ehristoforu/mistral-distil-test-3

Partial fine-tune vs unsloth/Mistral-Nemo-Instruct-2407. Partial change: 38% of the weights changed (attention, mlp, norms), in 13 of 29 layers.

Tensors changed: Embeddings 0/1, Norms 26/59, MLP 39/87, Attention 52/116, lm_head 0/1

On the Hub · Downloads on Model Pulse