File size: 4,807 Bytes
d10fd1a
 
 
 
 
 
 
 
 
 
 
d7ce378
 
d10fd1a
 
5d5af54
 
 
a31088e
 
 
dcc1c7f
 
 
 
 
 
 
 
 
 
d10fd1a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
dd361f0
a31088e
90b606d
a31088e
 
d10fd1a
c3233df
cfb0720
 
 
 
c3233df
d10fd1a
90b606d
 
dd361f0
 
 
 
d10fd1a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d7ce378
d10fd1a
 
 
 
 
 
 
d7ce378
d10fd1a
 
 
 
 
 
 
 
d7ce378
d10fd1a
 
 
 
 
 
 
d7ce378
d10fd1a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cfb0720
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
import gradio as gr
import numpy as np
import random

# import spaces #[uncomment to use ZeroGPU]
from diffusers import DiffusionPipeline
import torch

device = "cuda" if torch.cuda.is_available() else "cpu"
model_repo_id = "stabilityai/sdxl-turbo"  # Replace to the model you would like to use

# Force model to use float32 to avoid dtype mismatch
torch_dtype = torch.float32

pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)

# Explicitly convert submodules to float32 to prevent dtype mismatch
pipe.to(device)
pipe.text_encoder.to(device, dtype=torch.float32)
pipe.vae.to(device, dtype=torch.float32)
pipe.unet.to(device, dtype=torch.float32)

def force_float32(model):
    for param in model.parameters():
        param.data = param.data.to(torch.float32)
    for buffer in model.buffers():
        buffer.data = buffer.data.to(torch.float32)

force_float32(pipe.text_encoder)
force_float32(pipe.vae)
force_float32(pipe.unet)

MAX_SEED = np.iinfo(np.int32).max
MAX_IMAGE_SIZE = 1024

# @spaces.GPU #[uncomment to use ZeroGPU]
def infer(
    prompt,
    negative_prompt,
    seed,
    randomize_seed,
    width,
    height,
    guidance_scale,
    num_inference_steps,
    progress=gr.Progress(track_tqdm=True),
):
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)

    generator = torch.Generator(device).manual_seed(int(seed))

    # Ensure text inputs are strings
    prompt = str(prompt) if prompt else ""
    negative_prompt = str(negative_prompt) if negative_prompt else ""

    # Ensure text input IDs are of type LongTensor
    if isinstance(prompt, torch.Tensor):
        prompt = prompt.to(torch.long)
    if isinstance(negative_prompt, torch.Tensor):
        negative_prompt = negative_prompt.to(torch.long)

    image = pipe(
        prompt=prompt,
        negative_prompt=negative_prompt,
        guidance_scale=float(guidance_scale),
        num_inference_steps=int(num_inference_steps),
        width=int(width),
        height=int(height),
        generator=generator,
    ).images[0]

    return image, seed

examples = [
    "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
    "An astronaut riding a green horse",
    "A delicious ceviche cheesecake slice",
]

css = """
#col-container {
    margin: 0 auto;
    max-width: 640px;
}
"""

with gr.Blocks(css=css) as demo:
    with gr.Column(elem_id="col-container"):
        gr.Markdown(" # Text-to-Image Gradio Template")

        with gr.Row():
            prompt = gr.Text(
                label="Prompt",
                show_label=False,
                max_lines=1,
                placeholder="Enter your prompt",
                container=False,
            )

            run_button = gr.Button("Run", scale=0, variant="primary")

        result = gr.Image(label="Result", show_label=False)

        with gr.Accordion("Advanced Settings", open=False):
            negative_prompt = gr.Text(
                label="Negative prompt",
                max_lines=1,
                placeholder="Enter a negative prompt",
                visible=False,
            )

            seed = gr.Slider(
                label="Seed",
                minimum=0,
                maximum=MAX_SEED,
                step=1,
                value=0,
            )

            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)

            with gr.Row():
                width = gr.Slider(
                    label="Width",
                    minimum=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,
                )

                height = gr.Slider(
                    label="Height",
                    minimum=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,
                )

            with gr.Row():
                guidance_scale = gr.Slider(
                    label="Guidance scale",
                    minimum=0.0,
                    maximum=10.0,
                    step=0.1,
                    value=0.0,
                )

                num_inference_steps = gr.Slider(
                    label="Number of inference steps",
                    minimum=1,
                    maximum=50,
                    step=1,
                    value=2,
                )

        gr.Examples(examples=examples, inputs=[prompt])
    gr.on(
        triggers=[run_button.click, prompt.submit],
        fn=infer,
        inputs=[
            prompt,
            negative_prompt,
            seed,
            randomize_seed,
            width,
            height,
            guidance_scale,
            num_inference_steps,
        ],
        outputs=[result, seed],
    )

if __name__ == "__main__":
    demo.launch()