wassname/denoising-diffusion-pytorch
1
0
Fork 0
mirror of https://github.com/wassname/denoising-diffusion-pytorch.git synced 2026-08-22 12:01:02 +08:00
Code Issues Packages Projects Releases Wiki Activity
180 Commits 2 Branches 127 Tags
patch-1
Commit Graph
180 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
wassname edc6d5136a Update denoising_diffusion_pytorch_1d.py 2022-12-22 11:00:56 +08:00
wassname 207d23af86 Update denoising_diffusion_pytorch.py 
Shouldn't eta be 0 for DDIM sampling? In the DDIM paper in section 5.1 they mention the eta=0 is DDIM, and eta=1 approximates the normal DDPM. They also use [eta=0](https://github.com/openai/improved-diffusion/blob/e94489283bb876ac1477d5dd7709bbbd2d9902ce/improved_diffusion/gaussian_diffusion.py#L487) in the openai implementation of DDIM.
 2022-12-22 10:57:30 +08:00
Phil Wang cf4f44909f 0.1.5 0.1.5 2022-12-21 08:35:59 -08:00
Phil Wang a7f2d670bb always set results_folder on Trainer 2022-12-21 08:34:34 -08:00
Phil Wang 6ab29d5cea start saving version of library in the checkpoints, so in case of breaking changes, researchers can return to the right working version 0.1.0a 2022-11-28 12:11:11 -08:00
Phil Wang ddc31bc489 trust paper 0.32.0 2022-11-27 10:05:55 -08:00
Phil Wang e872ec3618 tweak readme 2022-11-21 10:08:56 -08:00
Phil Wang 60f760ead5 add classifier free guidance example for @ajrheng at https://github.com/lucidrains/denoising-diffusion-pytorch/issues/120 0.31.1 2022-11-11 20:01:42 -08:00
Phil Wang 6205e20e5e ocd 2022-11-05 17:09:34 -07:00
Phil Wang aece9c2a2a help researchers in need of 1d ddpm, ex. https://github.com/lucidrains/denoising-diffusion-pytorch/issues/116 2022-11-05 11:42:00 -07:00
Phil Wang 12079cadee add the v-parameterization from Salimans et al for the discrete case, allow for distillation 0.30.0 2022-10-30 09:53:28 -07:00
Phil Wang 0ffff59ca0 add option for random fourier features, given misinterpretation of Katherine's code, thanks to @tmabraham for addressing this in https://github.com/lucidrains/denoising-diffusion-pytorch/issues/112 0.29.1 2022-10-30 09:32:06 -07:00
Phil Wang aadaa7d288 bring in the continuous time v-parameterized ddpm, validated to work locally, and which will be used for imagen-video replication 0.29.0 2022-10-23 19:16:31 -07:00
Phil Wang 23fd887a5f switch back to regular attention, given @rromb results 2022-10-16 08:43:10 -07:00
Phil Wang dfbafee555 0.27.12 0.27.12 2022-10-05 13:50:54 -07:00
Phil Wang 40dd8ba1de Merge pull request #102 from npielawski/main 
Added gradient clipping.
