Guilherme34/Samantha-v6.5

Full fine-tune vs Guilherme34/Samantha-v6. Fine-tune: 93% of the weights changed, across 32 of 32 layers.

Tensors changed: lm_head 1/1, Embeddings 0/1, Norms 4/65, MLP 96/96, Attention 128/128

2 derivatives checked

• mradermacher/Samantha-v6.5-i1-GGUF: Quantization
• mradermacher/Samantha-v6.5-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse