cosmicoptima/computer-10

Full fine-tune vs cosmicoptima/computer-9. Fine-tune: 99% of the weights changed, across 80 of 80 layers.

Tensors changed: lm_head 1/1, Embeddings 0/1, Norms 94/161, MLP 240/240, Attention 320/320

3 derivatives checked

• bartowski/cosmicoptima_computer-10-GGUF: Same weights, converted
• multimodalart/computer-10-fp8: Quantization
• neuroscr/computer-10-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse