Commit History

Speed: 4-bit default on Spaces, SDPA option, lower token limits; CUDA greedy fix
b904a07

SebAustin commited on

V1.2
b6d4c5f

SebAustin commited on

V1.1
29e2ed7

SebAustin commited on

V1
3265b47

SebAustin commited on