--- language: - kk license: other license_name: bilibili-indextts-license license_link: https://huggingface.co/IndexTeam/IndexTTS-2/blob/main/INDEX_MODEL_LICENSE base_model: IndexTeam/IndexTTS-2 base_model_relation: finetune pipeline_tag: text-to-speech tags: - kazakh - indextts - tts datasets: - InflexionLab/ISSAI-KSC2-Structured --- # IndexTTS2 Kazakh Finetune Fine-tuned GPT component of [IndexTTS2](https://github.com/index-tts/index-tts) on Kazakh language data. Special thanks to **Nuraidar Mambetaly** and **Altair Balakhazy** for their collaboration and support throughout this project. ## Training Details - **Base model:** [IndexTeam/IndexTTS-2](https://huggingface.co/IndexTeam/IndexTTS-2) - **Dataset:** [InflexionLab/ISSAI-KSC2-Structured](https://huggingface.co/datasets/InflexionLab/ISSAI-KSC2-Structured) — 4.8GB subset - **Hardware:** Google Colab A100 40GB - **Training:** 3 epochs, 912 steps - **BPE vocab size:** 2000 tokens (Kazakh-specific) Validated on Tesla V100 16GB. Inference uses approximately 10GB VRAM. ## Audio Sample Text: `Менің досымның жақсы және жаман әдеттері бар. Жақсы әдетт ол маған көмектеседі. Ол мені әрқашан қолдайтын болады. Ал жаман әдетті кейде ол менің затымды рұқсатсыз алып кетуі мүмкін. Кейде ол мені ренжітеді. Бірақ артынан менден кешірім сұрайды.` [Listen to sample](https://huggingface.co/AnuarSv/IndexTTS2-Kazakh/resolve/main/preview/sample.mp3) ## Training Curves  ## Limitations The training data contains a single speaker throughout all 4.8GB. Voice cloning does not generalize to other voices at this stage. The model produces Kazakh speech but defaults to the training speaker regardless of the reference audio provided. This will improve with a multi-speaker dataset. Contributions and experiments are welcome. ## Installation Download the base model first: ```python from huggingface_hub import snapshot_download snapshot_download( repo_id="IndexTeam/IndexTTS-2", local_dir="checkpoints" ) ``` Then replace the GPT checkpoint and tokenizer with the Kazakh versions: ```bash # Download this repo hf download AnuarSv/IndexTTS2-Kazakh --local-dir kz_files # Replace cp kz_files/gpt.pth checkpoints/gpt.pth cp kz_files/bpe.model checkpoints/bpe.model cp kz_files/config.yaml checkpoints/config.yaml ``` Clone the training branch and run: ```bash git clone -b training_v2 https://github.com/JarodMica/index-tts.git cd index-tts uv sync --all-extras uv run webui.py --model_dir checkpoints --fp16 ``` ## Config Change The base model uses `number_text_tokens: 12000`. This finetune uses `2000` to match the Kazakh BPE vocabulary. The `config.yaml` in this repo already has this applied. ## What's in This Repo | File | Description | |------|-------------| | `gpt.pth` | Fine-tuned GPT checkpoint (Kazakh) | | `bpe.model` | SentencePiece BPE tokenizer trained on Kazakh text | | `config.yaml` | Model config with `number_text_tokens: 2000` | Files not included (`s2mel.pth`, `feat1.pt`, `feat2.pt`, `qwen0.6bemo4-merge`) are part of the base model and should be downloaded from `IndexTeam/IndexTTS-2`.
