Instructions to use adyoi/indigo.tf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use adyoi/indigo.tf with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://adyoi/indigo.tf") - Notebooks
- Google Colab
- Kaggle
| language: | |
| - id | |
| license: mit | |
| tags: | |
| - text-generation | |
| - from-scratch | |
| - gpt | |
| - transformer | |
| - tensorflow | |
| - keras | |
| # Indigo-TF | |
| Versi **TensorFlow/Keras** dari proyek [Indigo](https://huggingface.co/adyoi/indigo) — model bahasa GPT kecil dari nol. Repo utama (PyTorch): `D:\Documents\GitHub\Indigo`. | |
| Format checkpoint identik (`.safetensors` + `_meta.json`, penanda `backend: tensorflow`), sehingga tidak bisa tertukar dengan checkpoint PyTorch. | |
| ## Pakai | |
| ```bash | |
| pip install -r requirements.txt | |
| python train.py --data data/sample.txt --steps 2000 | |
| python generate.py --prompt "Indigo" | |
| ``` | |
| Fitur setara repo utama: pembersihan teks, split validasi per-file, best-checkpoint, resume (`--init-from`), tokenizer `char`/`bpe`, cosine LR + warmup, pilihan `--device`. | |
| Catatan: GPU TensorFlow di Windows native tidak didukung sejak TF 2.11 — gunakan WSL2/Colab/Linux. | |