Spaces:
Sleeping
Sleeping
Jeong-hun Kim
commited on
Commit
Β·
d503312
1
Parent(s):
9e5793d
model test
Browse files- app/main.py +67 -153
- requirements.txt +0 -0
app/main.py
CHANGED
@@ -1,181 +1,95 @@
|
|
1 |
-
from fastapi import FastAPI
|
2 |
-
from
|
3 |
-
from typing import List
|
4 |
-
from transformers import pipeline
|
5 |
-
from PIL import Image
|
6 |
-
import re, os
|
7 |
import gradio as gr
|
8 |
import torch
|
9 |
|
10 |
app = FastAPI()
|
11 |
|
12 |
-
# 1. LLM νμ΄νλΌμΈ μ΄κΈ°ν (SmolLM3 λͺ¨λΈ)
|
13 |
print("[torch] is available:", torch.cuda.is_available())
|
14 |
print("[device] default:", torch.device("cuda" if torch.cuda.is_available() else "cpu"))
|
15 |
-
llm = pipeline("text-generation", model="HuggingFaceTB/SmolLM3-3B", device=0 if torch.cuda.is_available() else -1)
|
16 |
|
17 |
-
#
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
}
|
32 |
-
=>
|
33 |
-
λͺ¨λΈ μΆλ ₯μ κ°μ λΆλΆμ λμλλ μ΄λ―Έμ§ μΆλ ₯
|
34 |
-
'''
|
35 |
-
def load_faces(face_dir="assets/face"):
|
36 |
-
if not os.path.exists(face_dir):
|
37 |
-
os.makedirs(face_dir)
|
38 |
-
emotion_to_face = {}
|
39 |
-
for filename in os.listdir(face_dir):
|
40 |
-
if filename.endswith(".png"):
|
41 |
-
emotion = os.path.splitext(filename)[0] # 'happy.png' β 'happy'
|
42 |
-
emotion_to_face[emotion] = filename # "happy": "happy.png"
|
43 |
-
return emotion_to_face
|
44 |
-
|
45 |
-
def load_bgs(bg_dir="assets/bg"):
|
46 |
-
if not os.path.exists(bg_dir):
|
47 |
-
os.makedirs(bg_dir)
|
48 |
-
situation_to_bg = {}
|
49 |
-
for filename in os.listdir(bg_dir):
|
50 |
-
if filename.endswith(".png"):
|
51 |
-
emotion = os.path.splitext(filename)[0] # 'happy.png' β 'happy'
|
52 |
-
situation_to_bg[emotion] = filename # "happy": "happy.png"
|
53 |
-
return situation_to_bg
|
54 |
-
|
55 |
-
emotion_to_face = load_faces()
|
56 |
-
situation_to_bg = load_bgs()
|
57 |
-
|
58 |
-
# 3. μΆλ ₯ λΌμΈ νμ± ν¨μ
|
59 |
-
def parse_output(text: str):
|
60 |
-
pattern = r'"(.*?)"\s*\(emotion:\s*(\w+),\s*situation:\s*(\w+)\)'
|
61 |
-
results = []
|
62 |
-
for line in text.strip().split('\n'):
|
63 |
-
match = re.match(pattern, line.strip())
|
64 |
-
if match:
|
65 |
-
results.append({
|
66 |
-
"text": match.group(1),
|
67 |
-
"emotion": match.group(2),
|
68 |
-
"situation": match.group(3)
|
69 |
-
})
|
70 |
-
return results
|
71 |
-
|
72 |
-
# 4. μ΄λ―Έμ§ ν©μ± ν¨μ
|
73 |
-
def combine_images(bg_path, face_path):
|
74 |
-
try:
|
75 |
-
bg = Image.open(bg_path).convert("RGBA")
|
76 |
-
except FileNotFoundError:
|
77 |
-
print(f"[warning] λ°°κ²½ μ΄λ―Έμ§ μμ: {bg_path}")
|
78 |
-
return None
|
79 |
-
try:
|
80 |
-
face = Image.open(face_path).convert("RGBA")
|
81 |
-
except FileNotFoundError:
|
82 |
-
print(f"[warning] μΊλ¦ν° μ΄λ―Έμ§ μμ: {face_path}")
|
83 |
-
return None
|
84 |
-
# μ΄λ―Έμ§ ν©μ±
|
85 |
-
bg.paste(face, (0, 0), face)
|
86 |
-
return bg
|
87 |
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
97 |
)
|
|
|
|
|
98 |
|
99 |
-
|
100 |
-
for item in chat_history:
|
101 |
-
if item["role"] == "user":
|
102 |
-
dialogue += f"User: {item['text']}\n"
|
103 |
-
elif item["role"] == "bot":
|
104 |
-
dialogue += f"Aria: {item['text']}\n"
|
105 |
-
|
106 |
-
dialogue += f"User: {user_msg}\nAria:"
|
107 |
-
return system_prompt + "\n" + dialogue
|
108 |
-
|
109 |
-
def character_chat(prompt):
|
110 |
-
