Modules:
| Name | Description |
|---|---|
base | Base worker class for vLLM-Omni with process-scoped GPU memory accounting. |
gpu_ar_model_runner | AR GPU Model Runner for vLLM-Omni. |
gpu_ar_worker | |
gpu_generation_model_runner | Code2Wav GPU Model Runner for vLLM-Omni. |
gpu_generation_worker | |
gpu_memory_utils | NVML-based per-process GPU memory utilities. |
gpu_model_runner | |
mixins | |
omni_connector_model_runner_mixin | Unified data-plane communication mixin for Model Runners. |
payload_span | Helpers for explicit thinker decode span metadata. |