Trained for 0 epochs and 4000 steps.

Trained with datasets ['text-embeds-sd3-nofilter', 'photo-concept-bucket', 'dalle3']
Learning rate 1e-06, batch size 6, 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: stabilityai/stable-diffusion-3-medium-diffusers
VAE: None

README.md

@@ -1164,7 +1164,7 @@ You may reuse the base model text encoder for inference.
 ## Training settings
 
 - Training epochs: 0
-- Training steps: 3500
+- Training steps: 4000
 - Learning rate: 1e-06
 - Effective batch size: 96
   - Micro-batch size: 6

optimizer.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:239c6915f592c76284a98e3afb6f99d730d8c6917ac21a546112a4baa666db16
+oid sha256:c3e5567d5bdb62ad2975a3a0d6f6cf00a7b3be70ce701ee597b190ce6409bb45
 size 12170595712

random_states_0.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7496944f61aaee062ef33c561862452947ca08f1ec46245cefaf08a9fccb6006
+oid sha256:884211de525601094c270d2871be55d1f1331171641cf4ab96fb1eda03184613
 size 16100

scheduler.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3481a9457c3f2703fda8fcd456dc1dcb50a3f6cdb6715ac04ab26f33db108ad4
+oid sha256:001fab7e7195f99a11605259cc60cec4bb90011df008f524f66eac9eb6aa8656
 size 1128

training_state.json

@@ -1 +1 @@
-{"global_step": 3500, "epoch_step": 3500, "epoch": 1, "exhausted_backends": [], "repeats": {}}
+{"global_step": 4000, "epoch_step": 4000, "epoch": 1, "exhausted_backends": [], "repeats": {}}

transformer/diffusion_pytorch_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e659981d629fbd35849bad0a7a6dcbfffe15d00abd8662896fe3b899504b75a
+oid sha256:d6bd9b713ac935e77eab60d47ed1c101b9a78fc5be4084a4c757fca113aad55d
 size 4169982088