IndexTTS2 MLX FP16
This is an fp16 artifact export of IndexTeam/IndexTTS-2 for Apple Silicon runtime evaluation.
Fonte do modelo
- Tipo
- Other
- Tamanho do arquivo
- 4,16 KB–1,62 GB
- Modelo base
- IndexTeam/IndexTTS-2
- Versão
- vRevision 208b3d6
- Atualizado
- 31 de out. de 2026
- Downloads
- 135
- Fonte
- Hugging Face
Descrição da fonte
IndexTTS2-MLX-fp16
This is an fp16 artifact export of IndexTeam/IndexTTS-2 for Apple Silicon runtime evaluation.
Runtime support is not implied by this model card. The bundle preserves upstream
key names and is intended for a speech-swift Swift/MLX port.
Runtime and Documentation
speech-swiftruntime repo: https://github.com/soniqo/speech-swiftspeech-modelsexporter repo: https://github.com/soniqo/speech-models- Voice cloning docs: https://soniqo.audio/voice-cloning
- CLI docs: https://soniqo.audio/cli
- Benchmark article: https://soniqo.audio/blog/voice-cloning-benchmarks
Model
| Field | Value |
|---|---|
| Display name | IndexTTS2 |
| Source | IndexTeam/IndexTTS-2 |
| Source revision | 740dcaff396282ffb241903d150ac011cd4b1ede |
| Publish repo | aufklarer/IndexTTS2-MLX-fp16 |
| Parameter count | 1.5B-class |
| Sample rate | 24000 Hz |
| Voice conditioning | reference audio plus optional emotion reference/control path |
| Streaming | needs runtime validation |
| License posture | check upstream/commercial terms before product use |
| Runtime status | artifact-export; Swift native inference not implemented |
Files
soniqo_manifest.json- export manifest with source, runtime status, and file metadataconfig.json- upstream or generated root config for runtime metadata*.safetensors- fp16-converted model weights with upstream key names preservedaux/- auxiliary model configs and fp16-converted weights required by this runtime, when applicable
Auxiliary Models
This expanded export includes auxiliary assets that upstream IndexTTS2 downloads at runtime:
| Component | Source | Purpose |
|---|---|---|
| w2v-BERT 2.0 | facebook/w2v-bert-2.0 | SeamlessM4T features and semantic hidden states for reference audio |
| MaskGCT semantic codec | amphion/MaskGCT | Quantizes reference semantic features and maps generated codes to embeddings |
| CAMPPlus | funasr/campplus | Extracts the 192-d global style vector from the speaker reference |
| BigVGAN | nvidia/bigvgan_v2_22khz_80band_256x | Decodes generated 80-band mel spectrograms to waveform |
Notes
- Public upstream weights include GPT/S2Mel checkpoints plus a Qwen emotion merge directory.
- The expanded export also bundles the upstream auxiliary models required by IndexTTS2 v2 synthesis.
- This exporter converts PyTorch pickle checkpoints to safetensors and preserves subdirectory layout.
- Native Swift inference still needs the graph ports for w2v-BERT, MaskGCT semantic codec, GPT/S2Mel, CAMPPlus, and BigVGAN.
Usage
These artifacts are for runtime implementation and evaluation. They are not a drop-in Transformers checkpoint.
git clone https://huggingface.co/aufklarer/IndexTTS2-MLX-fp16
Source
- Upstream model: https://huggingface.co/IndexTeam/IndexTTS-2
- Converted bundle: https://huggingface.co/aufklarer/IndexTTS2-MLX-fp16
Fontes
1 fonte
Fontes
1 fonteVerificado 31 de out.
Download files
Repository files
File inventory only; completeness is unverified.
