mradermacher/pythia-1b-13000-steps-polish-GGUF
Same weights, converted vs Lajonbot/pythia-1b-13000-steps-polish. Same weights stored in another dtype (F32/U8 → F16/F32).
Tensors changed: Embeddings 0/2, Norms 0/66, Attention 0/32, MLP 0/64
Same weights, converted vs Lajonbot/pythia-1b-13000-steps-polish. Same weights stored in another dtype (F32/U8 → F16/F32).
Tensors changed: Embeddings 0/2, Norms 0/66, Attention 0/32, MLP 0/64