 2022-10-05 13:50:39 -07:00
Nicolas Pielawski 2ac3f94a80 Added gradient clipping. 2022-10-05 11:29:49 -07:00
Phil Wang 98f2eeac35 link to flax implementation from @yiyixuxu 2022-09-27 11:21:23 -07:00
Phil Wang 6e8a0f2082 fix auto-conversion of images to mode in dataset 0.27.11 2022-09-20 19:29:35 -07:00
Phil Wang 8c36559295 0.27.10 0.27.10 2022-09-16 17:15:02 -07:00
Phil Wang f74f536339 Merge pull request #90 from kashif/patch-1 
fix torch.cumprod
 2022-09-16 17:14:48 -07:00
Kashif Rasul d85b8bbe2e fix torch.cumprod 2022-09-16 17:19:00 +02:00
Phil Wang e0a1bed31a 0.27.9 0.27.9 2022-09-05 02:43:11 -07:00
Phil Wang 82b67fc00a Merge pull request #85 from RyannDaGreat/main 
Trainer.load can use GPU's other than cuda:0
 2022-09-05 02:42:49 -07:00
Ryan Burgert 7c0cd05c27 Trainer.load can use GPU's other than cuda:0 2022-09-04 22:11:35 -04:00
Phil Wang 6dda508ff6 in ddim, clip x0 before calculation of predicted noise, thanks to @lukovnikov again for pointing out this inconsistency with glides implementation 0.27.8 2022-09-01 09:34:50 -07:00
Phil Wang 9ec8d27217 0.27.7 2022-08-31 07:36:19 -07:00
Phil Wang 4b4ebab7c3 Merge pull request #83 from lukovnikov/fix_ddim 
Fix ddim
 2022-08-31 07:02:26 -07:00
lukovnikov e4a4e4acaa Revert "Revert "fix ddim sampling"" 
This reverts commit cd8329cdd7.
 2022-08-31 15:28:35 +02:00
lukovnikov cd8329cdd7 Revert "fix ddim sampling" 
This reverts commit aec2a26984.
 2022-08-31 15:25:51 +02:00
lukovnikov aec2a26984 fix ddim sampling 2022-08-31 15:24:59 +02:00
Phil Wang c78709f887 0.27.6 0.27.6 2022-08-31 06:20:59 -07:00
Phil Wang 4436128a0b Merge pull request #82 from TheDudeFromCI/patch-1 
Update step before training checkpoint
 2022-08-31 06:20:31 -07:00
TheDudeFromCI e46a89e2bc Update step before training checkpoint 2022-08-31 02:19:12 -07:00
Phil Wang 42158d6248 fix ddim, for issue https://github.com/lucidrains/denoising-diffusion-pytorch/issues/81 0.27.5 2022-08-30 20:30:29 -07:00
Phil Wang 44f95e2e9d readme 2022-08-22 09:13:32 -07:00
Phil Wang d9275a744c add weight standardization prior to groupnorm, lessen cosine sim attention scale to 10 for fp16 0.27.4 2022-08-17 11:42:54 -07:00
Phil Wang beb2f2d8dd add self conditioning for elucidated ddpm 0.27.2 2022-08-10 13:15:45 -07:00
Phil Wang f0d59acdfd fix sampling ddpm tqdm 0.27.1 2022-08-10 12:02:56 -07:00
Phil Wang 689593a579 add the new self conditioning technique from hintons group from bit diffusion paper 
0.27.0
0.27.0 		2022-08-10 10:52:07 -07:00
Phil Wang eba44498d1 higher epsilon for fp16 in layernorm 0.26.5 2022-07-29 13:29:52 -07:00
Phil Wang 12f95b33d8 rescale values to prevent linear attention from overflowing in fp16 setting 0.26.4 2022-07-27 12:25:40 -07:00
Phil Wang 6b504c4ae9 fix accelerator prepare bug for dataloader 0.26.3 2022-07-25 08:13:46 -07:00
Phil Wang 37334ae824 fix a bug with ddim and predict x0 objective v0.26.1 2022-07-18 19:04:57 -07:00
Phil Wang 6eba6cdd50 refresh images 2022-07-18 11:30:34 -07:00
Phil Wang 555566c188 take a gamble on cosine sim attention v0.26.0 2022-07-18 11:29:20 -07:00
Phil Wang 2b742dd2cc move accelerator backward outside of autocast context, also calculate total loss correctly across gradient accumulated steps v0.25.3 2022-07-11 21:02:06 -07:00
Phil Wang 1345a8a41d do not noise at the last timestep for ddim v0.25.2 2022-07-09 18:36:45 -07:00
Phil Wang 931a5af2c3 bring in ddim sampling v0.25.1 2022-07-09 16:10:23 -07:00
Phil Wang a0c3443eaa optimizer should be saved and loaded v0.24.4 2022-07-08 17:43:41 -07:00