Qwen2.5 models with the tokenizers padded with extra unused special tokens to match the model embedding size. Workaround for various bugs.
Columba Labs
community
AI & ML interests
Birb make bots go brrr
Recent Activity
View all activity
Collections
1
models
36
Columbidae/gemma-3-12b-it-text
Updated
Columbidae/gemma-3-27b-half-full
Image-Text-to-Text
•
Updated
•
63
Columbidae/gemma-3-27b-half
Updated
•
139
Columbidae/Qwen2.5-72B
Text Generation
•
Updated
•
24
Columbidae/Qwen2.5-72B-Instruct
Text Generation
•
Updated
•
10
Columbidae/gemma-3-27b-special
Image-Text-to-Text
•
Updated
•
76
Columbidae/qwen-21b-base-trained
Updated
•
5
Columbidae/test-pruned-8
Updated
•
1
Columbidae/gemma-2-24b-pruned
Updated
•
4
Columbidae/gemma-2-27b
Text Generation
•
Updated
•
8
datasets
0
None public yet