Daniil-Domino/trocr-base-ru-dialectic-stackmix

LoRA-style vs raxtemur/trocr-base-ru. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of q_proj, query, v_proj, value changed, in 12 of 12 layers; norms, biases and embeddings are untouched.

Tensors changed: Embeddings 0/7, Attention 72/252, Norms 0/122, MLP 0/48, other 0/50

1 derivatives checked

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

On the Hub · Downloads on Model Pulse