Ana içeriğe geç

nari-labsDia-1.6B-0626

Dia

Uses Dia speaker tags in the text instead of an unrelated generic single-speaker prompt.

Text to speechVoiceHub-nativediaParameters: 1.6BLanguage: enTraining: nativeLicense: Checkpoint-specific

Parameter metadata: Exact learned-parameter total for VoiceHub's audited native primary graph at the registered default selection; separately loaded auxiliary models are excluded.

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: Uses Dia speaker tags in the text instead of an unrelated generic single-speaker prompt.

Inputs and controls: Keep [S1]/[S2] turns in the text when generating dialogue.

from pathlib import Path

from voicehub import AutoModelForTextToSpeech, TTSGenerationConfig

model = AutoModelForTextToSpeech.from_pretrained(
    'nari-labs/Dia-1.6B-0626',
    model_type='dia',
    device="cuda",
    lazy_load=True,
)
output = model.generate(
    '[S1] VoiceHub keeps the dialogue contract explicit. [S2] That makes review easier.',
    generation_config=TTSGenerationConfig(
        seed=42,
        output_file=Path("output.wav"),
    ),
)
print(output.file_path, output.sample_rate, output.metadata)

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

Overview

dia is a VoiceHub text to speech integration. This page is generated from its registry contract. Open the dia Colab notebook.

Property Value
Task Text to speech
Architecture dia
Runtime VoiceHub-native
Languages en
Capabilities text-to-speech, dialogue, safetensors, fine-tuning, voicehub-native, native-runtime
Reusable components dac
Normalized output TTSOutput

Language support

Supported language abbreviations

en

Paper and GitHub

Configuration

Load configuration without constructing the model:

from voicehub import AutoConfig

config = AutoConfig.for_model('dia')
print(config.model_type)
Property Value
Canonical model type dia
Configuration class DiaConfig
Architecture class DiaForTextToSpeech

Processing

Create the registered processor without allocating model weights:

from voicehub import AutoProcessor

processor = AutoProcessor.from_pretrained(
    'nari-labs/Dia-1.6B-0626',
    model_type='dia',
)
print(type(processor).__name__)

Inference

The Usage example returns TTSOutput through AutoModelForTextToSpeech.

Input and output contract

Property Value
Readiness integrated-raw
Data architecture sequence-to-sequence
Sample rate 44,100 Hz
Contract getter get_tts_dataset_spec('dia')
Variant Required fields One of Boundary Other rules
raw-audio text, audio — Source —
processor-ready input_ids, attention_mask, decoder_input_ids, decoder_attention_mask, labels — Prepared —

Encoder text plus teacher-forced acoustic or codec targets. 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 sequence-to-sequence
Recipe single-phase
Default phase codec_language_model
Training checkpoint nari-labs/Dia-1.6B-0626
Native training graph yes
Phase Kind Components Required inputs Loss keys
codec_language_model objective model input_ids, attention_mask, decoder_input_ids, decoder_attention_mask, labels 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 nari-labs/Dia-1.6B-0626
Hugging Face ID nari-labs/Dia-1.6B-0626
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.dia.modeling_dia.DiaForTextToSpeech
Configuration voicehub.models.dia.configuration_dia.DiaConfig
Source provenance voicehub/architectures/dia/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

DiaConfig

View source

DiaConfig(**config_kwargs)

Parameters

  • **config_kwargs — Configuration fields validated by DiaConfig.

Model

DiaForTextToSpeech

View source

AutoModelForTextToSpeech.from_pretrained(
    pretrained_model_name_or_path,
    *,
    model_type='dia',
    config=None,
    **model_kwargs,
)

Parameters

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

spec = get_model_spec('dia')
print(spec.display_name, spec.task.value)
Purpose Public object
Discover get_model_spec('dia')
Load and run AutoModelForTextToSpeech
Configure DiaConfig
Process AutoProcessor
Model implementation DiaForTextToSpeech
Normalized output TTSOutput
Training contract get_training_spec('dia')
Optimization lifecycle available_optimization_passes, apply_optimization_plan, optimization_manifest, restore_optimization_plan

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