Trained for 0 epochs and 5100 steps.

Trained with datasets ['text-embeds-sdxl', 'photocb-clip-embeds', 'celebrities', 'movieposters', 'normalnudes', 'propagandaposters', 'guys', 'pixel-art', 'signs', 'moviecollection', 'bookcovers', 'nijijourney', 'experimental', 'ethnic', 'sports', 'gay', 'architecture', 'shutterstock', 'cinemamix-1mp', 'nsfw-1024', 'anatomy', 'bg20k-1024', 'yoga', 'photo-aesthetics', 'text-1mp', 'photo-concept-bucket']
Learning rate 4e-07, batch size 8, and 4 gradient accumulation steps.
Used DDPM noise scheduler for training with v_prediction prediction type and rescaled_betas_zero_snr=True
Using 'trailing' timestep spacing.
Base model: ptx0/terminus-xl-velocity-v1
VAE: madebyollin/sdxl-vae-fp16-fix

README.md

@@ -44,7 +44,7 @@ You may reuse the base model text encoder for inference.
 ## Training settings
 
 - Training epochs: 0
-- Training steps: 3900
+- Training steps: 5100
 - Learning rate: 4e-07
 - Effective batch size: 32
   - Micro-batch size: 8
@@ -59,7 +59,7 @@ You may reuse the base model text encoder for inference.
 ## Datasets
 
 ### celebrities
-- Repeats: 0
+- Repeats: 4
 - Total number of images: 1184
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
@@ -67,7 +67,7 @@ You may reuse the base model text encoder for inference.
 - Crop style: random
 - Crop aspect: random
 ### movieposters
-- Repeats: 0
+- Repeats: 5
 - Total number of images: 1728
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
@@ -75,7 +75,7 @@ You may reuse the base model text encoder for inference.
 - Crop style: random
 - Crop aspect: random
 ### normalnudes
-- Repeats: 0
+- Repeats: 5
 - Total number of images: 1056
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
@@ -91,7 +91,7 @@ You may reuse the base model text encoder for inference.
 - Crop style: random
 - Crop aspect: random
 ### guys
-- Repeats: 0
+- Repeats: 5
 - Total number of images: 352
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
@@ -107,7 +107,7 @@ You may reuse the base model text encoder for inference.
 - Crop style: random
 - Crop aspect: random
 ### signs
-- Repeats: 0
+- Repeats: 5
 - Total number of images: 352
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
@@ -140,7 +140,7 @@ You may reuse the base model text encoder for inference.
 - Crop aspect: random
 ### experimental
 - Repeats: 0
-- Total number of images: 3042
+- Total number of images: 3040
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
 - Cropped: True
@@ -148,8 +148,8 @@ You may reuse the base model text encoder for inference.
 - Crop aspect: random
 ### ethnic
 - Repeats: 0
-- Total number of images: 3076
-- Total number of aspect buckets: 2
+- Total number of images: 3072
+- Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
 - Cropped: True
 - Crop style: random
@@ -173,7 +173,7 @@ You may reuse the base model text encoder for inference.
 ### architecture
 - Repeats: 0
 - Total number of images: 4324
-- Total number of aspect buckets: 3
+- Total number of aspect buckets: 2
 - Resolution: 1.0 megapixels
 - Cropped: True
 - Crop style: random
@@ -220,7 +220,7 @@ You may reuse the base model text encoder for inference.
 - Crop aspect: random
 ### yoga
 - Repeats: 0
-- Total number of images: 3589
+- Total number of images: 3584
 - Total number of aspect buckets: 3
 - Resolution: 1.0 megapixels
 - Cropped: True

optimizer.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:03994ace2362fd3324c01b26fa52bb2abd884114f00c88f8c07b934c76c965d7
+oid sha256:f20139f218525e7cbad6a58d990f3eb502f36b1a79054be50155bb747a9ec6fa
 size 15406336826

random_states_0.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9dc000e1a551d14b5cf09f008f740fa8709342e661a5ee79c78c5c4788530ba
+oid sha256:144c1fb57a49c438bb6e2fc69097c45c82e024a340bce368627684caf12e4769
 size 14344

scheduler.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea45ca160016e680217f942ded64e9e66d361c69096296e4841659f1ff692d54
+oid sha256:52ea9be04a7ddf81901a2d4dd9c15bebdb24a6ece94fe3f1432546bd67bf0bc6
 size 1000

training_state-bg20k-1024.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:07da6b69f0e7cda627e1d6a4cde0163f816232b27e4e269e3c6405641cbef8d8
-size 15117536
+oid sha256:b91eec8b7de2e6f1ffaa95bd4db6b33bfb12066d75434c6452e62edb54073b59
+size 15758639

training_state-photo-concept-bucket.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1a709533a4c4d4fec933bd8dc3a9976dca5f135209d30d96164939915b2dfb41
-size 15239853
+oid sha256:f3c5d798ebd280ea0ff5c33a34e90409236fed4a94af59848ec85073b6d6f062
+size 15385652

training_state.json

@@ -1 +1 @@
-{"global_step": 3900, "epoch_step": 309, "epoch": 1, "exhausted_backends": ["guys", "signs", "nijijourney", "propagandaposters", "bookcovers", "pixel-art", "normalnudes", "celebrities", "sports", "movieposters", "moviecollection", "gay"], "repeats": {"guys": 0, "signs": 0, "nijijourney": 0, "propagandaposters": 0, "bookcovers": 0, "pixel-art": 0, "normalnudes": 0, "celebrities": 0, "sports": 0, "movieposters": 0, "moviecollection": 0, "gay": 0}}
+{"global_step": 5100, "epoch_step": 166, "epoch": 1, "exhausted_backends": ["guys", "signs", "nijijourney", "propagandaposters", "bookcovers", "pixel-art", "normalnudes", "celebrities", "sports", "movieposters", "moviecollection", "gay", "ethnic", "experimental", "yoga"], "repeats": {"guys": 0, "signs": 0, "nijijourney": 0, "propagandaposters": 0, "bookcovers": 0, "pixel-art": 0, "normalnudes": 0, "celebrities": 0, "sports": 0, "movieposters": 0, "moviecollection": 0, "gay": 0, "ethnic": 0, "experimental": 0, "yoga": 0}}

unet/config.json

@@ -1,7 +1,7 @@
 {
   "_class_name": "UNet2DConditionModel",
   "_diffusers_version": "0.27.2",
-  "_name_or_path": "/notebooks/datasets/models/checkpoint-3700",
+  "_name_or_path": "/notebooks/datasets/models/checkpoint-5000",
   "act_fn": "silu",
   "addition_embed_type": "text_time",
   "addition_embed_type_num_heads": 64,

unet/diffusion_pytorch_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c1605d28ff0e21d71109453e897cca43c1d85be82b850de7895c7bae0dd7b3f
+oid sha256:828dcb130273b61f3da58aecae178a60aefb5ac7a1f5e7f2e45cff2fe9a2fb3b
 size 5135151440