Spaces:
Sleeping
Sleeping
macrdel
commited on
Commit
·
8fe6de7
1
Parent(s):
91ba022
update app.py
Browse files
app.py
CHANGED
@@ -2,7 +2,7 @@ import gradio as gr
|
|
2 |
import numpy as np
|
3 |
import random
|
4 |
|
5 |
-
from diffusers import DiffusionPipeline
|
6 |
from peft import PeftModel, PeftConfig
|
7 |
import torch
|
8 |
|
@@ -25,7 +25,7 @@ else:
|
|
25 |
# Cache to avoid re-initializing pipelines repeatedly
|
26 |
model_cache = {}
|
27 |
|
28 |
-
def load_pipeline(model_id: str):
|
29 |
"""
|
30 |
Loads or retrieves a cached DiffusionPipeline.
|
31 |
|
@@ -38,7 +38,7 @@ def load_pipeline(model_id: str):
|
|
38 |
if model_id == "macrdel/unico_proj":
|
39 |
# Use the specified base model for your LoRA adapter.
|
40 |
base_model = "CompVis/stable-diffusion-v1-4"
|
41 |
-
pipe =
|
42 |
# Load the LoRA weights
|
43 |
pipe.unet = PeftModel.from_pretrained(
|
44 |
pipe.unet,
|
@@ -53,7 +53,9 @@ def load_pipeline(model_id: str):
|
|
53 |
torch_dtype=torch_dtype
|
54 |
)
|
55 |
else:
|
56 |
-
pipe =
|
|
|
|
|
57 |
|
58 |
pipe.to(device)
|
59 |
model_cache[model_id] = pipe
|
@@ -76,7 +78,7 @@ def infer(
|
|
76 |
progress=gr.Progress(track_tqdm=True),
|
77 |
):
|
78 |
# Load the pipeline for the chosen model
|
79 |
-
pipe = load_pipeline(model_id)
|
80 |
|
81 |
if randomize_seed:
|
82 |
seed = random.randint(0, MAX_SEED)
|
@@ -84,12 +86,12 @@ def infer(
|
|
84 |
generator = torch.Generator(device=device).manual_seed(seed)
|
85 |
|
86 |
# If using the LoRA model, update the LoRA scale if supported.
|
87 |
-
if model_id == "macrdel/unico_proj":
|
88 |
# This assumes your pipeline's unet has a method to update the LoRA scale.
|
89 |
-
if hasattr(pipe.unet, "set_lora_scale"):
|
90 |
-
|
91 |
-
else:
|
92 |
-
|
93 |
|
94 |
image = pipe(
|
95 |
prompt=prompt,
|
|
|
2 |
import numpy as np
|
3 |
import random
|
4 |
|
5 |
+
from diffusers import StableDiffusionPipeline # DiffusionPipeline
|
6 |
from peft import PeftModel, PeftConfig
|
7 |
import torch
|
8 |
|
|
|
25 |
# Cache to avoid re-initializing pipelines repeatedly
|
26 |
model_cache = {}
|
27 |
|
28 |
+
def load_pipeline(model_id: str, lora_scale):
|
29 |
"""
|
30 |
Loads or retrieves a cached DiffusionPipeline.
|
31 |
|
|
|
38 |
if model_id == "macrdel/unico_proj":
|
39 |
# Use the specified base model for your LoRA adapter.
|
40 |
base_model = "CompVis/stable-diffusion-v1-4"
|
41 |
+
pipe = StableDiffusionPipeline.from_pretrained(base_model, torch_dtype=torch_dtype)
|
42 |
# Load the LoRA weights
|
43 |
pipe.unet = PeftModel.from_pretrained(
|
44 |
pipe.unet,
|
|
|
53 |
torch_dtype=torch_dtype
|
54 |
)
|
55 |
else:
|
56 |
+
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch_dtype, safety_checker=None).to(device)
|
57 |
+
pipe.unet.load_state_dict({k: lora_scale * v for k, v in pipe.unet.state_dict().items()})
|
58 |
+
pipe.text_encoder.load_state_dict({k: lora_scale * v for k, v in pipe.text_encoder.state_dict().items()})
|
59 |
|
60 |
pipe.to(device)
|
61 |
model_cache[model_id] = pipe
|
|
|
78 |
progress=gr.Progress(track_tqdm=True),
|
79 |
):
|
80 |
# Load the pipeline for the chosen model
|
81 |
+
pipe = load_pipeline(model_id, lora_scale)
|
82 |
|
83 |
if randomize_seed:
|
84 |
seed = random.randint(0, MAX_SEED)
|
|
|
86 |
generator = torch.Generator(device=device).manual_seed(seed)
|
87 |
|
88 |
# If using the LoRA model, update the LoRA scale if supported.
|
89 |
+
# if model_id == "macrdel/unico_proj":
|
90 |
# This assumes your pipeline's unet has a method to update the LoRA scale.
|
91 |
+
# if hasattr(pipe.unet, "set_lora_scale"):
|
92 |
+
# pipe.unet.set_lora_scale(lora_scale)
|
93 |
+
# else:
|
94 |
+
# print("Warning: LoRA scale adjustment method not found on UNet.")
|
95 |
|
96 |
image = pipe(
|
97 |
prompt=prompt,
|