Skip to content

Update number of mel bins for whisper model #25675

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ajindal1 merged 3 commits into from
Aug 7, 2025
Merged

Update number of mel bins for whisper model #25675

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c5a339e
get number of mel bins from config as it is differs based on whisper …
ajindal1 Aug 6, 2025
432d8e4
lint fix
ajindal1 Aug 6, 2025
9f82fc0
remove textwrap
ajindal1 Aug 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 9 additions & 20 deletions onnxruntime/python/tools/transformers/models/whisper/whisper_helper.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,6 @@
import json
import logging
import os
import textwrap
from pathlib import Path

import numpy as np
Expand Down Expand Up @@ -93,16 +92,10 @@ def save_processing(
if separate_encoder_and_decoder_init:
return

audio_processor_json = textwrap.dedent("""\
{
audio_processor_cfg = {
"feature_extraction": {
"sequence": [
{
"operation": {
"name": "audio_decoder",
"type": "AudioDecoder"
}
},
{"operation": {"name": "audio_decoder", "type": "AudioDecoder"}},
{
"operation": {
"name": "STFT",
Expand Down Expand Up @@ -511,27 +504,23 @@ def save_processing(
0.000986635684967041,
0.0005550682544708252,
0.0002467334270477295,
0.0000616908073425293
]
}
0.0000616908073425293,
],
},
}
},
{
"operation": {
"name": "log_mel_spectrogram",
"type": "LogMelSpectrum",
"attrs": {
"chunk_size": 30,
"hop_length": 160,
"n_fft": 400,
"n_mel": 80
}
"attrs": {"chunk_size": 30, "hop_length": 160, "n_fft": 400, "n_mel": config.num_mel_bins},
}
}
},
]
}
}
""")
audio_processor_json = json.dumps(audio_processor_cfg, indent=4)

with open(os.path.join(output_dir, "audio_processor_config.json"), "w") as f:
f.write(audio_processor_json)

Expand Down
Loading