In foreign bases for training AI analysts found hundreds of songs Basta and Bilan
Open foreign datasets for training neural networks contain hundreds of tracks of popular Russian artists. Analysts of the National Federation of the Music Industry (NFMI) came to such conclusions as a result of searching for songs in the database of the AI Watchdog project of the American magazine The Atlantic. AI Watchdog analyzes Laion-Disco-12M datasets and Sleeping-D 9M datasets from Sleeping AI. The first contains 12.3 million songs from YouTube, the second contains 9.7 million songs from the same video hosting site and their lyrics from the Genius site.
Open foreign datasets for training neural networks contain hundreds of tracks of popular Russian artists. Analysts of the National Federation of the Music Industry (NFMI) came to such conclusions as a result of searching for songs in the database of the AI Watchdog project of the American magazine The Atlantic. AI Watchdog analyzes Laion-Disco-12M datasets and Sleeping-D 9M datasets from Sleeping AI. The first contains 12.3 million songs from YouTube, the second contains 9.7 million songs from the same video hosting site and their lyrics from the Genius site.
