These language model checkpoints are trained at the 360M and 1.3Bn parameter scales for up to 50Bn tokens on the Pile corpus, for research purposes.
HazyResearch
community
AI & ML interests
None defined yet.
Collections
2
models
16
hazyresearch/M2-BERT-8k-Retrieval-Encoder-V1
Fill-Mask
•
Updated
•
36
•
1
hazyresearch/M2-BERT-2k-Retrieval-Encoder-V1
Fill-Mask
•
Updated
•
27
•
1
hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1
Fill-Mask
•
Updated
•
52
•
1
hazyresearch/M2-BERT-128-Retrieval-Encoder-V1
Fill-Mask
•
Updated
•
19
•
2
hazyresearch/based-1b-50b
Updated
•
177
•
1
hazyresearch/attn-1b-50bn
Updated
•
189
hazyresearch/mamba-360m
Updated
•
172
hazyresearch/attn-360m
Updated
•
189
hazyresearch/based-360m
Updated
•
306
•
3
hazyresearch/based-1b
Updated
•
181
•
6
datasets
9
hazyresearch/based-squad
Viewer
•
Updated
•
2.34k
hazyresearch/based-swde
Viewer
•
Updated
•
88
•
2
hazyresearch/based-fda
Viewer
•
Updated
•
1.57k
•
3
hazyresearch/LoCoV1-Queries
Viewer
•
Updated
•
325
hazyresearch/LoCoV1-Documents
Viewer
•
Updated
•
356
•
1
hazyresearch/based-swde-deprecated
Viewer
•
Updated
hazyresearch/fda
Viewer
•
Updated
•
1
•
1
hazyresearch/evaporate
Updated
•
4
•
5
hazyresearch/based-swde-old
Viewer
•
Updated
•
2