Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,83 +1,126 @@
|
|
1 |
-
# app.py
|
2 |
import gradio as gr
|
3 |
import torch
|
4 |
from transformers import pipeline, AutoTokenizer, T5ForConditionalGeneration
|
5 |
from diffusers import StableDiffusionPipeline
|
6 |
import speech_recognition as sr
|
7 |
-
|
|
|
8 |
|
9 |
-
#
|
10 |
-
|
11 |
-
|
|
|
|
|
12 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
def enhance_prompt(raw_input, style_choice):
|
14 |
template = f"Generate a detailed Stable Diffusion prompt about: {raw_input} in {style_choice} style."
|
15 |
inputs = prompt_tokenizer(template, return_tensors="pt")
|
16 |
outputs = prompt_model.generate(inputs.input_ids, max_length=100)
|
17 |
return prompt_tokenizer.decode(outputs[0], skip_special_tokens=True)
|
18 |
|
19 |
-
# ========== Step 2: Image Generation ==========
|
20 |
-
sd_pipe = StableDiffusionPipeline.from_pretrained(
|
21 |
-
"runwayml/stable-diffusion-v1-5",
|
22 |
-
torch_dtype=torch.float32,
|
23 |
-
use_safetensors=True
|
24 |
-
)
|
25 |
-
sd_pipe.enable_attention_slicing() # 降低内存消耗
|
26 |
-
|
27 |
def generate_image(enhanced_prompt, steps=20, guidance=7.5):
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
|
35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
recognizer = sr.Recognizer()
|
37 |
|
38 |
def audio_to_text(audio_file):
|
39 |
-
|
40 |
-
|
41 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
|
43 |
-
#
|
44 |
-
with gr.Blocks(title="AI Art Studio") as app:
|
45 |
-
gr.Markdown("## 🎨 AI
|
46 |
|
47 |
with gr.Row():
|
48 |
with gr.Column(scale=2):
|
49 |
-
#
|
50 |
-
input_type = gr.Radio(["
|
51 |
-
voice_input = gr.Audio(
|
52 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
|
|
|
54 |
style_choice = gr.Dropdown(
|
55 |
-
["
|
56 |
-
value="
|
57 |
label="艺术风格"
|
58 |
)
|
59 |
|
|
|
60 |
generate_btn = gr.Button("生成作品", variant="primary")
|
61 |
|
|
|
62 |
with gr.Accordion("高级设置", open=False):
|
63 |
steps_slider = gr.Slider(10, 30, value=20, step=1, label="生成步数")
|
64 |
guidance_slider = gr.Slider(5.0, 10.0, value=7.5, label="创意自由度")
|
65 |
-
|
66 |
with gr.Column(scale=3):
|
67 |
-
#
|
68 |
prompt_output = gr.Textbox(label="优化后的Prompt", interactive=False)
|
69 |
-
image_output = gr.Image(label="生成结果", show_label=False)
|
70 |
|
71 |
-
#
|
72 |
input_type.change(
|
73 |
-
fn=lambda x:
|
74 |
inputs=input_type,
|
75 |
-
outputs=
|
76 |
)
|
77 |
|
78 |
generate_btn.click(
|
79 |
-
fn=
|
80 |
-
inputs=
|
81 |
outputs=text_input
|
82 |
).success(
|
83 |
fn=enhance_prompt,
|
@@ -89,6 +132,5 @@ with gr.Blocks(title="AI Art Studio") as app:
|
|
89 |
outputs=image_output
|
90 |
)
|
91 |
|
92 |
-
# ========== Step 4: Huggingface Deployment ==========
|
93 |
if __name__ == "__main__":
|
94 |
app.launch(server_name="0.0.0.0", server_port=7860)
|
|
|
|
|
1 |
import gradio as gr
|
2 |
import torch
|
3 |
from transformers import pipeline, AutoTokenizer, T5ForConditionalGeneration
|
4 |
from diffusers import StableDiffusionPipeline
|
5 |
import speech_recognition as sr
|
6 |
+
import gc
|
7 |
+
from accelerate import init_empty_weights
|
8 |
|
9 |
+
# ===== 模型初始化 =====
|
10 |
+
def load_models():
|
11 |
+
# Prompt增强模型
|
12 |
+
prompt_model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-small")
|
13 |
+
prompt_tokenizer = AutoTokenizer.from_pretrained("google/flan-t5-small")
|
14 |
|
15 |
+
# Stable Diffusion管道
|
16 |
+
sd_pipe = StableDiffusionPipeline.from_pretrained(
|
17 |
+
"runwayml/stable-diffusion-v1-5",
|
18 |
+
torch_dtype=torch.float32,
|
19 |
+
use_safetensors=True,
|
20 |
+
variant="fp16",
|
21 |
+
device_map="auto",
|
22 |
+
offload_state_dict=True
|
23 |
+
)
|
24 |
+
sd_pipe.enable_attention_slicing()
|
25 |
+
sd_pipe.enable_sequential_cpu_offload()
|
26 |
+
|
27 |
+
return prompt_model, prompt_tokenizer, sd_pipe
|
28 |
+
|
29 |
+
prompt_model, prompt_tokenizer, sd_pipe = load_models()
|
30 |
+
|
31 |
+
# ===== 核心功能 =====
|
32 |
def enhance_prompt(raw_input, style_choice):
|
33 |
template = f"Generate a detailed Stable Diffusion prompt about: {raw_input} in {style_choice} style."
