llm_server/database/log_to_db.py

@@ -27,6 +27,4 @@ def log_to_db(ip: str, token: str, prompt: str, response: Union[str, None], gen_
             'is_error': is_error
         }
     }
-    for k, v in data['kwargs'].items():
-        print(k, type(v))
     r.publish('database-logger', pickle.dumps(data))

llm_server/workers/inferencer.py

@@ -66,6 +66,7 @@ def need_to_wait(backend_url: str):
     concurrent_gens = cluster_config.get_backend(backend_url).get('concurrent_gens', 1)
     s = time.time()
     while active_workers >= concurrent_gens:
+        print('worker waiting')
         time.sleep(0.01)
     e = time.time()
     if e - s > 0.5: