train_chatterbox is an open-weight model from Bibek. Its published files total 3.2 GB.
A modular, highly efficient infrastructure for fine-tuning both Chatterbox TTS (Standart) and Chatterbox Turbo models with your own dataset and generating high-quality speech synthesis.
Parameters—
Context—
Weights3.2 GB
License—
AccessOpen weights
Monthly Downloads—
Model Card
A modular, highly efficient infrastructure for fine-tuning both Chatterbox TTS (Standart) and Chatterbox Turbo models with your own dataset and generating high-quality speech synthesis. This kit is specially designed to support new languages and voices by intelligently extending the model's vocabulary. With the newly added LoRA (Low-Rank Adaptation) support, you can now train high-quality voices faster and with significantly less VRAM. This repository allows you to choose your training strategy using the islora flag in src/config.py. What is it? LoRA (Low-Rank Adaptation) freezes the massive base model and only trains tiny adapter layers alongside the new language embeddings. What is it?…
Excerpt from the card by Bibek.
Identity and Version
- Repository
- ibibek/train_chatterbox
- Publisher
- Bibek
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 94336a6f005183634dab90a75b482980942a756b
- First published
- 2026-09-28
- Last updated
- 2026-09-28
Files and Weights
85 files, 3.2 GB in total. The weights are 4 files totalling 3.2 GB in pt, safetensors.
Weights4 files · 3.2 GB
Configuration64 files · 390.6 KB
Tokenizer3 files · 85.3 KB
Documentation3 files · 43.5 KB
Other9 files · 13.7 MB
Repository2 files · 6.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| pretrained_models/conds.pt | Weights | 107.4 KB | 6552d7056883 |
| pretrained_models/s3gen.safetensors | Weights | 1.1 GB | 4a46190f3dcc |
| pretrained_models/t3_cfg.safetensors | Weights | 2.1 GB | 89fd813802e2 |
| pretrained_models/ve.safetensors | Weights | 5.7 MB | f0921cab452f |
| inference.py | Configuration | 8.2 KB | — |
| merge_lora.py | Configuration | 2.2 KB | — |
| prepare_dataset.py | Configuration | 3.6 KB | — |
| setup.py | Configuration | 7.0 KB | — |
| src/__init__.py | Configuration | — | |
| src/chatterbox_/__init__.py | Configuration | 127 B | — |
| src/chatterbox_/models/__init__.py | Configuration | — | |
| src/chatterbox_/models/s3gen/__init__.py | Configuration | 68 B | — |
| src/chatterbox_/models/s3gen/configs.py | Configuration | 220 B | — |
| src/chatterbox_/models/s3gen/const.py | Configuration | 34 B | — |
| src/chatterbox_/models/s3gen/decoder.py | Configuration | 13.2 KB | — |
| src/chatterbox_/models/s3gen/f0_predictor.py | Configuration | 2.0 KB | — |
| src/chatterbox_/models/s3gen/flow.py | Configuration | 8.2 KB | — |
| src/chatterbox_/models/s3gen/flow_matching.py | Configuration | 10.8 KB | — |
| src/chatterbox_/models/s3gen/hifigan.py | Configuration | 17.7 KB | — |
| src/chatterbox_/models/s3gen/matcha/decoder.py | Configuration | 14.4 KB | — |
| src/chatterbox_/models/s3gen/matcha/flow_matching.py | Configuration | 4.6 KB | — |
| src/chatterbox_/models/s3gen/matcha/text_encoder.py | Configuration | 15.0 KB | — |
| src/chatterbox_/models/s3gen/matcha/transformer.py | Configuration | 13.2 KB | — |
| src/chatterbox_/models/s3gen/s3gen.py | Configuration | 13.2 KB | — |
| src/chatterbox_/models/s3gen/transformer/__init__.py | Configuration | — | |
| src/chatterbox_/models/s3gen/transformer/activation.py | Configuration | 3.1 KB | — |
| src/chatterbox_/models/s3gen/transformer/attention.py | Configuration | 14.4 KB | — |
| src/chatterbox_/models/s3gen/transformer/convolution.py | Configuration | 5.2 KB | — |
| src/chatterbox_/models/s3gen/transformer/embedding.py | Configuration | 11.4 KB | — |
| src/chatterbox_/models/s3gen/transformer/encoder_layer.py | Configuration | 9.6 KB | — |
