netcat420/MFANN-liminerity-slerp-V4

Partial fine-tune vs liminerity/Phigments12. Partial change: 22% of the weights changed (attention), in 31 of 32 layers.

Tensors changed: Norms 0/66, MLP 0/128, Attention 94/256, lm_head 0/2, Embeddings 0/1

1 derivatives checked

• mradermacher/MFANN-liminerity-slerp-V4-i1-GGUF: Quantization

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse