Modelpublic

minjaechoi/qwen3next-80b-a3b-2p02bit-r22

sourceHugging Faceupdated 7d agoView on Hugging Face

0likes272downloads

Model Card

Qwen3-Next-80B-A3B-Thinking — 2.0161-bit routed experts (r22)

Internal research checkpoint. Routed experts average 2.0161 bits; every other weight is BF16. Weights are stored dequantized in BF16 tensors and load with stock transformers / vLLM.


Base model	Qwen/Qwen3-Next-80B-A3B-Thinking
Routed-expert average	2.0161 bits
Internal ID	r22

License follows the base model.