HuggingFaceBio/Carbon-8B

Support return_assistant_tokens_mask on slow HybridDNATokenizer for completion-only SFT (apply_chat_template token-boundary mask override)

Remove base fast-tokenizer files so loads without trust_remote_code raise instead of silently using base BPE

load chat_template in from_pretrained (vLLM apply_chat_template fix)

load chat_template in from_pretrained (vLLM apply_chat_template fix)

load chat_template in from_pretrained (vLLM apply_chat_template fix)

load chat_template in from_pretrained (vLLM apply_chat_template fix)

fix convert_tokens_to_string byte-level detok (vLLM)

Update README.md

Delete chat_template.jinja

tokenizer: expose .vocab property for fast-tokenizer-style callers (#1)

Update README.md

Update README.md

Update README.md

Update README.md

fix: correct bos/eos/pad token IDs in generation_config (eos=pad=151643, bos=null)

Update README.md

Upload figures/carbon-8b-banner.png

update tata and syn

tokenizer: sync with Carbon-500M/3B (warnings, add_special_tokens pass-through, tolist in decode)

Update README.md

Update README.md

Initial upload from HuggingFaceBio/carbon-8B-longctx-32k-from-1T-decay@step-227500 with new README

initial commit