MuXodious/gpt-oss-20b-RichardErkhov-heresy-GGUF

Same weights, converted vs MuXodious/gpt-oss-20b-RichardErkhov-heresy. Same weights stored in another dtype (U8/BF16 → MXFP4/BF16), plus 168 new tensors.

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/49, MLP 0/48, Attention 0/192

On the Hub · Downloads on Model Pulse