How much GPU memory does local-ai-toolkit need?

About 1.5 GB: 0.3 GB of weights and the runtime's own memory.

What is the cheapest GPU to run local-ai-toolkit?

Right now one RTX A2000 at $0.07 an hour; RTX 5090 at $0.73 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.

Can local-ai-toolkit run on an RTX 4090?

Yes: it needs about 1.5 GB and an RTX 4090 has 24 GB.

Models / local-ai-toolkit

m15dg/local-ai-toolkit

Text generation · served with ONNX Runtime · ↓ 6.4K downloads a month · View on Hugging Face

m15dg/local-ai-toolkit needs about 1.5 GB of GPU memory. It runs on one RTX A2000 from $0.07/h right now; best value: RTX 5090 at $0.73/h.

These ONNX files are an export for running in a browser: GridPup serves them as tensors in and out, not as a chat. For chat, pick the same model in its original form (served with vLLM).

Weights: 300 MB
GPU memory needed: about 1.5 GB
Memory · disk: 8 GB · 30 GB
Licence: other

Versions

Smaller versions (8-bit, 4-bit) need less GPU memory and answer a little less precisely.

full precision6-STT/Audio_Preprocessing_Filters/Studio_Vocal_Isolation_DemucsV4/Studio_Vocal_Isolation_DemucsV4.onnx300 MB · 1.5 GB GPU full precision6-STT/Audio_Preprocessing_Filters/Universal_Silence_Noise_Cutter_Silero/Universal_Silence_Noise_Cutter_Silero.onnx1 MB · 1 GB GPU

Hardware and deploy

Deploying ONNX models from the catalog is coming soon. Today, download the files from Hugging Face and upload them in Inference.

Questions

How much GPU memory does local-ai-toolkit need?: About 1.5 GB: 0.3 GB of weights and the runtime's own memory.
What is the cheapest GPU to run local-ai-toolkit?: Right now one RTX A2000 at $0.07 an hour; RTX 5090 at $0.73 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.
Can local-ai-toolkit run on an RTX 4090?: Yes: it needs about 1.5 GB and an RTX 4090 has 24 GB.

Loading…