Update f16c16_patch_1/log.txt
Browse files- f16c16_patch_1/log.txt +639 -0
f16c16_patch_1/log.txt
CHANGED
|
@@ -0,0 +1,639 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [09:56<00:00, 6.14s/it]
|
| 3 |
+
FID is 24.463851928710938
|
| 4 |
+
(512, 256, 256, 3)
|
| 5 |
+
Calc FID for CFG 1.0 and denoise_timesteps 64
|
| 6 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 7 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 8 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 9 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 10 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 11 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:08<00:00, 3.18s/it]
|
| 12 |
+
FID is 24.704532623291016
|
| 13 |
+
(512, 256, 256, 3)
|
| 14 |
+
Calc FID for CFG 1.0 and denoise_timesteps 32
|
| 15 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 16 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 17 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 18 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 19 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 20 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:14<00:00, 2.01s/it]
|
| 21 |
+
FID is 25.24359893798828
|
| 22 |
+
(512, 256, 256, 3)
|
| 23 |
+
Calc FID for CFG 1.0 and denoise_timesteps 16
|
| 24 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 25 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 26 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 27 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 28 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 29 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:17<00:00, 1.41s/it]
|
| 30 |
+
FID is 26.47027587890625
|
| 31 |
+
(512, 256, 256, 3)
|
| 32 |
+
Calc FID for CFG 1.0 and denoise_timesteps 8
|
| 33 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 34 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 35 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 36 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 37 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 38 |
+
100%|οΏ½οΏ½οΏ½ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:47<00:00, 1.11s/it]
|
| 39 |
+
FID is 29.478370666503906
|
| 40 |
+
(512, 256, 256, 3)
|
| 41 |
+
FID for CFG 1.0 and denoise_timesteps 4
|
| 42 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 43 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 44 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 45 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 46 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 47 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:33<00:00, 1.04it/s]
|
| 48 |
+
FID is 39.31206512451172
|
| 49 |
+
(512, 256, 256, 3)
|
| 50 |
+
Calc FID for CFG 1.0 and denoise_timesteps 2
|
| 51 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 52 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 53 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 54 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 55 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 56 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:26<00:00, 1.13it/s]
|
| 57 |
+
FID is 102.28096008300781
|
| 58 |
+
(512, 256, 256, 3)
|
| 59 |
+
Calc FID for CFG 1.0 and denoise_timesteps 1
|
| 60 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 61 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 62 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 63 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 64 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 65 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:21<00:00, 1.19it/s]
|
| 66 |
+
FID is 219.12063598632812
|
| 67 |
+
(512, 256, 256, 3)
|
| 68 |
+
Calc FID for CFG 1.25 and denoise_timesteps 128
|
| 69 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 70 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 71 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 72 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 73 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 74 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:36<00:00, 10.27s/it]
|
| 75 |
+
FID is 13.00404167175293
|
| 76 |
+
(512, 256, 256, 3)
|
| 77 |
+
Calc FID for CFG 1.25 and denoise_timesteps 64
|
| 78 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 79 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 80 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 81 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 82 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 83 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:56<00:00, 5.53s/it]
|
| 84 |
+
FID is 13.148388862609863
|
| 85 |
+
(512, 256, 256, 3)
|
| 86 |
+
Calc FID for CFG 1.25 and denoise_timesteps 32
|
| 87 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 88 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 89 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 90 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 91 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 92 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:08<00:00, 3.18s/it]
|
| 93 |
+
FID is 13.484004974365234
|
| 94 |
+
for CFG 1.25 and denoise_timesteps 16
|
| 95 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 96 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 97 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 98 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 99 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 100 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:14<00:00, 2.01s/it]
|
| 101 |
+
FID is 14.35787296295166
|
| 102 |
+
(512, 256, 256, 3)
|
| 103 |
+
Calc FID for CFG 1.25 and denoise_timesteps 8
|
| 104 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 105 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 106 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 107 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 108 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 109 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:16<00:00, 1.41s/it]
