thaloco/aman
1
Fork 0
Code Pull requests Releases Packages Activity Actions

Simplify editor cleanup and keep live ASR metadata
Some checks are pending
ci / test-and-build (push) Waiting to run
Details

Browse source 
Keep the daemon path on the full ASR result so word timings and detected language survive into the editor pipeline instead of falling back to a plain transcript string.

Add PipelineEngine.run_asr_result(), have aman call it when live ASR data is available, and cover the word-aware alignment behavior in the daemon tests.

Collapse the llama cleanup flow to a single JSON-shaped completion while leaving the legacy pass1/pass2 parameters in place as compatibility no-ops.

Validated with PYTHONPATH=src python3 -m unittest tests.test_aiprocess tests.test_aman.
This commit is contained in:
Thales Maciel 2026-03-12 13:24:36 -03:00
parent 8c1f7c1e13
commit fa91f313c4
No known key found for this signature in database
GPG key ID: 33112E6833C34679
5 changed files with 166 additions and 84 deletions
Download patch file Download diff file Expand all files Collapse all files

10
src/engine/pipeline.py
View file

@ -53,12 +53,20 @@ class PipelineEngine:
raise RuntimeError("asr stage is not configured")
started = time.perf_counter()
asr_result = self._asr_stage.transcribe(audio)
return self.run_asr_result(asr_result, started_at=started)
def run_asr_result(
self,
asr_result: AsrResult,
*,
started_at: float | None = None,
) -> PipelineResult:
return self._run_transcript_core(
asr_result.raw_text,
language=asr_result.language,
asr_result=asr_result,
words=asr_result.words,
started_at=started,
started_at=time.perf_counter() if started_at is None else started_at,
)
def run_transcript(self, transcript: str, *, language: str = "auto") -> PipelineResult: