Update app.py
Browse files
app.py
CHANGED
@@ -1,14 +1,31 @@
|
|
1 |
import gradio as gr
|
2 |
-
from
|
3 |
-
|
4 |
-
from transformers import pipeline
|
5 |
|
6 |
-
|
7 |
-
{"role": "user", "content": "Who are you?"},
|
8 |
-
]
|
9 |
-
pipe = pipeline("text-generation", model="Qwen/Qwen2.5-14B-Instruct-1M")
|
10 |
-
pipe(messages)
|
11 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
|
13 |
-
|
14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import gradio as gr
|
2 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
3 |
+
import torch
|
|
|
4 |
|
5 |
+
MODEL_NAME = "Qwen/Qwen2.5-14B-Instruct-1M"
|
|
|
|
|
|
|
|
|
6 |
|
7 |
+
# بارگذاری مدل و توکنایزر
|
8 |
+
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
|
9 |
+
model = AutoModelForCausalLM.from_pretrained(
|
10 |
+
MODEL_NAME,
|
11 |
+
torch_dtype=torch.float16,
|
12 |
+
device_map="auto"
|
13 |
+
)
|
14 |
|
15 |
+
# تابع تولید متن
|
16 |
+
def chat_with_qwen(prompt):
|
17 |
+
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
|
18 |
+
output = model.generate(**inputs, max_new_tokens=200)
|
19 |
+
response = tokenizer.decode(output[0], skip_special_tokens=True)
|
20 |
+
return response
|
21 |
+
|
22 |
+
# ایجاد رابط کاربری با Gradio
|
23 |
+
iface = gr.Interface(
|
24 |
+
fn=chat_with_qwen,
|
25 |
+
inputs=gr.Textbox(lines=2, placeholder="سوال خود را اینجا بنویسی"),
|
26 |
+
outputs="text",
|
27 |
+
title="Qwen 2.5 14B Chatbot",
|
28 |
+
description="یک چتبات مبتنی بر مدل Qwen/Qwen2.5-14B-Instruct-1M",
|
29 |
+
)
|
30 |
+
|
31 |
+
iface.launch()
|