MirilAI/Miril-DroneVLM-2B-2

Partial fine-tune vs google/gemma-4-E2B-it. Partial change: 37% of the weights changed (attention, embeddings, mlp, other), in 35 of 35 layers.

Tensors changed: MLP 105/585, Norms 0/433, other 70/240, Attention 100/687, Embeddings 1/6

1 derivatives checked

• mradermacher/Miril-DroneVLM-2B-2-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse