Search
Thanks to Apple's MLX framework, Ollama runs faster on Macs
9 hour ago /
Read about 0 minute
Author：小编
Ollama has recently been updated to incorporate Apple's machine learning framework, MLX, into its local large model operation solution, resulting in a significant performance boost on Macs equipped with Apple chips. Officials claim that the new version achieves approximately a 1.6-fold increase in processing speed during the prompt pre-filling stage and nearly doubles the speed during response generation, leading to an overall more responsive experience. Models equipped with the M5 series chips benefit the most, thanks to the GPU neural accelerator integrated into Apple's new-generation chips.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