|
| 110 |
+
FID is 16.668533325195312
|
| 111 |
+
(512, 256, 256, 3)
|
| 112 |
+
Calc FID for CFG 1.25 and denoise_timesteps 4
|
| 113 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 114 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 115 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 116 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 117 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 118 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:47<00:00, 1.11s/it]
|
| 119 |
+
FID is 24.52108383178711
|
| 120 |
+
(512, 256, 256, 3)
|
| 121 |
+
Calc FID for CFG 1.25 and denoise_timesteps 2
|
| 122 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 123 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 124 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 125 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 126 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 127 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:33<00:00, 1.04it/s]
|
| 128 |
+
FID is 80.90126037597656
|
| 129 |
+
(512, 256, 256, 3)
|
| 130 |
+
Calc FID for CFG 1.25 and denoise_timesteps 1
|
| 131 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 132 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 133 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 134 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 135 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 136 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:26<00:00, 1.12it/s]
|
| 137 |
+
FID is 210.71218872070312
|
| 138 |
+
(512, 256, 256, 3)
|
| 139 |
+
Calc FID for CFG 1.5 and denoise_timesteps 128
|
| 140 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 141 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 142 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 143 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 144 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 145 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββοΏ½οΏ½οΏ½βββββββββββββββββββββββββββββββ| 97/97 [16:35<00:00, 10.27s/it]
|
| 146 |
+
FID is 7.820997714996338
|
| 147 |
+
(512, 256, 256, 3)
|
| 148 |
+
Calc FID for CFG 1.5 and denoise_timesteps 64 [226/1988]
|
| 149 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 150 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 151 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 152 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 153 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 154 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:57<00:00, 5.55s/it]
|
| 155 |
+
FID is 7.919877052307129
|
| 156 |
+
(512, 256, 256, 3)
|
| 157 |
+
Calc FID for CFG 1.5 and denoise_timesteps 32
|
| 158 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 159 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 160 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 161 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 162 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 163 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:09<00:00, 3.19s/it]
|
| 164 |
+
FID is 8.184598922729492
|
| 165 |
+
(512, 256, 256, 3)
|
| 166 |
+
Calc FID for CFG 1.5 and denoise_timesteps 16
|
| 167 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 168 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 169 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 170 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 171 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 172 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:15<00:00, 2.02s/it]
|
| 173 |
+
FID is 8.815644264221191
|
| 174 |
+
(512, 256, 256, 3)
|
| 175 |
+
Calc FID for CFG 1.5 and denoise_timesteps 8
|
| 176 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 177 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 178 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 179 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 180 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 181 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:18<00:00, 1.43s/it]
|
| 182 |
+
FID is 10.650046348571777
|
| 183 |
+
(512, 256, 256, 3)
|
| 184 |
+
Calc FID for CFG 1.5 and denoise_timesteps 4
|
| 185 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 186 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 187 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 188 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 189 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 190 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:49<00:00, 1.13s/it]
|
| 191 |
+
FID is 16.99493408203125
|
| 192 |
+
(512, 256, 256, 3)
|
| 193 |
+
Calc FID for CFG 1.5 and denoise_timesteps 2
|
| 194 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 195 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 196 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 197 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 198 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 199 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:33<00:00, 1.04it/s]
|
| 200 |
+
FID is 68.02703857421875
|
| 201 |
+
Calc FID for CFG 1.5 and denoise_timesteps 1
|
| 202 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 203 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 204 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 205 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 206 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 207 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:26<00:00, 1.12it/s]
|
| 208 |
+
FID is 204.97640991210938
|
| 209 |
+
(512, 256, 256, 3)
|
| 210 |
+
Calc FID for CFG 1.75 and denoise_timesteps 128
|
| 211 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 212 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 213 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 214 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 215 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 216 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:35<00:00, 10.27s/it]