| src/chatterbox_/models/s3gen/transformer/positionwise_feed_forward.py | Configuration | 4.2 KB | — |
| src/chatterbox_/models/s3gen/transformer/subsampling.py | Configuration | 12.7 KB | — |
| src/chatterbox_/models/s3gen/transformer/upsample_encoder.py | Configuration | 13.7 KB | — |
| src/chatterbox_/models/s3gen/utils/class_utils.py | Configuration | 2.3 KB | — |
| src/chatterbox_/models/s3gen/utils/intmeanflow.py | Configuration | 951 B | — |
| src/chatterbox_/models/s3gen/utils/mask.py | Configuration | 7.7 KB | — |
| src/chatterbox_/models/s3gen/utils/mel.py | Configuration | 2.6 KB | — |
| src/chatterbox_/models/s3gen/xvector.py | Configuration | 14.1 KB | — |
| src/chatterbox_/models/s3tokenizer/__init__.py | Configuration | 595 B | — |
| src/chatterbox_/models/t3/__init__.py | Configuration | 19 B | — |
| src/chatterbox_/models/t3/inference/alignment_stream_analyzer.py | Configuration | 9.0 KB | — |
| src/chatterbox_/models/t3/inference/t3_hf_backend.py | Configuration | 4.3 KB | — |
| src/chatterbox_/models/t3/llama_configs.py | Configuration | 1.8 KB | — |
| src/chatterbox_/models/t3/modules/cond_enc.py | Configuration | 3.3 KB | — |
| src/chatterbox_/models/t3/modules/learned_pos_emb.py | Configuration | 1.1 KB | — |
| src/chatterbox_/models/t3/modules/perceiver.py | Configuration | 7.7 KB | — |
| src/chatterbox_/models/t3/modules/t3_config.py | Configuration | 1.2 KB | — |
| src/chatterbox_/models/t3/t3.py | Configuration | 19.8 KB | — |
| src/chatterbox_/models/tokenizers/__init__.py | Configuration | 48 B | — |
| src/chatterbox_/models/utils.py | Configuration | 148 B | — |
| src/chatterbox_/models/voice_encoder/__init__.py | Configuration | 56 B | — |
| src/chatterbox_/models/voice_encoder/config.py | Configuration | 406 B | — |
| src/chatterbox_/models/voice_encoder/melspec.py | Configuration | 1.9 KB | — |
| src/chatterbox_/models/voice_encoder/voice_encoder.py | Configuration | 10.6 KB | — |
| src/chatterbox_/mtl_tts.py | Configuration | 9.6 KB | — |
| src/chatterbox_/tts.py | Configuration | 9.6 KB | — |
| src/chatterbox_/tts_turbo.py | Configuration | 10.0 KB | — |
| src/chatterbox_/vc.py | Configuration | 3.5 KB | — |
| src/config.py | Configuration | 3.1 KB | — |
| src/dataset.py | Configuration | 4.6 KB | — |
| src/inference_callback.py | Configuration | 10.4 KB | — |
| src/model.py | Configuration | 6.2 KB | — |
| src/preprocess_file_based.py | Configuration | 5.0 KB | — |
| src/preprocess_json.py | Configuration | 5.1 KB | — |
| src/preprocess_ljspeech.py | Configuration | 4.4 KB | — |
| src/utils.py | Configuration | 4.3 KB | — |
| test_tts.py | Configuration | 9.8 KB | — |
| train.py | Configuration | 7.3 KB | — |
| LICENSE | Documentation | 11.4 KB | — |
| Memory.md | Documentation | 15.3 KB | — |
| README.md | Documentation | 16.8 KB | — |
| Inference.ipynb | Other | 6.7 MB | — |
| checkpoint1000_test.wav | Other | 65.3 KB | — |
| output.wav | Other | 157.5 KB | 173e70b4927d |
| requirements.txt | Other | 859 B | — |
| speaker_reference/reference.wav | Other | 3.2 MB | 8925a031b396 |
| test_Newari_dialect.wav | Other | 1.8 MB | 43c661523c34 |
| test_output.wav | Other | 495.2 KB | 2b62e5981fc8 |
| test_output2.wav | Other | 838.6 KB | 1a9252ff087a |
| test_output_fixed.wav | Other | 495.2 KB | 87a73c1adcc3 |
| .gitattributes | Repository | 1.9 KB | — |
| .gitignore | Repository | 4.7 KB | — |
| pretrained_models/tokenizer.json | Tokenizer | 70.0 KB | — |
| src/chatterbox_/models/s3tokenizer/s3tokenizer.py | Tokenizer | 5.2 KB | — |
| src/chatterbox_/models/tokenizers/tokenizer.py | Tokenizer | 10.2 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 3.2 GB
Download from Bibek
Released by Bibek through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 3.2 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.