(engine_dir)
| 838 | |
| 839 | |
| 840 | def supports_inflight_batching(engine_dir): |
| 841 | config_path = Path(engine_dir) / "config.json" |
| 842 | json_config = GptJsonConfig.parse_file(config_path) |
| 843 | model_config = json_config.model_config |
| 844 | return model_config.supports_inflight_batching |
| 845 | |
| 846 | |
| 847 | class QuantModeWrapper: |