19 Commits
Author SHA1 Message Date
Phil Wang 5ad56dda25 save samples and models to ./results path 2020-10-09 21:39:46 -07:00
Phil Wang ff451f697e update with new and improved cosine noise scheduler 2020-10-09 21:21:02 -07:00
Phil Wang ef2ca0b625 new paper suggests image linear attention is more effective without query normalization 2020-10-04 21:53:53 -07:00
Phil Wang 9f95a03c07 fix bug with rezero and linear attention 2020-09-21 20:11:34 -07:00
Phil Wang a4c68d3569 fix bug 2020-09-15 15:57:39 -07:00
Phil Wang b33a48e342 make sure when sampling, batch does not exceed training batch size 2020-09-15 15:15:10 -07:00
Phil Wang 4bf28914bc allow for mixed precision training with fp16 flag 2020-09-08 17:26:23 -07:00
Phil Wang 88f83d0ff2 fix loading checkpoint 2020-09-08 15:15:59 -07:00
Phil Wang 26b5cab6c8 update ema more frequently 2020-09-08 15:03:38 -07:00
Phil Wang 1307b3115d add exponential moving average of model, also allow passing in custom noise schedule 2020-09-08 14:58:09 -07:00
Phil Wang 698227ae13 fix small bug with gradient accumulation 2020-09-07 23:14:38 -07:00
Phil Wang c479adf960 add interpolation 2020-09-07 22:38:59 -07:00
Phil Wang 9c758662a3 fix another stray bug 2020-09-06 23:25:30 -07:00
Phil Wang 1f1e42e9f9 update readme 2020-09-06 22:35:37 -07:00
Phil Wang 11f27032ba offer training class to easily train model off an image directory 2020-09-06 14:22:44 -07:00
Phil Wang d59d8b05f6 fix a small bug 2020-09-06 01:50:59 -07:00
Phil Wang 0cc4aecfd5 first alpha release 2020-09-06 01:40:57 -07:00
Phil Wang c9855c8c74 complete model portion 2020-09-05 13:55:17 -07:00
Phil Wang dda425dc61 add scaffold 2020-08-31 12:45:27 -07:00