Update chat to use secrets

app.py

@@ -1,64 +1,60 @@
+import os
 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
-
-        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)",
-        ),
-    ],
+from openai import OpenAI
+
+title = None  # "ServiceNow-AI Chat"
+description = None
+
+modelConfig = {
+    "MODEL_NAME": os.environ.get("MODEL_NAME"),
+    "MODE_DISPLAY_NAME": os.environ.get("MODE_DISPLAY_NAME"),
+    "MODEL_HF_URL": os.environ.get("MODEL_HF_URL"),
+    "VLLM_API_URL": os.environ.get("VLLM_API_URL"),
+    "AUTH_TOKEN": os.environ.get("AUTH_TOKEN")
+}
+
+# Initialize the OpenAI client with the vLLM API URL and token
+client = OpenAI(
+    api_key=modelConfig.get('AUTH_TOKEN'),
+    base_url=modelConfig.get('VLLM_API_URL')
 )
 
 
-if __name__ == "__main__":
-    demo.launch()
+def chat_fn(message, history):
+    # Format history as OpenAI expects
+    formatted = [{"role": "user", "content": user} if i % 2 == 0 else {"role": "assistant", "content": assistant}
+                 for i, (user, assistant) in enumerate(history)]
+    formatted.append({"role": "user", "content": message})
+
+    # Create the streaming response
+    stream = client.chat.completions.create(
+        model=modelConfig.get('MODEL_NAME'),
+        messages=formatted,
+        temperature=0.8,
+        stream=True
+    )
+
+    output = ""
+    for chunk in stream:
+        # Extract the new content from the delta field
+        content = getattr(chunk.choices[0].delta, "content", "")
+        output += content
+        # Yield the current accumulated output, removing "<|end|>" if present
+        if output.endswith("<|end|>"):
+            yield {"role": "assistant", "content": output[:-7]}
+        else:
+            yield {"role": "assistant", "content": output}
+
+
+# Add the model display name and Hugging Face URL to the description
+# description = f"### Model: [{MODE_DISPLAY_NAME}]({MODEL_HF_URL})"
+
+print(f"Running model {modelConfig.get('MODE_DISPLAY_NAME')} ({modelConfig.get('MODEL_NAME')})")
+
+gr.ChatInterface(
+    chat_fn,
+    title=title,
+    description=description,
+    theme=gr.themes.Default(primary_hue="green"),
+    type="messages"
+).launch()