Hugging Face's logo

androiddrew
/
google-speech-embedding-pytorch

Feature Extraction
Safetensors
PyTorch
naptrap
audio
speech
keyword-spotting
wake-word
speech-embedding
Model card Files
xet
 Community
google-speech-embedding-pytorch / conversion /README.md
androiddrew's picture
androiddrew
Upload folder using huggingface_hub
ad7a3dd verified
preview code
|
Raw History Blame Contribute Delete
1.17 kB

Conversion record

How model.safetensors was produced from Google's original TensorFlow artifact. These scripts are kept as a record. They are not needed to use the model.

  1. inspect_google_tensorflow.py: lists the original SavedModel's graph and variables (TensorFlow 2.x, CPU).
  2. google_tensorflow_reference.py: extracts the 96 trained variables, the Hann window and the mel matrix exactly as the graph computes them, and runs the original model on reference clips. The reference clips came from the training project's private feature cache. That step cannot be reproduced from this repository, but its outputs are summarised in ../qualification/.
  3. convert_google_weights.py: maps the variables to the PyTorch module in naptrap.google (HWIO→OIHW kernels, and the first kernel's inference-time zero-mean subtraction applied ahead of time). It writes model.pt (sha256 20c27209…b246).
  4. to_safetensors.py: repackages model.pt as model.safetensors. Every tensor is checked to be bit-identical.

The ../qualification/ reports were measured against model.pt. Its tensors are identical to model.safetensors.