AyoubChLin/lfm2.5-8b-saudi-dialect

Partial fine-tune vs LiquidAI/LFM2.5-8B-A1B. Partial change: 4% of the weights changed (attention, other), in 24 of 24 layers.

Tensors changed: MLP 0/2162, Embeddings 0/2, other 36/54, Norms 0/60, Attention 24/24

1 derivatives checked

• mradermacher/LFM2.5-8b-saudi-dialect-GGUF: Partial fine-tune

On the Hub · Downloads on Model Pulse