What this repository contains
Four-bit bitsandbytes weights with BF16 quantization storage for FSDP compatibility. The model card explains component precision choices and the shard-by-shard conversion process. Inherits the base model license.
BITSANDBYTES / MODEL
Nemotron 120B · NF4
A pre-quantized NF4 build of NVIDIA Nemotron 3 Super.
THE PROJECT, IN PRACTICE
What you can do with it.
Choose the right runtime
Model formats are not interchangeable. Match the published artifact to the runtime, architecture support, and available memory documented in its card.
Inspect provenance
The model card identifies the base model, derivation, and license terms. A quantized build remains a derivative of its upstream model.
FROM IDEA TO USE
How it works.
- 01
Inspect
Read the model or dataset card and its provenance.
- 02
Prepare
Check the runtime or schema and your intended use.
- 03
Explore
Open the repository and follow its current usage instructions.
greghavens/nemotron-3-super-120b-bnb-4bitUse the source page for current files, requirements, and usage examples.
TAKE IT FOR A SPIN
Start here.
Four-bit bitsandbytes weights with BF16 quantization storage for FSDP compatibility. The model card explains component precision choices and the shard-by-shard conversion process. Inherits the base model license.
STATUS & REQUIREMENTS
Published artifact. Review its model card for runtime support and inherited license terms.