SpectraSuite/FloatLM_3.9B
Text Generation
•
4B
•
Updated
•
1
FloatLMs
FP16 LLMs with LLaMa architecture, trained on same 300B tokens as TriLMs.
-
-
SpectraSuite/FloatLM_2.4B
Text Generation • 2B • Updated • 4 -
SpectraSuite/FloatLM_1.5B
Text Generation • 2B • Updated • 471 -
SpectraSuite/FloatLM_1.1B
Text Generation • 1B • Updated -
SpectraSuite/FloatLM_830M
Text Generation • 0.8B • Updated -
SpectraSuite/FloatLM_560M
Text Generation • 0.6B • Updated -
SpectraSuite/FloatLM_390M
Text Generation • 0.4B • Updated • 554 -
SpectraSuite/FloatLM_190M
Text Generation • 0.2B • Updated • 2 -
SpectraSuite/FloatLM_99M
Text Generation • 99.8M • Updated