Markus Krimmel
Markus28
AI & ML interests
None yet
Organizations
None yet
Markus28's activity
feat: selective activation checkpointing
#16 opened 3 months ago
by
Markus28
Porting v2 models to flash attention
#15 opened 3 months ago
by
Markus28
feat: updated activation checkpointing
#14 opened 3 months ago
by
Markus28
feat: Allow LoRA to be merged into weights
#12 opened 3 months ago
by
Markus28
fix: remove cleaving
#13 opened 3 months ago
by
Markus28
feat: cleave off layers from encoder
#11 opened 3 months ago
by
Markus28
clean up embeddings.py
#7 opened 3 months ago
by
bwang0911
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63491dc83d8dc83a55cb749c/IoqJrOIaEnYO_S7si4KGp.jpeg)
Positional Interpolation
#14 opened 4 months ago
by
Markus28
support-multiple-task-ids
#5 opened 4 months ago
by
michael-guenther
![](https://cdn-avatars.huggingface.co/v1/production/uploads/6476ff2699a5ce743ccea3fc/zmFmF8tXXDaAGcl8RYiRr.jpeg)
Global CLS attention
#13 opened 4 months ago
by
Markus28
feat: implement task type embeddings
#1 opened 4 months ago
by
Markus28
Use attention dropout during training
1
#1 opened 4 months ago
by
Markus28
Use attention dropout during training
2
#10 opened 4 months ago
by
Markus28
Fix sorting heuristic
1
#3 opened 8 months ago
by
Markus28