fixed llamafile args quoting

flowaicom · Oct 8, 2024 · 2d7f16b · 2d7f16b
1 parent 1b56a93
commit 2d7f16b
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/flow_judge/models/llamafile.py b/flow_judge/models/llamafile.py
@@ -240,7 +240,7 @@ def start_llamafile_server(self):
                   f"--threads {self.generation_params.get('thread_count', os.cpu_count() or 1)} " \
                   f"--nobrowser -b {self.generation_params.get('batch_size', 32)} " \
                   f"--parallel {self.generation_params.get('max_concurrent_requests', 1)} " \
-                  f"--cont-batching'"
+                  f"--cont-batching"
 
         if self.disable_kv_offload:
             command += " -nkvo"
@@ -262,6 +262,8 @@ def start_llamafile_server(self):
             command += f" --{key} {value}"
             logging.info(f"Additional server argument added: --{key} {value}")
 
+        command += "'"
+
         logging.info(f"Starting llamafile server with command: {command}")
 
         def log_output(pipe, log_func):