Model reference · open weights

rubert-mini-frida

Available as managed deployment Embeddings sergeyzh · community Embeddings 1 variants 6k dl/mo

rubert-mini-frida is an open-weight embedding model from sergeyzh. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.

Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.

What it is

Released bysergeyzh
TypeEmbedding models
TaskEmbeddings
Parameters (lead)32M
Context2k tokens
Runs withsentence-transformers
Based onsergeyzh/rubert-mini-sts
Released2025-03-02
Popularity6k downloads / month
LicenceOpen weights

About

What rubert-mini-frida is

rubert-mini-frida - лёгкая и быстрая модификация FRIDA

Модель для расчетов эмбеддингов предложений на русском и английском языках получена методом дистилляции эмбеддингов ai-forever/FRIDA (размер эмбеддингов - 1536, слоёв - 24) в sergeyzh/rubert-mini-sts (размер эмбеддингов - 312, слоёв - 7). Основной режим использования FRIDA - CLS pooling заменен на mean pooling. Каких-либо других изменений поведения модели (модификации или фильтрации эмбеддингов, использования дополнительной модели) не производилось. Дистиляция выполнена в максимально возможном объеме - эмбеддинги русских и английских предложений, работа префиксов.

Read the full model card

Рекомендуемый размер контекста модели соответствует FRIDA и не превышает 512 токенов (фактический унаследованный от исходной модели - 2048).

Префиксы

Все префиксы унаследованы от FRIDA. Оптимальный (обеспечивающий средние результаты) для большинства задач - "categorize: " прописан по умолчанию в config_sentence_transformers.json

Перечень используемых префиксов и их влияние на оценки модели в encodechka:

ПрефиксSTSPINLISATI
-0.8390.7620.4750.8010.972
search_query:0.8460.7610.4980.8000.973
search_document:0.8300.7480.4680.7940.972
paraphrase:0.8350.7640.4750.7990.973
categorize:0.8500.7610.5160.8020.973
categorize_sentiment:0.7550.6560.4270.7980.959
categorize_topic:0.7340.5230.3890.7280.959
categorize_entailment:0.8370.7530.5440.8020.970

Задачи:

  • Semantic text similarity (STS);
  • Paraphrase identification (PI);
  • Natural language inference (NLI);
  • Sentiment analysis (SA);
  • Toxicity identification (TI).

Метрики

Оценки модели на бенчмарке ruMTEB:

Model NameMetricFridarubert-mini-fridamultilingual-e5-large-instructmultilingual-e5-large
CEDRClassificationAccuracy0.6460.5520.5000.448
GeoreviewClassificationAccuracy0.5770.4640.5590.497
GeoreviewClusteringP2PV-measure0.7830.6980.7430.605
HeadlineClassificationAccuracy0.8900.8800.8620.758
InappropriatenessClassificationAccuracy0.7830.6980.6550.616
KinopoiskClassificationAccuracy0.7050.5950.6610.566
RiaNewsRetrievalNDCG@100.8680.7210.8240.807
RuBQRerankingMAP@100.7710.7110.7170.756
RuBQRetrievalNDCG@100.7240.6540.6920.741
RuReviewsClassificationAccuracy0.7510.6580.6860.653
RuSTSBenchmarkSTSPearson correlation0.8140.8030.8400.831
RuSciBenchGRNTIClassificationAccuracy0.6990.6250.6510.582
RuSciBenchGRNTIClusteringP2PV-measure0.6700.5860.6220.520
RuSciBenchOECDClassificationAccuracy0.5460.4930.5020.445
RuSciBenchOECDClusteringP2PV-measure0.5660.5070.5280.450
SensitiveTopicsClassificationAccuracy0.3980.3730.3230.257
TERRaClassificationAverage Precision0.6650.6060.6390.584
Model NameMetricFridarubert-mini-fridamultilingual-e5-large-instructmultilingual-e5-large
ClassificationAccuracy0.7070.6310.6540.588
ClusteringV-measure0.6730.5970.631

From the published model card. Full card on the HuggingFace links in the sidebar.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys rubert-mini-frida for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (rubert-mini-frida below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/embeddings \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"rubert-mini-frida","input":"text to embed"}'

Create an account — your API key is available in the console. 3M free tokens every 30 days with every new account.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms