PseudoTerminal X
commited on
Commit
•
10e0640
1
Parent(s):
3e25706
Trained for 0 epochs and 500 steps.
Browse filesTrained with datasets ['text-embeds-pixart-nofilter', 'photo-concept-bucket']
Learning rate 1e-06, batch size 32, and 2 gradient accumulation steps.
Used DDPM noise scheduler for training with epsilon prediction type and rescaled_betas_zero_snr=False
Using 'trailing' timestep spacing.
Base model: PixArt-alpha/PixArt-Sigma-XL-2-1024-MS
VAE: madebyollin/sdxl-vae-fp16-fix
- README.md +6 -6
- optimizer.bin +1 -1
- random_states_0.pkl +1 -1
- scheduler.bin +1 -1
- training_state-photo-concept-bucket.json +0 -0
- transformer/diffusion_pytorch_model.safetensors +1 -1
README.md
CHANGED
|
@@ -28,7 +28,7 @@ a cute anime character named toast holding a sign that says SOON, sitting next t
|
|
| 28 |
- CFG: `6.5`
|
| 29 |
- CFG Rescale: `0.7`
|
| 30 |
- Steps: `30`
|
| 31 |
-
- Sampler: `
|
| 32 |
- Seed: `42`
|
| 33 |
- Resolutions: `1024x1024,1152x960,896x1152`
|
| 34 |
|
|
@@ -47,10 +47,10 @@ You may reuse the base model text encoder for inference.
|
|
| 47 |
|
| 48 |
- Training epochs: 0
|
| 49 |
- Training steps: 500
|
| 50 |
-
- Learning rate:
|
| 51 |
-
- Effective batch size:
|
| 52 |
-
- Micro-batch size:
|
| 53 |
-
- Gradient accumulation steps:
|
| 54 |
- Number of GPUs: 8
|
| 55 |
- Prediction type: epsilon
|
| 56 |
- Rescaled betas zero SNR: False
|
|
@@ -63,7 +63,7 @@ You may reuse the base model text encoder for inference.
|
|
| 63 |
|
| 64 |
### photo-concept-bucket
|
| 65 |
- Repeats: 0
|
| 66 |
-
- Total number of images: ~
|
| 67 |
- Total number of aspect buckets: 1
|
| 68 |
- Resolution: 1.0 megapixels
|
| 69 |
- Cropped: True
|
|
|
|
| 28 |
- CFG: `6.5`
|
| 29 |
- CFG Rescale: `0.7`
|
| 30 |
- Steps: `30`
|
| 31 |
+
- Sampler: `unipc`
|
| 32 |
- Seed: `42`
|
| 33 |
- Resolutions: `1024x1024,1152x960,896x1152`
|
| 34 |
|
|
|
|
| 47 |
|
| 48 |
- Training epochs: 0
|
| 49 |
- Training steps: 500
|
| 50 |
+
- Learning rate: 1e-06
|
| 51 |
+
- Effective batch size: 512
|
| 52 |
+
- Micro-batch size: 32
|
| 53 |
+
- Gradient accumulation steps: 2
|
| 54 |
- Number of GPUs: 8
|
| 55 |
- Prediction type: epsilon
|
| 56 |
- Rescaled betas zero SNR: False
|
|
|
|
| 63 |
|
| 64 |
### photo-concept-bucket
|
| 65 |
- Repeats: 0
|
| 66 |
+
- Total number of images: ~559104
|
| 67 |
- Total number of aspect buckets: 1
|
| 68 |
- Resolution: 1.0 megapixels
|
| 69 |
- Cropped: True
|
optimizer.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3665677155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96894a9be102cdd197738ebea83e3cc7489e0cbebae60357cc1d44204c5434ce
|
| 3 |
size 3665677155
|
random_states_0.pkl
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16100
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8007348491e1a6bd3d0a5ada2bb3f3d30ad9c1f47efc7c9bb714db31a36fd23f
|
| 3 |
size 16100
|
scheduler.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d03dd60c1db7705ccec8f331ba47fb077d91dd585f5cd968c47b1f84258b116
|
| 3 |
size 1000
|
training_state-photo-concept-bucket.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
transformer/diffusion_pytorch_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1221780352
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62077da035da6c6d14afb629eb1136c525494b7fcaf09033f4d826d0269cd05c
|
| 3 |
size 1221780352
|