gen

app.py

@@ -12,7 +12,36 @@ model = AutoModelForCausalLM.from_pretrained(
 print(f"load tokenizer {model_name}")
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 
+
+def text_gen(msg):
+    
+    prompt = "Give me a short introduction to large language models."
+    messages = [
+    {"role": "system", "content": "You are Ring, an assistant created by inclusionAI"},
+    {"role": "user", "content": prompt}
+    ]
+    text = tokenizer.apply_chat_template(
+    messages,
+    tokenize=False,
+    add_generation_prompt=True,
+    enable_thinking=True
+    )
+    model_inputs = tokenizer([text], return_tensors="pt", return_token_type_ids=False).to(model.device)
+
+    generated_ids = model.generate(
+    **model_inputs,
+    max_new_tokens=8192
+    )
+    generated_ids = [
+    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
+    ]
+
+    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
+    return response
+
+    
 from flask import Flask  
+from flask import request
 
 app = Flask(__name__)  
 print(f"Flask app")
@@ -31,5 +60,13 @@ def about():
 def contact():
     return "<h1>Контакты</h1><p>Свяжитесь с нами: email@example.com</p>"
 
+# Страница "Контакты"
+@app.route("/gen", methods=['POST', 'GET'])
+def gen_msg():
+    if request.form.get('msg'):
+        answer = text_gen(request.form['msg'])
+        
+    return f"<h1>{answer}</p>"
+
 if __name__ == "__main__":
     app.run(debug=False, host='0.0.0.0', port=7860)