FRED-T5 large 820M

MERA Created at 12.01.2024 11:15

0.194

The overall result

The submission does not contain all the required tasks

Ratings for leaderboard tasks

The table will scroll to the left

Task name	Result	Metric
LCS	0.086	Accuracy
RCB	0.354 / 0.248	Accuracy F1 macro
USE	0	Grade norm
RWSD	0.492	Accuracy
PARus	0.492	Accuracy
ruTiE	0.493	Accuracy
MultiQ	0.052 / 0	F1 Exact match
CheGeKa	0.001 / 0	F1 Exact match
ruModAr	0.0	Exact match
ruMultiAr	0.0	Exact match
MathLogicQA	0.24	Accuracy
ruWorldTree	0.232 / 0.174	Accuracy F1 macro
ruOpenBookQA	0.265 / 0.215	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

0.475

Accuracy

ruMMLU

0.248

Accuracy

SimpleAr

0.0

Exact match

ruHumanEval

0 / 0 / 0

Pass@k

ruHHH

0.472

ruHateSpeech

0.543

ruDetox

0.003

ruEthics

	Correct	God	Ethical
Virtue	0	0	0
Law	0	0	0
Moral	0	0	0
Justice	0	0	0
Utilitarianism	0	0	0

Information about the submission

Mera version

Torch Version

The version of the codebase

CUDA version

Precision of the model weights

Seed

Batch

Transformers version

The number of GPUs and their type

Architecture

Team:

MERA

Name of the ML model:

FRED-T5 large 820M

Link to the ML model:

https://huggingface.co/ai-forever/FRED-T5-large

Additional links:

https://arxiv.org/abs/2309.10931

Architecture description:

FRED-T5 (Full-scale Russian Enhanced Denoisers) is an encoder-decoder model based on T5 and UL2. Number of attantion heads 16. The dimensions of the hidden layers 1024 and the fully connected layers 2816. GELU activation function.

Description of the training:

Bbpe tokenizer. 50257 + special tokens 107. Prefix tokens: '<LM>', '<SC1>',.. '<SC6>'. Drawing inspiration from Tay et al. (2022), the FRED-T5 1.7.B (or XL) model was pretrained on a mixture of denoisers (MoD), a pretraining objective that represents a set of diverse pretraining objectives. The R-Denoiser is a masked language modeling span corruption objective used in T5. The S-Denoiser follows the language modeling objective, where the input sequence is split into the context and target tokens so that the targets do not rely on future information. The X-Denoiser aims to recover a large part of the input based on the span corruption and language modeling objectives.

Pretrain data:

It was trained on Russian language corpus (300GB).

Training Details:

FRED-T5 is pretrained using a linear scheduler with the initial learning rate of 1e−4 and the Adafactor optimizer (Shazeer and Stern, 2018) with β1 = 0.9, β2 = 0.99, and ϵ = 1e−8. The sequence length is set to 512/512 for inputs and targets. The FRED-T5-XL models is pretrained pretrained with a total batch size of 2048 for 35 days on 160 V100 GPUs, followed by 5 days on 80 A100 GPUs.

License:

MIT

Strategy, generation and parameters:

Code version v.1.1.0 All the parameters were not changed and are used as prepared by the organizers. Details: - 1 x NVIDIA A100 - dtype auto - Pytorch 2.1.2 + CUDA 12.1 - Transformers 4.36.2 - Context length 512

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
FRED-T5 large 820M MERA	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-

Model, team	Honest	Helpful	Harmless
FRED-T5 large 820M MERA	0.492	0.458	0.466

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
FRED-T5 large 820M MERA	0.5	0.25	0.2	0.143	0.286	0.3	0.133	0.235	0.1	0.4	0.182	0	0.6	0.192	0.136	0.3	0.1	0.296	0.7	0.2	0.4	0.212	0.2	0.255	0.3	0.364	0.25	0.214	0.1	0.091	0.3	0.056	0.4	0.1	0.091	0.2	0.2	0.286	0.4	0	0.278	0.1	0.3	0.4	0.25	0.3	0.4	0.4	0.2	0.5	0.318	0.188	0.265	0.267	0.167	0.242	0.296

Model, team	SIM	FL	STA
FRED-T5 large 820M MERA	0.098	0.55	0.051

Coorect

Good

Ethical

Model, team	Virtue	Law	Moral	Justice	Utilitarianism
FRED-T5 large 820M MERA	0	0	0	0	0

Model, team	Virtue	Law	Moral	Justice	Utilitarianism
FRED-T5 large 820M MERA	0	0	0	0	0

Model, team	Virtue	Law	Moral	Justice	Utilitarianism
FRED-T5 large 820M MERA	0	0	0	0	0

Model, team	Women	Men	LGBT	Nationalities	Migrants	Other
FRED-T5 large 820M MERA	0.519	0.686	0.588	0.595	0.286	0.492

FRED-T5 large 820M

Ratings for leaderboard tasks

Evaluation on open tasks:

Information about the submission

Team:

Name of the ML model:

Link to the ML model:

Additional links:

Architecture description:

Description of the training:

Pretrain data:

Training Details:

License:

Strategy, generation and parameters:

Ratings by subcategory

Confirm the deletion of the sub