llm_server/routes/v1/generate_stats.py

@@ -53,7 +53,7 @@ def generate_stats(regen: bool = False):
             'openaiKeys': '∞',
             'anthropicKeys': '∞',
         },
-        'backend_info': redis.get_dict('backend_info') if opts.show_backend_info else None,
+        'backends': {},
     }
 
     # TODO: have get_model_choices() return all the info so we don't have to loop over the backends ourself
@@ -64,7 +64,7 @@ def generate_stats(regen: bool = False):
             if not backend_info['online']:
                 continue
             backend_uptime = int((datetime.now() - datetime.fromtimestamp(backend_info['startup_time'])).total_seconds()) if opts.show_uptime else None
-            output['backend_info'][backend_info['hash']] = {
+            output['backends'][backend_info['hash']] = {
                 'uptime': backend_uptime,
                 'max_tokens': backend_info['model_config']['max_position_embeddings'],
                 'model': backend_info['model'],
@@ -72,8 +72,6 @@ def generate_stats(regen: bool = False):
                 'nvidia': backend_info['nvidia'],
                 'priority': backend_info['priority'],
             }
-    else:
-        output['backend_info'] = {}
 
     result = deep_sort(output)