鈿狅笍 WORK IN PROGRESS

This repository is currently a work in progress and is undergoing active calibration/testing for deployment on Blackwell hardware architectures.

Current Status

  • Quantization Format: NVFP4 (compressed-tensors layout)
  • Status: Testing inference engine runtime and tensor mapping alignment.
Downloads last month
514
Safetensors
Model size
34B params
Tensor type
F32
BF16
F8_E4M3
U8
Inference Providers NEW
This model isn't deployed by any Inference Provider. 馃檵 Ask for provider support