Sakalti/ultiima-14B-v0.4

Full fine-tune vs sometimesanotion/Lamarck-14B-v0.7. Fine-tune: 100% of the weights changed, across 48 of 48 layers.

Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 97/97, MLP 144/144, Attention 192/336

3 derivatives checked

• mradermacher/ultiima-14B-v0.4-i1-GGUF: Quantization
• mradermacher/ultiima-14B-v0.4-GGUF: Quantization
• Sakalti/SJT-24B-Alpha: Partial fine-tune (fewer layers)

On the Hub · Downloads on Model Pulse