қосымша
/usr/local/lib/python3.12/dist-packages/torch/cuda/init.py:65: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report this to the maintainers of the package that installed pynvml for you.
import pynvml # type: ignore[import]
WARNING:nv_one_logger.api.config:OneLogger: Setting error_handling_strategy to DISABLE_QUIETLY_AND_REPORT_METRIC_ERROR for rank (rank=0) with OneLogger disabled. To override: explicitly set error_handling_strategy parameter.
WARNING:nv_one_logger.training_telemetry.api.training_telemetry_provider:No exporters were provided. This means that no telemetry data will be collected.
WARNING:torchao.kernel.intmm:Warning: Detected no triton, on systems without Triton certain kernels will not work
[NeMo W 2026-03-24 21:21:24 nemo_logging:364] /usr/local/lib/python3.12/dist-packages/pydub/utils.py:300: SyntaxWarning: invalid escape sequence '('
m = re.match('(sup?) (([0-9]{1,2}) bit)$', token)
[NeMo W 2026-03-24 21:21:24 nemo_logging:364] /usr/local/lib/python3.12/dist-packages/pydub/utils.py:301: SyntaxWarning: invalid escape sequence '('
m2 = re.match('(sup?)( (default))?$', token)
[NeMo W 2026-03-24 21:21:24 nemo_logging:364] /usr/local/lib/python3.12/dist-packages/pydub/utils.py:310: SyntaxWarning: invalid escape sequence '('
elif re.match('(flt)p?( (default))?$', token):
[NeMo W 2026-03-24 21:21:24 nemo_logging:364] /usr/local/lib/python3.12/dist-packages/pydub/utils.py:314: SyntaxWarning: invalid escape sequence '('
elif re.match('(dbl)p?( (default))?$', token):
[NeMo W 2026-03-24 21:21:25 modelPT:188] If you intend to do training or fine-tuning, please call the ModelPT.setup_training_data() method and provide a valid configuration file to setup the train data loader.
Train config :
manifest_filepath: ksc/ksc-train.json
sample_rate: 16000
labels:
- ' '
- а
- б
- в
- г
- д
- е
- ж
- з
- и
- й
- к
- л
- м
- н
- о
- п
- р
- с
- т
- у
- ф
- х
- ц
- ч
- ш
- щ
- ъ
- ы
- ь
- э
- ю
- я
- і
- ғ
- қ
- ң
- ү
- ұ
- һ
- ә
- ө
batch_size: 64
trim_silence: true
max_duration: 16.7
shuffle: true
is_tarred: false
tarred_audio_filepaths: /asr_set_1.2/train/train_{0..1023}.tar
num_workers: 8
normalize_transcripts: false
pin_memory: true
[NeMo W 2026-03-24 21:21:25 modelPT:195] If you intend to do validation, please call the ModelPT.setup_validation_data() or ModelPT.setup_multiple_validation_data() method and provide a valid configuration file to setup the validation data loader(s).
Validation config :
manifest_filepath: ksc/ksc-dev.json
sample_rate: 16000
labels:
- ' '
- а
- б
- в
- г
- д
- е
- ж
- з
- и
- й
- к
- л
- м
- н
- о
- п
- р
- с
- т
- у
- ф
- х
- ц
- ч
- ш
- щ
- ъ
- ы
- ь
- э
- ю
- я
- і
- ғ
- қ
- ң
- ү
- ұ
- һ
- ә
- ө
batch_size: 4
shuffle: false
normalize_transcripts: false
num_workers: 4
pin_memory: true
trim_silence: true
[NeMo I 2026-03-24 21:21:27 save_restore_connector:285] Model EncDecCTCModel was successfully restored from /content/stt_kz_quartznet15x5.nemo.
[NeMo W 2026-03-24 21:21:27 conv_asr:73] Turned off 170 masked convolutions
[NeMo I 2026-03-24 21:21:52 exportable:135] Successfully exported EncDecCTCModel to stt_kz_quartznet.onnx
✅ Стандартты ONNX моделі сақталды: stt_kz_quartznet.onnx
WARNING:root:Please consider to run pre-processing before quantization. Refer to example: https://github.com/microsoft/onnxruntime-inference-examples/blob/main/quantization/image_classification/cpu/ReadMe.md
🔥 Квантталған (жылдам) модель дайын: stt_kz_quartznet_int8.onnx