Update demo_watermark.py

demo_watermark.py

@@ -477,18 +477,14 @@ def run_gradio(args, model=None, device=None, tokenizer=None):
         # note that state obj automatically calls value if it's a callable, want to avoid calling tokenizer at startup
         session_tokenizer = gr.State(value=lambda : tokenizer) 
 
-        with gr.Tab("Welcome"):
+        with gr.Tab("欢迎"):
             with gr.Row():
                 with gr.Column(scale=2):
                     gr.Markdown(
                         """
-                        Potential harms of large language models can be mitigated by *watermarking* a model's output. 
-                        *Watermarks* are embedded signals in the generated text that are invisible to humans but algorithmically 
-                        detectable, that allow *anyone* to later check whether a given span of text 
-                        was likely to have been generated by a model that uses the watermark.
+                         大型语言模型的潜在危害可以通过*水印*来减轻模型的输出。 *水印*是嵌入到生成的文本中的信号，对人类是不可见的，但是可以通过算法检测到的，可以让*任何人*以后检查给定的文本片段是否很可能是由使用了水印的模型生成的。
 
-                        This space showcases a watermarking approach that can be applied to _any_ generative language model. 
-                        For demonstration purposes, the space demos a selection of multi-billion parameter models (see the following note for caveats).
+                此空间展示了一种可以应用于_任何_生成语言模型的水印方法。 为了演示目的，该空间演示了一些十亿级参数模型的选择（有关注意事项，请参见以下说明）。
                         """
                         )
                     with gr.Accordion("A note on the available models:",open=False):