mradermacher/cosmos-reason2-8b-abliterated-GGUF
Same weights, converted (fewer layers) vs RowanBird779/cosmos-reason2-8b-abliterated. mradermacher/cosmos-reason2-8b-abliterated-GGUF matches part of RowanBird779/cosmos-reason2-8b-abliterated (399 of its 750 tensors; the rest of RowanBird779/cosmos-reason2-8b-abliterated has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Embeddings 0/1, Norms 0/145, MLP 0/108, Attention 0/144, lm_head 0/1