|
| 217 |
+
FID is 6.368059158325195
|
| 218 |
+
(512, 256, 256, 3)
|
| 219 |
+
Calc FID for CFG 1.75 and denoise_timesteps 64
|
| 220 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 221 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 222 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 223 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 224 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 225 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:57<00:00, 5.55s/it]
|
| 226 |
+
FID is 6.437324523925781
|
| 227 |
+
(512, 256, 256, 3)
|
| 228 |
+
Calc FID for CFG 1.75 and denoise_timesteps 32
|
| 229 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 230 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 231 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 232 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 233 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 234 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:08<00:00, 3.18s/it]
|
| 235 |
+
FID is 6.6300506591796875
|
| 236 |
+
(512, 256, 256, 3)
|
| 237 |
+
Calc FID for CFG 1.75 and denoise_timesteps 16
|
| 238 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 239 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 240 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 241 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 242 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 243 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:14<00:00, 2.01s/it]
|
| 244 |
+
FID is 7.124289512634277
|
| 245 |
+
alc FID for CFG 1.75 and denoise_timesteps 8
|
| 246 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 247 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 248 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 249 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 250 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 251 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:18<00:00, 1.42s/it]
|
| 252 |
+
FID is 8.551801681518555
|
| 253 |
+
(512, 256, 256, 3)
|
| 254 |
+
Calc FID for CFG 1.75 and denoise_timesteps 4
|
| 255 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 256 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 257 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 258 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 259 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 260 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:48<00:00, 1.12s/it]
|
| 261 |
+
FID is 13.70229721069336
|
| 262 |
+
(512, 256, 256, 3)
|
| 263 |
+
Calc FID for CFG 1.75 and denoise_timesteps 2
|
| 264 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 265 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 266 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 267 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 268 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 269 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:35<00:00, 1.02it/s]
|
| 270 |
+
FID is 61.87627410888672
|
| 271 |
+
(512, 256, 256, 3)
|
| 272 |
+
Calc FID for CFG 1.75 and denoise_timesteps 1
|
| 273 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 274 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 275 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 276 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 277 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 278 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:27<00:00, 1.11it/s]
|
| 279 |
+
FID is 200.99813842773438
|
| 280 |
+
(512, 256, 256, 3)
|
| 281 |
+
Calc FID for CFG 2.0 and denoise_timesteps 128
|
| 282 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 283 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 284 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 285 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 286 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 287 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:37<00:00, 10.28s/it]
|
| 288 |
+
FID is 6.765195846557617
|
| 289 |
+
(512, 256, 256, 3)
|
| 290 |
+
Calc FID for CFG 2.0 and denoise_timesteps 64
|
| 291 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 292 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 293 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 294 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 295 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 296 |
+
100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββοΏ½οΏ½ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:58<00:00, 5.56s/it]
|
| 297 |
+
FID is 6.835395336151123
|
| 298 |
+
(512, 256, 256, 3)
|
| 299 |
+
Calc FID for CFG 2.0 and denoise_timesteps 32
|
| 300 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 301 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 302 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 303 |
+
FID is 7.002254962921143
|
| 304 |
+
(512, 256, 256, 3)
|
| 305 |
+
Calc FID for CFG 2.0 and denoise_timesteps 16
|
| 306 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 307 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 308 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 309 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 310 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 311 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:14<00:00, 2.01s/it]
|
| 312 |
+
FID is 7.40873384475708
|
| 313 |
+
(512, 256, 256, 3)
|
| 314 |
+
Calc FID for CFG 2.0 and denoise_timesteps 8
|
| 315 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 316 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 317 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 318 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 319 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 320 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:18<00:00, 1.43s/it]
|
| 321 |
+
FID is 8.563901901245117
|
| 322 |
+
(512, 256, 256, 3)
|
| 323 |
+
Calc FID for CFG 2.0 and denoise_timesteps 4
