0.3.1.2

Added:

• speaker identification (test_local_speaker.py)
• Windows binaries (let me know if they work for you!)
• Model class:
  • int SampleFrequency() const - sample rate (e.g. 16000) that the model expects its input to be in
  • void SetAllowDownsample(bool val) - allow down-sampling in case input sample rate is too high
  • void SetAllowUpsample(bool val) - allow up-sampling in case input sample rate is too low
• microphone example (test_microphone.py)