Instructions to use SJ-Ray/Re-Punctuate with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use SJ-Ray/Re-Punctuate with Transformers:
# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("SJ-Ray/Re-Punctuate") model = AutoModelForSeq2SeqLM.from_pretrained("SJ-Ray/Re-Punctuate") - Notebooks
- Google Colab
- Kaggle
error loading the model in python 3.12
#3
by Lou1sM - opened
So it seems this is a TF checkpoint, and we have to pass the argument from_tf=True when loading the model. We also can't use python 3.12 and have to install sentencepiece for the tokenizer and have to install tf-keras for the loading the model. This produces errors though:tensorflow.python.framework.errors_impl.UnknownError: Exception encountered when calling layer 'layer_norm' (type TFT5LayerNorm).