Microsoft India Proposes Varuna: Scalable, Low-Cost Training of Massive Deep Learning ... - Synced

Retrieved on: 2021-11-10 15:47:02

Tags for this article:

Click the tags to see associated articles and topics

Microsoft India Proposes Varuna: Scalable, Low-Cost Training of Massive Deep Learning ... - Synced. View article details on hiswai:

Excerpt

... was a game-changer, but even BERT-large's 340 million parameters have since been eclipsed by OpenAI's GPT-3 with its 175 billion parameters.

Article found on: syncedreview.com