Multi-domain, multi-language SFT dataset pushes LLM performance to the next level

BigCode project: Code-generating LLMs boosted by Toloka's crowd

Object detection and matching, image-to-text, and photo moderation for geoservices

Perplexity enhances LLMs with holistic quality evaluation

Toloka helps ServiceNow increase evaluation throughput multiple times

LLM for code generation: a scalable pipeline to gather SFT data

Boosting revenue via DSAT: offline evaluation to enhance e-commerce search performance

Calculating evaluation metrics for a voice assistant

Product moderation for e-commerce: a case study