LiquidAI/LFM2.5-ColBERT-350M

Full fine-tune vs LiquidAI/LFM2.5-350M-Base. Fine-tune: 100% of the weights changed, across 16 of 16 layers.

Tensors changed: Embeddings 1/1, other 30/30, MLP 48/48, Norms 44/44, Attention 24/24

3 derivatives checked

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

On the Hub · Downloads on Model Pulse