4 Commits

Author	SHA1	Message	Date
Prince Canuma	b5e18ef1e3	Add Phi-3.5-MoE (#946) ... * add phimoe * add phimoe to tunner * add switch_mlp * fix SuScaled args * nits --------- Co-authored-by: Awni Hannun <awni@apple.com>	2024-08-24 06:52:33 -07:00
Awni Hannun	6731254e76	Use fast rope (#945) ... * use fast rope * fix llama * use fast rope for llama3.1 * requires unreleased mlx * fix su * fix deepseek v2 * only one of base or freqs * nit * fix * hard code freqs	2024-08-23 13:18:51 -07:00
Awni Hannun	7be292c0c9	Handle longer prompt/generation (#931) ... * rebase * nits * nit * fix rotating cache with step prefill * update version	2024-08-16 15:28:39 -07:00
JosefAlbers	fda41545a6	Su-RoPE(Rotary Position Embedding) for Phi-3 (#813) ... * Su-RoPE * nits * Update su_rope.py * Update su_rope.py Per GPT4: "The error TypeError: 'type' object is not subscriptable is caused by using the type hint list[float] in a version of Python that does not support it. This syntax is only available in Python 3.9 and later." * Ran isort --------- Co-authored-by: Awni Hannun <awni@apple.com>	2024-06-11 06:20:04 -07:00