metadata
tags:
- autotrain
- transformers
- automatic-speech-recognition
base_model: openai/whisper-base
widget:
- src: >-
https://huggingface.co/datasets/mishig/sample_audio/resolve/main/sample1.wav
example_title: Sample Audio 1
- src: >-
https://huggingface.co/datasets/mishig/sample_audio/resolve/main/sample2.wav
example_title: Sample Audio 2
Model Trained Using AutoTrain
- Problem type: Automatic Speech Recognition
Validation Metrics
No validation metrics available
Training Configuration
- Base Model: base_model: openai/whisper-base
- Training Data: autotrain-mfnfj-9bczx\autotrain-data
- Validation Data: None
- Epochs: 3
- Batch Size: 8
- Learning Rate: 3e-05
- Optimizer: adamw_torch
- Scheduler: linear
- Mixed Precision: no
Usage
from transformers import AutoModelForCTC, Wav2Vec2Processor
import torch
import librosa
# Load model and processor
model = AutoModelForCTC.from_pretrained("Mathur08/autotrain-mfnfj-9bczx")
processor = Wav2Vec2Processor.from_pretrained("Mathur08/autotrain-mfnfj-9bczx")
audio, sr = librosa.load("path_to_audio.wav", sr=16000)
inputs = processor(audio, sampling_rate=16000, return_tensors="pt", padding=True)
with torch.no_grad():
logits = model(inputs.input_values).logits
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)