Zum Inhalt

Trelistiron

Tiron

Enables Tiron constrained decoding and caps diarized speakers for a meeting recording.

Automatic speech recognitionVoiceHub-nativewhisperParameters: 1.5BLanguages: en, zhTraining: nativeLicense: Checkpoint-specific

Parameter metadata: Exact Safetensors total reported by the Hugging Face model API for the registered default checkpoint, retrieved 2026-08-13.

Use this model Checkpoint
Resources
Upstream GitHub VoiceHub source Open in Colab

Usage

Complete the VoiceHub installation once, then run this repository-authored example. Model pages intentionally contain no package-install command.

This example is maintained against VoiceHub's public API; it is not copied from an upstream demo or package README.

Model-specific path: Enables Tiron constrained decoding and caps diarized speakers for a meeting recording.

Inputs and controls: Choose max_speakers from the recording context rather than treating diarization as unbounded.

from pathlib import Path

from voicehub import AutoModelForSpeechRecognition

AUDIO_FILE = Path("speech.wav")
if not AUDIO_FILE.is_file():
    raise FileNotFoundError(AUDIO_FILE)

model = AutoModelForSpeechRecognition.from_pretrained(
    'Trelis/tiron',
    model_type='asr_tiron',
    device="cuda",
    lazy_load=True,
)
output = model.transcribe(
    AUDIO_FILE,
    return_timestamps=True,
    max_speakers=4,
    constrained_decoding=True,
)
print(output.text)
for segment in output.segments:
    print(segment.start, segment.end, segment.text, segment.confidence)

Use authorized recordings. Verify hardware needs and pin a revision in production.

Overview

asr_tiron is a VoiceHub automatic speech recognition integration. This page is generated from its registry contract. Open the asr_tiron Colab notebook.

Property Value
Task Automatic speech recognition
Architecture whisper
Runtime VoiceHub-native
Languages en, zh
Capabilities automatic-speech-recognition, multilingual, speaker-attribution, timestamps, safetensors, fine-tuning, constrained-decoding, voicehub-native
Reusable components —
Normalized output ASROutput

Language support

Supported language abbreviations

en, zh

Paper and GitHub

Configuration

Load configuration without constructing the model:

from voicehub import AutoConfig

config = AutoConfig.for_model('asr_tiron')
print(config.model_type)
Property Value
Canonical model type asr_tiron
Configuration class TironASRConfig
Architecture class TironForSpeechRecognition

Processing

Create the registered processor without allocating model weights:

from voicehub import AutoProcessor

processor = AutoProcessor.from_pretrained(
    'Trelis/tiron',
    model_type='asr_tiron',
)
print(type(processor).__name__)

Inference

The Usage example returns ASROutput through AutoModelForSpeechRecognition.

Input and output contract

Property Value
Readiness integrated-raw
Data architecture speech-sequence-to-sequence
Sample rate 16,000 Hz
Contract getter get_asr_dataset_spec('asr_tiron')
Variant Required fields One of Boundary Other rules
raw-audio audio text / transcription / transcript Source at most one: text / transcription / transcript
tiron-model-ready input_features, labels — Prepared —

Speaker-aware Whisper fine-tuning with Tiron's inline timestamp grammar. See the data workflow.

Training and optimization

Use available_optimization_passes() to discover reversible public passes. Unsupported runtime or hardware fails closed before mutation.

Training contract

Property Value
Support native
Family speech-sequence-to-sequence
Recipe single-phase
Default phase speech_recognition
Training checkpoint Trelis/tiron
Native training graph yes
Phase Kind Components Required inputs Loss keys
speech_recognition objective model — loss

The integration accepts its declared source or prepared contract directly. Call model.validate_training_support() first, then follow the training workflow.

Checkpoints, provenance, license, and limitations

Property Value
Default checkpoint Trelis/tiron
Hugging Face ID Trelis/tiron
Repository availability verified through the Hugging Face model API on 2026-08-11; pin a revision before production use.
Checkpoint status Registry default; pin an immutable revision for production and reproducible evidence
Optional dependency extra Core package
Hardware and runtime Usage selects cuda; verify checkpoint-specific requirements
Real-checkpoint evidence Release evidence; a registry default alone is not execution evidence
Implementation voicehub.models.asr_tiron.modeling_asr_tiron.TironForSpeechRecognition
Configuration voicehub.models.asr_tiron.configuration_asr_tiron.TironASRConfig
Source provenance voicehub/models/asr_tiron/source/SOURCE.json
License Checkpoint-specific

No VoiceHub-specific license override is registered. Verify the checkpoint and upstream source terms before use.

Confirm the checkpoint revision, access terms, provenance, and license.

Limitations

  • No integration-specific checkpoint limitation is registered. Verify the selected checkpoint revision and its documented runtime requirements.
  • Validate memory, precision, and optional dependencies on the target system.
  • Public optimizations fail closed when the runtime or hardware cannot satisfy their validation contract; an unavailable pass is not reported as applied.
  • Contract tests do not replace the linked released-checkpoint evidence.

Public API

Use the stable configuration, processor, and task-model facades below.

Configuration

TironASRConfig

View source

TironASRConfig(**config_kwargs)

Parameters

  • **config_kwargs — Configuration fields validated by TironASRConfig.

Model

TironForSpeechRecognition

View source

AutoModelForSpeechRecognition.from_pretrained(
    pretrained_model_name_or_path,
    *,
    model_type='asr_tiron',
    config=None,
    **model_kwargs,
)

Parameters

  • pretrained_model_name_or_path — Hub ID or compatible local directory.
  • model_type — Canonical model type; use 'asr_tiron'.
  • config — Optional preloaded TironASRConfig instance.
  • **model_kwargs — Model-specific loading arguments.
from voicehub import get_model_spec

spec = get_model_spec('asr_tiron')
print(spec.display_name, spec.task.value)
Purpose Public object
Discover get_model_spec('asr_tiron')
Load and run AutoModelForSpeechRecognition
Configure TironASRConfig
Process AutoProcessor
Model implementation TironForSpeechRecognition
Normalized output ASROutput
Training contract get_training_spec('asr_tiron')
Optimization lifecycle available_optimization_passes, apply_optimization_plan, optimization_manifest, restore_optimization_plan

See all model guides, inference, and the training matrix.