Instructions to use philschmid/gemma-tokenizer-chatml with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use philschmid/gemma-tokenizer-chatml with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("philschmid/gemma-tokenizer-chatml", dtype="auto") - Notebooks
- Google Colab
- Kaggle
replacing existing tokens?
#3
by deleted - opened
Hi @philschmid thanks for sharing this.
Can you share how you updated the vocab with the replacement of the chatml tokens, amongst other values? I'm trying to use some of the <unusedXX> tokens, but my changes are vanishing after saving the tokenizer and then re-loading those changes.