[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

+7
View File
@@ -241,6 +241,13 @@ def modify_stage_config(
# get_chunk_config and get_batch_token_config). int(time.time()) would collide
# and the later write would overwrite the earlier YAML on disk.
# Keep generated configs outside the repo and delete them when pytest exits.
# The generated YAML lives in a temp dir, so a relative ``base_config`` (e.g.
# ``../qwen3_tts.yaml`` in ``deploy/edge/``) must be anchored to the source
# file's directory before writing (same rule as ``_materialize_ci_overlay``).
base_config = config.get("base_config")
if isinstance(base_config, str) and base_config and not os.path.isabs(base_config):
config["base_config"] = str((path.parent / base_config).resolve())
output_fd, output_path = tempfile.mkstemp(prefix=f"{path.stem}_", suffix=".yaml")
atexit.register(Path(output_path).unlink, missing_ok=True)