Lucyfer1718/Spartan_v3

Full fine-tune vs meta-llama/Llama-3.1-8B-Instruct. Fine-tune: 97% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 1/1, Norms 65/65, MLP 93/96, Attention 124/128, lm_head 1/1

2 derivatives checked

• mradermacher/Spartan_v3-i1-GGUF: Quantization
• mradermacher/Spartan_v3-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse