Trained for 0 epochs and 9500 steps.

Trained with datasets ['text-embeds-sdxl-nofilter', 'dalle3', 'photo-concept-bucket']
Learning rate 4e-07, batch size 16, 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-v2
VAE: madebyollin/sdxl-vae-fp16-fix

README.md

@@ -14,7 +14,7 @@ inference: true
 
 # terminus-xl-velocity-training
 
-This is a full rank finetuned model derived from [ptx0/terminus-xl-velocity-v2](https://huggingface.co/ptx0/terminus-xl-velocity-v2).
+This is a full rank finetune derived from [ptx0/terminus-xl-velocity-v2](https://huggingface.co/ptx0/terminus-xl-velocity-v2).
 
 The main validation prompt used during training was:
 
@@ -44,40 +44,33 @@ You may reuse the base model text encoder for inference.
 ## Training settings
 
 - Training epochs: 0
-- Training steps: 9000
+- Training steps: 9500
 - Learning rate: 4e-07
-- Effective batch size: 16
-  - Micro-batch size: 4
+- Effective batch size: 512
+  - Micro-batch size: 16
   - Gradient accumulation steps: 4
+  - Number of GPUs: 8
 - Prediction type: v_prediction
 - Rescaled betas zero SNR: True
 - Optimizer: AdamW, stochastic bf16
 - Precision: Pure BF16
-- Xformers: Not used
+- Xformers: Enabled
 
 
 ## Datasets
 
-### training-test
-- Repeats: 1
-- Total number of images: 192
-- Total number of aspect buckets: 17
-- Resolution: 1.0 megapixels
-- Cropped: True
-- Crop style: corner
-- Crop aspect: preserve
 ### dalle3
-- Repeats: 1
-- Total number of images: 57744
-- Total number of aspect buckets: 1
+- Repeats: 0
+- Total number of images: ~461848
+- Total number of aspect buckets: 27
 - Resolution: 1.0 megapixels
 - Cropped: False
 - Crop style: None
 - Crop aspect: None
 ### photo-concept-bucket
 - Repeats: 0
-- Total number of images: 70616
-- Total number of aspect buckets: 4
+- Total number of images: ~557568
+- Total number of aspect buckets: 5
 - Resolution: 1.0 megapixels
 - Cropped: True
 - Crop style: random

optimizer.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1c5fb4a52cbf58218b3e0ca3261904c233bf6895f9f9cfffbfeabd203026c139
+oid sha256:d1f1ddd05d4c41863dabfca0c5e90e90c525836a259cfccb431643878329641f
 size 15406336826

random_states_0.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df132266404cad7e3daf43793b9b1bbccb66243137d04410e452d5b75dcc63f7
+oid sha256:904dd262888fa3761ef496229f2619ad05b67a5fa23e3cf2cf78409455186bfe
 size 16036

scheduler.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8eb730670117893a941498760651c7bb31af112047a438fe7509a55386824f73
+oid sha256:d3abc2992b4caa0f0747a6977a22983beb24275ae52a4edb2e93fda245503a0d
 size 1000

training_state-photo-concept-bucket.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c35d2aba7a0651ff48f997b1cdfdedd4e2209063a28dfe9e04df3b60dd2975c4
-size 9128699
+oid sha256:dfad78eb826fdd11841cad1087efdc6fc1869be251a3862ae05f4f0fe88448cf
+size 4368036

training_state.json

@@ -1 +1 @@
-{"global_step": 9000, "epoch_step": 6923, "epoch": 1, "exhausted_backends": ["celebrities", "pixel-art", "movieposters", "moviecollection", "training-test"], "repeats": {"celebrities": 0, "pixel-art": 0, "movieposters": 0, "moviecollection": 0, "training-test": 0, "dalle3": 1}}
+{"global_step": 9500, "epoch_step": 2231, "epoch": 1, "exhausted_backends": ["celebrities", "pixel-art", "movieposters", "moviecollection", "training-test", "photo-concept-bucket"], "repeats": {"celebrities": 0, "pixel-art": 0, "movieposters": 0, "moviecollection": 0, "training-test": 0, "dalle3": 1, "photo-concept-bucket": 0}}

unet/config.json

@@ -1,7 +1,7 @@
 {
   "_class_name": "UNet2DConditionModel",
   "_diffusers_version": "0.29.0.dev0",
-  "_name_or_path": "/home/user/training/lite-models/checkpoint-2200",
+  "_name_or_path": "/home/user/training/lite-models/checkpoint-9000",
   "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 sha256:81e8ca245a5180d0279bc92299a27a77095827e33c9a933f8e9ed8edb230daf7
+oid sha256:080d173458538fecbf28799833e51197017f0ab2f6ee23291369ce1a8ecf4720
 size 5135151416