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
File size: 872 Bytes
f2fbf5f 47dfda4 f2fbf5f 47dfda4 f2fbf5f 47dfda4 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 | ---
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.
|