mradermacher/bigscience-small-testing-GGUF
Same weights, converted vs bigscience/bigscience-small-testing. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Norms 0/10, MLP 0/8, Attention 0/4, Embeddings 0/3
Same weights, converted vs bigscience/bigscience-small-testing. Same weights stored in another dtype (BF16 → F16/F32).
Tensors changed: Norms 0/10, MLP 0/8, Attention 0/4, Embeddings 0/3