12 commits

Author	SHA1	Message	Date
Ayo	07672aaf2b	refactor: remove unneeded torch.no_grad()	2025-09-03 17:06:11 +02:00
Ayo	1cb2a50ce5	feat: load to XPU	2025-09-03 09:02:23 +02:00
Ayo	aeb26374df	feat: use same config as cuda for xpu	2025-09-02 23:26:22 +02:00
Ayo	66b32b2e4e	feat: use xpu as default if it exists	2025-09-02 23:13:54 +02:00
gregory-fanous	6dd49cb720	feat: add Apple MPS device support (dtype + attention handling) to demos (#67) ... * Enhance model loading with device support and error handling Updated device handling for model loading and added support for MPS. Improved error handling and fallback mechanisms for attention implementations. * Improve device handling and model loading logic Updated device argument handling to support MPS and added validation for MPS availability. Enhanced model loading logic based on the selected device type. * fallback only when flash_attention_2 and add some comments back --------- Co-authored-by: YaoyaoChang <cyy574006791@qq.com>	2025-09-01 17:16:01 +08:00
YaoyaoChang	3074f898ca	support CPU	2025-09-01 16:32:50 +08:00
YaoyaoChang	4b8b6f7700	update	2025-08-28 01:28:27 -07:00
_	e611deafac	FIX: adjust quote type in inference_from_file (#33) ... adds full_script = full_script.replace("’", "'") to the data preparation	2025-08-28 15:24:42 +08:00
YaoyaoChang	ac0104c65e	try sdpa if error in flash_attention_2	2025-08-27 18:46:15 -07:00
YaoyaoChang	8ed05ba7de	update to sdpa	2025-08-27 18:21:30 -07:00
YaoyaoChang	056bb5b0fa	add args to use_eager	2025-08-26 19:44:34 -07:00
pengzhiliang	d5895e76d4	init	2025-08-25 15:28:13 +00:00