wangzhang/LFM2-24B-A2B-abliterated
Partial fine-tune vs LiquidAI/LFM2-24B-A2B. Partial change: 13% of the weights changed (attention, mlp, other), in 38 of 40 layers.
Tensors changed: MLP 377/2768, Embeddings 0/2, other 10/90, Norms 0/100, Attention 7/40
2 derivatives checked
- mradermacher/LFM2-24B-A2B-abliterated-i1-GGUF: Quantization
- mradermacher/LFM2-24B-A2B-abliterated-GGUF: Quantization
1 other repos hold byte-for-byte identical weight files.