.96 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 0.979 Accuracy
ruMMLU 0.821 Accuracy
SimpleAr 0.997 Exact match
ruHumanEval 0.626 / 0.818 / 0.848 Pass@k
ruHHH 0.843
ruHateSpeech 0.902
ruDetox 0.315
ruEthics
Correct God Ethical
Virtue 0.561 0.546 0.637
Law 0.566 0.514 0.587
Moral 0.6 0.581 0.664
Justice 0.511 0.481 0.564
Utilitarianism 0.491 0.481 0.54

Information about the submission

Mera version
v1.2.0
Torch Version
2.9.1
The version of the codebase
7030e16
CUDA version
12.8
Precision of the model weights
bfloat16
Seed
1234
Batch
1
Transformers version
4.56.1
The number of GPUs and their type
4 x NVIDIA A100-SXM4-80GB
Architecture
local-chat-completions

Team:

MERA

Name of the ML model:

Qwen3-32B

Model size

32.0B

Model type:

Opened

SFT

Additional links:

https://arxiv.org/abs/2505.09388

Architecture description:

Qwen3-32B is a decoder-only transformer language model from the Qwen3 family with approximately 32 billion parameters. The model is designed for general language understanding and generation tasks, with strong capabilities in reasoning, coding, and multilingual tasks. It supports long-context inputs and improved efficiency compared to previous Qwen models.

Description of the training:

The model is trained using a multi-stage pipeline consisting of large-scale pretraining followed by post-training. The post-training stage includes instruction tuning and alignment techniques to improve instruction following, reasoning ability, and response quality.

Pretrain data:

The model was pretrained on a large multilingual corpus of approximately 36 trillion tokens covering 119 languages. It was further post-trained on instruction-following and reasoning-oriented datasets to improve instruction following and reasoning performance.

License:

Apache License 2.0

Inference parameters

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:
{%- if tools %} {{- '<|im_start|>system\n' }} {%- if messages[0].role == 'system' %} {{- messages[0].content + '\n\n' }} {%- endif %} {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }} {%- for tool in tools %} {{- "\n" }} {{- tool | tojson }} {%- endfor %} {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }} {%- else %} {%- if messages[0].role == 'system' %} {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }} {%- endif %} {%- endif %} {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} {%- for message in messages[::-1] %} {%- set index = (messages|length - 1) - loop.index0 %} {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %} {%- set ns.multi_step_tool = false %} {%- set ns.last_query_index = index %} {%- endif %} {%- endfor %} {%- for message in messages %} {%- if message.content is string %} {%- set content = message.content %} {%- else %} {%- set content = '' %} {%- endif %} {%- if (message.role == "user") or (message.role == "system" and not loop.first) %} {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} {%- elif message.role == "assistant" %} {%- set reasoning_content = '' %} {%- if message.reasoning_content is string %} {%- set reasoning_content = message.reasoning_content %} {%- else %} {%- if '</think>' in content %} {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %} {%- set content = content.split('</think>')[-1].lstrip('\n') %} {%- endif %} {%- endif %} {%- if loop.index0 > ns.last_query_index %} {%- if loop.last or (not loop.last and reasoning_content) %} {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }} {%- else %} {{- '<|im_start|>' + message.role + '\n' + content }} {%- endif %} {%- else %} {{- '<|im_start|>' + message.role + '\n' + content }} {%- endif %} {%- if message.tool_calls %} {%- for tool_call in message.tool_calls %} {%- if (loop.first and content) or (not loop.first) %} {{- '\n' }} {%- endif %} {%- if tool_call.function %} {%- set tool_call = tool_call.function %} {%- endif %} {{- '<tool_call>\n{"name": "' }} {{- tool_call.name }} {{- '", "arguments": ' }} {%- if tool_call.arguments is string %} {{- tool_call.arguments }} {%- else %} {{- tool_call.arguments | tojson }} {%- endif %} {{- '}\n</tool_call>' }} {%- endfor %} {%- endif %} {{- '<|im_end|>\n' }} {%- elif message.role == "tool" %} {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} {{- '<|im_start|>user' }} {%- endif %} {{- '\n<tool_response>\n' }} {{- content }} {{- '\n</tool_response>' }} {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} {{- '<|im_end|>\n' }} {%- endif %} {%- endif %} {%- endfor %} {%- if add_generation_prompt %} {{- '<|im_start|>assistant\n' }} {%- if enable_thinking is defined and enable_thinking is false %} {{- '<think>\n\n</think>\n\n' }} {%- endif %} {%- endif %}

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
Qwen3-32B
MERA
0.6 0.367 0.867 0 0.367 0.5 0.333 - 0.067 0.1 0.333 0.233 0.267 0.167 0.133 0.35 0.1 0.133 0.467 0.133 0.467 0.733 0.3 0.6 0.4 0.742 0.367 0.533 0.667 0.667 0.733
Model, team Honest Helpful Harmless
Qwen3-32B
MERA
0.754 0.831 0.948
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
Qwen3-32B
MERA
0.807 0.56 0.934 0.88 0.863 0.856 0.854 0.733 0.861 0.753 0.772 0.865 0.58 0.824 0.904 0.786 0.79 0.951 0.944 0.832 0.703 0.877 0.92 0.832 0.848 0.94 0.579 0.694 0.727 0.71 0.83 0.826 0.822 0.828 0.868 0.953 0.87 0.939 0.914 0.936 0.899 0.901 0.841 0.958 0.936 0.889 0.926 0.941 0.777 0.815 0.9 0.873 0.908 0.958 0.95 0.842 0.922
Model, team SIM FL STA
Qwen3-32B
MERA
0.752 0.617 0.706
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
Qwen3-32B
MERA
0.822 0.921 0.817 0.759 0.961 0.845 0.793 0.772 0.846 0.738 0.885 0.825 0.6 0.798 0.807 0.741 0.832 0.844 0.877 0.877 0.807 0.898 0.933 0.935 0.978 0.894 0.808 0.649 0.947 0.933 0.889 0.872 0.866 0.895 0.773 0.929 0.911 0.911 0.912 0.892 0.902 0.889 0.867 1 0.931 0.889 0.931 0.932 0.862 0.947 0.933 0.884 0.835 0.818 0.744 0.76 0.878
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Qwen3-32B
MERA
0.561 0.566 0.6 0.511 0.491
Model, team Virtue Law Moral Justice Utilitarianism
Qwen3-32B
MERA
0.546 0.514 0.581 0.481 0.481
Model, team Virtue Law Moral Justice Utilitarianism
Qwen3-32B
MERA
0.637 0.587 0.664 0.564 0.54
Model, team Women Men LGBT Nationalities Migrants Other
Qwen3-32B
MERA
0.907 0.829 0.941 0.946 0.857 0.902