Hugging Face's logo

shunyalabs
/
pingala-v1-en-verbatim

Automatic Speech Recognition
English
faster-whisper
whisper
speech
audio
asr
ct2
shunyalabs
gated
english
pingala-shunya
Eval Results
Model card Files
xet
Community
1
pingala-v1-en-verbatim / README.md
ayush-shunyalabs's picture
ayush-shunyalabs
Update README.md
ff79992 verified
preview code
|
raw
history blame contribute delete
13.9 kB
metadata
language:
  - en
license: other
library_name: faster-whisper
tags:
  - speech
  - audio
  - automatic-speech-recognition
  - asr
  - ct2
  - faster-whisper
  - shunyalabs
  - gated
  - english
  - pingala-shunya
license_name: pingala-v1-english-verbatim-rail-m
license_link: https://huggingface.co/shunyalabs/pingala-v1-en-verbatim/blob/main/LICENSE.md
metrics:
  - wer
model-index:
  - name: pingala-v1-en-verbatim
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Composite
          type: internal
        metrics:
          - name: Overall WER
            type: wer
            value: 2.94
          - name: Average RTFx
            type: rtfx
            value: 14.61
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: AMI
          type: ami
        metrics:
          - name: WER
            type: wer
            value: 3.52
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Earnings22
          type: earnings22
        metrics:
          - name: WER
            type: wer
            value: 4.36
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: GigaSpeech
          type: gigaspeech
        metrics:
          - name: WER
            type: wer
            value: 4.26
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: LibriSpeech Test Clean
          type: librispeech_asr
          args: test.clean
        metrics:
          - name: WER
            type: wer
            value: 1.84
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: LibriSpeech Test Other
          type: librispeech_asr
          args: test.other
        metrics:
          - name: WER
            type: wer
            value: 2.81
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: SPGISpeech
          type: spgispeech
        metrics:
          - name: WER
            type: wer
            value: 1.13
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: TedLium
          type: tedlium
        metrics:
          - name: WER
            type: wer
            value: 2.14
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: VoxPopuli
          type: voxpopuli
        metrics:
          - name: WER
            type: wer
            value: 3.47
pipeline_tag: automatic-speech-recognition
extra_gated_prompt: >
  ## Access Request for pingala-v1-en-verbatim


  This model is distributed under the Shunya Labs RAIL-M License with use-based
  restrictions.


  By requesting access, you agree to:

  - Use the model only for permitted purposes as defined in the license

  - Not redistribute or create derivative works

  - Comply with all use-based restrictions

  - Use the model responsibly and ethically


  Please provide the following information:
extra_gated_fields:
  Name: text
  Email: text
  Phone Number: text
  Organization: text
  Intended Use: text
  I agree to the Shunya Labs RAIL-M License terms, confirm I will not use this model for prohibited purposes, and understand this model cannot be redistributed: checkbox

Pingala V1 English Verbatim

A high-performance English speech recognition model optimized for verbatim transcription by Shunyalabs, converted to CT2 format for efficient inference with FasterWhisper.

Try the demo at https://www.shunyalabs.ai

License

This model is distributed under the Shunya Labs RAIL-M License, which includes specific use-based restrictions and commercial licensing requirements.

License Summary

  • Free Use: Up to 10,000 hours of audio transcription per calendar month
  • Distribution: Model cannot be redistributed to third parties
  • Derivatives: Creation of derivative works is not permitted
  • Attribution: Required when outputs are made public or shared

Key Restrictions

The license prohibits use for discrimination, military applications, disinformation, privacy violations, unauthorized medical advice, and other harmful purposes. Please refer to the complete LICENSE file for detailed terms and conditions.

For inquiries, contact: [email protected]

Model Overview

Pingala V1 English Verbatim is a state-of-the-art automatic speech recognition (ASR) model that delivers exceptional accuracy across diverse English audio domains. The model has been optimized for production use with CTranslate2 and FasterWhisper, providing both high accuracy and fast inference speeds.

This model is specifically designed for verbatim transcription, ensuring accurate word-for-word capture of spoken English content across various domains including meetings, earnings calls, broadcast media, and educational content.

Performance Benchmarks

image/png

OpenASR Leaderboard Results

The model has been extensively evaluated on the OpenASR leaderboard across multiple English datasets, demonstrating superior performance compared to larger open-source models:

Dataset WER (%) RTFx
AMI Test 3.52 18.38
Earnings22 Test 4.36 25.67
GigaSpeech Test 4.26 24.62
LibriSpeech Test Clean 1.84 29.20
LibriSpeech Test Other 2.81 25.01
SPGISpeech Test 1.13 11.67
TedLium Test 2.14 11.03
VoxPopuli Test 3.47 31.81

Composite Results

  • Overall WER: 2.94%
  • Average RTFx: 14.61

RTFx (Real-Time Factor) indicates inference speed relative to audio duration. Higher values mean faster processing.

