ByteDance
/

Ouro-2.6B-Thinking

Text Generation

looped-language-model

recurrent-depth

chain-of-thought

Model card Files Files and versions

Resources

View closed (1)

Fix UniversalTransformerCache.get_mask_sizes for batched generation

#8 opened 2 days ago by

Fix bos/eos token IDs + add enable_thinking to chat template

#7 opened 3 days ago by

rope_type='default' excluded from ROPE_INIT_FUNCTIONS in transfomers >=5.0

#6 opened 3 days ago by

Added 'pad_token_id'.

#5 opened 3 days ago by

Updated ids for bos_id, eos_id

#4 opened 3 days ago by

Upd tokenizer bos_token, eos_token, pad_token

#3 opened 3 days ago by

OuroRotaryEmbedding ROPE_INIT_FUNCTIONS does not contains a 'default' key

#2 opened about 2 months ago by