1× RTX 308010 GB VRAM
Nosana network, hosts worldwide
Details
Stock and price checked 9 min ago.
Self-service: you start it yourself, with no order or sales contact.
nvidia-3080
Full price
€0.09
per hour, whole instance, 1 GPU
medgemma-1.5-4b-it
image text to textotherAccess request needed206.7K downloadsHugging Face
Cheapest GPU in stock that fits
A single GPU with at least 9.6 GB for inference.
View offer
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)
9.6 GB
Inference (8-bit)
4.8 GB
Inference (4-bit)
2.4 GB
LoRA fine-tuning
10 GB
QLoRA (8-bit base)
5.2 GB
Full fine-tuning
38 GB