Text Generation
Transformers
PyTorch
Safetensors
English
gpt_neox
causal-lm
pythia
text-generation-inference

PAD tokens is present in `tokenizer.json` but not in `special_tokens_map.json`.

#4
by sadra-barikbin - opened

Hi there!
PAD token is present in tokenizer's added tokens in tokenizer.json as a special token with content "<|padding|>", while it's not in special_tokens_map.json and also in tokenizer_config.json which results in tokenizer.pad_token being None.

Sign up or log in to comment