anton-l/wav2vec2-base-ft-keyword-spotting model quantized with Optimum OpenVINO.
Accuracy on eval (baseline) | Accuracy on eval (quantized) |
---|---|
0.9828 | 0.9553 (-0.0274) |
anton-l/wav2vec2-base-ft-keyword-spotting model quantized with Optimum OpenVINO.
Accuracy on eval (baseline) | Accuracy on eval (quantized) |
---|---|
0.9828 | 0.9553 (-0.0274) |