toandev/Spark-X2.5-4B-Uncensored

Partial fine-tune vs XHToken/Spark-X2.5-4B. Partial change: 31% of the weights changed (attention, embeddings, mlp), in 26 of 36 layers.

Tensors changed: Embeddings 1/1, Norms 0/73, MLP 26/108, Attention 26/108

1 derivatives checked

• toandev/Spark-X2.5-4B-Uncensored-GGUF: Quantization

On the Hub · Downloads on Model Pulse