Text generation · served with ONNX Runtime GenAI · ↓ 0 downloads a month · View on Hugging Face
keystone-systems/Qwen2.5-1.5B-Instruct-onnx-dml-int4 needs about 3.3 GB of GPU memory. It runs on one GTX 1660 SUPER from $0.07/h right now; best value: RTX PRO 4000 at $0.24/h.
Weights
1.4 GB
GPU memory needed
about 3.3 GB
Memory · disk
8 GB · 33 GB
Licence
apache-2.0
Versions of Qwen2.5 1.5B
The same model packaged differently: a quantized version needs less GPU memory and runs on a cheaper machine.
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 Qwen2.5-1.5B-Instruct-onnx-dml-int4 need?
About 3.3 GB: 1.36 GB of weights and the runtime's own memory.
What is the cheapest GPU to run Qwen2.5-1.5B-Instruct-onnx-dml-int4?
Right now one GTX 1660 SUPER at $0.07 an hour; RTX PRO 4000 at $0.24 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.
Can Qwen2.5-1.5B-Instruct-onnx-dml-int4 run on an RTX 4090?
Yes: it needs about 3.3 GB and an RTX 4090 has 24 GB.
Prices are per hour for the whole machine and its disk, right now. Speed is Vast's DLPerf score of the machine, relative to the cheapest. Best value: the most speed per dollar. Only machines with a Vast reliability above 95% that download at 400 Mbit/s or more are offered: the machine fetches the model itself.