Comparative Performance

Pingala V1 significantly outperforms larger open-source models on 8 common speech benchmarks:

Model AMI Earnings22 GigaSpeech LS Clean LS Other SPGISpeech TedLium Voxpopuli Avg WER
nvidia/canary-qwen-2.5b 10.19 10.45 9.43 1.61 3.10 1.90 2.71 5.66 5.63
ibm-granite/granite-speech-8b 9.12 9.53 10.33 1.42 2.99 3.86 3.50 6.00 5.85
nvidia/parakeet-tdt-0.6b-v2 11.16 11.15 9.74 1.69 3.19 2.17 3.38 5.95 6.05
microsoft/Phi-4-multimodal 11.45 10.50 9.77 1.67 3.82 3.11 2.89 5.93 6.14
nvidia/canary-1b-flash 13.11 12.77 9.85 1.48 2.87 1.95 3.12 5.63 6.35
shunyalabs/pingala-v1-en-verbatim 3.52 4.36 4.26 1.84 2.81 1.13 2.14 3.47 2.94

Authentication with Hugging Face Hub

This model require authentication with Hugging Face Hub. Here's how to set up and use your Hugging Face token.

Getting Your Hugging Face Token

  1. Create a Hugging Face Account: Go to huggingface.co and sign up
  2. Generate a Token:

Setting Up Authentication

Method 1: Environment Variable (Recommended) 

# Set your token as an environment variable
export HUGGINGFACE_HUB_TOKEN="hf_your_token_here"

# Or add to your ~/.bashrc or ~/.zshrc for persistence
echo 'export HUGGINGFACE_HUB_TOKEN="hf_your_token_here"' >> ~/.bashrc
source ~/.bashrc

Method 2: Hugging Face CLI Login 

# Install Hugging Face CLI if not already installed
pip install huggingface_hub

# Login using CLI
huggingface-cli login
# Enter your token when prompted

Method 3: Programmatic Authentication 

from huggingface_hub import login

# Login programmatically
login(token="hf_your_token_here")

Installation

Basic Installation 

pip install pingala-shunya

Usage

Quick Start 

from pingala_shunya import PingalaTranscriber

# Initialize with default Shunya Labs model and auto-detected backend
transcriber = PingalaTranscriber()

# Simple transcription
segments = transcriber.transcribe_file_simple("audio.wav")

for segment in segments:
    print(f"[{segment.start:.2f}s -> {segment.end:.2f}s] {segment.text}")

Backend Selection 

from pingala_shunya import PingalaTranscriber

# Explicitly choose backends with Shunya Labs model
transcriber_ct2 = PingalaTranscriber(model_name="shunyalabs/pingala-v1-en-verbatim", backend="ct2")

# Auto-detection (recommended)
transcriber_auto = PingalaTranscriber()  # Uses default Shunya Labs model with ct2

Model Details

  • Architecture: ctranslate2-based model optimized for verbatim English transcription
  • Format: CTranslate2 compatible for efficient inference
  • Framework: FasterWhisper optimized
  • Language: English (en)
  • Sampling Rate: 16 kHz
  • Model Size: Production-optimized for deployment
  • Optimization: Real-time inference capable with GPU acceleration

Key Features

  • Exceptional Accuracy: Achieves 2.94% WER across diverse English test sets
  • Real-Time Performance: Average RTFx of 14.61 enables real-time applications
  • Verbatim Transcription: Optimized for accurate, word-for-word transcription
  • Production Ready: CTranslate2 optimization ensures efficient deployment
  • Multi-Domain Excellence: Superior performance across conversational, broadcast, and read English speech
  • Voice Activity Detection: Built-in VAD for better handling of silence

Performance Optimization Tips

  • GPU Acceleration: Use device="cuda" for significantly faster inference
  • Precision: Set compute_type="float16" for optimal speed on modern GPUs
  • Threading: Adjust cpu_threads and num_workers based on your hardware configuration
  • VAD Filtering: Enable vad_filter=True for improved performance on long audio files
  • Language Specification: Set language="en" for English audio to improve accuracy and speed
  • Beam Size: Use beam_size=5 for best accuracy, reduce for faster inference
  • Batch Processing: Process multiple files with a single model instance for efficiency

Use Cases

The model excels in various English speech recognition scenarios:

  • Meeting Transcription: High accuracy on conversational English speech (AMI: 3.52% WER)
  • Financial Communications: Specialized performance on earnings calls and financial content (Earnings22: 4.36% WER)
  • Broadcast Media: Excellent results on news, podcasts, and media content
  • Educational Content: Optimized for lectures, presentations, and educational material transcription
  • Customer Support: Accurate transcription of support calls and customer interactions
  • Legal Documentation: Professional-grade accuracy for legal proceedings and depositions
  • Medical Transcription: High-quality transcription for medical consultations and documentation

Support and Contact

For technical support, licensing inquiries, or commercial partnerships:

Acknowledgments

Built with FasterWhisper and CTranslate2 for optimal inference performance. Special thanks to the open-source community for providing the foundational tools that make this model possible.

Version History

  • v1.0: Initial release with state-of-the-art performance across multiple English domains
    • Optimized for verbatim transcription with 2.94% composite WER
    • CTranslate2 format for efficient inference
    • Production-ready deployment capabilities

This model is provided under the Shunya Labs RAIL-M License. Please ensure compliance with all license terms before use.