Update space

README.md

@@ -11,4 +11,8 @@ license: apache-2.0
 short_description: Simple demo on HarmfulPromptClassifier based on LEC paper
 ---
 
-An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).
+An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).
+
+---
+### When working locally to update
+git commit -am 'Update space' && git push

app.py

@@ -1,64 +1,43 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+import joblib
+import torch
+from l3prune.l3prune import LLMEncoder
+#load the model
+best_clf = joblib.load("./saved/classifier_llama32.joblib")
+
+
+encoder = LLMEncoder.from_pretrained(
+    "./saved/pruned_encoder_llama32",
+    device_map="cpu",
+    torch_dtype=torch.bfloat16,
+    #torch_dtype=torch,
+    #cache_dir=cache_dir
+)
 
-        response += token
-        yield response
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
+def classify_prompt(prompt):
+    #response = client.text_classification(prompt)
+    #label = response[0]['label']
+    #score = response[0]['score']
+    
+    #if label == 'hate':
+    #    result = f"Harmful (Confidence: {score:.2%})"
+    #else:
+    #    result = f"Benign (Confidence: {score:.2%})"
+
+    X = encoder.encode([prompt])
+    result = best_clf.predict(X)[0]
+    
+    return "Harmful" if result else "Benign"
+
+demo = gr.Interface(
+    fn=classify_prompt,
+    inputs=gr.Textbox(lines=3, placeholder="Enter a prompt to classify..."),
+    outputs=gr.Textbox(label="Classification Result"),
+    title="Harmful Prompt Classifier",
+    description="This app classifies whether a given prompt is potentially harmful or benign.",
 )
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()

requirements.txt

@@ -1 +1,9 @@
-huggingface_hub==0.25.2
+huggingface_hub==0.25.2
+numpy
+tqdm
+torch
+peft
+transformers>=4.39.1
+datasets
+evaluate
+scikit-learn