SpectraSuite/FloatLM_3.9B
Text Generation
•
Updated
•
7
•
1
FloatLMs
FP16 LLMs with LLaMa architecture, trained on same 300B tokens as TriLMs.
-
-
SpectraSuite/FloatLM_2.4B
Text Generation • Updated • 5 -
SpectraSuite/FloatLM_1.5B
Text Generation • Updated • 5 -
SpectraSuite/FloatLM_1.1B
Text Generation • Updated • 7 -
SpectraSuite/FloatLM_830M
Text Generation • Updated • 6 -
SpectraSuite/FloatLM_560M
Text Generation • Updated • 5 -
SpectraSuite/FloatLM_390M
Text Generation • Updated • 7 -
SpectraSuite/FloatLM_190M
Text Generation • Updated • 6 -
SpectraSuite/FloatLM_99M
Text Generation • Updated • 16