§ feed · storyline
Google speeds up Gemma 4 threefold with multi-token prediction
Google releases multi-token prediction drafters for Gemma 4, enabling a small auxiliary model to suggest multiple tokens at once and accelerating text generation by up to three times.
Google has released multi-token prediction drafters for its Gemma 4 open model family that speed up text generation by up to three times. A small auxiliary model suggests several tokens at once while the main model checks them in a single pass.
The article Google speeds up Gemma 4 threefold with multi-token prediction appeared first on The Decoder.
§ sources1 publication · timeline below