NexesMess/Llama_3.1_70b_Luminoblivion_CW_V1.01

Full fine-tune vs NobodySpecial/Llama-3.1-70B-Instruct-Lorablated-Creative-Writer. Fine-tune: 97% of the weights changed, across 80 of 80 layers.

Tensors changed: Embeddings 0/1, Norms 1/161, MLP 240/240, Attention 320/320, lm_head 0/1

3 derivatives checked

• mradermacher/Llama_3.1_70b_Luminoblivion_CW_V1.01-GGUF: Quantization
• NexesMess/Llama_3.x_70b_Triads_V7: Full fine-tune
• Nexesenex/Llama_3.x_70b_Triads_V5: Full fine-tune

On the Hub · Downloads on Model Pulse