πŸŽ™οΈ Whisper Base Karakalpak ASR

This model is a fine-tuned version of openai/whisper-base for Automatic Speech Recognition (ASR) in the Karakalpak language.


πŸ‘€ Founder

Quyashbek Allanazarov


🀝 Acknowledgements

  • πŸŽ“ New Uzbekistan University β€” for knowledge, research environment, and academic direction
  • 🏦 Xalq Banki AI Lab β€” for providing GPU resources and supporting the technical direction of the project

πŸ“Š Model Performance

Evaluation was performed on a held-out test set.

Metric Score
WER (Word Error Rate) 38.59%
CER (Character Error Rate) 9.85%

Test Details

  • Total samples: 504
  • Sampling rate: 16 kHz
  • Task: Transcribe

πŸ§ͺ Inference Example

import torch
from transformers import WhisperForConditionalGeneration, WhisperProcessor

model_id = "your-username/whisper-base-karakalpak"
processor = WhisperProcessor.from_pretrained(model_id)
model = WhisperForConditionalGeneration.from_pretrained(model_id)

# speech, sr = librosa.load("audio.wav", sr=16000)
# inputs = processor(speech, sampling_rate=16000, return_tensors="pt")
# generated_ids = model.generate(inputs.input_features)
# transcription = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
Downloads last month
18
Safetensors
Model size
72.6M params
Tensor type
F32
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for Quyashbek/whisper-base-karakalpak

Finetuned
(686)
this model

Collection including Quyashbek/whisper-base-karakalpak