Update app.py

app.py

@@ -1,64 +1,49 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from huggingface_hub import hf_hub_download
+from llama_cpp import Llama
+import os
 
-"""
-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("AXCXEPT/llm-jp-3-3.7b-instruct-EZO-Humanities")
+# Hugging Face Hub上のモデルを指定
+repo_id = "mmnga/ELYZA-japanese-Llama-2-7b-instruct-gguf"
+filename = "ELYZA-japanese-Llama-2-7b-instruct-q4_K_M.gguf"
 
+# モデルをダウンロード（キャッシュされている場合はキャッシュを使用）
+model_path = hf_hub_download(repo_id=repo_id, filename=filename)
 
-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
+CONTEXT_SIZE = 4096
 
+llm = Llama(
+    model_path=model_path,
+    n_threads=os.cpu_count(),
+    n_batch=32,
+    verbose=False,
+    n_ctx=CONTEXT_SIZE,
+)
 
-"""
-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 get_llama_response(prompt):
+    return llm(prompt, max_tokens=2048, temperature=0.7, top_p=0.95, repeat_penalty=1.1, stream=True)
+
+def greet(prompt, intensity):
+    full_response = ""
+    for output in get_llama_response(prompt):
+        if len(output['choices']) > 0:
+            text_chunk = output['choices'][0]['text']
+            full_response += text_chunk
+            yield full_response
+    
+    return full_response + "！" * int(intensity)
+
+demo = gr.Interface(
+    title="Llama.cpp-python-sample (Streaming)",
+    description=f"MODEL: {filename} from {repo_id}",
+    fn=greet,
+    inputs=[
+        gr.Textbox(label="Enter your prompt"),
+        gr.Slider(minimum=0, maximum=10, step=1, label="Intensity")
     ],
+    outputs=gr.Textbox(label="Generated Response"),
+    live=False
 )
 
-
-if __name__ == "__main__":
-    demo.launch()
+demo.queue()
+demo.launch()