Spaces:
Running
Running
Commit
·
d6f9ffa
1
Parent(s):
339f63e
fresh pipeline for each concept
Browse files
app.py
CHANGED
@@ -5,6 +5,7 @@ from diffusers import StableDiffusionPipeline
|
|
5 |
import random
|
6 |
from huggingface_hub import hf_hub_download
|
7 |
import os
|
|
|
8 |
|
9 |
# Initialize the model
|
10 |
model_id = "CompVis/stable-diffusion-v1-4"
|
@@ -52,18 +53,14 @@ def generate_images(prompt):
|
|
52 |
images = []
|
53 |
failed_concepts = []
|
54 |
|
55 |
-
# Load base pipeline
|
56 |
-
pipe = StableDiffusionPipeline.from_pretrained(
|
57 |
-
model_id,
|
58 |
-
torch_dtype=torch.float16 if device == "cuda" else torch.float32
|
59 |
-
).to(device)
|
60 |
-
|
61 |
-
# Store original tokenizer and text encoder state
|
62 |
-
original_vocab_size = len(pipe.tokenizer)
|
63 |
-
original_embedding_weights = pipe.text_encoder.get_input_embeddings().weight.data.clone()
|
64 |
-
|
65 |
for concept in concepts:
|
66 |
try:
|
|
|
|
|
|
|
|
|
|
|
|
|
67 |
# Download and load concept embedding
|
68 |
embed_path = download_concept_embedding(concept)
|
69 |
if embed_path is None:
|
@@ -94,10 +91,9 @@ def generate_images(prompt):
|
|
94 |
|
95 |
images.append(image)
|
96 |
|
97 |
-
#
|
98 |
-
|
99 |
-
|
100 |
-
pipe.text_encoder.get_input_embeddings().weight.data.copy_(original_embedding_weights)
|
101 |
|
102 |
except Exception as e:
|
103 |
print(f"Error processing concept {concept}: {str(e)}")
|
|
|
5 |
import random
|
6 |
from huggingface_hub import hf_hub_download
|
7 |
import os
|
8 |
+
from transformers import CLIPTokenizer
|
9 |
|
10 |
# Initialize the model
|
11 |
model_id = "CompVis/stable-diffusion-v1-4"
|
|
|
53 |
images = []
|
54 |
failed_concepts = []
|
55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
for concept in concepts:
|
57 |
try:
|
58 |
+
# Create a fresh pipeline for each concept
|
59 |
+
pipe = StableDiffusionPipeline.from_pretrained(
|
60 |
+
model_id,
|
61 |
+
torch_dtype=torch.float16 if device == "cuda" else torch.float32
|
62 |
+
).to(device)
|
63 |
+
|
64 |
# Download and load concept embedding
|
65 |
embed_path = download_concept_embedding(concept)
|
66 |
if embed_path is None:
|
|
|
91 |
|
92 |
images.append(image)
|
93 |
|
94 |
+
# Clean up to free memory
|
95 |
+
del pipe
|
96 |
+
torch.cuda.empty_cache()
|
|
|
97 |
|
98 |
except Exception as e:
|
99 |
print(f"Error processing concept {concept}: {str(e)}")
|