Automatic Speech Recognition
Transformers
ONNX
Safetensors
English
multilingual
whisper
windyword
english
Instructions to use WindyWord/listen-windy-edge with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use WindyWord/listen-windy-edge with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="WindyWord/listen-windy-edge")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("WindyWord/listen-windy-edge", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Commit History
26 days ago
-
Model card: correct licence metadata, add LICENSE, link windytranslate.com
Verified signatureMetadata/documentation only: weights, configs, tokenizers and file paths are unchanged.393b093
-
Add missing preprocessor_config.json to the CT2 folder
Verified signatureWithout it faster-whisper defaults to an 80-mel feature extractor. These weights report n_mels=128, so every transcription failed with 'Invalid input features shape: expected (1,128,3000)'. feature_size is read from the CT2 model itself and verified by transcribing a real clip before upload. Weights untouched. Dr. J (Opus 5), clinic log doctor-logs/2026-08-05_stt-fleet-qa.md16d23c3
-
correct licence to mit (base is MIT upstream)
Verified signature1ebc5d9
-
Refresh README — uniform WindyWord template with WER tier + dialect notes
Verified signature9566640 -
Refresh README — uniform WindyWord template with WER tier + dialect notes
Verified signature210d6f1
-
Add model card
Verified signature7b50f8b -
Add onnx-int8 variant
Verified signature0dc9b2c -
Add onnx variant
Verified signature7e8f51c -
Add ct2-int8 variant
Verified signaturedb9f440 -
Add safetensors variant
Verified signaturee6dc7e0 -
initial commit
Verified signature8b64cde