Model not work, no feedback

[MohsenNz]

Hi, I have currently created a model by rustpotter-cli with about 180 wake sample and about, 17000 non-wake sample (the same count used in test-train).
I got acc: 99.4% or some-thing like that, in the training phase.

After that, when I tested model through rustpotter-cli spont and rustpotter-cli test, I got no feedback even against the files used for training, e.g.:

λ> ./rustpotter-cli test trained-small2.rpw train/\[hey_echo\]accents-3_148.wav
Testing file train/[hey_echo]accents-3_148.wav against model trained-small2.rpw!
Loading wakeword file: trained-small2.rpw

With debug enabled:

λ> ./rustpotter-cli test trained-small2.rpw train/\[hey_echo\]accents-3_148.wav -d
Testing file train/[hey_echo]accents-3_148.wav against model trained-small2.rpw!
Rustpotter config:
RustpotterConfig { fmt: AudioFmt { sample_rate: 24000, sample_format: I16, channels: 1, endianness: Little }, detector: DetectorConfig { avg_threshold: 0.2, threshold: 0.5, min_scores: 10, eager: false, score_ref: 0.22, band_size: 5, score_mode: Max, vad_mode: None, record_path: None }, filters: FiltersConfig { gain_normalizer: GainNormalizationConfig { enabled: false, gain_ref: None, min_gain: 0.1, max_gain: 1.0 }, band_pass: BandPassConfig { enabled: false, low_cutoff: 80.0, high_cutoff: 400.0 } } }
Loading wakeword file: trained-small2.rpw

After I clone the code and add some debug, it is the output:

λ> ./rustpotter-cli-latest test trained-small2.rpw train/\[hey_echo\]accents-3_148.wav
Testing file train/[hey_echo]accents-3_148.wav against model trained-small2.rpw!
Loading wakeword file: trained-small2.rpw
wav_specs.sample_format=Int
chunk_size=720
detection=None
partial_detection=None
detection=None
partial_detection=None
...
detection=None
partial_detection=None

I need help.