|
| 324 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 325 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 326 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 327 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 328 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 329 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:48<00:00, 1.12s/it]
|
| 330 |
+
FID is 12.72092056274414
|
| 331 |
+
(512, 256, 256, 3)
|
| 332 |
+
Calc FID for CFG 2.0 and denoise_timesteps 2
|
| 333 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 334 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 335 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 336 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 337 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 338 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:34<00:00, 1.03it/s]
|
| 339 |
+
FID is 60.08670425415039
|
| 340 |
+
(512, 256, 256, 3)
|
| 341 |
+
Calc FID for CFG 2.0 and denoise_timesteps 1
|
| 342 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 343 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 344 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 345 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 346 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 347 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:28<00:00, 1.10it/s]
|
| 348 |
+
FID is 198.37855529785156
|
| 349 |
+
(512, 256, 256, 3)
|
| 350 |
+
Calc FID for CFG 2.25 and denoise_timesteps 128
|
| 351 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 352 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 353 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 354 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 355 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 356 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:37<00:00, 10.28s/it]
|
| 357 |
+
FID is 7.977522850036621
|
| 358 |
+
Calc FID for CFG 2.25 and denoise_timesteps 64 [10/1988]
|
| 359 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 360 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 361 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 362 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 363 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 364 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:59<00:00, 5.56s/it]
|
| 365 |
+
FID is 8.055110931396484
|
| 366 |
+
(512, 256, 256, 3)
|
| 367 |
+
Calc FID for CFG 2.25 and denoise_timesteps 32
|
| 368 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 369 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 370 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 371 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 372 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 373 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββοΏ½οΏ½βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:10<00:00, 3.20s/it]
|
| 374 |
+
FID is 8.19555377960205
|
| 375 |
+
(512, 256, 256, 3)
|
| 376 |
+
Calc FID for CFG 2.25 and denoise_timesteps 16
|
| 377 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 378 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 379 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 380 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 381 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 382 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:15<00:00, 2.01s/it]
|
| 383 |
+
FID is 8.558436393737793
|
| 384 |
+
(512, 256, 256, 3)
|
| 385 |
+
Calc FID for CFG 2.25 and denoise_timesteps 8
|
| 386 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 387 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 388 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 389 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 390 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 391 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:18<00:00, 1.42s/it]
|
| 392 |
+
FID is 9.505577087402344
|
| 393 |
+
(512, 256, 256, 3)
|
| 394 |
+
Calc FID for CFG 2.25 and denoise_timesteps 4
|
| 395 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 396 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 397 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 398 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 399 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 400 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:49<00:00, 1.13s/it]
|
| 401 |
+
FID is 12.873838424682617
|
| 402 |
+
(512, 256, 256, 3)
|
| 403 |
+
Calc FID for CFG 2.25 and denoise_timesteps 2
|
| 404 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 405 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 406 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 407 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 408 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 409 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:35<00:00, 1.02it/s]
|
| 410 |
+
FID is 60.937774658203125
|
| 411 |
+
(512, 256, 256, 3)
|
| 412 |
+
Calc FID for CFG 2.25 and denoise_timesteps 1
|
| 413 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 414 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 415 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 416 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 417 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 418 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:28<00:00, 1.10it/s]
|
| 419 |
+
FID is 196.56155395507812
|
| 420 |
+
(512, 256, 256, 3)
|
| 421 |
+
Calc FID for CFG 2.5 and denoise_timesteps 128
|
| 422 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 423 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 424 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 425 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 426 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 427 |
+
Calc FID for CFG 2.5 and denoise_timesteps 128 [160/1801]
|
