Jim Lai

grimjim

AI & ML interests

Experimenting primarily with 7B-12B parameter text completion models. Not all models are intended for direct use, but aim for research and/or educational purposes.

Recent Activity

updated a dataset 2 days ago

grimjim/role_meta_info_multilingual

updated a dataset 2 days ago

grimjim/empatheticdialogues

updated a dataset 4 days ago

grimjim/ontario_baby_names_1917-2022

View all activity

Organizations

grimjim's activity

updated 2 datasets 2 days ago

grimjim/role_meta_info_multilingual

Preview • Updated 2 days ago • 52

grimjim/empatheticdialogues

Updated 2 days ago • 94

updated a dataset 4 days ago

grimjim/ontario_baby_names_1917-2022

Viewer • Updated 4 days ago • 9.37k • 35

published a dataset 4 days ago

grimjim/role_meta_info_multilingual

Preview • Updated 2 days ago • 52

updated a model 6 days ago

grimjim/MagTie-v1-12B-GGUF

Text Generation • Updated 6 days ago • 189 • 1

published a model 6 days ago

grimjim/MagTie-v1-12B-GGUF

Text Generation • Updated 6 days ago • 189 • 1

updated a model 7 days ago

grimjim/MagTie-v1-12B

Text Generation • Updated 7 days ago • 4 • 1

published a model 7 days ago

grimjim/MagTie-v1-12B

Text Generation • Updated 7 days ago • 4 • 1

commented on Exploring SLERP Abliteration 8 days ago

To be clear, the parameter t allows for tuning. t=-1 could be used for reversal even. Fractional settings like t=0.7 should better respect model encodings, as is the case for SLERP merger.

published an article 8 days ago

Article