|
34 |
inputs = prompt_tokenizer(template, return_tensors="pt")
|
35 |
outputs = prompt_model.generate(inputs.input_ids, max_length=100)
|
36 |
return prompt_tokenizer.decode(outputs[0], skip_special_tokens=True)
|
37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
def generate_image(enhanced_prompt, steps=20, guidance=7.5):
|
39 |
+
try:
|
40 |
+
image = sd_pipe(
|
41 |
+
enhanced_prompt,
|
42 |
+
num_inference_steps=int(steps),
|
43 |
+
guidance_scale=guidance,
|
44 |
+
generator=torch.Generator().manual_seed(42)
|
45 |
+
).images[0]
|
46 |
+
finally:
|
47 |
+
# 清理内存
|
48 |
+
gc.collect()
|
49 |
+
with init_empty_weights():
|
50 |
+
reload_models()
|
51 |
+
return image
|
52 |
|
53 |
+
def reload_models():
|
54 |
+
global sd_pipe
|
55 |
+
del sd_pipe
|
56 |
+
sd_pipe = StableDiffusionPipeline.from_pretrained(
|
57 |
+
"runwayml/stable-diffusion-v1-5",
|
58 |
+
torch_dtype=torch.float32,
|
59 |
+
device_map="auto",
|
60 |
+
offload_folder="offload"
|
61 |
+
)
|
62 |
+
|
63 |
+
# ===== 语音处理 =====
|
64 |
recognizer = sr.Recognizer()
|
65 |
|
66 |
def audio_to_text(audio_file):
|
67 |
+
if not audio_file:
|
68 |
+
return ""
|
69 |
+
try:
|
70 |
+
with sr.AudioFile(audio_file) as source:
|
71 |
+
audio = recognizer.record(source)
|
72 |
+
return recognizer.recognize_whisper(audio, model="tiny.en")
|
73 |
+
except Exception as e:
|
74 |
+
print(f"语音识别错误: {e}")
|
75 |
+
return ""
|
76 |
|
77 |
+
# ===== Gradio界面 =====
|
78 |
+
with gr.Blocks(title="AI Art Studio", css=".gradio-container {max-width: 800px !important}") as app:
|
79 |
+
gr.Markdown("## 🎨 AI 艺术生成器 (CPU优化版)")
|
80 |
|
81 |
with gr.Row():
|
82 |
with gr.Column(scale=2):
|
83 |
+
# 输入控件
|
84 |
+
input_type = gr.Radio(["文字", "语音"], label="输入方式", value="文字")
|
85 |
+
voice_input = gr.Audio(
|
86 |
+
sources=["upload"],
|
87 |
+
type="filepath",
|
88 |
+
visible=False,
|
89 |
+
label="上传语音文件",
|
90 |
+
elem_classes="voice-input"
|
91 |
+
)
|
92 |
+
text_input = gr.Textbox(label="输入描述", placeholder="例:空中的魔法树屋...", lines=3)
|
93 |
|
94 |
+
# 风格选择
|
95 |
style_choice = gr.Dropdown(
|
96 |
+
["数字艺术", "油画", "动漫", "照片写实"],
|
97 |
+
value="数字艺术",
|
98 |
label="艺术风格"
|
99 |
)
|
100 |
|
101 |
+
# 生成按钮
|
102 |
generate_btn = gr.Button("生成作品", variant="primary")
|
103 |
|
104 |
+
# 高级设置
|
105 |
with gr.Accordion("高级设置", open=False):
|
106 |
steps_slider = gr.Slider(10, 30, value=20, step=1, label="生成步数")
|
107 |
guidance_slider = gr.Slider(5.0, 10.0, value=7.5, label="创意自由度")
|
108 |
+
|
109 |
with gr.Column(scale=3):
|
110 |
+
# 输出展示
|
111 |
prompt_output = gr.Textbox(label="优化后的Prompt", interactive=False)
|
112 |
+
image_output = gr.Image(label="生成结果", show_label=False, elem_id="output-image")
|
113 |
|
114 |
+
# 交互逻辑
|
115 |
input_type.change(
|
116 |
+
fn=lambda x: gr.update(visible=x == "语音"),
|
117 |
inputs=input_type,
|
118 |
+
outputs=voice_input
|
119 |
)
|
120 |
|
121 |
generate_btn.click(
|
122 |
+
fn=audio_to_text,
|
123 |
+
inputs=voice_input,
|
124 |
outputs=text_input
|
125 |
).success(
|
126 |
fn=enhance_prompt,
|
|
|
132 |
outputs=image_output
|
133 |
)
|
134 |
|
|
|
135 |
if __name__ == "__main__":
|
136 |
app.launch(server_name="0.0.0.0", server_port=7860)
|