mradermacher/cosmos3-reason-32b-GGUF
Quantization (fewer layers) vs TensorAuto/cosmos3-reason-32b. mradermacher/cosmos3-reason-32b-GGUF matches part of TensorAuto/cosmos3-reason-32b (707 of its 1058 tensors; the rest of TensorAuto/cosmos3-reason-32b has no counterpart). On the shared part: mradermacher/cosmos3-reason-32b-GGUF is a GGUF quantization (Q8_0) of the same weights: 257 tensors match exactly and 450 within quantization error.
Tensors changed: Embeddings 0/1, Norms 0/257, MLP 0/192, Attention 0/256, lm_head 0/1