BPE Tokenizer fitted on a custom corpus, with digit separation, byte fallback and other features from LlamaTokenizer.
Only fitted on 1,000,000 samples.
Unable to determine this model's library. Check the
docs
.
BPE Tokenizer fitted on a custom corpus, with digit separation, byte fallback and other features from LlamaTokenizer.
Only fitted on 1,000,000 samples.