ai-forever / MERA
MERA (Multimodal Evaluation for Russian-language Architectures) is a new open benchmark for the Russian language for evaluating fundamental models.
☆58Updated last month
Related projects ⓘ
Alternatives and complementary repositories for MERA
- SAGE: Spelling correction, corruption and evaluation for multiple languages☆131Updated last month
- Modified Arena-Hard-Auto LLM evaluation toolkit with an emphasis on Russian language☆24Updated 3 weeks ago
- Бенчмарк сравнивает русские аналоги ChatGPT: Saiga, YandexGPT, Gigachat☆57Updated last year
- Augmentex — a library for augmenting texts with errors☆52Updated 4 months ago
- Effective LLM Alignment Toolkit☆87Updated 3 weeks ago
- ☆58Updated 9 months ago
- Framework for processing and filtering datasets☆25Updated 3 months ago
- "Руформеры" - список популярных базовых моделей на основе трансформеров для решения задач по автоматической обработке русского языка☆36Updated last year
- RuLeanALBERT is a pretrained masked language model for the Russian language that uses a memory-efficient architecture.☆93Updated last year
- Telegram bot for different language models. Supports system prompts and images☆39Updated 3 weeks ago
- Top ML papers of the week.☆21Updated this week
- The tiniest sentence encoder for Russian language☆189Updated 3 months ago
- Deep Learning for Speech☆80Updated last week
- Foundational Model for Speech Recognition Tasks☆113Updated 5 months ago
- MMLU eval for RU/EN☆14Updated last year
- Question answering on russian with XLMRobertaLarge as a service☆21Updated 3 years ago
- Russian Corpus of Linguistic Acceptability☆41Updated last month
- RuTransform: python framework for adversarial attacks and text data augmentation for Russian☆17Updated last year
- Простой нормализатор текстов перед синтезом речи☆20Updated 6 months ago
- Russian SuperGLUE benchmark☆108Updated last year
- Библиотека для извлечения статистик из текстов на русском языке.☆103Updated last year
- A new second practical assignment for Huawei's NLP course☆16Updated 8 months ago
- ☆78Updated last year
- ML Course created for Bauman Moscow State Technical University☆59Updated 2 years ago
- CLIP implementation for Russian language☆139Updated last year
- Pipeline for easy fine-tuning of BERT architecture for sequence classification☆22Updated last year
- RuBLiMP: Russian Benchmark of Linguistic Minimal Pairs☆15Updated last month
- Question-answer bot, using Retrieval-Augmented Generation method.☆14Updated last year
- ☆21Updated last year