---

2026-09-11 12:05:46.241 [syste] [info ] available RAM MiB=1042; available commit MiB=2442
2026-09-11 12:05:46.241 [syste] [info ] fatal exception log=C:\Users\Admin\AppData\Local/NeuroReport/logs/call-20260911-120546.log.crash.log
2026-09-11 12:05:46.242 [trial] [info ] Тестовая версия: обрабатываются только первые 60 секунд каждой записи
2026-09-11 12:05:46.242 [main ] [info ] NeuroReport (Нейроотчёт) 0.4.6 (Windows, COREI3)
2026-09-11 12:05:46.242 [main ] [debug] args: input=C:\TEMP\call.wav model=large-v3 speakers=auto
2026-09-11 12:05:46.249 [main ] [debug] input sha256=3dfe7ffcc6e0ebc7aedf665873e9c9a1434d5d3b709b37e561c29b853a59fdf0
2026-09-11 12:06:09.004 [asr  ] [debug] model ggml-large-v3.bin sha256=64d182b440b9
2026-09-11 12:06:09.017 [audio] [debug] resample 11025Hz -> 16000Hz
2026-09-11 12:06:09.019 [audio] [info ] аудио: 1 кан., 16 кГц, длительность 00:20
2026-09-11 12:06:09.023 [audio] [info ] предобработка: high-pass 80 Гц + нормализация громкости
2026-09-11 12:06:09.025 [audio] [info ] audio stage: 0.018640s
2026-09-11 12:06:09.025 [diar ] [info ] поиск модели определения собеседников: sortformer-v2-q8_0
2026-09-11 12:06:09.029 [diar ] [info ] загрузка движка Sortformer
2026-09-11 12:06:09.029 [diar ] [debug] загрузка библиотеки: C:\Program Files\neuroreport/nemo_speech_asr_c.dll
2026-09-11 12:06:09.061 [diar ] [info ] загрузка модели Sortformer (gpu=-1)
2026-09-11 12:06:09.067 [diar-] [debug] D:\a\neuroreport\neuroreport\build\nemo-speech\source\src\runtime\ggml\backend.cpp:43:<init_backends> Found 1 devices.
2026-09-11 12:06:09.067 [diar-] [debug] D:\a\neuroreport\neuroreport\build\nemo-speech\source\src\runtime\ggml\backend.cpp:101:<init_backends> Using CPU backend
2026-09-11 12:06:09.067 [diar-] [debug] D:\a\neuroreport\neuroreport\build\nemo-speech\source\src\runtime\ggml\backend.cpp:120:<init_backends> Buffer type: CPU_REPACK
2026-09-11 12:06:09.067 [diar-] [debug] D:\a\neuroreport\neuroreport\build\nemo-speech\source\src\runtime\ggml\backend.cpp:120:<init_backends> Buffer type: CPU