Update

4 files changed, 12 insertions, 9 deletions

diff --git a/data/csv.py b/data/csv.py
index 85b98f8..b4c81d7 100644
--- a/data/csv.py
+++ b/data/csv.py
@@ -282,7 +282,7 @@ class VlpnDataModule():
         collate_fn_ = partial(collate_fn, self.dtype, self.tokenizer, self.num_class_images != 0)
 
         if valid_set_size == 0:
-            data_train, data_val = items, items[:1]
+            data_train, data_val = items, items[:self.batch_size]
         else:
             data_train, data_val = random_split(items, [train_set_size, valid_set_size], generator=generator)
 
diff --git a/models/clip/embeddings.py b/models/clip/embeddings.py
index 9a23a2a..1cc59d9 100644
--- a/models/clip/embeddings.py
+++ b/models/clip/embeddings.py
@@ -100,6 +100,9 @@ class ManagedCLIPTextEmbeddings(CLIPTextEmbeddings):
         return embeds
 
     def normalize(self, target: float = 0.4, lambda_: float = 1.0):
+        if lambda_ == 0:
+            return
+
         w = self.temp_token_embedding.weight
         pre_norm = w[self.temp_token_ids, :].norm(dim=-1, keepdim=True)
         w[self.temp_token_ids] = F.normalize(
diff --git a/training/strategy/dreambooth.py b/training/strategy/dreambooth.py
index bc26ee6..d813b49 100644
--- a/training/strategy/dreambooth.py
+++ b/training/strategy/dreambooth.py
@@ -88,7 +88,7 @@ def dreambooth_strategy_callbacks(
         ema_unet = None
 
     def ema_context():
-        if use_ema:
+        if ema_unet is not None:
             return ema_unet.apply_temporary(unet.parameters())
         else:
             return nullcontext()
@@ -102,7 +102,7 @@ def dreambooth_strategy_callbacks(
         text_encoder.text_model.embeddings.persist()
         text_encoder.text_model.embeddings.temp_token_embedding.requires_grad_(False)
 
-        if use_ema:
+        if ema_unet is not None:
             ema_unet.to(accelerator.device)
 
     @contextmanager
@@ -134,11 +134,11 @@ def dreambooth_strategy_callbacks(
 
     @torch.no_grad()
     def on_after_optimize(lr: float):
-        if use_ema:
+        if ema_unet is not None:
             ema_unet.step(unet.parameters())
 
     def on_log():
-        if use_ema:
+        if ema_unet is not None:
             return {"ema_decay": ema_unet.decay}
         return {}
 
diff --git a/training/strategy/ti.py b/training/strategy/ti.py
index 597abd0..081180f 100644
--- a/training/strategy/ti.py
+++ b/training/strategy/ti.py
@@ -88,7 +88,7 @@ def textual_inversion_strategy_callbacks(
         ema_embeddings = None
 
     def ema_context():
-        if use_ema:
+        if ema_embeddings is not None:
             return ema_embeddings.apply_temporary(
                 text_encoder.text_model.embeddings.temp_token_embedding.parameters()
             )
@@ -101,7 +101,7 @@ def textual_inversion_strategy_callbacks(
     def on_prepare():
         text_encoder.text_model.embeddings.temp_token_embedding.requires_grad_(True)
 
-        if use_ema:
+        if ema_embeddings is not None:
             ema_embeddings.to(accelerator.device)
 
         if gradient_checkpointing:
@@ -120,7 +120,7 @@ def textual_inversion_strategy_callbacks(
             yield
 
     def on_after_optimize(lr: float):
-        if use_ema:
+        if ema_embeddings is not None:
             ema_embeddings.step(text_encoder.text_model.embeddings.temp_token_embedding.parameters())
 
     @torch.no_grad()
@@ -132,7 +132,7 @@ def textual_inversion_strategy_callbacks(
             )
 
     def on_log():
-        if use_ema:
+        if ema_embeddings is not None:
             return {"ema_decay": ema_embeddings.decay}
         return {}