jhu-clsp/mmBERT-base
Fill-Mask
•
Updated
•
31.8k
•
•
154
mmBERT: a modern multilingual encoder
mmBERT is trained on 3T tokens from over 1800 languages, showing SoTA scores on benchmarks and exceptional low-resource performance
-
jhu-clsp/mmBERT-small
Fill-Mask • Updated • 7.31k • • 51Note Intermediate checkpoints for continued pre-training (MosiacML Composer format)
jhu-clsp/mmBERT-checkpoints
Updated • 2Note Pre-training Data
-
jhu-clsp/mmBERT-pretrain-p1-fineweb2-langs
Updated • 1.32k • 4 -
jhu-clsp/mmBERT-pretrain-p2-fineweb2-remaining
Updated • 3.22k -
jhu-clsp/mmBERT-pretrain-p3-others
Updated • 20k -
jhu-clsp/mmBERT-midtraining-data
Updated • 23.7k • 1
jhu-clsp/mmBERT-decay-data
Updated • 6.1k • 2Note Randomized data for training (not recommended unless you are using the same data mix)
-
orionweller/mmBERT-data-decay-all
Updated • 341 -
orionweller/mmBERT-data-decay-cont
Updated • 211 -
orionweller/mmBERT-data-decay-eng
Updated • 345 -
orionweller/mmBERT-data-midtraining
Updated • 387 -
orionweller/mmBERT-pretraining-data-chunk0
Preview • Updated • 3.95k -
orionweller/mmBERT-pretraining-data-chunk1
Preview • Updated • 1.71k -
orionweller/mmBERT-pretraining-data-chunk2
Preview • Updated • 475 -
orionweller/mmBERT-pretraining-data-chunk3
Preview • Updated • 1.24k