Update app.py
Browse files
app.py
CHANGED
@@ -1,11 +1,9 @@
|
|
1 |
import gradio as gr
|
2 |
-
from
|
3 |
import os
|
4 |
-
from typing import List, Tuple
|
5 |
-
import concurrent.futures
|
6 |
|
7 |
# Hugging Face ํ ํฐ ์ค์
|
8 |
-
os.environ["TOKENIZERS_PARALLELISM"] = "false" # ๊ฒฝ๊ณ ๋ฉ์์ง ๋ฐฉ์ง
|
9 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
10 |
|
11 |
# Available LLM models
|
@@ -26,18 +24,11 @@ DEFAULT_MODELS = [
|
|
26 |
"mistralai/Mistral-Nemo-Instruct-2407"
|
27 |
]
|
28 |
|
29 |
-
#
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
"text-generation",
|
35 |
-
model=model_name,
|
36 |
-
token=HF_TOKEN,
|
37 |
-
device_map="auto"
|
38 |
-
)
|
39 |
-
except Exception as e:
|
40 |
-
print(f"Failed to load model {model_name}: {str(e)}")
|
41 |
|
42 |
def process_file(file) -> str:
|
43 |
if file is None:
|
@@ -46,7 +37,15 @@ def process_file(file) -> str:
|
|
46 |
return file.read().decode('utf-8')
|
47 |
return f"Uploaded file: {file.name}"
|
48 |
|
49 |
-
def
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
messages = [{"role": "system", "content": system_message}]
|
51 |
|
52 |
for user, assistant in history:
|
@@ -56,35 +55,18 @@ def format_messages(message: str, history: List[Tuple[str, str]], system_message
|
|
56 |
messages.append({"role": "assistant", "content": assistant})
|
57 |
|
58 |
messages.append({"role": "user", "content": message})
|
59 |
-
|
60 |
-
|
61 |
-
def generate_response(
|
62 |
-
pipe,
|
63 |
-
messages: List[dict],
|
64 |
-
max_tokens: int,
|
65 |
-
temperature: float,
|
66 |
-
top_p: float
|
67 |
-
) -> Generator[str, None, None]:
|
68 |
try:
|
69 |
-
|
70 |
-
|
71 |
-
response = pipe(
|
72 |
-
formatted_prompt,
|
73 |
max_new_tokens=max_tokens,
|
|
|
74 |
temperature=temperature,
|
75 |
top_p=top_p,
|
76 |
-
|
77 |
-
|
78 |
-
|
79 |
-
streaming=True
|
80 |
-
)
|
81 |
-
|
82 |
-
generated_text = ""
|
83 |
-
for output in response:
|
84 |
-
new_text = output[0]['generated_text'][len(formatted_prompt):].strip()
|
85 |
-
generated_text = new_text
|
86 |
-
yield generated_text
|
87 |
-
|
88 |
except Exception as e:
|
89 |
yield f"Error: {str(e)}"
|
90 |
|
@@ -99,7 +81,7 @@ def respond_all(
|
|
99 |
max_tokens: int,
|
100 |
temperature: float,
|
101 |
top_p: float,
|
102 |
-
)
|
103 |
if file:
|
104 |
file_content = process_file(file)
|
105 |
message = f"{message}\n\nFile content:\n{file_content}"
|
@@ -107,16 +89,25 @@ def respond_all(
|
|
107 |
while len(selected_models) < 3:
|
108 |
selected_models.append(selected_models[-1])
|
109 |
|
110 |
-
def generate(
|
111 |
-
|
112 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
|
114 |
return (
|
115 |
-
generate(
|
116 |
-
generate(
|
117 |
-
generate(
|
118 |
)
|
119 |
|
|
|
|
|
120 |
css = """
|
121 |
footer {
|
122 |
visibility: hidden;
|
@@ -126,6 +117,7 @@ footer {
|
|
126 |
|
127 |
|
128 |
with gr.Blocks(theme="Yntec/HaleyCH_Theme_Orange", css=css) as demo:
|
|
|
129 |
with gr.Row():
|
130 |
model_choices = gr.Checkboxgroup(
|
131 |
choices=list(LLM_MODELS.values()),
|
@@ -212,7 +204,7 @@ with gr.Blocks(theme="Yntec/HaleyCH_Theme_Orange", css=css) as demo:
