Update app.py

app.py

@@ -1,57 +1,62 @@
 import os
-from typing import Iterator
-
-from text_generation import Client
-
-model_id = 'HuggingFaceH4/zephyr-7b-beta'
-
-API_URL = "https://api-inference.huggingface.co/models/" + model_id
-HF_TOKEN = os.environ.get("HF_READ_TOKEN", None)
-
-client = Client(
-    API_URL,
-    headers={"Authorization": f"Bearer {HF_TOKEN}"},
-)
-EOS_STRING = "</s>"
-EOT_STRING = "<EOT>"
-
-
-def get_prompt(message: str, chat_history: list[tuple[str, str]],
-               system_prompt: str) -> str:
-    texts = [f'<s>[INST] <<SYS>>\n{system_prompt}\n<</SYS>>\n\n']
-    # The first user input is _not_ stripped
-    do_strip = False
-    for user_input, response in chat_history:
-        user_input = user_input.strip() if do_strip else user_input
-        do_strip = True
-        texts.append(f'{user_input} [/INST] {response.strip()} </s><s>[INST] ')
-    message = message.strip() if do_strip else message
-    texts.append(f'{message} [/INST]')
-    return ''.join(texts)
-
-
-def run(message: str,
-        chat_history: list[tuple[str, str]],
-        system_prompt: str,
-        max_new_tokens: int = 1024,
-        temperature: float = 0.1,
-        top_p: float = 0.9,
-        top_k: int = 50) -> Iterator[str]:
-    prompt = get_prompt(message, chat_history, system_prompt)
-
-    generate_kwargs = dict(
-        max_new_tokens=max_new_tokens,
-        do_sample=True,
-        top_p=top_p,
-        top_k=top_k,
-        temperature=temperature,
-    )
-    stream = client.generate_stream(prompt, **generate_kwargs)
-    output = ""
-    for response in stream:
-        if any([end_token in response.token.text for end_token in [EOS_STRING, EOT_STRING]]):
-            return output
-        else:
-            output += response.token.text
-        yield output
-    return output
+import gradio as gr
+from transformers import pipeline
+
+# auth_token = os.environ.get("access_token")
+pipeline_en = pipeline(task="text-classification", model="Hello-SimpleAI/chatgpt-detector-roberta")
+pipeline_zh = pipeline(task="text-classification", model="Hello-SimpleAI/chatgpt-detector-roberta-chinese")
+
+
+
+def predict_en(text):
+    res = pipeline_en(text)[0]
+    return res['label'],res['score']
+
+def predict_zh(text):
+    res = pipeline_zh(text)[0]
+    return res['label'],res['score']
+
+
+
+
+with gr.Blocks() as demo:
+    gr.Markdown("""
+                ## ChatGPT Detector 🔬 (Sinlge-text version)
+                Visit our project on Github: [chatgpt-comparison-detection project](https://github.com/Hello-SimpleAI/chatgpt-comparison-detection)<br>
+                欢迎在 Github 上关注我们的 [ChatGPT 对比与检测项目](https://github.com/Hello-SimpleAI/chatgpt-comparison-detection)
+                We provide three kinds of detectors, all in Bilingual / 我们提供了三个版本的检测器，且都支持中英文:
+                - [**QA version / 问答版**](https://huggingface.co/spaces/Hello-SimpleAI/chatgpt-detector-qa)<br>
+                    detect whether an **answer** is generated by ChatGPT for certain **question**, using PLM-based classifiers / 判断某个**问题的回答**是否由ChatGPT生成，使用基于PTM的分类器来开发;
+                - [Sinlge-text version / 独立文本版 (👈 Current / 当前使用)](https://huggingface.co/spaces/Hello-SimpleAI/chatgpt-detector-single)<br>
+                    detect whether a piece of text is ChatGPT generated, using PLM-based classifiers / 判断**单条文本**是否由ChatGPT生成，使用基于PTM的分类器来开发;
+                - [Linguistic version / 语言学版](https://huggingface.co/spaces/Hello-SimpleAI/chatgpt-detector-ling)<br>
+                    detect whether a piece of text is ChatGPT generated, using linguistic features / 判断**单条文本**是否由ChatGPT生成，使用基于语言学特征的模型来开发;
+                
+                
+                """)
+    with gr.Tab("English"):
+        gr.Markdown("""
+                    Note: Providing more text to the `Text` box can make the prediction more accurate!
+                    """)
+        t1 = gr.Textbox(lines=5, label='Text',value="There are a few things that can help protect your credit card information from being misused when you give it to a restaurant or any other business:\n\nEncryption: Many businesses use encryption to protect your credit card information when it is being transmitted or stored. This means that the information is transformed into a code that is difficult for anyone to read without the right key.")
+        button1 = gr.Button("🤖 Predict!")
+        label1 = gr.Textbox(lines=1, label='Predicted Label 🎃')
+        score1 = gr.Textbox(lines=1, label='Prob')
+    with gr.Tab("中文版"):
+        gr.Markdown("""
+                    注意: 在`文本`栏中输入更多的文本，可以让预测更准确哦！
+                    """)
+        t2 = gr.Textbox(lines=5, label='文本',value="对于OpenAI大力出奇迹的工作，自然每个人都有自己的看点。我自己最欣赏的地方是ChatGPT如何解决 “AI校正(Alignment)“这个问题。这个问题也是我们课题组这两年在探索的学术问题之一。")
+        button2 = gr.Button("🤖 预测!")
+        label2 = gr.Textbox(lines=1, label='预测结果 🎃')
+        score2 = gr.Textbox(lines=1, label='模型概率')
+
+    button1.click(predict_en, inputs=[t1], outputs=[label1,score1], api_name='predict_en')
+    button2.click(predict_zh, inputs=[t2], outputs=[label2,score2], api_name='predict_zh')
+
+    # Page Count
+    gr.Markdown("""
+                <center><a href='https://clustrmaps.com/site/1bsdc'  title='Visit tracker'><img src='//clustrmaps.com/map_v2.png?cl=080808&w=a&t=tt&d=NXQdnwxvIm27veMbB5F7oHNID09nhSvkBRZ_Aji9eIA&co=ffffff&ct=808080'/></a></center>
+                """)
+
+demo.launch()