Adapted textual inversion distillation for quantization example to la…

…test transformers and diffusers packages (#1471)
intel · Apr 15, 2024 · 0ec83b1 · 0ec83b1
1 parent 733697c
commit 0ec83b1
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 23 deletions.
diff --git a/...ples/huggingface/pytorch/textual-inversion/distillation_for_quantization/requirements.txt b/...ples/huggingface/pytorch/textual-inversion/distillation_for_quantization/requirements.txt
@@ -1,6 +1,6 @@
 datasets
-diffusers==0.4.1
-accelerate<=0.19.0
+diffusers
+accelerate
 torchvision
 ftfy
 tensorboard

diff --git a/examples/huggingface/pytorch/textual-inversion/distillation_for_quantization/text2images.py b/examples/huggingface/pytorch/textual-inversion/distillation_for_quantization/text2images.py
@@ -104,10 +104,9 @@ def generate_images(
     unet=unet,
     tokenizer=tokenizer,
     scheduler=PNDMScheduler.from_config("CompVis/stable-diffusion-v1-4", subfolder="scheduler"),
-    safety_checker=StableDiffusionSafetyChecker.from_pretrained("CompVis/stable-diffusion-safety-checker"),
+    safety_checker=None,
     feature_extractor=CLIPFeatureExtractor.from_pretrained("openai/clip-vit-base-patch32"),
 )
-pipeline.safety_checker = lambda images, clip_input: (images, False)
 int8_model_path = os.path.join(args.pretrained_model_name_or_path, "pytorch_model.bin")
 if os.path.exists(int8_model_path):
     unet = load(int8_model_path, model=unet)
@@ -120,7 +119,7 @@ def generate_images(
 grid.save(
     os.path.join(args.pretrained_model_name_or_path, "{}.png".format("_".join(args.caption.split())))
 )
-dirname = os.path.join(args.pretrained_model_name_or_path, "_".join(args.caption.split()))
-os.makedirs(shlex.quote(dirname), exist_ok=True)
+dirname = shlex.quote(os.path.join(args.pretrained_model_name_or_path, "_".join(args.caption.split())))
+os.makedirs(dirname, exist_ok=True)
 for idx, image in enumerate(images):
     image.save(os.path.join(dirname, "{}.png".format(idx+1)))
diff --git a/.../huggingface/pytorch/textual-inversion/distillation_for_quantization/textual_inversion.py b/.../huggingface/pytorch/textual-inversion/distillation_for_quantization/textual_inversion.py
@@ -476,7 +476,7 @@ def main():
         gradient_accumulation_steps=args.gradient_accumulation_steps,
         mixed_precision=args.mixed_precision,
         log_with="tensorboard",
-        logging_dir=logging_dir,
+        project_dir=logging_dir,
     )
 
     # If passed along, set the training seed now.
@@ -782,18 +782,18 @@ def train_func(model):
             layer_mappings=[
                 [['conv_in', ]],
                 [['time_embedding', ]],
-                [['down_blocks.0.attentions.0',]],
-                [['down_blocks.0.attentions.1',]],
+                [['down_blocks.0.attentions.0', '0']],
+                [['down_blocks.0.attentions.1', '0']],
                 [['down_blocks.0.resnets.0',]],
                 [['down_blocks.0.resnets.1',]],
                 [['down_blocks.0.downsamplers.0',]],
-                [['down_blocks.1.attentions.0',]],
-                [['down_blocks.1.attentions.1',]],
+                [['down_blocks.1.attentions.0', '0']],
+                [['down_blocks.1.attentions.1', '0']],
                 [['down_blocks.1.resnets.0',]],
                 [['down_blocks.1.resnets.1',]],
                 [['down_blocks.1.downsamplers.0',]],
-                [['down_blocks.2.attentions.0',]],
-                [['down_blocks.2.attentions.1',]],
+                [['down_blocks.2.attentions.0', '0']],
+                [['down_blocks.2.attentions.1', '0']],
                 [['down_blocks.2.resnets.0',]],
                 [['down_blocks.2.resnets.1',]],
                 [['down_blocks.2.downsamplers.0',]],
@@ -803,27 +803,27 @@ def train_func(model):
                 [['up_blocks.0.resnets.1', ]],
                 [['up_blocks.0.resnets.2', ]],
                 [['up_blocks.0.upsamplers.0', ]],
-                [['up_blocks.1.attentions.0', ]],
-                [['up_blocks.1.attentions.1', ]],
-                [['up_blocks.1.attentions.2', ]],
+                [['up_blocks.1.attentions.0', '0']],
+                [['up_blocks.1.attentions.1', '0']],
+                [['up_blocks.1.attentions.2', '0']],
                 [['up_blocks.1.resnets.0', ]],
                 [['up_blocks.1.resnets.1', ]],
                 [['up_blocks.1.resnets.2', ]],
                 [['up_blocks.1.upsamplers.0', ]],
-                [['up_blocks.2.attentions.0', ]],
-                [['up_blocks.2.attentions.1', ]],
-                [['up_blocks.2.attentions.2', ]],
+                [['up_blocks.2.attentions.0', '0']],
+                [['up_blocks.2.attentions.1', '0']],
+                [['up_blocks.2.attentions.2', '0']],
                 [['up_blocks.2.resnets.0', ]],
                 [['up_blocks.2.resnets.1', ]],
                 [['up_blocks.2.resnets.2', ]],
                 [['up_blocks.2.upsamplers.0', ]],
-                [['up_blocks.3.attentions.0', ]],
-                [['up_blocks.3.attentions.1', ]],
-                [['up_blocks.3.attentions.2', ]],
+                [['up_blocks.3.attentions.0', '0']],
+                [['up_blocks.3.attentions.1', '0']],
+                [['up_blocks.3.attentions.2', '0']],
                 [['up_blocks.3.resnets.0', ]],
                 [['up_blocks.3.resnets.1', ]],
                 [['up_blocks.3.resnets.2', ]],
-                [['mid_block.attentions.0', ]],
+                [['mid_block.attentions.0', '0']],
                 [['mid_block.resnets.0', ]],
                 [['mid_block.resnets.1', ]],
                 [['conv_out', ]],