Hugging Face's logo

bghira
/
sd3-reality-mix

Text-to-Image
Diffusers
Safetensors
StableDiffusion3Pipeline
stable-diffusion
stable-diffusion-diffusers
full
Model card Files Community
2
PseudoTerminal X commited on
Commit
bf7b676
•
1 Parent(s): 531404b

Trained for 0 epochs and 6500 steps.

Browse files

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

Files changed (9) hide show
  1. README.md +3 -3
  2. optimizer.bin +1 -1
  3. random_states_0.pkl +2 -2
  4. scheduler.bin +1 -1
  5. training_state-dalle3.json +0 -0
  6. training_state-photo-concept-bucket.json +0 -0
  7. training_state.json +1 -1
  8. transformer/config.json +1 -1
  9. transformer/diffusion_pytorch_model.safetensors +1 -1
README.md CHANGED
@@ -1164,7 +1164,7 @@ You may reuse the base model text encoder for inference.
1164
  ## Training settings
1165
 
1166
  - Training epochs: 0
1167
- - Training steps: 6000
1168
  - Learning rate: 1e-06
1169
  - Effective batch size: 96
1170
  - Micro-batch size: 6
@@ -1182,7 +1182,7 @@ You may reuse the base model text encoder for inference.
1182
  ### photo-concept-bucket
1183
  - Repeats: 0
1184
  - Total number of images: ~557568
1185
- - Total number of aspect buckets: 2
1186
  - Resolution: 1.0 megapixels
1187
  - Cropped: False
1188
  - Crop style: None
@@ -1190,7 +1190,7 @@ You may reuse the base model text encoder for inference.
1190
  ### dalle3
1191
  - Repeats: 0
1192
  - Total number of images: ~984960
1193
- - Total number of aspect buckets: 1
1194
  - Resolution: 1.0 megapixels
1195
  - Cropped: False
1196
  - Crop style: None
 
1164
  ## Training settings
1165
 
1166
  - Training epochs: 0
1167
+ - Training steps: 6500
1168
  - Learning rate: 1e-06
1169
  - Effective batch size: 96
1170
  - Micro-batch size: 6
 
1182
  ### photo-concept-bucket
1183
  - Repeats: 0
1184
  - Total number of images: ~557568
1185
+ - Total number of aspect buckets: 38
1186
  - Resolution: 1.0 megapixels
1187
  - Cropped: False
1188
  - Crop style: None
 
1190
  ### dalle3
1191
  - Repeats: 0
1192
  - Total number of images: ~984960
1193
+ - Total number of aspect buckets: 44
1194
  - Resolution: 1.0 megapixels
1195
  - Cropped: False
1196
  - Crop style: None
optimizer.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:06e7b3bd3c2e41169a6a32ca435e94825f5121ec67eb3e792f1d1786810b9bf6
3
  size 12170595712
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9aab67951716f33adfd694940e180f747a83ba801a23e2acb646508316b732d1
3
  size 12170595712
random_states_0.pkl CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f49480729ff3ab718b4a2a8d9256aaa1ce5882bd8f387a6d421b3cd262e24f71
3
- size 16036
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33e4b3c9237c0b74e86807e9ac4ab5db460928655ad2d2a314062dfee1df7386
3
+ size 16100
scheduler.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7f30c9fbfa479a9edb8fa992b83695b0ce1122eace54970a7b35dbcc7f5d4a75
3
  size 1128
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d39965bd721b9296b5f05e0397e1d0c4330f924effd6a7e7d246e17fef150c6
3
  size 1128
training_state-dalle3.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_state-photo-concept-bucket.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_state.json CHANGED
@@ -1 +1 @@
1
- {"global_step": 6000, "epoch_step": 6000, "epoch": 1, "exhausted_backends": [], "repeats": {}}
 
1
+ {"global_step": 6500, "epoch_step": 3500, "epoch": 1, "exhausted_backends": [], "repeats": {}}
transformer/config.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "_class_name": "SD3Transformer2DModel",
3
  "_diffusers_version": "0.30.0.dev0",
4
- "_name_or_path": "stabilityai/stable-diffusion-3-medium-diffusers",
5
  "attention_head_dim": 64,
6
  "caption_projection_dim": 1536,
7
  "in_channels": 16,
 
1
  {
2
  "_class_name": "SD3Transformer2DModel",
3
  "_diffusers_version": "0.30.0.dev0",
4
+ "_name_or_path": "/home/ubuntu/training/models/checkpoint-6000",
5
  "attention_head_dim": 64,
6
  "caption_projection_dim": 1536,
7
  "in_channels": 16,
transformer/diffusion_pytorch_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c3332da85b0460b065149f8212f7b56da455a1a9981f087f31ed53a48ca97dfb
3
  size 4169982088
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72c6cffd03e6981e8384b2356401d7f8e15a693e147e8d6fc1c05f818e52e3e5
3
  size 4169982088