RU EN
Return to the leaderboard

Yi-1.5-34B-Chat

MERA Created at 20.09.2024 12:37
0.413
The overall result
257
Place in the rating

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.116 Accuracy
RCB 0.564 / 0.506 Accuracy F1 macro
USE 0.13 Grade norm
RWSD 0.446 Accuracy
PARus 0.738 Accuracy
ruTiE 0.69 Accuracy
MultiQ 0.416 / 0.266 F1 Exact match
CheGeKa 0.049 / 0.017 F1 Exact match
ruModAr 0.39 Exact match
MaMuRAMu 0.572 Accuracy
ruMultiAr 0.287 Exact match
ruCodeEval 0 / 0 / 0 Pass@k
MathLogicQA 0.458 Accuracy
ruWorldTree 0.796 / 0.795 Accuracy F1 macro
ruOpenBookQA 0.673 / 0.669 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 0.924 Accuracy
ruMMLU 0.52 Accuracy
SimpleAr 0.977 Exact match
ruHumanEval 0.006 / 0.006 / 0.006 Pass@k
ruHHH 0.579
ruHateSpeech 0.642
ruDetox 0.147
ruEthics
Correct God Ethical
Virtue 0.264 0.267 0.213
Law 0.239 0.285 0.201
Moral 0.278 0.286 0.221
Justice 0.228 0.214 0.175
Utilitarianism 0.212 0.223 0.181

Information about the submission:

Mera version
v.1.2.0
Torch Version
2.4.0
The version of the codebase
9b26db97
CUDA version
12.1
Precision of the model weights
bfloat16
Seed
1234
Butch
1
Transformers version
4.43.2
The number of GPUs and their type
4 x NVIDIA H100 80GB HBM3
Architecture
vllm

Team:

MERA

Name of the ML model:

Yi-1.5-34B-Chat

Link to the ML model:

https://huggingface.co/01-ai/Yi-1.5-34B-Chat

Model size

34.4B

Model type:

Opened

SFT

Architecture description:

The Yi series models adopt the same model architecture as Llama but are NOT derivatives of Llama.

Description of the training:

Yi has independently created its own high-quality training datasets, efficient training pipelines, and robust training infrastructure entirely from the ground up.

Pretrain data:

Yi-1.5 is an upgraded version of Yi (which was trained on 3T multilingual corpus). It is continuously pre-trained on Yi with a high-quality corpus of 500B tokens and fine-tuned on 3M diverse fine-tuning samples.

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, bps, lcs, chegeka, mathlogicqa, parus, rcb, rudetox, ruhatespeech, rummlu, ruworldtree, ruopenbookqa, rumultiar, use, rwsd, mamuramu, multiq, rumodar, ruethics, ruhhh, ruhumaneval, rucodeeval, rutie, rutie - 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:
{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|im_start|>user\n' + content + '<|im_end|>\n<|im_start|>assistant\n' }}{% elif message['role'] == 'assistant' %}{{ content + '<|im_end|>' + '\n' }}{% endif %}{% endfor %}

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
Yi-1.5-34B-Chat
MERA 		0.2 0.067 0.667 0.2 0 0.133 0 - 0 0.033 0.033 0.033 0.167 0.033 0.133 0.383 0 0.033 0 0 0.033 0.567 0.067 0.033 0.1 0.142 0.033 0.067 0.4 0.033 0.033
Model, team Honest Helpful Harmless
Yi-1.5-34B-Chat
MERA 		0.59 0.576 0.569
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
Yi-1.5-34B-Chat
MERA 		0.415 0.452 0.572 0.752 0.565 0.647 0.66 0.63 0.503 0.57 0.465 0.484 0.37 0.639 0.581 0.555 0.6 0.521 0.444 0.511 0.33 0.485 0.44 0.416 0.393 0.56 0.399 0.574 0.657 0.38 0.7 0.678 0.589 0.687 0.506 0.547 0.37 0.594 0.464 0.463 0.646 0.379 0.552 0.621 0.588 0.477 0.603 0.404 0.411 0.52 0.53 0.679 0.595 0.525 0.77 0.642 0.575
Model, team SIM FL STA
Yi-1.5-34B-Chat
MERA 		0.531 0.497 0.637
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
Yi-1.5-34B-Chat
MERA 		0.444 0.545 0.483 0.546 0.553 0.655 0.431 0.579 0.462 0.462 0.641 0.6 0.433 0.589 0.497 0.444 0.645 0.556 0.491 0.596 0.316 0.576 0.756 0.574 0.689 0.5 0.603 0.526 0.754 0.556 0.778 0.654 0.554 0.737 0.5 0.554 0.578 0.689 0.579 0.462 0.592 0.46 0.778 0.822 0.845 0.8 0.552 0.773 0.615 0.754 0.733 0.348 0.671 0.571 0.535 0.404 0.622
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Yi-1.5-34B-Chat
MERA 		0.264 0.239 0.278 0.228 0.212
Model, team Virtue Law Moral Justice Utilitarianism
Yi-1.5-34B-Chat
MERA 		0.267 0.285 0.286 0.214 0.223
Model, team Virtue Law Moral Justice Utilitarianism
Yi-1.5-34B-Chat
MERA 		0.213 0.201 0.221 0.175 0.181
Model, team Women Men LGBT Nationalities Migrants Other
Yi-1.5-34B-Chat
MERA 		0.704 0.629 0.471 0.757 0.286 0.557

Пользовательское соглашение

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