Hibernates/Hiber-Multi-10B-Instruct
Full fine-tune (fewer layers) vs meta-llama/Llama-3.1-8B-Instruct. meta-llama/Llama-3.1-8B-Instruct matches part of Hibernates/Hiber-Multi-10B-Instruct (289 of its 435 tensors; the rest of Hibernates/Hiber-Multi-10B-Instruct has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Norms 65/65, MLP 96/96, Attention 128/128
2 derivatives checked
- mradermacher/Hiber-Multi-10B-Instruct-i1-GGUF: Quantization
- mradermacher/Hiber-Multi-10B-Instruct-GGUF: Same weights, converted
1 other repos hold byte-for-byte identical weight files.