Merge pull request #36 from neph1/openai_zone_generation_hotifx

fix openai start zone generation

Author: neph1

tale/llm/llm_utils.py

@@ -516,9 +516,11 @@ def generate_start_zone(self, location_desc: str, story_type: str, story_context
         if self.backend == 'kobold_cpp':
             request_body = self._kobold_generation_prompt(request_body)
             request_body['prompt'] = prompt
+            request_body['max_length'] = 750
         elif self.backend == 'openai':
             request_body['messages'][1]['content'] = prompt
-        request_body['max_length'] = 750
+            request_body['max_tokens'] = 750
+        
         result = self.io_util.synchronous_request(request_body)
         try:
             json_result = json.loads(parse_utils.sanitize_json(result))