| 428 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 429 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 430 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 431 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 432 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 433 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:37<00:00, 10.28s/it]
|
| 434 |
+
FID is 9.51274299621582 (512, 256, 256, 3)
|
| 435 |
+
Calc FID for CFG 2.5 and denoise_timesteps 64
|
| 436 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 437 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 438 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 439 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 440 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 441 |
+
100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββοΏ½οΏ½οΏ½ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:58<00:00, 5.55s/it]
|
| 442 |
+
FID is 9.59133243560791 (512, 256, 256, 3)
|
| 443 |
+
Calc FID for CFG 2.5 and denoise_timesteps 32
|
| 444 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 445 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 446 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 447 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 448 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 449 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:10<00:00, 3.20s/it]
|
| 450 |
+
FID is 9.75399398803711 (512, 256, 256, 3)
|
| 451 |
+
Calc FID for CFG 2.5 and denoise_timesteps 16
|
| 452 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 453 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 454 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 455 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 456 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 457 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:14<00:00, 2.01s/it]
|
| 458 |
+
FID is 10.051287651062012 (512, 256, 256, 3)
|
| 459 |
+
Calc FID for CFG 2.5 and denoise_timesteps 8
|
| 460 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 461 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 462 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 463 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 464 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 465 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:18<00:00, 1.42s/it]
|
| 466 |
+
FID is 10.817917823791504
|
| 467 |
+
(512, 256, 256, 3)
|
| 468 |
+
Calc FID for CFG 2.5 and denoise_timesteps 4
|
| 469 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 470 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 471 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 472 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 473 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 474 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:49<00:00, 1.13s/it]
|
| 475 |
+
FID is 13.553150177001953
|
| 476 |
+
(512, 256, 256, 3)
|
| 477 |
+
Calc FID for CFG 2.5 and denoise_timesteps 2
|
| 478 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 479 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 480 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 481 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 482 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 483 |
+
: Conditioning of shape (512, 768) dtype float32
|
| 484 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:35<00:00, 1.02it/s]
|
| 485 |
+
FID is 63.57429885864258
|
| 486 |
+
(512, 256, 256, 3)
|
| 487 |
+
Calc FID for CFG 2.5 and denoise_timesteps 1
|
| 488 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 489 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 490 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 491 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 492 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 493 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:27<00:00, 1.10it/s]
|
| 494 |
+
FID is 195.20191955566406
|
| 495 |
+
(512, 256, 256, 3)
|
| 496 |
+
Calc FID for CFG 2.75 and denoise_timesteps 128
|
| 497 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 498 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 499 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 500 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 501 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 502 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:37<00:00, 10.28s/it]
|
| 503 |
+
FID is 11.13734245300293
|
| 504 |
+
(512, 256, 256, 3)
|
| 505 |
+
Calc FID for CFG 2.75 and denoise_timesteps 64
|
| 506 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 507 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 508 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 509 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 510 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 511 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:58<00:00, 5.55s/it]
|
| 512 |
+
FID is 11.210472106933594
|
| 513 |
+
(512, 256, 256, 3)
|
| 514 |
+
Calc FID for CFG 2.75 and denoise_timesteps 32
|
| 515 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 516 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 517 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 518 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 519 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 520 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:09<00:00, 3.19s/it]
|
| 521 |
+
FID is 11.345915794372559
|
| 522 |
+
(512, 256, 256, 3)
|
| 523 |
+
Calc FID for CFG 2.75 and denoise_timesteps 16
|
| 524 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 525 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 526 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 527 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 528 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 529 |
+
100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββοΏ½οΏ½ββββββββββββ| 97/97 [03:15<00:00, 2.02s/it]
|
| 530 |
+
FID is 11.617697715759277
|
| 531 |
+
(512, 256, 256, 3)
|
| 532 |
+
Calc FID for CFG 2.75 and denoise_timesteps 8
