Vykos
/

Barbados

Text Generation

mixture-of-experts

looped-transformer

Model card Files Files and versions

Barbados / bundle /source /models

262 kB

Ctrl+K

Ctrl+K

1 contributor

History: 1 commit

Vykos's picture

Release Barbados looped MoE research prototype

d5147ba verified 15 days ago

baselines
Release Barbados looped MoE research prototype 15 days ago
__init__.py

0 Bytes
Release Barbados looped MoE research prototype 15 days ago
adam_atan2.py

4.14 kB
Release Barbados looped MoE research prototype 15 days ago
attn_res.py

3.58 kB
Release Barbados looped MoE research prototype 15 days ago
common.py

864 Bytes
Release Barbados looped MoE research prototype 15 days ago
dual_axis_carry.py

45.8 kB
Release Barbados looped MoE research prototype 15 days ago
flash_attention_prefixlm_v2.py

11.6 kB
Release Barbados looped MoE research prototype 15 days ago
layers.py

10.9 kB
Release Barbados looped MoE research prototype 15 days ago
lm_head.py

19 kB
Release Barbados looped MoE research prototype 15 days ago
melt_memory.py

8.26 kB
Release Barbados looped MoE research prototype 15 days ago
moe.py

32.3 kB
Release Barbados looped MoE research prototype 15 days ago
prefixlm_attention.py

22.6 kB
Release Barbados looped MoE research prototype 15 days ago
transformer.py

24.3 kB
Release Barbados looped MoE research prototype 15 days ago