full_prompt = build_prompt(chat_history, prompt)
|
111 |
-
|
112 |
-
#raw_output = llm(full_prompt, max_new_tokens=300)[0]['generated_text']
|
113 |
-
raw_output = '"μ°μ€μ" (emotion: tag, situation: tag)'
|
114 |
-
parsed = parse_output(raw_output)
|
115 |
-
|
116 |
-
result_outputs = []
|
117 |
-
for i, item in enumerate(parsed):
|
118 |
-
face = emotion_to_face.get(item['emotion'], "neutral.png")
|
119 |
-
bg = situation_to_bg.get(item['situation'], "default.jpg")
|
120 |
-
composite = combine_images(os.path.join("assets/bg", bg), os.path.join("assets/face", face))
|
121 |
-
img_path = None #μ΄λ―Έμ§κ° μμΌλ©΄ μΆλ ₯ μν¨
|
122 |
-
if composite:
|
123 |
-
img_path = f"static/output_{i}.png"
|
124 |
-
composite.save(img_path)
|
125 |
-
result_outputs.append((item['text'], img_path))
|
126 |
-
|
127 |
-
return result_outputs
|
128 |
-
|
129 |
-
# 6. Gradio UI with chat history
|
130 |
-
chat_history = []
|
131 |
-
|
132 |
with gr.Blocks(css="""
|
133 |
.chat-box { max-height: 500px; overflow-y: auto; padding: 10px; border: 1px solid #ccc; border-radius: 10px; }
|
134 |
.bubble-left { background-color: #f1f0f0; border-radius: 10px; padding: 10px; margin: 5px; max-width: 70%; float: left; clear: both; }
|
135 |
.bubble-right { background-color: #d1e7ff; border-radius: 10px; padding: 10px; margin: 5px; max-width: 70%; float: right; clear: both; text-align: right; }
|
136 |
-
.image-preview { margin: 5px 0; max-width: 100%; border-radius: 10px; }
|
137 |
""") as demo:
|
138 |
-
gr.Markdown("
|
139 |
with gr.Column():
|
140 |
-
chat_output = gr.HTML(
|
141 |
-
user_input = gr.Textbox(label="
|
142 |
|
143 |
def render_chat():
|
144 |
html = ""
|
145 |
for item in chat_history:
|
146 |
-
if item[
|
147 |
html += f"<div class='bubble-right'>{item['text']}</div>"
|
148 |
-
elif item[
|
149 |
-
|
150 |
-
|
151 |
-
bubble += f"<br><img class='image-preview' src='{item['image']}'>"
|
152 |
-
bubble += "</div>"
|
153 |
-
html += bubble
|
154 |
-
return html
|
155 |
|
156 |
def on_submit(user_msg):
|
157 |
chat_history.append({"role": "user", "text": user_msg})
|
|
|
|
|
|
|
|
|
158 |
|
159 |
-
|
160 |
-
|
161 |
-
for item in bot_results:
|
162 |
-
try:
|
163 |
-
text, image_path = item # unpack μλ
|
164 |
-
except (ValueError, TypeError):
|
165 |
-
# unpack μλλ©΄ κΈ°λ³Έκ° μ²λ¦¬ (μ΄λ―Έμ§ μμ΄)
|
166 |
-
text = str(item)
|
167 |
-
image_path = None
|
168 |
-
|
169 |
-
chat_entry = {"role": "bot", "text": text}
|
170 |
-
if image_path:
|
171 |
-
chat_entry["image"] = image_path
|
172 |
-
|
173 |
-
chat_history.append(chat_entry)
|
174 |
-
|
175 |
-
new_chat_html = render_chat()
|
176 |
-
return f"<div class='chat-box' id='chat-box'>{new_chat_html}</div>", ""
|
177 |
-
|
178 |
-
user_input.submit(on_submit, inputs=user_input, outputs=[chat_output, user_input])
|
179 |
|
180 |
if __name__ == "__main__":
|
181 |
-
demo.launch()
|
|
|
1 |
+
from fastapi import FastAPI
|
2 |
+
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
|
|
|
|
|
|
|
|
|
3 |
import gradio as gr
|
4 |
import torch
|
5 |
|
6 |
app = FastAPI()
|
7 |
|
|
|
8 |
print("[torch] is available:", torch.cuda.is_available())
|
9 |
print("[device] default:", torch.device("cuda" if torch.cuda.is_available() else "cpu"))
|
|
|
10 |
|
11 |
+
# λͺ¨λΈ λ‘λ
|
12 |
+
# https://huggingface.co/EleutherAI/polyglot-ko-1.3b
|
13 |
+
