Z-L-Leo/my-sft-model-20260930-220416
Full fine-tune vs LiquidAI/LFM2.5-350M. Fine-tune: 100% of the weights changed, across 16 of 16 layers.
Tensors changed: Embeddings 1/2, other 30/30, MLP 48/48, Norms 19/44, Attention 24/24
Full fine-tune vs LiquidAI/LFM2.5-350M. Fine-tune: 100% of the weights changed, across 16 of 16 layers.
Tensors changed: Embeddings 1/2, other 30/30, MLP 48/48, Norms 19/44, Attention 24/24