Yi-Coder-9B

MERA Created at 04.11.2024 18:05
0.374
The overall result
363
Place in the rating
Weak tasks:
238
RWSD
382
PARus
340
RCB
388
ruEthics
549
MultiQ
466
ruWorldTree
472
ruOpenBookQA
440
CheGeKa
467
ruMMLU
383
ruHateSpeech
430
ruDetox
344
ruHHH
483
ruTiE
70
ruHumanEval
380
USE
318
MathLogicQA
374
ruMultiAr
364
SimpleAr
135
LCS
109
BPS
345
ruModAr
443
MaMuRAMu
83
ruCodeEval
+19
Hide

Ratings for leaderboard tasks

The table will scroll to the left

Task name Result Metric
LCS 0.144 Accuracy
RCB 0.505 / 0.42 Accuracy F1 macro
USE 0.085 Grade norm
RWSD 0.55 Accuracy
PARus 0.662 Accuracy
ruTiE 0.509 Accuracy
MultiQ 0.135 / 0.082 F1 Exact match
CheGeKa 0.023 / 0.005 F1 Exact match
ruModAr 0.404 Exact match
MaMuRAMu 0.462 Accuracy
ruMultiAr 0.188 Exact match
ruCodeEval 0.259 / 0.611 / 0.707 Pass@k
MathLogicQA 0.379 Accuracy
ruWorldTree 0.602 / 0.595 Accuracy F1 macro
ruOpenBookQA 0.515 / 0.51 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 0.977 Accuracy
ruMMLU 0.379 Accuracy
SimpleAr 0.893 Exact match
ruHumanEval 0.293 / 0.618 / 0.707 Pass@k
ruHHH 0.596
ruHateSpeech 0.6
ruDetox 0.095
ruEthics
Correct God Ethical
Virtue 0.142 0.172 0.14
Law 0.144 0.167 0.152
Moral 0.139 0.188 0.158
Justice 0.144 0.179 0.116
Utilitarianism 0.114 0.126 0.148

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
Batch
1
Transformers version
4.44.2
The number of GPUs and their type
1 x NVIDIA H100 80GB HBM3
Architecture
vllm

Team:

MERA

Name of the ML model:

Yi-Coder-9B

Model size

9.0B

Model type:

Opened

Pretrain

Additional links:

https://01-ai.github.io/blog.html?post=en/2024-09-05-A-Small-but-Mighty-LLM-for-Code.md

Architecture description:

Yi-Coder-9B builds upon Yi-9B with an additional 2.4T high-quality tokens, meticulously sourced from a repository-level code corpus on GitHub and code-related data filtered from CommonCrawl.

Description of the training:

-

Pretrain data:

Continue pretrained on 2.4 Trillion high-quality tokens over 52 major programming languages.

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 - 131072

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
Yi-Coder-9B
MERA
0.167 0.067 0.467 0.1 0 0.033 0 - 0 0.067 0 0.033 0.133 0 0.133 0.167 0.033 0 0.033 0.033 0 0.133 0 0 0.033 0.142 0.033 0.133 0.167 0.1 0.1
Model, team Honest Helpful Harmless
Yi-Coder-9B
MERA
0.574 0.644 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-Coder-9B
MERA
0.296 0.367 0.414 0.611 0.395 0.453 0.534 0.444 0.398 0.372 0.36 0.349 0.31 0.435 0.437 0.393 0.46 0.333 0.311 0.359 0.24 0.333 0.3 0.37 0.304 0.35 0.299 0.426 0.453 0.24 0.56 0.537 0.417 0.545 0.4 0.415 0.29 0.423 0.291 0.36 0.399 0.313 0.524 0.438 0.414 0.333 0.431 0.352 0.387 0.346 0.39 0.435 0.341 0.412 0.64 0.424 0.363
Model, team SIM FL STA
Yi-Coder-9B
MERA
0.814 0.561 0.252
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-Coder-9B
MERA
0.489 0.446 0.383 0.435 0.421 0.466 0.448 0.386 0.308 0.385 0.603 0.533 0.417 0.457 0.392 0.333 0.486 0.422 0.298 0.456 0.281 0.441 0.489 0.391 0.533 0.364 0.487 0.526 0.754 0.511 0.711 0.423 0.464 0.439 0.364 0.518 0.511 0.622 0.404 0.462 0.457 0.317 0.578 0.6 0.707 0.711 0.483 0.727 0.631 0.632 0.711 0.246 0.532 0.532 0.419 0.292 0.467
Coorect
Good
Ethical
Model, team Virtue Law Moral Justice Utilitarianism
Yi-Coder-9B
MERA
0.142 0.144 0.139 0.144 0.114
Model, team Virtue Law Moral Justice Utilitarianism
Yi-Coder-9B
MERA
0.172 0.167 0.188 0.179 0.126
Model, team Virtue Law Moral Justice Utilitarianism
Yi-Coder-9B
MERA
0.14 0.152 0.158 0.116 0.148
Model, team Women Men LGBT Nationalities Migrants Other
Yi-Coder-9B
MERA
0.583 0.571 0.647 0.649 0.571 0.607