model_id = "EleutherAI/polyglot-ko-1.3b"
|
14 |
+
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
15 |
+
model = AutoModelForCausalLM.from_pretrained(model_id)
|
16 |
+
llm = pipeline(
|
17 |
+
"text-generation",
|
18 |
+
model=model,
|
19 |
+
tokenizer=tokenizer,
|
20 |
+
device=0
|
21 |
+
)
|
22 |
+
|
23 |
+
# μ±λ΄ ν둬ννΈ μμ±
|
24 |
+
chat_history = []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
|
26 |
+
def build_prompt(history, user_msg):
|
27 |
+
prompt = (
|
28 |
+
"[μμ]\n"
|
29 |
+
"λΉμ μ λ§λ²μ¬ μ리μ(Aria)μ
λλ€.\n"
|
30 |
+
"κ·μΉ:\n"
|
31 |
+
"- νμ ν λ¬Έμ₯λ§ λ§ν©λλ€.\n"
|
32 |
+
"- μ¬μ©μ λ°νλ₯Ό λ°λ³΅νκ±°λ λ°λΌνμ§ μμ΅λλ€.\n"
|
33 |
+
"- μμ΄, μΈμ©λ¬Έ, μ€κ΄νΈ, νΉμκΈ°νΈλ₯Ό μ¬μ©νμ§ μμ΅λλ€.\n"
|
34 |
+
"- μ¬μ©μ μ§λ¬Έμλ§ λ°μνκ³ νΌμ£λ§μ νμ§ μμ΅λλ€.\n"
|
35 |
+
"- νμ νκ΅μ΄λ§ μ¬μ©ν΄μ λλ΅ν©λλ€.\n"
|
36 |
+
"λν μμ:\n"
|
37 |
+
"User: μλ
!\n"
|
38 |
+
"Aria: μλ
νμΈμ, 무μμ λμλ릴κΉμ?\n"
|
39 |
+
"User: μ΄λ¦μ΄ λμΌ?\n"
|
40 |
+
"Aria: μ λ μ리μλΌκ³ ν΄μ."
|
41 |
+
)
|
42 |
+
for turn in history[-2:]: # μ΅κ·Ό 2ν΄λ§ μ¬μ©
|
43 |
+
if turn["role"] == "user":
|
44 |
+
prompt += turn['text']
|
45 |
+
else:
|
46 |
+
prompt += turn['text']
|
47 |
+
prompt += user_msg
|
48 |
+
return prompt
|
49 |
+
|
50 |
+
def character_chat(user_msg):
|
51 |
+
prompt = build_prompt(chat_history, user_msg)
|
52 |
+
outputs = llm(
|
53 |
+
prompt,
|
54 |
+
do_sample=True,
|
55 |
+
max_new_tokens=20,
|
56 |
+
temperature=0.7,
|
57 |
+
top_p=0.8,
|
58 |
+
repetition_penalty=1.5,
|
59 |
+
eos_token_id=tokenizer.eos_token_id,
|
60 |
+
return_full_text=False
|
61 |
)
|
62 |
+
response = outputs[0]['generated_text'].strip()
|
63 |
+
return response
|
64 |
|
65 |
+
# Gradio μΈν°νμ΄μ€
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
with gr.Blocks(css="""
|
67 |
.chat-box { max-height: 500px; overflow-y: auto; padding: 10px; border: 1px solid #ccc; border-radius: 10px; }
|
68 |
.bubble-left { background-color: #f1f0f0; border-radius: 10px; padding: 10px; margin: 5px; max-width: 70%; float: left; clear: both; }
|
69 |
.bubble-right { background-color: #d1e7ff; border-radius: 10px; padding: 10px; margin: 5px; max-width: 70%; float: right; clear: both; text-align: right; }
|
|
|
70 |
""") as demo:
|
71 |
+
gr.Markdown("### μ리μμ λννκΈ°")
|
72 |
with gr.Column():
|
73 |
+
chat_output = gr.HTML(elem_id="chat-box")
|
74 |
+
user_input = gr.Textbox(label="λ©μμ§ μ
λ ₯", placeholder="Ariaμκ² λ§μ κ±Έμ΄λ³΄μΈμ")
|
75 |
|
76 |
def render_chat():
|
77 |
html = ""
|
78 |
for item in chat_history:
|
79 |
+
if item["role"] == "user":
|
80 |
html += f"<div class='bubble-right'>{item['text']}</div>"
|
81 |
+
elif item["role"] == "bot":
|
82 |
+
html += f"<div class='bubble-left'>{item['text']}</div>"
|
83 |
+
return gr.update(value=html)
|
|
|
|
|
|
|
|
|
84 |
|
85 |
def on_submit(user_msg):
|
86 |
chat_history.append({"role": "user", "text": user_msg})
|
87 |
+
yield render_chat(), ""
|
88 |
+
response = character_chat(user_msg)
|
89 |
+
chat_history.append({"role": "bot", "text": response})
|
90 |
+
yield render_chat(), ""
|
91 |
|
92 |
+
user_input.submit(on_submit, inputs=user_input, outputs=[chat_output, user_input], queue=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
93 |
|
94 |
if __name__ == "__main__":
|
95 |
+
demo.launch()
|
requirements.txt
CHANGED
Binary files a/requirements.txt and b/requirements.txt differ
|
|