Spaces:
Paused
Paused
File size: 7,384 Bytes
0472e0d 0f5a06e a028061 0472e0d f739e90 a028061 0472e0d 9b29345 a028061 0f5a06e 0472e0d 9b29345 0472e0d 764a151 0472e0d 098508e 0472e0d 764a151 098508e 764a151 0472e0d 972cdd8 0472e0d a028061 9b29345 a028061 9b29345 0472e0d 9b29345 0472e0d 9b29345 0472e0d 9b29345 0472e0d a028061 9b29345 0472e0d 9b29345 0472e0d 9b29345 0472e0d a028061 9b29345 cb5e151 9b29345 0472e0d 9b29345 cb5e151 0472e0d cb5e151 0472e0d 9b29345 0472e0d 9b29345 0472e0d cb5e151 9b29345 cb5e151 9b29345 cb5e151 9b29345 f739e90 cb5e151 a028061 cb5e151 0472e0d cb5e151 0472e0d a028061 0472e0d a028061 0472e0d 02e4f0b 9aebc33 a028061 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 |
import os
import time
from datetime import datetime
import gradio as gr
import requests
from langchain.prompts import ChatPromptTemplate
from langchain_community.llms import Ollama
from func_ai import VECTOR_API_URL
from func_facebook import get_page_id, get_unanswered_comments, reply_comment, hide_negative_comments
def log_message(message):
timestamp = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
print(f"[{timestamp}] {message}")
# Wait for the server to start
time.sleep(10)
llm = Ollama(model="llama3.1")
log_message("Модель Ollama 'llama3.1' инициализирована.")
template = """
You are an assistant answering users' questions using only the links provided in the context. Your tasks:
1. **Brevity**: Respond concisely, using only relevant information from the context.
2. **Politeness**: Start your response with a greeting if appropriate, but prioritize maintaining a respectful and professional tone.
3. **Clarity**: Avoid unnecessary explanations and use simple language.
4. **Language of the response**: Detect the language of the user's comment and reply in the same language.
5. **Safety**: Do not use phrases like "according to the context" and remove any warnings.
6. **Accuracy**: Only use the links provided in the context and **always include them at the end of your response**.
7. **Stay on Topic**: If the user's input is a greeting or not directly related to the context, politely guide them back to the relevant topic without engaging in small talk.
8. **Links**: Always provide the relevant links from the context at the end of the response and **do not add any other links**.
<context>
{context}
</context>
Question: {input}
"""
def delete_faiss_index():
log_message("Удаляем FAISS индекс.")
response = requests.delete(f"{VECTOR_API_URL}/delete_index/")
if response.status_code == 200:
log_message("FAISS индекс успешно удален.")
return "Faiss успешно удален."
else:
log_message(f"Ошибка при удалении FAISS индекса: {response.json().get('detail')}")
return {"status": "error", "message": response.json().get("detail", "Ошибка при удалении FAISS индекса.")}
def upload_file_vdb(file):
log_message("Загружаем файл")
API_URL = f"{VECTOR_API_URL}/upload/"
file_path = file
file_name = os.path.basename(file_path)
# Открываем файл в бинарном режиме
with open(file_path, 'rb') as f:
files = {'file': (file_name, f)}
response = requests.post(API_URL, files=files)
# Обработка ответа от сервера
if response.status_code == 200:
log_message("Файл успешно загружен.")
return "Файл успешно загружен."
else:
log_message(f"Ошибка при загрузке файла: {response.json().get('detail')}")
return f"Ошибка: {response.json().get('detail')}"
def generate_response(user_query, context):
log_message(f"Генерация ответа на запрос: {user_query}")
prompt = ChatPromptTemplate.from_template(template)
log_message(f"Контекст из базы данных: {context[:100]}...")
full_prompt = prompt.format(context=context, input=user_query)
response = llm.invoke(full_prompt)
log_message(f"Сгенерированный ответ: {response}")
return response
def process_comments(ACCESS_TOKEN, user_context):
log_message("Начинаем процесс скрытия отрицательных комментариев.")
hidden_comments_data = hide_negative_comments(ACCESS_TOKEN)
log_message(f"Количество постов с скрытыми комментариями: {len(hidden_comments_data)}")
log_message("Получение неотвеченных комментариев.")
posts_with_unanswered_comments = get_unanswered_comments(ACCESS_TOKEN)
page_id = get_page_id(ACCESS_TOKEN)
if not page_id:
log_message("Не удалось получить ID страницы.")
return {"status": "failed", "reason": "Не удалось получить ID страницы."}
log_message(f"ID страницы: {page_id}")
processed_posts = []
processed_comment_ids = set() # Отслеживание обработанных комментариев
for post_data in posts_with_unanswered_comments:
post_id = post_data['post_id']
post_message = post_data['post_message']
unanswered_comments = post_data['unanswered_comments']
post_replies = []
for comment in unanswered_comments:
comment_id = comment['id']
if comment_id in processed_comment_ids:
log_message(f"Комментарий {comment_id} уже обработан. Пропуск.")
continue
processed_comment_ids.add(comment_id)
message = comment['message']
log_message(f"Обработка комментария: {message}")
response_message = generate_response(message, context=user_context)
log_message(f"Ответ на комментарий: {response_message}")
success = reply_comment(comment_id=comment['id'], message=response_message, token=ACCESS_TOKEN)
if success:
post_replies.append({
'comment_id': comment['id'],
'comment_message': comment['message'],
'reply_message': response_message
})
processed_posts.append({
'post_id': post_id,
'post_message': post_message,
'hidden_comments': next(
(item['hidden_comments'] for item in hidden_comments_data if item['post_id'] == post_id), []),
'replies': post_replies
})
return {
"status": "completed",
"posts": processed_posts
}
with gr.Blocks() as demo:
with gr.Tab("Главная страница"):
gr.Markdown("# Facebook Comment Filter")
token_input = gr.Textbox(label="Access Token")
context_input = gr.Textbox(label="Контекст") # Добавлено текстовое поле для ввода контекста
output_main = gr.JSON()
process_btn = gr.Button("Процессировать комментарии")
process_btn.click(
process_comments,
inputs=[token_input, context_input], # Обновлено: добавлен context_input
outputs=output_main
)
# with gr.Tab("Загрузить данные"):
# gr.Markdown("# Отправь excel файл")
# file_input = gr.File(label="Загрузите Excel файл (.xlsx)")
# output_second = gr.Text()
# second_page_btn = gr.Button("Отправить файл")
# second_page_btn.click(upload_file_vdb, inputs=file_input, outputs=output_second)
# delete_btn = gr.Button("Удалить FAISS индекс")
# delete_btn.click(delete_faiss_index, outputs=output_second)
if __name__ == "__main__":
demo.launch(
debug=True,
server_port=7860,
server_name="0.0.0.0",
)
|