Fixed model compilation

2 files changed, 9 insertions, 3 deletions

diff --git a/train_lora.py b/train_lora.py
index 64346bc..74afeed 100644
--- a/train_lora.py
+++ b/train_lora.py
@@ -456,6 +456,11 @@ def parse_args():
         ),
     )
     parser.add_argument(
+        "--compile_unet",
+        action="store_true",
+        help="Compile UNet with Torch Dynamo.",
+    )
+    parser.add_argument(
         "--lora_rank",
         type=int,
         default=256,
@@ -892,6 +897,7 @@ def main():
         noise_scheduler=noise_scheduler,
         dtype=weight_dtype,
         seed=args.seed,
+        compile_unet=args.compile_unet,
         guidance_scale=args.guidance_scale,
         prior_loss_weight=args.prior_loss_weight if args.num_class_images != 0 else 0,
         sample_scheduler=sample_scheduler,
diff --git a/training/functional.py b/training/functional.py
index e7cc20f..68ea40c 100644
--- a/training/functional.py
+++ b/training/functional.py
@@ -672,6 +672,7 @@ def train(
     optimizer: torch.optim.Optimizer,
     lr_scheduler: torch.optim.lr_scheduler._LRScheduler,
     strategy: TrainingStrategy,
+    compile_unet: bool = False,
     no_val: bool = False,
     num_train_epochs: int = 100,
     gradient_accumulation_steps: int = 1,
@@ -699,9 +700,8 @@ def train(
     vae.requires_grad_(False)
     vae.eval()
 
-    unet = torch.compile(unet)
-    text_encoder = torch.compile(text_encoder)
-    vae = torch.compile(vae)
+    if compile_unet:
+        unet = torch.compile(unet, backend='hidet')
 
     callbacks = strategy.callbacks(
         accelerator=accelerator,