mlx-community/granite-4.1-3b-bf16

Same weights, repacked vs ibm-granite/granite-4.1-3b. Same weights in different files (re-saved, re-sharded or new metadata).

Tensors changed: Embeddings 0/1, Norms 0/81, MLP 0/120, Attention 0/160

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

On the Hub · Downloads on Model Pulse