|
|
212 |
)
|
213 |
|
214 |
if __name__ == "__main__":
|
215 |
-
# Hugging Face ํ ํฐ์ด ์ค์ ๋์ด ์๋์ง ํ์ธ
|
216 |
if not HF_TOKEN:
|
217 |
print("Warning: HF_TOKEN environment variable is not set")
|
218 |
-
demo.launch()
|
|
|
|
1 |
import gradio as gr
|
2 |
+
from huggingface_hub import InferenceClient
|
3 |
import os
|
4 |
+
from typing import List, Tuple
|
|
|
5 |
|
6 |
# Hugging Face ํ ํฐ ์ค์
|
|
|
7 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
8 |
|
9 |
# Available LLM models
|
|
|
24 |
"mistralai/Mistral-Nemo-Instruct-2407"
|
25 |
]
|
26 |
|
27 |
+
# Initialize clients with token
|
28 |
+
clients = {
|
29 |
+
model: InferenceClient(model, token=HF_TOKEN)
|
30 |
+
for model in LLM_MODELS.values()
|
31 |
+
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
def process_file(file) -> str:
|
34 |
if file is None:
|
|
|
37 |
return file.read().decode('utf-8')
|
38 |
return f"Uploaded file: {file.name}"
|
39 |
|
40 |
+
def respond_single(
|
41 |
+
client,
|
42 |
+
message: str,
|
43 |
+
history: List[Tuple[str, str]],
|
44 |
+
system_message: str,
|
45 |
+
max_tokens: int,
|
46 |
+
temperature: float,
|
47 |
+
top_p: float,
|
48 |
+
):
|
49 |
messages = [{"role": "system", "content": system_message}]
|
50 |
|
51 |
for user, assistant in history:
|
|
|
55 |
messages.append({"role": "assistant", "content": assistant})
|
56 |
|
57 |
messages.append({"role": "user", "content": message})
|
58 |
+
|
59 |
+
response = ""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
60 |
try:
|
61 |
+
for msg in client.text_generation(
|
62 |
+
prompt=message,
|
|
|
|
|
63 |
max_new_tokens=max_tokens,
|
64 |
+
stream=True,
|
65 |
temperature=temperature,
|
66 |
top_p=top_p,
|
67 |
+
):
|
68 |
+
response += msg
|
69 |
+
yield response
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
70 |
except Exception as e:
|
71 |
yield f"Error: {str(e)}"
|
72 |
|
|
|
81 |
max_tokens: int,
|
82 |
temperature: float,
|
83 |
top_p: float,
|
84 |
+
):
|
85 |
if file:
|
86 |
file_content = process_file(file)
|
87 |
message = f"{message}\n\nFile content:\n{file_content}"
|
|
|
89 |
while len(selected_models) < 3:
|
90 |
selected_models.append(selected_models[-1])
|
91 |
|
92 |
+
def generate(client, history):
|
93 |
+
return respond_single(
|
94 |
+
client,
|
95 |
+
message,
|
96 |
+
history,
|
97 |
+
system_message,
|
98 |
+
max_tokens,
|
99 |
+
temperature,
|
100 |
+
top_p,
|
101 |
+
)
|
102 |
|
103 |
return (
|
104 |
+
generate(clients[selected_models[0]], history1),
|
105 |
+
generate(clients[selected_models[1]], history2),
|
106 |
+
generate(clients[selected_models[2]], history3),
|
107 |
)
|
108 |
|
109 |
+
|
110 |
+
|
111 |
css = """
|
112 |
footer {
|
113 |
visibility: hidden;
|
|
|
117 |
|
118 |
|
119 |
with gr.Blocks(theme="Yntec/HaleyCH_Theme_Orange", css=css) as demo:
|
120 |
+
|
121 |
with gr.Row():
|
122 |
model_choices = gr.Checkboxgroup(
|
123 |
choices=list(LLM_MODELS.values()),
|
|
|
204 |
)
|
205 |
|
206 |
if __name__ == "__main__":
|
|
|
207 |
if not HF_TOKEN:
|
208 |
print("Warning: HF_TOKEN environment variable is not set")
|
209 |
+
demo.launch()
|
210 |
+
|