mradermacher/OpenELM-1_1B-Instruct-GGUF
Same weights, converted vs apple/OpenELM-1_1B-Instruct. Same weights stored in another dtype (BF16 → F16/F32), plus 56 new tensors.
Tensors changed: Norms 0/113, Attention 0/56, MLP 0/28, Embeddings 0/1
Same weights, converted vs apple/OpenELM-1_1B-Instruct. Same weights stored in another dtype (BF16 → F16/F32), plus 56 new tensors.
Tensors changed: Norms 0/113, Attention 0/56, MLP 0/28, Embeddings 0/1