13 Commits

Author	SHA1	Message	Date
Ting-Jun Wang	17949bd1a6	feat: add training Dockerfile	2024-05-29 04:15:35 +08:00
Ting-Jun Wang	0a287e3b46	feat: save()	2024-05-22 19:58:13 +08:00
Ting-Jun Wang	20fd2fbe08	fix: move trainer	2024-05-20 15:29:02 +00:00
TING-JUN WANG	86e0c50a65	fix: global rank	2024-05-16 23:31:24 +08:00
TING-JUN WANG	d4b9aaa1d6	feat: change rank when multi machine training	2024-05-16 23:17:35 +08:00
TING-JUN WANG	874c160eae	fix: delete matplotlib dependence	2024-05-16 23:09:35 +08:00
TING-JUN WANG	233bec6d1c	feat: torchrun on single machine	2024-05-16 20:55:18 +08:00
TING-JUN WANG	24240f1c3a	feat: trainer class for single/multi GPU	2024-05-16 20:25:56 +08:00
TING-JUN WANG	8f3253ff24	fix: single machine parallel training success	2024-05-16 16:28:22 +08:00
Ting-Jun Wang	aedc6b46e9	feat: single machine DDP (test)	2024-05-16 15:59:20 +08:00
Ting-Jun Wang	939aa6d92e	docs: add some hint comment	2024-05-15 20:55:08 +08:00
snsd0805	fc01163995	feat: train on single GPU	2024-05-12 01:48:39 +08:00
snsd0805	bf905e9e03	feat: dataset	2024-05-12 00:43:50 +08:00