giers10/Heimgeist
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Refactor Whisper and Torch imports for lazy loading in corpus_builder.py

Browse Source
This commit is contained in:
Victor Giers
2026-03-19 23:33:31 +01:00
parent 1e50fcedac
commit 1cacc6e119
1 changed files with 23 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
backend/rag/corpus_builder.py
View File

@@ -46,6 +46,8 @@ Python deps (install as needed):
from __future__ import annotations from __future__ import annotations
import argparse import argparse
import concurrent.futures as cf import concurrent.futures as cf
import importlib
import importlib.util
import json import json
import os import os
import re import re
@@ -244,17 +246,8 @@ try:
except ImportError: except ImportError:
cv2 = None cv2 = None
# Whisper (OpenAI) # Whisper/Torch are imported lazily inside the AV path so the long-lived backend
try: # process does not keep OpenMP-heavy runtimes loaded after a media build.
import whisper
except ImportError:
whisper = None
# Optional: device hinting for Whisper
try:
import torch
except Exception:
torch = None
# Optional language detection (either works) # Optional language detection (either works)
try: try:
@@ -1425,8 +1418,20 @@ def slice_audio(audio_path: Path, out_dir: Path, num_slices: int, overlap_sec: f
_WHISPER_MODEL = None _WHISPER_MODEL = None
def _load_whisper_module():
try:
return importlib.import_module("whisper")
except Exception:
return None
def _load_torch_module():
try:
return importlib.import_module("torch")
except Exception:
return None
def whisper_runtime_error() -> Optional[str]: def whisper_runtime_error() -> Optional[str]:
if whisper is None: if importlib.util.find_spec("whisper") is None:
return ( return (
"Audio/video transcription requires the optional 'openai-whisper' package. " "Audio/video transcription requires the optional 'openai-whisper' package. "
"Install it in backend/.venv, for example: pip install -U openai-whisper" "Install it in backend/.venv, for example: pip install -U openai-whisper"
@@ -1437,7 +1442,8 @@ def _resolve_whisper_device(flag: str) -> Optional[str]:
if flag and flag != "auto": if flag and flag != "auto":
return flag return flag
try: try:
if torch is not None and getattr(torch.cuda, "is_available", lambda: False)(): torch_mod = _load_torch_module()
if torch_mod is not None and getattr(torch_mod.cuda, "is_available", lambda: False)():
return "cuda" return "cuda"
except Exception: except Exception:
pass pass
@@ -1445,15 +1451,16 @@ def _resolve_whisper_device(flag: str) -> Optional[str]:
def _whisper_pool_init(model_name: str, device: Optional[str] = None): def _whisper_pool_init(model_name: str, device: Optional[str] = None):
global _WHISPER_MODEL global _WHISPER_MODEL
if whisper is None: whisper_mod = _load_whisper_module()
if whisper_mod is None:
raise RuntimeError("Whisper package is required (pip install -U openai-whisper)") raise RuntimeError("Whisper package is required (pip install -U openai-whisper)")
warnings.filterwarnings("ignore", message="FP16 is not supported on CPU; using FP32 instead") warnings.filterwarnings("ignore", message="FP16 is not supported on CPU; using FP32 instead")
if device in (None, "auto"): if device in (None, "auto"):
device = _resolve_whisper_device("auto") device = _resolve_whisper_device("auto")
try: try:
_WHISPER_MODEL = whisper.load_model(model_name, device=device) _WHISPER_MODEL = whisper_mod.load_model(model_name, device=device)
except TypeError: except TypeError:
_WHISPER_MODEL = whisper.load_model(model_name) _WHISPER_MODEL = whisper_mod.load_model(model_name)
def _transcribe_slice(task: str, tup: Tuple[Path, int, str]) -> Tuple[int, str]: def _transcribe_slice(task: str, tup: Tuple[Path, int, str]) -> Tuple[int, str]:
global _WHISPER_MODEL global _WHISPER_MODEL