Upload 2 files

app.py

@@ -3,6 +3,7 @@ import os
 import re
 import json
 import threading
+from  hf_api import restart_space
 
 import gradio as gr
 
@@ -467,15 +468,24 @@ with gr.Blocks() as chatting:
     @btn_com2.click(inputs=setting_cache_path,
                     outputs=[s_info, btn_submit])
     def btn_com2(_cache_path):
-        with lock:
-            _tmp = model.load_session(setting_cache_path.value)
-            print(f'load cache from {setting_cache_path.value} {_tmp}')
-            global vo_idx
-            vo_idx = 0
-            model.venv = [0]
-            global session_active
-            session_active = False
-            return str((model.n_tokens, model.venv)), gr.update(interactive=True)
+        try:
+            with lock:
+                _tmp = model.load_session(setting_cache_path.value)
+                print(f'load cache from {setting_cache_path.value} {_tmp}')
+                global vo_idx
+                vo_idx = 0
+                model.venv = [0]
+                global session_active
+                session_active = False
+                return str((model.n_tokens, model.venv)), gr.update(interactive=True)
+        except Exception as e:
+            restart_space()
+            raise e
+
+    @btn_com3.click()
+    def btn_com3():
+        restart_space()
+
 
     # ========== 开始运行 ==========
 demo = gr.TabbedInterface([chatting, setting, role],

hf_api.py

@@ -0,0 +1,10 @@
+import os
+
+from huggingface_hub import HfApi
+
+API = HfApi(token=os.environ.get("HF_TOKEN"))
+REPO_ID = "Limour/llama-python-streamingllm"
+
+
+def restart_space():
+    API.restart_space(repo_id=REPO_ID, token=os.environ.get("HF_TOKEN"))