bunnycore/LLama-3.1-8B-HyperNova-abliteration
Partial fine-tune vs bunnycore/LLama-3.1-8B-HyperNova. Partial change: 28% of the weights changed (attention, mlp, norms), in 30 of 32 layers.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 7/65, MLP 30/96, Attention 27/128
6 derivatives checked
- QuantFactory/Llama-3.1-8B-OmniMatrix-v2-GGUF: Full fine-tune
- mradermacher/LLama-3.1-8B-HyperNova-abliteration-i1-GGUF: Quantization
- mradermacher/LLama-3.1-8B-HyperNova-abliteration-GGUF: Same weights, converted
- QuantFactory/LLama-3.1-8B-Matrix-GGUF: Full fine-tune
- bunnycore/LLama-3.1-8B-Matrix: Full fine-tune
- bunnycore/Llama-3.1-8B-OmniMatrix-v2: Full fine-tune