Update app.py

app.py

@@ -1,64 +1,60 @@
+import os
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from typing import List, Tuple
 
-"""
-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")
+# Пути и настройки
+model_repo = "cody82/innopolis_bot_model"
+cache_dir = "/data/model"  # Путь к Persistent Storage
 
+if not os.path.exists(cache_dir):
+    os.makedirs(cache_dir)
 
+# Загружаем токенизатор и модель из локального кеша / Persistent Storage
+tokenizer = AutoTokenizer.from_pretrained(model_repo, cache_dir=cache_dir)
+model = AutoModelForCausalLM.from_pretrained(model_repo, cache_dir=cache_dir)
+model.to("cpu")
+
+# Функция обработки сообщений
 def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
+    message: str,
+    history: List[Tuple[str, str]],
+    system_message: str,
+    max_tokens: int,
+    temperature: float,
+    top_p: float,
 ):
-    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,
+    full_input = f"{system_message}\n"
+    for user_msg, assistant_msg in history:
+        full_input += f"User: {user_msg}\nAssistant: {assistant_msg}\n"
+    full_input += f"User: {message}\nAssistant:"
+
+    inputs = tokenizer(full_input, return_tensors="pt").to("cpu")
+    outputs = model.generate(
+        **inputs,
+        max_new_tokens=max_tokens,
+        do_sample=True,
         temperature=temperature,
         top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
+        pad_token_id=tokenizer.eos_token_id,
+    )
+    output_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    response = output_text.split("Assistant:")[-1].strip()
 
+    history.append((message, response))
+    return response, history
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Gradio интерфейс
 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)",
-        ),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
     ],
 )
 
-
 if __name__ == "__main__":
     demo.launch()