[Edge][WP1] Add edge deploy profile and deploy_profile selector

* vllm_omni/deploy/edge/qwen3_tts.yaml: single-device Qwen3-TTS profile with a
  ramped codec chunk schedule (2->4->8->16->25 frames) that removes the ~80 ms
  playback stall of the default 1->25 schedule, decode batch 1, max_num_seqs 4.
* deploy_profile="edge" (Omni/AsyncOmni kwarg, --deploy-profile CLI, headless
  serve) resolves vllm_omni/deploy/<profile>/<model_type>.yaml; mutually
  exclusive with deploy_config; explicit profiles never fall back silently.
* tests/helpers/stage_config.py: modify_stage_config anchors a relative
  base_config before writing the temp YAML so overlay-derived configs resolve.
* async_omni_engine.py also carries the init-timeline hooks used by WP2.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This commit is contained in:
tzhouamandClaude Fable 5.1 committed 2026-09-09 10:27:33 +08:00
1 parent b70207f57f
commit b22dcc59f7
7 files changed
+278

No files matched your search

+11
View File
@@ -275,6 +275,13 @@ class OmniServeCommand(CLISubcommand):
default=None,
help="Path to a deploy config YAML (new format with stages/engine_args).",
)
omni_config_group.add_argument(
"--deploy-profile",
type=str,
default=None,
help="Named deploy profile resolved to vllm_omni/deploy/<profile>/<model_type>.yaml "
"(e.g. 'edge'). Mutually exclusive with --deploy-config.",
)
omni_config_group.add_argument(
"--strategy-config",
type=str,
@@ -1002,6 +1009,10 @@ def run_headless(args: TrackingNamespace) -> None:
# standard launches resolve to the same per-stage device layout.
stage_overrides = parse_stage_overrides(args_dict.get("stage_overrides"))
from vllm_omni.entrypoints.utils import apply_deploy_profile
apply_deploy_profile(model, args_dict)
config_path, stage_configs, _ = load_and_resolve_stage_configs(
model,
args_dict,