Hugging Face's logo

ByteDance
/
Ouro-2.6B-Thinking

Text Generation
Transformers
Safetensors
ouro
looped-language-model
reasoning
recurrent-depth
thinking
chain-of-thought
conversational
custom_code
Model card Files
xet
Community
8

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

#6
by sirorezka - opened
base: refs/heads/main
←
from: refs/pr/6
Discussion Files changed
+28
-2
sirorezka

Rope_type='default' excluded from ROPE_INIT_FUNCTIONS in transfomers >=5.0. Thus each model must implement it's own default method.

 rope_type='default' excluded from ROPE_INIT_FUNCTIONS in transfomers >=5.0c49969e8
ridger changed pull request status to merged

Sign up or log in to comment