PONY training: added Head to warmup, increase max steps count
Browse files- MANIFEST.knf.xlsx +2 -2
- configs/PONY.knf.yaml +7 -7
MANIFEST.knf.xlsx
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bed6ff45a1234238cc06e6055b7cd9d696c88ce44712abd6c747fd2c6a2ff7ce
|
| 3 |
+
size 14600753
|
configs/PONY.knf.yaml
CHANGED
|
@@ -206,10 +206,10 @@ training:
|
|
| 206 |
species_token: felkin
|
| 207 |
negative_control_species: wolf
|
| 208 |
custom:
|
| 209 |
-
pony_safe_portrait: "rating_safe, Rook_Kaefer, felkin, close-up portrait, detailed lighting"
|
| 210 |
-
pony_safe_fullbody_reference: "rating_safe, Rook_Kaefer, felkin, full-body character reference, standing, solo"
|
| 211 |
-
pony_safe_fur_detail: "rating_safe, Rook_Kaefer, felkin, fur texture reference, ladybug_fur_markings, mane, tail, paws"
|
| 212 |
-
pony_explicit_anatomy_reference: "rating_explicit, Rook_Kaefer, felkin, nude anatomy reference, sheath detail, clean lighting"
|
| 213 |
|
| 214 |
# Model and LoRA settings merged into the top-level SimpleTuner config.
|
| 215 |
model:
|
|
@@ -241,8 +241,8 @@ training:
|
|
| 241 |
# == Curricular warmup
|
| 242 |
subset_warmup:
|
| 243 |
enabled: false
|
| 244 |
-
subsets: [fullbody]
|
| 245 |
-
steps:
|
| 246 |
|
| 247 |
# Full-run trainer config. trainer_testrun below overrides selected keys
|
| 248 |
# only when knf train prepare/start is called with --testrun.
|
|
@@ -263,7 +263,7 @@ training:
|
|
| 263 |
train_batch_size: 4 # Per-device batch size.
|
| 264 |
gradient_accumulation_steps: 1 # Effective batch multiplier.
|
| 265 |
num_train_epochs: 0 # Step-based training; max_train_steps controls run length.
|
| 266 |
-
max_train_steps:
|
| 267 |
|
| 268 |
caption_dropout_probability: 0.0 # Caption dropout for regularization.
|
| 269 |
data_backend_sampling: uniform # Uniform dataloader sampling.
|
|
|
|
| 206 |
species_token: felkin
|
| 207 |
negative_control_species: wolf
|
| 208 |
custom:
|
| 209 |
+
pony_safe_portrait: "rating_safe, Rook_Kaefer, felkin, fur, red fur, black spots, spotted fur, fur markings, musclegut, close-up portrait, detailed lighting"
|
| 210 |
+
pony_safe_fullbody_reference: "rating_safe, Rook_Kaefer, felkin, fur, red fur, black spots, spotted fur, fur markings, musclegut, full-body character reference, standing, solo"
|
| 211 |
+
pony_safe_fur_detail: "rating_safe, Rook_Kaefer, felkin, fur, red fur, black spots, spotted fur, fur markings, musclegut, fur texture reference, ladybug_fur_markings, mane, tail, paws"
|
| 212 |
+
pony_explicit_anatomy_reference: "rating_explicit, Rook_Kaefer, felkin, fur, red fur, black spots, spotted fur, fur markings, musclegut, nude anatomy reference, sheath detail, clean lighting"
|
| 213 |
|
| 214 |
# Model and LoRA settings merged into the top-level SimpleTuner config.
|
| 215 |
model:
|
|
|
|
| 241 |
# == Curricular warmup
|
| 242 |
subset_warmup:
|
| 243 |
enabled: false
|
| 244 |
+
subsets: [fullbody, head]
|
| 245 |
+
steps: 800 # !! First steps of the total run are ONLY listed subsets when enabled!
|
| 246 |
|
| 247 |
# Full-run trainer config. trainer_testrun below overrides selected keys
|
| 248 |
# only when knf train prepare/start is called with --testrun.
|
|
|
|
| 263 |
train_batch_size: 4 # Per-device batch size.
|
| 264 |
gradient_accumulation_steps: 1 # Effective batch multiplier.
|
| 265 |
num_train_epochs: 0 # Step-based training; max_train_steps controls run length.
|
| 266 |
+
max_train_steps: 2500 # Current short Pony run length.
|
| 267 |
|
| 268 |
caption_dropout_probability: 0.0 # Caption dropout for regularization.
|
| 269 |
data_backend_sampling: uniform # Uniform dataloader sampling.
|