Update app.py

app.py

@@ -1,21 +1,35 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
+import requests
+import os
+
+url = "http://47.94.86.196:8080/chat_completion"
 
 """
-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
+class ChatRequest(BaseModel):
+    messages: typing.List[typing.Dict]
+    max_new_tokens: int
+    max_length: int = 4096
+    temperature: int = 1.0
+    top_k: int = 0
+    top_p: float = 1.0
+    repetition_penalty: float = 1.03
+    do_sample: bool = True
+    seed: int = 42
 """
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
 
 def respond(
     message,
     history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
+    do_sample: bool,
+    seed: int,
+    max_new_tokens,
     temperature,
     top_p,
+    top_k,
+    repetition_penalty
 ):
-    messages = [{"role": "system", "content": system_message}]
+    messages = []
 
     for val in history:
         if val[0]:
@@ -26,18 +40,22 @@ def respond(
     messages.append({"role": "user", "content": message})
 
     response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
+    request_data = dict(
+        messages=messages,
+        max_new_tokens=max_new_tokens,
+        do_sample=do_sample,
+        seed=seed,
         top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
+        top_k=top_k,
+        temperature=temperature,
+        repetition_penalty=repetition_penalty
+    )
+    print(request_data)
+    with requests.post(url, json=request_data, stream=True, headers={"Authorization": f"Bearer {os.environ['HF_TOKEN']}"}) as r:
+        # printing response of each stream
+        for chunk in r.iter_content(1024):
+            response += chunk.decode("utf8")
+            yield response
 
 """
 For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
@@ -45,19 +63,36 @@ For information on how to customize the ChatInterface, peruse the gradio docs: h
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        # gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Checkbox(True, label="do sample"),
+        gr.Number(42, precision=0, label="seed"),
         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=0.01, maximum=4.0, value=0.7, step=0.01, label="Temperature"),
         gr.Slider(
             minimum=0.1,
             maximum=1.0,
-            value=0.95,
+            value=1.0,
             step=0.05,
             label="Top-p (nucleus sampling)",
         ),
+        gr.Slider(
+            minimum=0,
+            maximum=100,
+            value=0,
+            step=1,
+            label="Top-K (Top-K sampling)",
+        ),
+        gr.Slider(
+            minimum=1,
+            maximum=2,
+            value=1.03,
+            step=0.01,
+            label="repetition penalty",
+        ),
     ],
 )
 
 
 if __name__ == "__main__":
+    demo.queue(default_concurrency_limit=1, max_size=5)
     demo.launch()