|
| 533 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 534 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 535 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 536 |
+
ditioning of shape (512, 768) dtype float32
|
| 537 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:17<00:00, 1.42s/it]
|
| 538 |
+
FID is 12.210426330566406
|
| 539 |
+
(512, 256, 256, 3)
|
| 540 |
+
Calc FID for CFG 2.75 and denoise_timesteps 4
|
| 541 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 542 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 543 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 544 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 545 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 546 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:49<00:00, 1.13s/it]
|
| 547 |
+
FID is 14.486330032348633
|
| 548 |
+
(512, 256, 256, 3)
|
| 549 |
+
Calc FID for CFG 2.75 and denoise_timesteps 2
|
| 550 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 551 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 552 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 553 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 554 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 555 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:34<00:00, 1.03it/s]
|
| 556 |
+
FID is 67.45709228515625
|
| 557 |
+
(512, 256, 256, 3)
|
| 558 |
+
Calc FID for CFG 2.75 and denoise_timesteps 1
|
| 559 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 560 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 561 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 562 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 563 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 564 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:28<00:00, 1.10it/s]
|
| 565 |
+
FID is 194.17886352539062
|
| 566 |
+
(512, 256, 256, 3)
|
| 567 |
+
Calc FID for CFG 3.0 and denoise_timesteps 128
|
| 568 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 569 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 570 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 571 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 572 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 573 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [16:36<00:00, 10.27s/it]
|
| 574 |
+
FID is 12.682769775390625
|
| 575 |
+
(512, 256, 256, 3)
|
| 576 |
+
Calc FID for CFG 3.0 and denoise_timesteps 64
|
| 577 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 578 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 579 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 580 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 581 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 582 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [08:59<00:00, 5.56s/it]
|
| 583 |
+
FID is 12.768280029296875
|
| 584 |
+
(512, 256, 256, 3)
|
| 585 |
+
Calc FID for CFG 3.0 and denoise_timesteps 32
|
| 586 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 587 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 588 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 589 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 590 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 591 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [05:09<00:00, 3.19s/it]
|
| 592 |
+
FID is 12.913314819335938
|
| 593 |
+
(512, 256, 256, 3)
|
| 594 |
+
Calc FID for CFG 3.0 and denoise_timesteps 16
|
| 595 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 596 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 597 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 598 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 599 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 600 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [03:15<00:00, 2.01s/it]
|
| 601 |
+
FID is 13.155349731445312
|
| 602 |
+
(512, 256, 256, 3)
|
| 603 |
+
Calc FID for CFG 3.0 and denoise_timesteps 8
|
| 604 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 605 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 606 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 607 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 608 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 609 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [02:17<00:00, 1.42s/it]
|
| 610 |
+
FID is 13.564606666564941
|
| 611 |
+
(512, 256, 256, 3)
|
| 612 |
+
Calc FID for CFG 3.0 and denoise_timesteps 4
|
| 613 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 614 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 615 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 616 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 617 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 618 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:49<00:00, 1.13s/it]
|
| 619 |
+
FID is 15.401412010192871
|
| 620 |
+
(512, 256, 256, 3)
|
| 621 |
+
Calc FID for CFG 3.0 and denoise_timesteps 2
|
| 622 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 623 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 624 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 625 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 626 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 627 |
+
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:35<00:00, 1.01it/s]
|
| 628 |
+
FID is 72.08023834228516
|
| 629 |
+
(512, 256, 256, 3)
|
| 630 |
+
Calc FID for CFG 3.0 and denoise_timesteps 1
|
| 631 |
+
0%| | 0/97 [00:00<?, ?it/s]
|
| 632 |
+
DiT: Input of shape (512, 16, 16, 16) dtype float32
|
| 633 |
+
DiT: After patch embed, shape is (512, 256, 768) dtype bfloat16
|
| 634 |
+
DiT: Patch Embed of shape (512, 256, 768) dtype bfloat16
|
| 635 |
+
DiT: Conditioning of shape (512, 768) dtype float32
|
| 636 |
+
100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββοΏ½οΏ½ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 97/97 [01:28<00:00, 1.10it/s]
|
| 637 |
+
FID is 193.4287567138672
|
| 638 |
+
|
| 639 |
+
|