- LICENSE.txt
- LICENSE_ZH.txt
- aux/bigvgan/bigvgan_generator.safetensors
- aux/bigvgan/config.json
- aux/campplus/campplus_cn_common.safetensors
- aux/campplus/configuration.json
- aux/maskgct/config.json
- aux/maskgct/semantic_codec/model.safetensors
- aux/w2v-bert-2.0/config.json
- aux/w2v-bert-2.0/model.safetensors
- aux/w2v-bert-2.0/preprocessor_config.json
- bpe.model
- config.json
- feat1.safetensors
- feat2.safetensors
- gpt.safetensors
- qwen0.6bemo4-merge/added_tokens.json
- qwen0.6bemo4-merge/config.json
- qwen0.6bemo4-merge/generation_config.json
- qwen0.6bemo4-merge/merges.txt
- qwen0.6bemo4-merge/model.safetensors
- qwen0.6bemo4-merge/special_tokens_map.json
- qwen0.6bemo4-merge/tokenizer.json
- qwen0.6bemo4-merge/tokenizer_config.json
- qwen0.6bemo4-merge/vocab.json
- s2mel.safetensors
- soniqo_manifest.json
- wav2vec2bert_stats.safetensors
Artefatos de modelo
10 artefatos
Artefatos de modelo
10 artefatosaux/bigvgan/bigvgan_generator.safetensors
safetensors · FP16 (unverified) · 214 MB · SHA-256 62d63d69265f…8569 · Hugging Face
Download fileaux/campplus/campplus_cn_common.safetensors
safetensors · 13,3 MB · SHA-256 eb29858cf9eb…adc4 · Hugging Face
Download fileaux/maskgct/semantic_codec/model.safetensors
safetensors · 84,5 MB · SHA-256 a665396a04de…eaaa · Hugging Face
Download fileaux/w2v-bert-2.0/model.safetensors
safetensors · 1,08 GB · SHA-256 e56eab7a7c9c…b6eb · Hugging Face
Download fileqwen0.6bemo4-merge/model.safetensors
safetensors · 1,11 GB · SHA-256 74a54db600ca…d72c · Hugging Face
Download filewav2vec2bert_stats.safetensors
safetensors · 4,16 KB · SHA-256 2ec34a778c9d…5903 · Hugging Face
Download fileTrechos de fonte
3 trechos
Trechos de fonte
3 trechos--- license: other language: - zh - en tags: - text-to-speech - voice-cloning - indextts2 - mlx - fp16 base_model: IndexTeam/IndexTTS-2 library_name: mlx pipeline_tag: text-to-speech --- # IndexTTS2-MLX-fp16 This is an fp16 artifact export of [IndexTeam/IndexTTS-2](https://huggingface.co/IndexTeam/IndexTTS-2) for Apple Silicon runtime evaluation. Runtime support is not implied by this model card. The bundle preserves upstream key names and is intended for a `speech-swift` Swift/MLX port. ## Runtime and Documentation - `speech-swift` runtime repo: https://github.com/soniqo/speech-swift - `speech-models` exporter repo: https://github.com/soniqo/speech-models - Voice cloning docs: https://soniqo.audio/voice-cloning - CLI docs: https://soniqo.audio/cli - Benchmark article: https://soniqo.audio/blog/voice-cloning-benchmarks ## Model | Field | Value | |---|---| | Display name | IndexTTS2 | | Source | `IndexTeam/IndexTTS-2` | | Source revision | `740dcaff396282ffb241903d150ac011cd4b1ede` | | Publish repo | `aufklarer/IndexTTS2-MLX-fp16` | | Parameter count | 1.5B-class | | Sample rate | 24000 Hz | | Voice conditioning | reference audio plus optional emotion reference/control path | | Streaming | needs runtime validation | | License posture | check upstream/commercial terms before product use | | Runtime status | artifact-export; Swift native inference not implemented | ## Files - `soniqo_manifest.json` - export manifest with source, runtime status, and file metadata - `config.json` - upstream or generated root config for runtime metadata - `*.safetensors` - fp16-converted model weights with upstream key names preserved - `aux/` - auxiliary model configs and fp16-converted weights required by this runtime, when applicable ## Auxiliary Models This expanded export includes auxiliary assets that upstream IndexTTS2 downloads at runtime: | Component | Source | Purpose | |---|---|---| | w2v-BERT 2.0 | `facebook/w2v-bert-2.0` | SeamlessM4T features and semantic hidden states for reference audio | | MaskGCT semantic codec | `amphion/MaskGCT` | Quantizes reference semantic features and maps generated codes to embeddings | | CAMPPlus | `funasr/campplus` | Extracts the 192-d global style vector from the speaker reference | | BigVGAN | `nvidia/bigvgan_v2_22khz_80band_256x` | Decodes generated 80-band mel spectrograms to waveform | ## Notes - Public upstream wei...
Source context: 135 downloads · 0 likes · Pipeline text-to-speech · Library mlx · Repo aufklarer/IndexTTS2-MLX-fp16
Source context: 1167 downloads · 0 likes · Pipeline text-to-speech · Library mlx · Repo aufklarer/IndexTTS2-MLX-fp16
