gorbatjovy/qwen3.8-flash-next-abliterated-NVFP4-plefp8

Same weights, repacked vs windowsxp811203/Qwen3.8-Flash-Next-Abliterated-NVFP4. Same weights in different files (re-saved, re-sharded or new metadata), plus 129 new tensors.

Tensors changed: Embeddings 0/9, MLP 0/1943, Attention 0/608, Norms 0/302, other 0/8, lm_head 0/1

On the Hub · Downloads on Model Pulse