John T Davies ๐Ÿ‡บ๐Ÿ‡ฆ๐Ÿ‡ช๐Ÿ‡บ๐ŸŒ (@jtdavies) 's Twitter Profile
John T Davies ๐Ÿ‡บ๐Ÿ‡ฆ๐Ÿ‡ช๐Ÿ‡บ๐ŸŒ

@jtdavies

Entrepreneur, CTO in Gen-AI, investor, father to 3 grown boys, husband to Rachel, astrophysicist, keen photographer, cyclist, รผber-geek, travelled a lot.

ID: 15784290

linkhttp://www.johntdavies.com calendar_today08-08-2008 23:20:02

3,3K Tweet

1,1K Takipรงi

531 Takip Edilen

John T Davies ๐Ÿ‡บ๐Ÿ‡ฆ๐Ÿ‡ช๐Ÿ‡บ๐ŸŒ (@jtdavies) 's Twitter Profile Photo

310 Tokens / second with an unquantised (bf16) model using MLX, this is CRAZY fast! I've just tried the new MLX-VML beta from Prince Canuma (using an M4 Max). DeepSeek-OCR (3B) on MLX is a game-changer.

310 Tokens / second with an unquantised (bf16) model using MLX, this is CRAZY fast!
I've just tried the new MLX-VML beta from <a href="/Prince_Canuma/">Prince Canuma</a> (using an M4 Max). DeepSeek-OCR (3B) on MLX is a game-changer.