mradermacher/pythia-160m-deduped-GGUF
Quantization vs EleutherAI/pythia-160m-deduped. mradermacher/pythia-160m-deduped-GGUF is a GGUF quantization (Q8_0) of the same weights: 86 tensors match exactly and 38 within quantization error.
Tensors changed: Embeddings 0/2, Norms 0/50, Attention 0/24, MLP 0/48