ZeroXClem/Llama3.1-Hermes3-SuperNova-8B-L3.1-Purosani-2-8B
Full fine-tune vs djuna/L3.1-Purosani-2-8B. Fine-tune: 97% of the weights changed, across 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 65/65, MLP 93/96, Attention 124/128
3 derivatives checked
- mradermacher/Llama3.1-Hermes3-SuperNova-8B-L3.1-Purosani-2-8B-i1-GGUF: Quantization
- mradermacher/Llama3.1-Hermes3-SuperNova-8B-L3.1-Purosani-2-8B-GGUF: Same weights, converted
- ZeroXClem/L3SAO-Mix-SuperHermes-NovaPurosani-8B: Full fine-tune