deep-principle
Hydrogen-1.0
text generationmitAccess request needed26K downloadsHugging Face
No server in stock right now has the 1684 GB this model needs for inference.
GPU memory it needs (estimate)
Rule of thumb from the model's size. Real needs depend on batch size, context length and framework.
Inference (full precision)
1684 GB
Inference (8-bit)
842 GB
Inference (4-bit)
421 GB
LoRA fine-tuning
1824 GB
QLoRA (8-bit base)
912 GB
Full fine-tuning
6735 GB